You are on page 1of 80

R E V I E W

Treatment With Thyroid Hormone

Bernadette Biondi and Leonard Wartofsky


Department of Clinical Medicine and Surgery (B.B.), University of Naples Federico II, 80131 Naples, Italy; and
Washington Hospital Center (L.W.), Washington, D.C. 20010

Thyroid hormone deficiency can have important repercussions. Treatment with thyroid hormone in replacement
doses is essential in patients with hypothyroidism. In this review, we critically discuss the thyroid hormone formu-
lations that are available and approaches to correct replacement therapy with thyroid hormone in primary and
central hypothyroidism in different periods of life such as pregnancy, birth, infancy, childhood, and adolescence as
well as in adult patients, the elderly, and in patients with comorbidities. Despite the frequent and long term use of
L-T4, several studies have documented frequent under- and overtreatment during replacement therapy in hypo-

thyroid patients. We assess the factors determining L-T4 requirements (sex, age, gender, menstrual status, body
weight, and lean body mass), the major causes of failure to achieve optimal serum TSH levels in undertreated patients
(poor patient compliance, timing of L-T4 administration, interferences with absorption, gastrointestinal diseases, and
drugs), and the adverse consequences of unintentional TSH suppression in overtreated patients. Opinions differ
regarding the treatment of mild thyroid hormone deficiency, and we examine the recent evidence favoring treat-
ment of this condition. New data suggesting that combined therapy with T3 and T4 could be indicated in some
patients with hypothyroidism are assessed, and the indications for TSH suppression with L-T4 in patients with eu-
thyroid multinodular goiter and in those with differentiated thyroid cancer are reviewed. Lastly, we address the
potential use of thyroid hormones or their analogs in obese patients and in severe cardiac diseases, dyslipidemia, and
nonthyroidal illnesses. (Endocrine Reviews 35: 433512, 2014)

I. Introduction B. Poor patient compliance


II. Methods C. Interference with absorption of L-T4
A. Identification of sources VII. Potential Adverse Effects During L-T4 Therapy
B. Methods of evaluation A. Adverse effects of overtreatment with L-T4
III. Hypothyroidism B. Cardiovascular risk and TSH suppression
A. Definition C. Fracture risk and TSH suppression
B. Etiology D. Adverse effects of increased serum FT4 during L-T4
C. Epidemiology therapy
D. Diagnosis VIII. Combined Treatment With L-T4 Plus Liothyronine in
IV. Treatment With Thyroid Hormone Hypothyroid Patients
A. Brief historical overview of T4 treatment for A. Persistence of symptoms in hypothyroid patients
hypothyroidism with normal serum TSH during L-T4 therapy
B. Feedback regulation of thyroid secretion B. The effects of low-normal serum TSH during L-T4
C. Thyroid hormone formulations: pharmacokinet- therapy
ics and pharmacodynamics
Abbreviations: AACE, American Association of Clinical Endocrinologists; AAP, American
V. Replacement Therapy With L-T4 in Hypothyroid Patients
Academy of Pediatrics; AF, atrial fibrillation; APS, autoimmune polyglandular syndrome;
A. Treatment of primary hypothyroidism ATA, American Thyroid Association; AUC, area under the curve; BMD, bone mineral den-
B. Determinants of L-T4 requirements and starting sity; BMI, body mass index; BP, blood pressure; CH, central hypothyroidism; CHD, coronary
dose of L-T4 heart disease; CI, confidence interval; CKD, chronic kidney disease; CoH, congenital hy-
pothyroidism; D1, deiodinase type 1; D2, deiodinase type 2; D3, deiodinase type 3; DITPA,
C. Follow-up and target serum TSH in patients re-
3,5-diiodothyropropionic acid; DTC, differentiated thyroid cancer; ECG, electrocardio-
ceiving replacement therapy with L-T4 gram; eGFR, estimated glomerular filtration rate; EndoSHyper, endogenous subclinical
D. Drugs interfering with L-T4 administration hyperthyroidism; ES, Endocrine Society; ETA, European Thyroid Association; ExoSHyper,
VI. Management of Persistent TSH Elevation in Patients exogenous subclinical hyperthyroidism; FT4, free T4; GI, gastrointestinal; HF, heart failure;
HR, hazard ratio; IQ, intelligence quotient; LDL, low-density lipoprotein; LV, left ventricular;
on High-Dose Replacement L-T4 Therapy
MC, myxedema coma; MCT8, moncarboxylase transporter 8; MI, myocardial infarction;
A. Timing of L-T4 administration MNG, multinodular goiter; NHANES, National Health and Nutrition Survey; OR, odds ratio;
PTC, papillary thyroid cancer; QOL, quality of life; RCT, randomized controlled trial; rhGH,
ISSN Print 0163-769X ISSN Online 1945-7189 recombinant GH; RR, relative risk; SHypo, subclinical hypothyroidism; SVR, systemic vas-
Printed in U.S.A. cular resistance; TBG, T4-binding globulin; T1DM, type 1 diabetes mellitus; Tg, thyroglob-
Copyright 2014 by the Endocrine Society ulin; THST, thyroid hormone TSH suppression therapy; TKI, tyrosine kinase inhibitor;
Received July 27, 2013. Accepted December 3, 2013. TPOAb, thyroid peroxidase antibody; TR, thyroid hormone receptor; TRIAC, 3,5,3-triiodo-
First Published Online January 16, 2014 thyroacetic acid.

doi: 10.1210/er.2013-1083 Endocrine Reviews, June 2014, 35(3):433512 edrv.endojournals.org 433

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
434 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

C. L-T4 therapy in thyroidectomized patients lion prescriptions in 2011 (IMS Institute for Healthcare
D. Polymorphisms in the genes of the deiodinases Informatics) (7).
E. Effects of combination treatment with T3 and T4 in Adequacy of L-T4 treatment is generally monitored by
adults
measurements of serum T4 and TSH. Although manage-
F. Combination treatment with T3 and T4 in children
with CoH ment of L-T4 therapy has been considered relatively
IX. Replacement Therapy in Central Hypothyroidism straightforward, several studies in the past 2 decades have
A. Treatment of CH in adults indicated aspects of previously unrecognized complexity
B. Limitations in peripheral markers of thyroid hor- despite the improvement in TSH and thyroid hormone
mone action assays (4, 8 11). Epidemiological studies have reported
C. Interactions with other pituitary hormone defi- that inadequate thyroid hormone replacement is present in
ciencies and hormone replacement
over a third of L-T4treated hypothyroid patients (1214)
D. Long-term cardiovascular risk in CH
E. Treatment of CH in pediatric patients despite frequent biochemical monitoring (15, 16). Under-
F. Combination treatment with T3 and T4 in CH or overtreatment with L-T4 leads to subclinical hypo- or
X. Replacement Therapy in Specific Hypothyroid Conditions hyperthyroidism, respectively (9, 10). These conditions
A. Pregnancy may induce adverse clinical consequences, especially in
B. SHypo in the elderly regard to specific periods of life (8 11).
C. Congenital hypothyroidism Consequently, new guidelines have clarified recom-
D. Treatment of SHypo in children and adolescents
mendations for replacement treatment with L-T4 in pa-
E. Replacement therapy with L-T4 in hypothyroid pa-
tients with comorbidities tients with persistent overt and subclinical hypothyroid-
XI. L-T4 Therapy in Benign and Malignant Thyroid Nod- ism (SHypo) (3), including specific suggestions in
ular Disease pregnant women (1719) and in children with congenital
A. L-T4 therapy in benign hypofunctioning thyroid hypothyroidism (CoH) (20 22). And more precise guide-
nodules lines have been formulated for the degrees of TSH-sup-
B. L-T4 therapy in DTC pressive therapy, with L-T4 indicated in patients with be-
XII. Appropriate and Inappropriate Use of Replacement
nign and malignant thyroid nodules (2325).
Therapy in Other Conditions
A. Fatigue (Wilsons syndrome) Controversy persists whether L-T4 monotherapy is ad-
B. Replacement therapy with thyroid hormone in equate or optimal in all patients because some hypothy-
obese patients roid patients may remain symptomatic despite biochem-
C. Effects of replacement therapy with thyroid hor- ical euthyroidism during treatment with L-T4 alone (26).
mone in severe cardiac diseases and dyslipidemia As a consequence, several studies have suggested a ratio-
D. Replacement therapy with thyroid hormone for nale and potential indication for combined treatment with
postoperative nonthyroidal illnesses
L-T4 and L-T3 in selected hypothyroid patients (2731).
E. Thyroid hormone therapy for the surgical or peri-
operative patient Recent commentaries have reviewed and evaluated the
F. Treatment of MCT8 deficiency potential role, limits, and adverse effects of this combina-
XIII. The Future of Replacement Therapy With Thyroid tion treatment (26, 32).
Hormone A final important point related to thyroid hormone
therapy is the potential usefulness of thyroid hormone
analogs for treatment of heart dysfunction and metabolic
I. Introduction disorders, opening a new scenario for pioneering indica-
tions or modulating applications of replacement therapy
eplacement therapy with thyroid hormone is indicated
R as a lifelong treatment when the diagnosis of persis-
tent thyroid hormone deficiency is confirmed (13). L-T4
with thyroid hormone (3335). Given the continuing con-
troversies in the management of replacement therapy with
thyroid hormone and its importance for clinical endocri-
is considered to be the treatment of choice in hypothyroid nologists, cardiologists, gynecologists, and pediatricians,
patients and has long been used in TSH-suppressive doses the aim of this review is to update advances in this treat-
in patients with nontoxic uninodular or multinodular goi- ment with all of its implications, limits, and adverse effects
ter (MNG) and differentiated thyroid cancer (DTC) to as well as potential future indications.
improve their prognosis (4).
L-T4 represents one of the most commonly adminis-
II. Methods
tered drugs in the world, and its use has been increasing in
the last several years (5, 6). In a report by the IMS Institute A. Identification of sources
of Healthcare Informatics, L-T4 was listed as the second We searched personal files, MEDLINE articles, meta-
most prescribed drug in the United States with 104.7 mil- analyses, and references of relevant articles and textbooks

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 435

27published from 1968 to 2013, as well as citations from mary hypothyroidism caused by underactivity of the thyroid
recently published international guidelines. Although the gland accounts for perhaps 99% of cases of hypothyroidism
primary focus was on articles in English, selected relevant (9, 10, 36). Secondary hypothyroidism, also known as CH,
translated articles were also included. Some pre-1968 re- is due to insufficient stimulation of a normal thyroid gland
ports of studies were also included absent more recent due to a deficiency in TSH as a result of hypothalamic or
information on a given issue. The following search terms pituitary disease (37, 38). CH is characterized by low thyroid
were used: L-thyroxine, L-triiodothyronine, thyroid hormone levels and inappropriately low-normal or slightly
hormone analogs, hypothyroidism, subclinical thyroid elevated serum TSH. It is estimated that less than 1% of cases
disease, replacement therapy, thyroid hormone TSH of hypothyroidism are due to TSH deficiency.
suppression therapy (TSHT), pregnancy, children, ad- Hypothyroidism can be further categorized on the basis
olescents, elderly, CoH, central hypothyroidism (CH), of its time of onset, eg, congenital or acquired, and on the
and myxedema coma (MC). basis of its severity, as in overt, subclinical, and mild dis-
B. Methods of evaluation ease (Table 1).
A critical assessment of the literature was performed. Traditionally, the term myxedema is reserved for severe
The 2 authors agreed on criteria for the inclusion or ex- and/or complicated thyroid hormone deficiency in adult
clusion of the studies considered. Preference was given to patients; the term is also used to indicate a nonpitting edema
high-quality papers, randomized controlled trials (RCTs), caused by the accumulation of glycosaminoglycans in sc and
longitudinal trials, and studies that appeared to have been interstitial tissues (3, 39, 40) On the other hand, cretinism
performed with correct statistical analysis and accurate refers to the complex severe syndrome of thyroid hormone
methods. When identified, shortcomings or limitations in deficiency (mental retardation, short stature, facial deformi-
study design or execution are described. ties, deafness, etc,) occurring in untreated CoH (41).
Overt hypothyroidism is a clinical condition in which
TSH is increased and free thyroid hormones (especially
III. Hypothyroidism
free T4 [FT4]) are low (36). SHypo is present when TSH is
A. Definition above the upper limit of the normal reference range and
Hypothyroidism is a pathological condition in which free thyroid hormones are within their reference range (9,
there is an insufficient production of thyroid hormones. Pri- 10). Patients with SHypo are often classified into 2 groups:

Table 1. Definition of Primary and Secondary Hypothyroidism According to Age and Physiological Conditions
Condition Definition
In adult patients
Primary overt hypothyroidism TSH 10 mU/L with low FT4 concentration
Severe hypothyroidism TSH 10 mU/L with low FT4 and FT3 concentrations
SHypo
Severe SHypo TSH levels 10 mU/L with normal FT4 concentrations
Mild SHypo TSH levels 4.59.9 mU/L with normal FT4 concentrations
Mild SHypo in elderly patients TSH levels 7.0 mU/L with normal FT4 concentrations
In CH Low serum TSH and low thyroid hormones
In pregnant women
Overt hypothyroidism Serum TSH 2.5 mU/L with decreased FT4 concentrations during the
first trimester of pregnancy
Serum TSH 3 mU/L with decreased FT4 concentrations during the
second and third trimester of pregnancy
Serum TSH 10 mU/L irrespective of FT4 concentrations
SHypo Serum TSH 2.510mU/L with normal FT4 concentrations
Maternal isolated hypothyroxinemia Normal TSH concentration with FT4 concentrations in the lower 5th or
10th percentile of the reference range.
In CoH
Suspicious CoH at first screening T4 below the 10th percentile and/or TSH 30mU/L
program
Confirmatory serum thyroid testing
Primary hypothyroidism TSH level 9 mU/L and free T4 0.6 ng/dL
SHypo TSH level 9 mU/L and normal free T4 (0.9 2.3 ng/dL)
CoH TSH level 9 mU/L and free T4 0.6 ng/dL
Persistent hyperthyrotropinemia Moderate TSH increase (510 mU/L) during follow-up
Transient CoH Normal thyroid function and normal TSH after L-T4 withdrawal and
during the subsequent follow-up

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
436 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

those with mild SHypo in whom TSH is mildly increased by 0.3 mU/L every 10 years after the age of 30 to 39 years
(TSH 4.59.9 mU/L) and those with a more severe dys- (49 51).
function when TSH is 10 mU/L (9, 10). The reanalysis of of NHANES III data 5 years later
In the past few years, the definition of hypothyroidism showed that the upper limit of normal serum TSH at the
and the target serum TSH for treatment of thyroid hor- 97.5th percentile was approximately 3.5 mU/L in individ-
mone deficiency have been modified by the results of as- uals 20 to 29 years old, 4.5 mU/L in people 50 to 59 years
sessments of normal thyroid function in important pop- old, 5.9 mU/L in elderly subjects 70 to 79 years old, and 7.5
ulation-based studies (12, 42, 43) and by the opinion of mU/L in those at 80 years and older (51). Therefore, it
experts regarding the upper limit of the normal serum TSH seems conceivable that a slightly increased serum TSH
concentration (44 47). The National Health and Nutri- might not always reflect mild thyroid hormone deficiency
tion Survey (NHANES) III, a population-based study of a in healthy elderly subjects because the distribution of se-
disease-free population in the United States, indicated that rum TSH shifts to higher concentrations with age due to
95% of adult subjects have a serum TSH concentration a presumed change in the set-point of the hypothalamic-
within the range of 0.45 to 4.12 mU/L, after the exclusion pituitary-thyroid axis with aging (47, 59). Consequently,
of subjects with a personal or family history of thyroid the diagnosis of SHypo in elderly patients should be con-
disease and /or positive thyroid autoantibodies (42). How- sidered only when serum TSH is above a cutoff limit for
ever, TSH values did not have a Gaussian distribution age, which will be higher in older subjects, whereas serum
because the curve was skewed by individuals with occult free and total T4 will be by definition within their reference
autoimmune thyroid dysfunction despite negative thyroid range. Moreover, age-adjusted serum TSH levels should
peroxidase antibodies (TPOAbs) and the inclusion of el- be considered during L-T4 replacement therapy (3, 10).
BMI may also influence serum TSH levels. TSH is
derly patients (42). Similar findings were reported by the
higher in overweight and obese individuals than in lean
more recent Hanford Thyroid Disease Study in a cohort
subjects. This appears to be an effect of the hormone, lep-
without evidence of thyroid disease, negative thyroid au-
tin, produced by adipocytes, which acts to stimulate re-
toantibodies, and normal thyroid ultrasound examination
lease of TRH by the paraventricular nucleus of the hypo-
(43). On the contrary, the National Academy of Clinical
thalamus (52, 60). The altered thyroid hormone pattern in
Biochemists in the United States indicates that 95% of
obese patients is reversible with weight loss, suggesting
individuals without evidence of thyroid disease will have
that the mild increased serum TSH level is not indicative
TSH concentrations below 2.5 mU/L (46). This last opin-
of SHypo in obese patients, especially when thyroid hor-
ion is supported by a recent meta-analysis that offered
mones are in their reference range and thyroid autoanti-
evidence for the association of high-normal serum TSH
bodies are negative (52).
with negative cardiovascular and metabolic effects (48).
During pregnancy, the thyroid gland increases its size
This meta-analysis indicates that there is an increased risk by about 10% in iodine-replete countries and by 20% to
of adverse cardiovascular outcomes (odds ratio [OR] 40% in the presence of iodine deficiency. The production
1.21; 95% confidence interval [CI], 1.151.27) and of of T4 and T3 is increased by 50%, and the daily iodine
adverse metabolic outcomes (OR 1.37; 95% CI, 1.27 requirement is increased by 50% (61). The TSH normal
1.48) in individuals with TSH levels in the upper part of reference range in pregnancy is influenced by high T4-
the reference range than in subjects with TSH levels in the binding globulin (TBG), estrogens, human chorionic go-
lower part of the reference range. nadotropin levels, increased iodine clearance, and en-
To further complicate the issue of what constitutes the hanced type III 5-deiodinase activity of the placenta (61).
normal TSH reference range, epidemiological studies have Recent guidelines have been updated to reflect recent data
emphasized that age, ethnicity, body mass index (BMI), bearing on establishing a correct diagnosis of thyroid hor-
iodine intake, genetic and environmental factors, and mone deficiency in pregnancy (1719). According to these
some particular physiological and pathological conditions guidelines, the upper limit for TSH should be 2.5 mU/L in
(pregnancy or illness) may also change serum TSH values the first trimester and 3.0 mU/L in the second and third
(49 55). Genetic variants in the phosphodiesterase 8B trimester (1719).
gene and in the TSH receptor may also influence the set- Consequently, hypothyroidism in pregnancy is defined
point of the hypothalamic-pituitary-thyroid axis (56 58). as an elevated serum TSH 2.5 to 3 mU/L with a decreased
Serum TSH level is higher in white populations than in serum FT4 concentration. Moreover, women with TSH
black ones, which suggests a genetic and ethnic/race in- levels of 10 mU/L or higher should be considered as having
fluence (49). Several studies have indicated an age effect, overt hypothyroidism, irrespective of their FT4 levels (18).
with the 97.5th percentile of serum TSH noted to increase On the basis of these criteria, SHypo in pregnant women

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 437

is defined as a serum TSH between 2.5 and 10 mU/L with inition of previously unrecognized entities such as sub-
a normal FT4 concentration (18). Isolated hypothyrox- clinical, mild hypothyroidism, isolated hypothyroxine-
inemia is defined as a normal maternal TSH concentration mia, and delayed TSH rise. The TSH and FT4 reference
in conjunction with FT4 concentrations in the lower 5th or limits defining these conditions should take into account
10th percentile of the reference range (18, 19). age, sex, race, BMI, iodine intake, method-specific varia-
CoH is a condition of thyroid hormone deficiency di- tions in the assay assessments, and physiological and path-
agnosed at birth (20 22). Primary CoH is defined by a low ological conditions. Different TSH cutoff levels should be
FT4 or total T4 concentration and elevated serum TSH used to define and treat hypothyroidism in children with
concentration, whereas the diagnosis of SHypo can be CoH, in pregnant women, in elderly patients, and in pa-
made as in adults, when TSH is increased and total and tients with comorbidities.
FT4 is normal (62). Infants with a low serum total T4 and The recent American Thyroid Association (ATA)
normal TSH concentrations may have CH, hypothalamic guidelines suggest that when upper and lower values of
immaturity, TBG deficiency, or maternal hyperthyroidism third-generation TSH assays are not available, the TSH
(62). A repeated serum FT4 with measurement by the equi- normal reference range should be 0.45 to 4.12 mU/L based
librium dialysis method is recommended to confirm a diag-
on the NHANES III reference population (3).
nosis of CH (62, 63). On occasion, a delayed TSH rise may
be seen and is defined as a normal TSH level with a low T4 B. Etiology
level at an initial screening for CoH, but a persistently low T4
level and an elevated TSH level at the subsequent screening 1. Transient and persistent adult hypothyroidism
evaluation (62, 63). This condition, which is frequent in pre- Transient hypothyroidism is defined as a period of re-
term infants, low-birth-weight infants and sick full-term duced thyroid function with elevated TSH, which is fol-
newborns, may reflect an altered hypothalamic-pituitary- lowed by recovery to a euthyroid state (9, 10). The major
thyroid axis secondary to immaturity (64). causes of acquired transient hypothyroidism are listed in
Finally, the entity of consumptive hypothyroidism re- Table 2. Transient causes of a TSH increase in adults are
fers to a rare condition that may occur in patients with frequently due to viral or autoimmune thyroiditis, drugs
large hepatic hemangiomas and other tumors in which that interfere with thyroid function, or a toxic injury af-
type 3 iodothyronine deiodinase is expressed, resulting in fecting the thyroid gland.
accelerated degradation of T4 and T3 (64). The term persistent subclinical and overt hypothyroid-
In conclusion, the definition of hypothyroidism has ism is defined by the presence of a permanent reduced
changed somewhat over the past several years, with def- secretion of thyroid hormones by the thyroid gland (9, 10,

Table 2. Etiology of Congenital and Acquired Transient Hypothyroidism


Condition Causes
Transient primary hypothyroidism in Transient hypothyroidism following subacute, painless, or postpartum thyroiditis
adults Drugs impairing thyroid function
Chemotherapy in patients with lymphoma and women with breast cancer
Toxic injury: pesticides, herbicides, industrial chemicals, and naturally occurring
environmental chemicals
Administration of antithyroid drugs
Transient CH in adults After withdrawal of thyroid hormone therapy in euthyroid patients
Drugs that inhibit TSH secretion: glucocorticoid, dopamine, cocaine, bexarotene
Nonthyroidal illnesses
Major surgery
Bone marrow transplantation
Severe trauma
Infections
Depressive disorders
Anorexia nervosa
Malnutrition
Transient CoH Maternal or neonatal excess iodine exposure
Maternal or neonatal mild iodine deficiency
Maternal treatment of hyperthyroidism with anti-thyroid drugs
Maternal transfer of TR-blocking antibodies
Heterozygous thyroid oxidase (THOX2) and DUOXA2 mutations
Congenital hepatic hemangioma/ hemangioendothelioma

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
438 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

36). The major causes of acquired persistent hypothyroid- Graves disease (68). Treatment with radioiodine for hy-
ism are listed in Table 3. perthyroidism may induce thyroid hormone deficiency ul-
Iodine deficiency is the single most common cause of per- timately in about 82% of patients with Graves disease
sistent hypothyroidism in the world (65, 66). Hashimotos over a 25-year period (69, 70). A lower percentage of
thyroiditis, an autoimmune disease often able to progres- hypothyroidism has been reported in patients with auton-
sively destroy the thyroid gland, represents the most com- omous thyroid nodules treated with radioiodine with a
mon cause of acquired subclinical or overt hypothyroid- progressive development of clinical hypothyroidism or
ism in adults in areas of iodine sufficiency (9, 10, 36). Its SHypo in 60% of patients with toxic adenoma after 20
incidence is about 7-fold in women with an increased risk years (71) and in 64% of patients with toxic multinodular
of development during the middle years of life (9, 10, 36). goiter after 24 years (72). Acquired hypothyroidism de-
The frequency of Hashimotos thyroiditis increases with velops after total and partial thyroidectomy (73, 74), ex-
age and is more common in people with a familial or per- ternal radiotherapy of the head and neck (75),and radia-
sonal history of other autoimmune endocrine or nonen- tion release from nuclear accidents (76) and in about 20%
docrine disorders (9, 10, 36). Hashimotos thyroiditis may to 25% of patients treated for malignancy with [131I]meta-
be associated with autoimmune polyglandular syndrome iodobenzylguanidine (77) or with monoclonal antibodies
(APS) (67). Autoimmune thyroiditis is present in about radiolabeled with 131I (78).
10% to 15% of cases with type 1 APS, and it is caused by Several drugs (iodine-containing compounds, lithium
mutations in the autoimmune regulator gene (AIRE), re- carbonate, cytokines, tyrosine kinase inhibitors (TKIs),
sulting in production of defective autoimmune regulator amiodarone, aminoglutethimide, ethionamide, sulfon-
protein (67); it may be associated with hypoparathyroid- amides, sulfonylureas, and thalidomide) can induce per-
ism, Addisons disease, and mucocutaneous candidiasis sistent subclinical or overt hypothyroidism especially in
(67). Type 2 APS, usually known as Schmidts syndrome, patients with preexisting thyroid autoimmunity (79).
is characterized by Addisons disease, autoimmune thy- Infiltrative or infectious disorders (amyloidosis, sclero-
roiditis, and type 1 diabetes (67). Hypothyroidism may derma, hemochromatosis and cystinosis, AIDS, and Ka-
spontaneously develop in 5% to 20% of patients with posis sarcoma) and inflammatory or infiltrative thyroid

Table 3. Etiology of Persistent Congenital and Acquired Primary Hypothyroidism


Condition Causes
Persistent primary hypothyroidism Severe iodine deficiency
in adults Chronic autoimmune thyroiditis
Non-autoimmune thyroiditis
Partial or total thyroidectomy and/or neck surgery
Radioactive iodine therapy for treatment of hyperthyroidism
External radiotherapy of the head and neck in patients with Hodgkins lymphoma, leukemia, aplastic anemia,
brain tumors, or bone marrow transplantation
Radiation release from nuclear accidents
Treatment with 131I meta-iodobenzylguanidine or with 131I-radiolabeled monoclonal antibodies
Infiltrative disorders of the thyroid gland (amyloidosis, sarcoidosis, hemochromatosis, Riedels thyroiditis,
cystinosis, AIDS, primary thyroid lymphoma)
Drugs impairing thyroid function in patients with autoimmune thyroiditis (iodine excess, iodine-containing
medications such as amiodarone and radiographic contrast agents, lithium carbonate, cytokines (especially
interferon-)
Inadequate replacement therapy
Toxic substances and industrial and environmental agents
Adult onset of partial defect of hormonogenesis
Persistent primary congenital Thyroid dysgenesis ( athyreosis, hemiagenesis, thyroid ectopia, thyroid hypoplasia)
hypothyroidism Thyroid dyshormonogenesis: sodium-iodide symporter (trapping) defect, thyroid peroxidase defect, hydrogen
peroxide generation or maturation defects, Tg synthesis defect, deiodinase defect
Resistance to TSH binding or signaling
TSH receptor defect
G protein defect
Familial infantile hypothyroidism Dyshormonogenesis
Generalized resistance to thyroid hormone
Acquired juvenile and Iodine deficiency
adolescent hypothyroidism Autoimmune thyroiditis
External irradiation of the neck
Thyroidectomy

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 439

Table 4. Etiology of Persistent Congenital and Table 5. Patients at High Risk of


Acquired CH Developing Hypothyroidism
Condition Causes High-Risk Patients
Persistent CH in Hypothalamic or suprasellar mass Individuals with a personal or family history of autoimmune
adults (germinoma, glioma and meningioma, thyroid disorders
craniopharyngioma, etc) Associated autoimmune endocrine conditions (T1DM,
Pituitary macroadenoma adrenal insufficiency, ovarian failure, etc)
Iatrogenic causes: surgery, irradiation, Associated nonendocrine autoimmune disorders (coelic
drugs, head trauma disease, vitiligo, pernicious anemia, Sjogren syndrome,
Infiltrative disorders causing multiple sclerosis, primary pulmonary hypertension, etc)
hypothalamic dysfunction (sarcoidosis, Postpartum women
hemochromatosis, cell hystiocytosis, Underlying thyroid, pituitary, or hypothalamic disorders
Langerhans cell) Previous treatment that can destroy thyroid, pituitary, or
Immunological disorders: lymphocytic hypothalamic tissue
hypophysitis Previous head and neck external radiation therapy
Pituitary apoplexy (including Sheehan Downs and Turner syndromes
syndrome) Subjects with a history of having given up smoking
Infectious disorders: tuberculosis, Patients submitted to bone marrow transplantation
syphilis, mycoses Smokers
Empty sella
Metastases
Persistent Isolated TSH deficiency (synthesis defect
congenital CH or hypoactivity)
is increased in Europe (1 in 100) compared with the United
Congenital hypopituitarism (multiple States and frequently occurs in preterm infants born in
pituitary hormone deficiencies) iodine-deficient areas (41). Frequent causes of transient
CoH are maternal thyroid autoantibodies, which can
cross the placenta and block the TSH receptor, fetal ex-
diseases (Riedels thyroiditis, subacute, postpartum, and posure to maternal antithyroid drugs, and maternal or
painless thyroiditis) may also induce persistent subclinical neonatal iodine exposure (41) (Table 2). Rare congenital
and overt hypothyroidism (9, 10, 36). Synthetic and nat- liver hemangiomas that produce large amounts of type 3
ural chemicals and industrial agents (pesticides, herbi- iodothyronine deiodinase may induce a consumptive hy-
cides, biphenyls, and resorcinol) may increase the risk of pothyroidism for which large doses of L-T4 are required
developing hypothyroidism (80). (41). Mutations in dual oxidase 2 (DUOX2) can be re-
CH can be due to pituitary gland adenomas, empty sponsible for transient or permanent CoH (41).
sella, pituitary surgery, radiotherapy, extrapituitary tu- Permanent CoH is defined as persistent thyroid hor-
mors, or inflammatory and infiltrative disorders of the mone deficiency that requires life-long treatment. It can be
pituitary gland (81) (Table 4). further classified into primary hypothyroidism caused by
Some special populations are at higher risk of develop- thyroid gland disease and secondary or CH due to a con-
ing hypothyroidism (Table 5), such as postpartum women genital TSH deficiency. Permanent primary hypothyroid-
and individuals with a personal and/or family history of ism may be caused by a defect in thyroid gland develop-
autoimmune thyroid disorders, autoimmune endocrine ment (thyroid dysgenesis) or a defect in thyroid hormone
conditions (type 1 diabetes mellitus, adrenal insufficiency, production (dyshormonogenesis) (Table 3) (41, 62). Thy-
ovarian failure, etc), or nonendocrine autoimmune disor- roid dysgenesis ( thyroid ectopy, athyreosis, and thyroid
ders (celiac disease, vitiligo, pernicious anemia, Sjogren hypoplasia) accounts for about 85% of cases of perma-
syndrome, multiple sclerosis, primary pulmonary hyper- nent primary CH, with thyroid ectopy being the most com-
tension, etc) (9, 10, 36). Subjects with Downs and Turner mon, accounting for approximately two-thirds of thyroid
syndromes are at increased risk of developing thyroid hor- dysgenesis (41, 62). Some gene mutations may be impli-
mone deficiency (9, 10, 36). Inexplicably, subjects with a cated in the development of thyroid dysgenesis, including
history of having given up smoking are at an increased risk PAX8, NKX2-1, and FOXE-1 (41, 62).
of hypothyroidism (82). Thyroid dyshormonogenesis accounts for about 10%
to 15% of cases of permanent CH, leading to goitrous
2. Transient and persistent CoH hypothyroidism (41, 62). These hereditary defects are au-
CoH can be classified as permanent and transient. tosomal recessive and include mutations in the genes cod-
Transient CoH refers to a temporary thyroid hormone ing for the sodium-iodide symporter, TPO, hydrogen per-
deficiency discovered at birth with subsequent recovery of oxide generation (thyroid oxidase and dual oxidase),
thyroid function (41, 62). The incidence of transient CoH transcription factors, thyroglobulin (Tg), and iodoty-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
440 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

rosine deiodinase. CH may be associated with other con- hypothyroidism has been reported in some cases. Some
genital malformations leading to syndromic forms of CH patients with an inactivating mutation in TR1 have been
such as Pendred, Bamford-Lazarus, and Kocher-Debr- reported; these patients had abnormal thyroid function
Semelaigne syndromes (41, 62). Pendred syndrome is a tests (low FT4, high T3, and normal TSH levels), with a
genetic syndrome characterized by a defect in the trans- phenotype characterized by mildly delayed cognitive de-
membrane protein pendrin, which leads to impaired velopment, growth retardation, delayed bone develop-
iodine organification. This syndrome is defined by the as- ment, and constipation (85, 86). Treatment with L-T4 im-
sociation of hypothyroidism, goiter, and deafness. Bam- proved certain features of this phenotype, although the
ford-Lazarus syndrome is a genetic condition that results impairment in cognitive and motor function did not im-
in thyroid dysgenesis and is due to recessive mutations in prove (87).
forkhead/winged-helix domain transcription factor FKLH15
or TTF2. It is characterized by hypothyroidism, cleft pal- 3. Hypothyroidism during infancy, childhood, and adolescence
ate, and spiky hair. Kocher-Debr-Semelaigne syndrome Infantile hypothyroidism is defined by the appearance
is characterized by myopathy and hypothyroidism asso- of hypothyroid symptoms after 6 months of age in infants
ciated with pseudohypertrophy in infancy or childhood. with CH that are not detected by screening (88, 89). Ec-
Rarely, a homozygous mutation in the human iodoty- topic thyroid dysgenesis, thyroid dyshormonogenesis, and
rosine deiodinase gene (DHEAL1) may lead to CH and generalized resistance to thyroid hormone action repre-
goiter. sent the most frequent causes of these late-onset congenital
Permanent secondary CH is usually caused by a muta- causes of hypothyroidism (88, 89). The clinical presenta-
tion of the TSH subunit gene or may be associated with tion is similar to that of infants with CH, and the intel-
congenital hypopituitarism (Table 4). Some mutations in lectual and neuropsycological impairment depends on to
genes regulating pituitary gland development may be re- the age of the onset of hypothyroidism and its duration (
sponsible for familial hypopituitarism (81). Mutations in Table 7).
transcription factor genes involved in pituitary gland de- Acquired juvenile hypothyroidism is defined by the on-
velopment or function (HESX1, LHX3, LHX4, PHF6, set of hypothyroidism after 3 years of age (88, 89). The
PITX1, PITX2, OTX2, SOX2, SOX3, PROP1, and appearance of thyroid hormone deficiency in this period of
POU1F1) have been reported. Peripheral CH is defined as the life is not associated with a permament impairment of
a defect in thyroid hormone transport, metabolism, or the central nervous system (Table 7).
action (Table 6). Uncommon causes of peripheral CH in- Acquired hypothyroidism during adolescence is usually
clude defects in thyroid hormone transport (mutations in a less severe form of thyroid hormone deficiency (88, 89)
the gene for monocarboxylase transporter 8 [MCT8]), (Table 7). Acquired hypothyroidism may develop after
metabolism (selenocysteine insertion sequence-binding external irradiation or bone marrow transplantations to
protein 2), or resistance to thyroid hormone action (mu- treat tumors or primary immunodeficiency during infancy
tations in the thyroid hormone receptor [TR]). and adolescence. The incidence of autoimmune hypothy-
A mutation in a gene encoding MCT8 is a rare cause of roidism during adolescence is approximately 1% to 2%
X-linked hypothyroidism with mental retardation and and may occur in association with other autoimmune dis-
neurological abnormalities (Allan-Herndon-Dudley syn- eases, such as type 1diabetes mellitus, alopecia, vitiligo, or
drome) (83). Addisons disease or in association with other disorders
Peripheral resistance to thyroid hormone is caused by a such as Downs syndrome, Turner syndrome, cystinosis,
mutation in the gene encoding for TR in 90% of cases and thalassemia. Children whose family members have
(84). These patients are usually euthyroid with elevated autoimmune thyroid diseases are at increased risk of au-
thyroid hormones without suppression of TSH; however, toimmune hypothyroidism.

C. Epidemiology
Table 6. Causes of Persistent Peripheral Hypothyroidism
Hypothyroidism is one of the most common endocrine
Cause disorders. The prevalence of clinical hypothyroidism in
Resistance to thyroid hormone adults is 1.9% in women and 0.1% in men (90) with an
Mutation in the gene encoding for TR in 90% of cases annual incidence of 0.4% in women and 0.06% in men
Inactivating mutation in TR1 (rare cases)
Thyroid hormone transport defect (MCT8, Allan-Herndon-Dudley
(91). Its prevalence is increased in men and women older
syndrome) than 60 years of age, being 0.5% to 5.0% (92). The prev-
Thyroid hormone metabolism defect (selenocysteine insertion sequence alence of SHypo is about 3% to 10% in iodine-replete
binding protein)
population (8 10, 90, 91). This prevalence increases to

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 441

Table 7. Symptoms and Signs of Primary and CH in Different Periods of the Life
Condition Symptoms and Signs
Primary overt hypothyroidism
In young and middle-aged patients Fatigue
Weight gain
Dry skin and cold intolerance
Slow thinking
Puffy eyes
Coarseness or loss of hair
Goiter
Constipation
Cold intolerance
Hoarseness
Muscle weakness and cramps
Delayed relaxation of deep tendon reflexes
Memory impairment
Decreased concentration
Neurocognitive deficits
Irregular or heavy menses
Infertility
Myalgias
Hyperlipidemia
Bradycardia
Hypothermia
CHD
Mixedema in severe hypothyroidism
In elderly patients Memory and mental impairment
Dementia
Anemia
Depression
Heart failure
CHD
Neonatal appearance Prolonged icterus
Edema of the eyelids, hands, and feet
Hypotonia
Inactivity
Gestation 42 wk
Birth weight 4 kg
Poor feeding
Hypothermia
Abdominal distention
Open posterior fontanelle (5 mm)
During the first month of age Respiratory distress
Failure to gain weight, poor sucking ability
Decreased stool frequency
Decreased activity and lethargy
During the first 3 mo of age Umbilical hernia
Dry skin with carotenemia
Macroglossia
Infrequent and hard stools
Generalized swelling or myxedema
Hoarse cry
Between 6 mo and 3 y of age Deceleration of linear growth
Coarse facial features
Dry skin with carotenemia
Hoarse cry and large tongue
Umbilical hernia
Muscular pseudohypertrophy
In childhood Deceleration of linear growth with or without short stature
Delayed eruption of teeth and in shedding of primary teeth
Skeletal retardation
Muscle weakness
Muscular pseudohypertrophy
(Continued)

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
442 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

Table 7. Continued
Condition Symptoms and Signs
Infrequent and hard stools
Dry skin with carotenemia
Precocious sexual development with incomplete sexual maturation
Isolated menarche ,
Hyperprolactinaemia and galactorrhea
Changes in mental and physical performance
Enlargement of the sella turcica and the pituitary gland
Generalized swelling or mixedema
During adolescence Delayed onset of puberty
Deceleration of linear growth with or without short stature
Delayed eruption of permanent teeth
Infrequent and hard stools
Dry skin with carotenemia
Galactorrhea (girls)
Generalized swelling or myxedema
CH Mild-moderate symptoms of hypothyroidism
Signs and symptoms of other pituitary deficits
Manifestations of concomitant hypothalamic-pituitary disease

about 10% in women older than 55 to 60 years of age and thyroiditis was found in 55% of patients with SHypo and
rises further to 5% to 20% in men and women older than 80% of those with overt disease (98).
60 years of age (8 10, 12).
This wide range of prevalence of hypothyroidism re- D. Diagnosis
flects the heterogeneity in race, age, sex, BMI, and dietary 1. In adult patients
iodine intake among different populations and the differ- Overt hypothyroidism is a clinical condition and can be
ent TSH cutoff values used for the definition of SHypo in suspected in the presence of specific symptoms of thyroid
different studies (9, 10). About 75% of patients with
hormone deficiency (Table 7) (99). However, the onset of
SHypo have mild thyroid deficiency (TSH 10 mU/L) (9,
clinical symptoms in hypothyroid patients may be influ-
10). Patients with SHypo may progress to overt disease,
enced by the severity of the disease, duration, age, and the
with the rate and frequency of progression affected by
individual sensitivity to thyroid hormone deficiency (9).
iodine intake, sex, the underlying cause of thyroid hor-
Elderly patients with thyroid hormone deficiency may be
mone deficiency, basal TSH, FT4 levels, and the patients
minimally or completely asymptomatic (100, 101).
age (9, 10). A serum TSH above 2.5 mU/L and positive
In symptomatic patients, the onset of symptoms may be
thyroid autoantibodies were predictors of long-term risk
attributed to the aging process or concomitant diseases
of hypothyroidism in some longitudinal studies (89, 93
(9). Moreover, atypical clinical presentations of hypothy-
95). The annual rate of progression to overt disease is
particularly increased (4.3%) in women with elevated se- roidism are prevalent in elderly subjects (Table 7) (102
rum TSH and antithyroid antibodies (9, 10). Other factors 105). Pericardial and pleural effusion, hypothermia, and
associated with an increased risk of progression to overt coma may develop in patients with more severe disease
hypothyroidism include high iodine intake, a greater in- (103107) (Table 7).
crease in serum TSH (10 mU/L), positive thyroid auto- Although several rating scales have been proposed for
antibodies, low-normal FT4 levels, and adult age (9396). the diagnosis of hypothyroidism and for the assessment of
The incidence of overt hypothyroidism in pregnant the severity of specific symptoms (108 110), it is difficult
women is approximately 0.3% to 0.5%, whereas the in- to diagnose thyroid hormone deficiency on the basis of
cidence of SHypo is approximately 2% to 3% (9, 10, 97). clinical symptoms alone because none of the symptoms or
Worldwide iodine deficiency is the main cause of hypo- signs of hypothyroidism is sufficiently sensitive or specific
thyroidism in pregnancy (65). Chronic autoimmune to distinguish euthyroid subjects from patients with mild
thyroiditis represents the most frequent cause of thyroid thyroid hormone deficiency (99). The Nijmegen Biomed-
hormone deficiency in pregnant women living in iodine- ical Study, which assessed the relationship between fatigue
sufficient areas. Thyroid autoantibodies are positive in and serum TSH, FT4, and TPOAbs in 5897 participants,
5% to 15% of women during pregnancy. In a prospective showed that subjects with a history of thyroid disease ex-
population study on 9471 pregnant women, autoimmune perienced more fatigue than the general population de-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 443

spite the normal TSH and FT4 values (111). In the Colo- TSH should be excluded before commitment of patients to
rado study, the positive predictive value of individual a lifetime of T4 replacement therapy. A laboratory pattern
hypothyroid symptoms was only 8% to 12% (97). How- indistinguishable from SHypo may be rarely seen in pa-
ever, this study showed that patients who report many or tients with TSH receptor mutations causing mild TSH re-
newly developed symptoms are most likely to have thyroid sistance. These subjects usually relate a family history of
hormone deficiency, although absence of symptoms does raised serum TSH concentrations but an absence of
not exclude the diagnosis of thyroid hormone deficiency. personal or familial history of thyroid autoimmunity. Pi-
As a result, the presence of specific and suggestive symp- tuitary resistance to thyroid hormones and thyrotropin
toms may be useful only to select patients who need thy- secreting pituitary adenoma may be associated with in-
roid function testing, with the diagnosis of hypothyroid- creased serum TSH; however, both of these conditions are
ism in adults made on the basis of the biochemical test characterized by raised serum free thyroid hormone levels.
results. Although TSH is the first-line test to perform when
Hypercholesterolemia, hypertriglyceridemia, hyperpro- hypothyroidism is suspected, measurement of thyroid
lactinemia, hyperhomocysteinemia, hyponatremia, anemia, hormones can provide additional information as to the
and/or elevated creatine phosphokinase levels may raise sus- cause of the hypothyroidism and is necessary to achieve a
picion of hypothyroidism sufficiently to warrant thyroid correct diagnosis before replacement therapy is initiated.
function testing (36, 112). Consequently, serum TSH assay alone may be insuffi-
Primary hypothyroidism is the major cause of hypo- ciently sensitive for the diagnosis of 1) patients with CH
thyroidism, and serum TSH measurement is the most sen- due to hypothalamic or pituitary disorders, 2) subjects
sitive and specific test in the presence of an intact pituitary- recovering from nonthyroidal illnesses, 3) administration
hypothalamic axis, with immunometric assays for TSH of some drugs that can suppress TSH, 4) overweight and
having a 99% sensitivity and specificity (36). Serum TSH obese subjects, 5) short-term withdrawal of thyroid hor-
is the first-line diagnostic test for the identification of thy- mone therapy in euthyroid subjects, 6) presence of het-
roid hormone deficiency, even in patients with mild thy- erophilic antibodies against mouse proteins in some im-
roid hormone deficiency (9, 10, 36, 113). Elevated serum munoassays that may falsely raise serum TSH, and 7)
TSH and decreased FT4 levels represent the classical com- patients with untreated adrenal insufficiency.
bination indicating primary hypothyroidism. Serum FT4 The concomitant presence of other disorders and/or
RIA and serum FT4 index are adequate but less sensitive nonthyroidal illnesses may severely limit the diagnosis of
methods to detect hypothyroidism, especially SHypo (9, hypothyroidism in the elderly. Elderly patients with raised
10, 36). In fact, FT4 levels are in the lower-normal refer- serum TSH are hypothyroid only after excluding the ad-
ence range in patients with SHypo and are associated with ministration of drugs that can interfere with thyroid func-
raised TSH values; thus, the rising TSH in this condition tion or the state of recovery from severe illness. Some drugs
is based on the sensitivity of the hypothalamic pituitary may interfere with TSH synthesis and secretion; glucocor-
thyroid axis to the slight decreases from baseline normal ticoids, dopamine, and dopamine agonists (bromocriptine
T4 levels to the low-normal range. and cabergoline), somatostatin analogs (octreotide), do-
Serum FT3 evaluation is only marginally useful in di- butamine, and retinoids may all suppress TSH at the level
agnosing hypothyroidism and may achieve low serum lev- of the hypothalamus or pituitary (115). Moreover, met-
els only in severe hypothyroidism. Significant decreases in formin can lower serum TSH and was shown to do so in
FT3 levels are not seen until patients are in an advanced one small study in type 2 persons with diabetes who also
state of thyroid hormone deficiency due to the compen- had hypothyroidism (116) and in euthyroid patients with
satory increase in hepatic T4 to T3 conversion and to re- higher baseline TSH levels independent of the presence of
sidual thyroidal T3 secretion (36). Recent data have shown positive TPOAb (117).
that a single evaluation of thyroid-function tests and bi- During the recovery phase from nonthyroidal illness,
ological variation in thyroid test results may severely limit serum TSH levels may be increased to levels above normal,
the diagnosis of true persistent SHypo, supporting the although generally no higher than 20 mU/L (118).
need for a periodic follow-up without treatment before About 90% of patients with Hashimotos thyroiditis
diagnosing this disorder in presence of mild TSH increase have detectable anti-TPO and anti-Tg antibodies. The
(114). Serum TSH concentration should be reevaluated evaluation of thyroid antibody titers (which include an-
after 3 to 6 months in the event of slightly increased values ti-Tg antibodies, antimicrosomal/TPOAbs, and anti-TSH
to exclude a laboratory error or a transient TSH increase receptor antibodies) is useful to confirm the diagnosis of
due to temporary thyroiditis recovery from systemic ill- autoimmune hypothyroidism and predict the subsequent
ness or drugs (9, 10). Transient causes of increased serum development of overt hypothyroidism in patients with

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
444 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

mild disease, in pregnant women, in the postpartum pe- and their interpretation is complicated by the increased
riod, in the presence of other autoimmune disease, and TBG levels and the decreased albumin concentrations.
during treatment with drugs such as lithium, interferon-, These changes may affect and limit the validity of the FT4
and amiodarone or exposure to excessive amounts of measurement by immunoassay methods during preg-
iodine. nancy. In addition, there are large method-dependent vari-
The hypoechogenicity of the thyroid gland at the ul- ations in FT4 assessments during pregnancy, and there-
trasound evaluation may allow clinicians to identify indi- fore, incorrect abnormal FT4 levels can be detected with
viduals with autoimmune thyroiditis (9). The ultrasono- the different FT4 assays in the laboratories. The best rec-
graphic pattern typical of autoimmune thyroiditis may be ognized technique for FT4 measurement in pregnancy is
useful to provide a correct diagnosis in about 10% of the separation of the free hormone from binding globulins
patients with no detectable thyroid autoantibodies (9). by equilibrium dialysis or ultrafiltration of serum, fol-
CH can be suspected in the presence of specific symp- lowed by thyroid hormone measurement by either RIA or
toms of thyroid hormone deficiency in patients with clin- liquid chromatography tandem mass spectrometry (120).
ical findings of other pituitary hormone deficiency or in However, this latter method is expensive. As an alterna-
the presence of a history of pituitary gland tumors and tive, the recent Endocrine Society (ES) guidelines recom-
surgery or pituitary inflammatory and infiltrative disor- mend the use of the FT4 index or the total T4 adjusted for
ders (38, 81). TSH measurement is of little value in guiding pregnancy to estimate FT4 values in pregnancy (19). Ac-
management in CH because the deficiency in central dis- cording to these suggestions, total T4 should be adjusted
ease is due to the low levels of TSH. Instead, SHBG, ankle by a factor of 1.5 in pregnancy. The FT4 index should be
reflex relaxation time, heart rate, and Doppler echocar- measured as total T4 mathematically corrected for TBG
diography may be useful surrogate peripheral tissue pa- (19).
rameters to assess the restoration of euthyroidism during
treatment with thyroid hormone (81). Although poten- 3. Diagnosis of CoH
tially useful to guide replacement hormone dosage, none Infants with CoH can be detected by newborn screen-
of these tissue parameters is sufficiently sensitive or spe- ing programs. Congenital hypothyroidism meets the pub-
cific for actual diagnosis. lished accepted criteria for a mass population screening
because of the following points: 1) the disorder is infre-
2. Diagnosis of hypothyroidism in pregnancy quent (the incidence in Europe is 1 in 2708 births and in
The diagnosis of thyroid hormone deficiency in preg- the United States was 1 in 2372 in 2002), 2) the clinical
nancy can be difficult. Serum TSH decreases in the first diagnosis is made in only 10% of cases within the first
trimester of pregnancy secondary to the physiological in- month of life, 3) the delay or lack of diagnosis may induce
crease in human chorionic gonadotropin and the resultant severe mental retardation, 4) the cost to benefit ratio is
stimulation of thyroid hormone release. The use of a nor- favorable because it is a preventable disorder with an ef-
mal reference range for nonpregnant women did not iden- fective method of screening, and 5) an effective therapy
tify 10.4% of women with thyroid dysfunction in the first (with L-T4) is available. The first newborn screening pro-
trimester and 6.7% in the second trimester (119). There- gram was introduced in Canada in the late 1970s, and now
fore, the current guidelines suggest considering a trimes- this screening is available for all newborns in most devel-
ter-specific reference range for both TSH and thyroid hor- oped countries (United States, Europe, Australia, and Ja-
mones during pregnancy (1719). If trimester-specific pan) (121, 122). Changes in recent clinical screening prac-
reference ranges for TSH are not available in the labora- tices (reduction of TSH cutoff) and assay techniques (from
tory, the following reference ranges are recommended: 0.1 a radiochemical-based assay to fluroimmunoassay or
to 2.5 mU/L in the first trimester, 0.2 to 3.0 mU/L in the chemiluminescence for determining both TSH and T4 in
second trimester, and 0.3 to 3.0 mU/L in the third trimester dried blood spots) have allowed detection of milder cases
(1719). of CoH. The typical screening test for CoH is performed
The ATA guidelines (18) defined maternal hypothy- by heel-prick blood that is spotted on special filter paper
roxinemia as low FT4 concentrations (5th or 10th per- cards. It is usually collected between 2 and 4 days after
centile) with normal TSH levels. Moreover, the recently birth on cord blood. Earlier collection may lead to a false-
updated guidelines have recommended being prudent in positive result due to the physiological postnatal rise in
interpreting serum FT4 levels during pregnancy. They serum TSH.
have suggested the use of a trimester-specific FT4 reference Some screening programs use a primary T4 test with a
range, which should be method specific (19). Serum FT4 follow-up TSH test in infants with a serum T4 below the
concentrations decrease with progression of gestation, cutoff value; this strategy helps detect cases with primary

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 445

and secondary hypothyroidism and infants with delayed To summarize the issue of diagnosis, although hypo-
TSH rise (123). Some other screening programs perform thyroidism can be suspected by the onset of specific symp-
an initial TSH test that has the advantages of detecting toms of thyroid hormone deficiency, it should be specifi-
infants with mild or SHypo. Simultaneous T4 and TSH cally confirmed on the basis of results of thyroid function
measurement appears to represent the best strategy, al- tests. The diagnosis of hypothyroidism can be difficult in
though it is more expensive and used in a minority of children with CoH, in CH, in elderly patients, in pregnant
screening centers (124). women, in subjects with comorbidities or in those receiv-
With the more sensitive TSH assays currently in use, the ing drugs that can interfere with thyroid function. Periph-
new TSH cutoff has declined from 40 to 50 mU/L to serum eral parameters of thyroid hormone action may help to
values of 20 to 30 mU/L (corresponding to 10 15 mU/L in diagnose CH.
whole blood). However, age-specific normal reference
values for TSH and FT4 should be considered to establish
specific cutoff levels. Diagnosis and treatment should not IV. Treatment With Thyroid Hormone
be based on screening test results alone. Generally, if the
screening T4 is below the 10th percentile and/or the TSH A. Brief historical overview of T4 treatment
is greater than 30 mU/L, an infant is recalled for confir- for hypothyroidism
matory serum testing (41, 62) (Table 1). The confirmatory A thyroid hormone preparation for the treatment of
serum tests may be a TSH and freeT4 or a total T4 com- hypothyroidism was first administered by Dr George
bined with some measures of binding proteins such as a T3 Murray in 1891 by the im injection of sheep thyroid ex-
resin uptake. These tests can be performed after 1 or 2 tract into a patient with myxedema. In just a few years
weeks of life when the upper limit of serum TSH is ap- thereafter, preparations of thyroid extracts were admin-
proximately 10 mU/L and FT4 in the range of 10 to 26 istered by mouth (125). In 1914, Edward Kendall isolated
pmol/L (41, 62) (Table 1). A routine second screening test and crystallized the active substance in the thyroid that he
at 2 to 6 weeks of age may help diagnose a delayed TSH initially named thyroxin, and it was Harrington who es-
rise in preterm and acutely ill term infants who may not tablished its crystalline structure in 1926 (126). Desic-
have an elevated serum TSH at the first screening test (41, cated thyroid, made from animal thyroid glands, remained
62). It is important to establish the etiology of the hypo- the main form of therapy until about 1960 (126). This
thyroidism for the future management of children with extract contains a combination of T4 and T3 in a ratio 2-
CoH diagnosed at birth. In most cases, the underlying to 3-fold higher than that found in human thyroid. The
defect in thyroid function may be uncovered with a thy- development of the sodium salt of L-T4 in the 1950s pro-
roid radionuclide uptake and scan, thyroid ultrasonogra- vided the compound that today is the basis for the therapy
phy, serum Tg measurement, thyroid autoantibodies, and of hypothyroidism. Simultaneous publications in 1952 by
urinary iodine measurement (62). 123I or sodium pertech- Gross and Pitt-Rivers in the United Kingdom and by
netate 99m (Tc99m) rather than 131I should be used in Roche, Lissitsky, and Michel in France, reported the dis-
neonates to avoid a high dose of radioactivity in the thy- covery of T3 (126). Almost 20 years later in 1970, Braver-
roid and total body. Radionuclide uptake and scan may man, Sterling, and Ingbar demonstrated that circulating
identify thyroid aplasia, hypoplasia (decreased uptake, T3 is largely derived from T4 deiodination in extrathyroi-
small gland in a eutopic location), or an ectopic gland. dal tissues by detecting T3 in the serum of athyreotic pa-
Infants with absent uptake should be further evaluated by tients receiving L-T4 (127). Synthetic preparations of L-T3
thyroid ultrasonography. A large gland or goiter with in- were available in 1956, whereas synthetic preparations of
creased uptake suggests dyshormonogenesis. A perchlor- L-T4 became accessible in 1958. Until 1962, the U.S. Food

ate discharge test may be helpful in the identification of and Drug Administration (FDA) in the United States did
defective oxidation and organification. Genetic counsel- not require a new drug application for L-T4, based on the
ing is necessary in familial forms of CoH to diagnose the belief that it was not a new drug. Replacement doses of
specific gene mutation. In the absence of a screening pro- L-T4 were higher in 1960 (about 200 400 g daily) com-
gram, the diagnosis of CoH will be made after the onset of pared with the doses recommended today; the poorly sen-
clinical manifestations (Table 7). Clinical manifestations sitive hormone assays in the past did not allow clinicians
usually depend on the severity of the thyroid hormone to distinguish between low-normal and suppressed TSH
deficiency and may progressively develop after birth. Neu- levels (9). Total T4 and FT4 levels were frequently in-
rological abnormalities and genitourinary and cardiac creased at that time in patients being treated with L-T4 (9).
congenital malformations should be investigated in in- The more highly sensitive TSH assay became available in
fants with CoH. 1980, and by the early 1990s, the first studies showed the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
446 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

negative effects of TSH suppression on the heart, liver, and is synthesized and secreted exclusively by the thyroid
bone (9). Between the years 1990 and 2000, the first dou- gland, whereas the major source of human extrathyroid-
ble-blind controlled studies evaluated the effects of L-T4 in produced T3 in the euthyroid state is derived by D2 activity
patients with SHypo (9). In 1970, Smith et al (128) pub- (133). The major role of D2 is to control the intracellular
lished the first double-blind crossover study to assess the T3 concentration to protect tissues from the detrimental
effects of combination therapy with T3 and T4 in patients effects of hypothyroidism (131133).
with hypothyroidism. The efficiency of conversion of T4 to T3 by D2 increases
as the serum T4 decreases (26). Consequently, in the pres-
A. Feedback regulation of thyroid secretion ence of a low level of T4 or in case of a hypothyroid state,
Thyroid hormones are produced by the thyroid gland D2 expression and activity are increased and can generate
through the iodination of tyrosine residues in the glyco- a proportionately greater quantity of plasma T3 (26).
protein Tg. The anterior pituitary releases TSH in re-
sponse to TRH, which is secreted by hypothalamus (129). C. Thyroid hormone formulations: pharmacokinetics
TSH acts directly on the TSH receptor expressed on the and pharmacodynamics
thyroid follicular cell membrane. TSH regulates iodide Thyroid hormone preparations available for therapy fit
uptake mediated by the sodium/iodide symporter and into 2 categories: natural hormonal preparations derived
stimulates thyroid hormone synthesis and secretion (26, from animal thyroid and synthetic preparations.
129). The thyroid gland secretes both T4 and T3, which
exert a negative feedback on TRH and TSH secretion. 1. Thyroid extracts (thyroid USP)
TSH secretion by the pituitary gland is the result of a com- Natural preparations include desiccated thyroid and
plex feedback interaction between central hypothalamic Tg. Desiccated thyroid products (dried and powdered) are
TRH control and thyroidal production of thyroid hor- derived from the thyroid glands of domesticated animals
mones to the periphery. This negative feedback loop main- that are used for food by man (either beef, hog, or sheep),
tains levels of the circulating thyroid hormones and TSH and Tg is derived from the thyroid glands of the hog. The
in a physiological inverse relationship that defines the hy- content of thyroid hormone and the ratio of T4 to T3 may
pothalamic-pituitary-thyroid axis set-point, which is ge- vary in desiccated thyroid preparations, depending on the
netically determined and variable among individuals brand employed and whether it is of porcine or bovine
(130). The set-point for TSH secretion depends on both origin. Desiccated thyroid contains approximately 80%
circulating serum T3 and intracellular pituitary T3. It may T4 and 20% T3 as well as other iodinated compounds
be influenced by environmental factors, iodine intake, (eg, diiodotyrosine and monoiodotyrosine). Absorp-
age, and systemic illness as the organism autoregulates tion studies indicate that the bioavailability of T3 in
deiodinase activity and the resultant conversion of T4 to desiccated thyroid is comparable to that of orally ad-
T3 (26). ministered synthetic T3.
More than 99% of circulating thyroid hormones are The most commonly used form of desiccated thyroid,
bound to plasma proteins TBG, T4-binding prealbumin known as Armour Thyroid, is of porcine origin and may
(transthyreitin), and albumin. The higher affinity of both be viewed as a mixture of T3 and T4 (1 grain, about 60 mg
TBG and T4-binding prealbumin for T4 vs T3 partially desiccated pig thyroid extract is approximately equivalent
explains the higher serum T4 levels and its slower meta- to 88 g L-T4) (134). Other commercial preparations de-
bolic clearance and longer half-life compared with T3. rived from animal thyroid extracts are Prothyroid, Novo-
Only the unbound or free thyroid hormone concentration thyral, Thyreotom, Thyrolar-3, and Diotroxin.
is metabolically active. Some formulations of thyroid extracts have a charac-
The availability of T3 at the tissue level is regulated by teristic odor and may be antigenic being an animal protein-
3 deiodinase isoforms termed deiodinase type 1 (D1), type derived product. A recent randomized, double-blind,
2 (D2), and type 3 (D3) (131, 132). D1 is expressed mainly crossover study has assessed the efficacy of desiccated thy-
in the thyroid gland, liver, and kidney, where it converts T4 roid treatment vs L-T4 in 70 patients (age 18 65 years)
to T3 and thus contributes significantly to the pool of cir- with primary hypothyroidism who were on a stable L-T4
culating T3. D3 protects tissues from thyroid hormone dosage for 6 months (134). The authors evaluated symp-
excess by decreasing local T3 concentrations (131, 132). toms, cognitive function, and the quality of life (QOL).
D2 catalyzes 5-deiodination and converts T4 to T3; it is The results have shown that although the treatment with
present in the brain, pituitary gland, skeletal muscle, desiccated thyroid hormone did not improve the QOL;
brown adipose tissue, thyroid gland, osteoblasts, aortic 48.6% of these patients preferred this therapy vs L-T4 be-
smooth muscle cells, and the human heart (131, 132). T4 cause it was associated with weight loss (134).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 447

The inappropriate use of thyroid extracts in euthyroid same clinical effect and safety profile as the reference prod-
and hypothyroid patients can result in thyrotoxic symp- uct to which they are compared (145148). In 1997, the
toms and severe adverse effects (135). Some cases of thy- FDA changed the process for approving L-T4 preparations
roid storm have been reported during the inappropriate after receiving 58 adverse drug event reports over a 7-year
use of thyroid hormone extracts (136). The recent Amer- period. However, since 1997, whether or not different
ican Association of Clinical Endocrinologists (AACE) and L-T4 preparations can be substituted has been a matter of
ATA guidelines advise that there is no evidence to support great controversy (149 151). Two medicinal products
using desiccated thyroid hormone in preference to syn- containing the same active substance are considered bio-
thetic L-T4 monotherapy and advise that desiccated thy- equivalent if they are pharmaceutically equivalent or phar-
roid hormone not be used for treatment of hypothyroid- maceutical alternatives with a similar record of safety and
ism particularly due to content of a nonphysiological efficacy. Bioavailability refers to the rate and extent to
proportion of T3 (3) (recommendation 22.4). which the active ingredient or therapeutic ingredient is
absorbed from a drug product and becomes available at
2. Levothyroxine the site of drug action. Bioequivalence, in turn, establishes
L-T4 represents the first-line treatment in patients with therapeutic equivalence and, therefore, the interchange-
hypothyroidism. Although plasma T4 reaches a peak con- ability of 2 similar products. Therapeutic equivalence per-
centration 2 to 4 hours after oral administration, a once- mits substitution of one drug for another with the expec-
daily dose of L-T4 each morning provides stable and rel- tation of similar clinical effects with the expectation that
atively constant blood levels of T4 because of its long half- strict follow-up testing would not be required.
life (about 7 days) (26). The absorption of oral L-T4 takes The FDA approves a generic substitute only if it has
place through the intestinal mucosa within the first 90 been proven to be identical or bioequivalent to a brand-
minutes after tablet ingestion with 21% absorbed in the name drug in dosage form, safety, strength, route of ad-
duodenum, 45% in the upper jejunum, and 34% in the ministration, quality, performance characteristics, and in-
lower jejunum and ileum (137139). Absorption is in- tended use. The FDA requires that the 90% CIs of the
complete because only 70% to 80% of the administered generics pharmacokinetic properties fall within the 80%
dose is absorbed. Normal gastric acid secretion is impor- to 125% range of that of the brand. In 2004, the FDA
tant for the subsequent intestinal absorption of thyroxine determined bioequivalence of L-T4 formulations on the
(140 142). T4 absorption is increased by fasting and de- basis of 2 pharmacokinetic parameters (the area under the
creased in malabsorption syndromes and by certain foods curve [AUC] and the maximum T4 concentration) in eu-
and drugs (143, 144). Thyroid hormones are metabolized thyroid adult volunteers who received a supraphysiologi-
by the liver and mainly eliminated by the kidneys. A por- cal dose (600 g) of the L-T4 formulation. The analysis of
tion of the conjugated hormone reaches the colon un- data provided to the FDA indicated that the reference L-T4
changed. Approximately 20% of T4 is eliminated in the preparations (Unithroid, Levoxyl, Synthroid, or Levo-
stool. Physiological amounts of T3 are generated by the throid) were comparable because the 90% CIs of the dif-
monodeiodination of T4 in target tissues, thereby main- ference in AUC and maximum T4 concentration fell be-
taining normal serum T3 levels in patients receiving L-T4 tween 80% and 125% of the reference formulation.
monotherapy. With an appropriate individual dosage ad- Therefore, these products were considered therapeutically
justment, treatment with L-T4 is generally considered safe equivalent and interchangeable, and the FDA approved
and well tolerated, and its use is commonly associated with generic substitution for L-T4 (146). The designation AB
good patient compliance (9, 10, 26). In several countries, was used for interchangeable products meeting a positive
L-T4 is available in tablets with different dosage strengths standard for bioequivalence, whereas products not meet-
of 25, 50, 75, 88, 100, 112, 125, 137,150, 175, 200, and ing the standard were rated as BX. However, the latter
300 g. These multiple various tablets allow clinicians to method of determining bioequivalence by the FDA has
obtain a correct titration for individualized dosing. been consistently criticized by professional societies of en-
docrinologists, including the ES, the AACE, and the ATA.
3. Different T4 formulations: branded vs generic preparations The physicians expressed concern that potential harm
Several different branded and generic formulations of could come to patients being either overtreated or under-
L-T4 are commercially available. As a result, 10 to 12 mil- treated when nonbioequivalent preparations of L-T4 were
lion Americans affected by hypothyroidism are treated interchanged. Clinical data were collected to demonstrate
with different formulations of L-T4. The FDA issued stan- this issue in 2007 in a pharmacovigilance study that re-
dards for L-T4 bioequivalence and established that only ported adverse events that were related to interchanges of
L-T4 preparations that meet these standards may have the L-T4 preparations (151). The results of this survey sus-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
448 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

tained the opinion that although L-T4 products were os- product, Synthroid (160). On the other hand, the opposite
tensibly standardized, there existed subtle differences be- result was noted in another recent prospective randomized
tween various brand and generic L-T4 products because crossover study that evaluated the bioequivalence of a
the methodology to determine interchangeability was brand-name L-T4 (Synthroid) vs an AB-rated generic for-
faulty and that the various preparations could not neces- mulation (Sandoz) in children with severe hypothyroidism
sarily be considered therapeutically equivalent (152). Pre- (161). The patients received 8 weeks of one L-T4 formu-
sumably, the pharmacokinetic method used by the FDA to lation followed by 8 weeks of the other. The results sug-
assess bioequivalence was too insensitive to assess thera- gested that Synthroid and the AB-rated generic L-T4 are
peutic equivalence for L-T4 (152, 153), with the major not bioequivalent for patients with severe hypothyroidism
criticism being that pharmacodynamic considerations due to CoH (161). Reasons for the discrepancy between
such as serum TSH levels accompanying administration of these two studies have not been proposed.
the various L-T4 preparations were not assessed. There are numerous different branded formulations of
In a pharmacokinetic study of 36 normal volunteers, L-T4 that are available in the world. They include Eltroxin,
Blakesley et al (154) demonstrated that using current FDA Euthyrox, Letrox, Levaxin, L-T4, Thyrax, and Thyrax
methodology, a difference of 25% to 33% of the admin- Duotab in Europe; Thyrox in South Asia; Eutirox, Syn-
istered L-T4 dose might not be detected, and even after throid, and Tirosint in Europe and North and South
correction for baseline endogenous T4 levels, a difference America; and Thyrolar in Bangladesh (148). Given the
of 12.5% would not be identified. Based upon these ob- above concerns and considerations regarding true bio-
servations, it was conceivable that the pharmacokinetic equivalence or interchangeability of different prepara-
criteria that have been used by the FDA to assess the bio- tions, patients and physicians who are not thyroid spe-
equivalence of L-T4 products might not distinguish dose cialists should be aware of the possible consequences of a
differences of up to 33% (eg, 400 vs 600 g) and doses that change in L-T4 dosage or commercial formulation to avoid
differ by 12.5% (eg, 400 vs 450 g). A log-linear rela- potential adverse events (150, 162). Patients should be
tionship exists between the serum concentrations of T4 instructed to avoid switching between branded L-T4 prod-
and TSH such that small changes in the serum FT4 con- ucts and should consult their physician should any related
centration are associated with much larger changes in the adverse effects be suspected. Moreover, should there occur
TSH level (155). As a result, L-T4 is a drug with a narrow a switch in the T4 product, thyroid function testing should
therapeutic index or toxic-to-therapeutic ratio with the be repeated within 4 to 8 weeks and dosage retitrated to
potential of significant clinical consequences with minor achieve the therapeutic target (serum TSH level) with the
degrees of excessive or inadequate dosage (26). Variability new preparation.
in dosage due to poor interchangeability was considered In conclusion, until it is certain that the degree of any
potentially especially dangerous in vulnerable popula- differences in branded L-T4 preparations has little clinical
tions of patients who require precise dose titration such as significance, the available literature supports a prudent
children, patients with thyroid cancer, pregnant women, approach to interchange or substitution of L-T4 formula-
elderly patients, and patients with heart or bone disease (9, tions in certain clinical situations, including severe hypo-
10). If these concerns regarding noninterchangeability thyroidism, children, pregnant women, the elderly, and
were valid, even switching L-T4 preparations with a patients with major comorbidities.
12.5-g difference could produce clinically relevant ef-
fects due to iatrogenic subclinical hyperthyroidism and 4. Liothyronine
hypothyroidism (9, 10). However, despite widespread In individuals with normal thyroid function, the major
concern about the validity of the current FDA methodol- surge of circulating T3 arises from the 5-deiodination of
ogy for determining L-T4 bioequivalence, there are few T4 in the liver and other peripheral tissues (133). T3 is the
prospective RCTs that support that criticism (156 159). most active thyroid hormone because its affinity for the
The few available data extant today are conflicting and nuclear receptor is 10- to 20-fold that of T4 (131, 132).
controversial and often marred by their own methodolog- The ratio of T4 to T3 in the human thyroid gland is ap-
ical flaws. proximately 15:1 (26). L-T3 tablets are commercially
It is conceivable that improved manufacturing tech- available in the United States as Cytomel, a synthetic form
niques resulting in products of more uniform potency and of the sodium salt available in the United States and Eu-
content have assuaged many of these earlier concerns. In rope in 3 dosage strengths of 5, 12.5, and 25 g. Most
a 5-year retrospective study of children with CoH, the use authorities discourage the routine use of L-T3 containing
of generic L-T4 treatment appeared to result in better con- preparations for thyroid hormone replacement therapy
trol of their hypothyroidism than the branded comparator because T3 has a relatively short half-life and the available

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 449

formulations of L-T3 are rapidly absorbed, resulting in formulation by pharmacokinetic analysis (165). The same
wide nonphysiological fluctuations in serum T3 levels (3, commercial formulation is also available in Europe in an
26, 32). Typically after T3 administration, supraphysi- oral liquid formulation of either L-T3 or L-T4 that can be
ological serum T3 concentrations were maintained for sev- administered as drops and have been shown to be bio-
eral hours followed by a rapid decline (26). equivalent to tablets of the two hormones (166). Some
A recent clinical trial of T3 therapy for hypothyroid studies have suggested a different dissolution and absorp-
patients by Celi et al (27) has aroused new interest in re- tion profile of these liquid formulations compared with
examining the role for T3 replacement therapy. These au- solid formulations. Initial studies have suggested their po-
thors performed a randomized double-blind crossover tential utility in patients affected by changes in gastric pH
study in 10 thyroidectomized patients. Steady-state phar- such as those with chronic gastritis or lactose intolerance
macodynamic equivalence of T3 with T4 was achieved by or those receiving histamine H2 receptor blockers and
completely substituting L-T4 with L-T3, using a thrice-daily proton pump inhibitors (167169). This liquid L-T4 prep-
regimen of T3 in an approximate ratio of 1:3 designed to aration may allow 100% absorption from the gastroin-
obtain a target serum TSH of 0.5 to 1.5 mU/L (27). The testinal (GI) tract with relatively immediate dissolution
assessment of the dynamic pituitary response to escalating apparently relatively unaffected by pH (169).
doses in a TRH stimulation test showed that this T3 ratio
obtained a near-identical degree of pituitary euthyroidism
compared with equivalent doses of L-T4 (163). Patients V. Replacement Therapy With L-T4 in
receiving T3 achieved reduced body weight and an im- Hypothyroid Patients
proved lipid profile after 6 weeks of treatment with T3
with no significant adverse effects on cardiac function, A. Treatment of primary hypothyroidism
insulin sensitivity, or QOL (164). Additional prospective 1. Overt hypothyroidism: evidence for treatment
double-blind randomized and controlled larger studies Although the normal thyroid gland secretes both T4
will be necessary to determine whether potential beneficial and T3, currently, only L-T4 is recommended as a replace-
effects of L-T3 therapy may be achieved without accom- ment therapy for hypothyroidism. Guidelines from all
panying adverse effects. professional societies, including the European Thyroid
Currently, the clinical use of L-T3 as monotherapy is Association (ETA), the ATA, the AACE, and the ES rec-
quite limited. L-T3 generally is reserved for use as a second- ommend L-T4 monotherapy as the treatment of choice for
line drug with specific indications for conditions such as all patients with persistent overt hypothyroidism (13,
MC or as short-term therapy in patients with DTC when 1721, 32). Patients with hypothyroidism should be
L-T4 therapy is being withdrawn or restarted in prepara-
treated to prevent the risk of progression to a more severe
tion for radioiodine therapy. The rationale for the latter disease and to avoid the risk of adverse cardiovascular
brief interval of T3 therapy is to reduce the duration of
events. The goal of replacement therapy in hypothyroid
hypothyroidism and its associated symptoms and improve
patients is to restore biochemical and clinical euthyroid-
the QOL of patients with DTC undergoing radioiodine
ism. Treatment with L-T4 is generally considered safe and
therapy (26).
well tolerated. It is important to normalize the serum
TSH during replacement doses of L-T4 because a high-
5. Preparations containing both L-T3 and L-T4
Some commercially available tablets are available normal and undetectable low-normal serum TSH have
throughout the world of mixtures of synthetic L-T3 and been associated with an increased cardiovascular risk
L-T4 doses in various ratios of content. However, such
(9, 10, 47, 48).
fixed-combination products do not allow personalization Replacement treatment with L-T4 to achieve a target
of dosing to attempt to provide the most physiological TSH level should be individualized in patients with hypo-
ratio of the two hormones for the treatment of hypothy- thyroidism, and the desired TSH level may vary in differ-
roid patients. ent patient populations. In the NHANES III disease-free
population, 86.7% of young people from 20 to 29 years of
6. Liquid thyroid hormone preparations age had TSH concentrations in the category of 0.4 to 2.5
Recently, oral soft gelatin capsules (TIROSINT; L-T4 mIU/L and only 2.5% had TSH values greater than 4.5
sodium capsules), containing L-T4 dissolved in inactive mIU/L (51). Therefore, in our opinion, the target TSH level
ingredients gelatin, glycerin, and water have become avail- with replacement therapy should be from 1 to 2.5 mU/L in
able in the United States. This new formulation is thera- young patients, although data supporting a target TSH in
peutically equivalent with the same preparation in a tablet this range in hypothyroid young patient are lacking. An

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
450 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

age-adjusted serum TSH should be targeted in middle- During pregnancy, hypothyroidism should be treated
aged and elderly patients (9, 10, 47, 48). to avoid adverse obstetric outcomes and the impaired neu-
A double-blind, randomized clinical study with a cross- ropsychological development that may occur in the off-
over design has shown that the target TSH for treatment spring of untreated women (1719, 178).
of primary hypothyroidism in the lower part of the refer- After delivery of an infant with CH, treatment should
ence range does not improve the persistence of symptoms be treated with optimal doses as soon as possible to avoid
or optimize the QOL in middle-aged and elderly patients the risk of an adverse neurological outcome (20 22).
with primary hypothyroidism (170).
After establishing appropriate initial dosage, TSH con- 2. Subclinical hypothyroidism
centrations should be reassessed every 6 to 12 months to a. Evidence for treatment. Patients with SHypo have a sig-
ensure that they remain within the normal range (9, 10). nificantly increased risk of progression to overt hypothy-
Untreated overt hypothyroidism can lead to an in- roidism, are more frequently symptomatic, and may have
creased risk of atherosclerosis, coronary heart disease an increased risk of HF and CHD events and mortality (9,
(CHD), heart failure (HF), pericardial and pleural effu- 10). The serum TSH concentration represents the most
sion, and ventricular arrhythmias because thyroid hor- important predictive factor of progression to overt hypo-
mone has profound effects on the vascular system, cardiac thyroidism. In the study by Dez et al (93), TSH levels
function, and lipid profiles (171175). Overtly hypothy- greater than 10 to 15 mU/L were associated with a hazard
roid patients may have elevated levels of serum total cho- ratio (HR) of 9.69 for the development of overt disease;
lesterol, low-density lipoprotein (LDL)-cholesterol, apo- this risk increased even more (HR, 27.9) when serum TSH
lipoprotein B, lipoprotein(a), and triglycerides, which are was 15 to 19.9 mU/L.
reversible with L-T4 therapy (9, 10, 175, 176). Postmortem Although the degree of cardiovascular risk associated
studies in patients with hypothyroidism have documented with SHypo remains controversial, some clarity may be
the link between atherosclerosis and thyroid hormone de- gleaned from the various meta-analyses that have critically
ficiency (175). evaluated the results of all of the available prospective
One meta-analysis involved a pooled examination of studies (179 183) (Table 8). A modest increased risk of
17 417 individuals with overt hypothyroidism to assess CHD and mortality was reported in 3 of these analyses
mortality risk (177). Although the results of this analysis (181183), although heterogeneous results were found
did not show an increased mortality in patients with hy- among the selected epidemiological studies. The risk of
pothyroidism compared with the control group, the au- CHD increased with the severity of thyroid hormone de-
thors concluded that the available studies are not ade- ficiency and was even higher when TSH values were above
quately designed to determine whether hypothyroidism 10 mU/L (182, 183). Furthermore, as assessed in 2 meta-
contributes to mortality. In fact, many patients were analyses, the risk of all-cause mortality (181) and mortal-
treated with L-T4 during the follow-up interval, thereby ity from CHD (183) were particularly increased in patients
potentially affecting the study results. The authors of this with comorbid conditions. However, the prospective
meta-analysis estimated that a Cox regression analysis studies that were included in these analyses differ in re-
would require a study population of 900 cases to detect a spect to several important points: 1) the inclusion of in-
20% increase in mortality. As a consequence, most of the dividuals who varied in terms of age, sex, race/ethnicity,
studies extant could be considered underpowered to assess and lifestyle; 2) the duration of follow-up; 3) cutoff values
the cardiovascular and total mortality in overt hypothy- of TSH levels considered to define SHypo; 4) the methods
roidism (177). employed to assess the risk of CHD; and 5) confounding
Treatment with L-T4 will prevent progression of hypo- factors used for adjustment of cardiovascular risk. The
thyroidism to a more profoundly hypothyroid state (myx- validity of these meta-analyses to accurately assess the risk
edema) and may improve symptoms, QOL, and the risk of of CHD events and mortality in patients with SHypo is
cardiovascular complications (9, 10, 36). Normalization limited or precluded by these important limits and the lack
of a variety of clinical and metabolic end points (heart rate, of individual participant data from the selected cohort
systolic and diastolic function, diastolic blood pressure studies.
[BP], systemic vascular resistance [SVR], arterial stiffness, A recent meta-analysis assessed the risk of CHD events,
serum cholesterol levels, serum creatine kinase, or other CHD mortality, and total mortality from an evaluation of
muscle or hepatic enzymes, sleep pattern, menstrual cycle the data derived from 55 287 participants from 11 pro-
abnormalities, fertility, etc) have been documented when spective studies on SHypo from United States, Europe,
a euthyroid state is reached during optimal replacement Australia, Brazil, and Japan (184). The risk of CHD events
therapy with L-T4 (36, 173175). and CHD deaths was examined in 25 977 participants

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 451

Table 8. Summary of Meta-analyses Assessing the Risk of Coronary Artery Disease Events, Cardiovascular Mortality,
and Total Mortality in Patients With SHypo
Studies TSH,
Study Ref. Year Included mU/L RR or HRa
CHD events
Singh et al 180 2008 3 RR 1.18 (1.021.37)
Razvi et alb 182 2008 6 HR 1.27 (0.951.22)
Ochs et alc 183 2008 10 HR 1.20 (0.971.49)
Rodondi et ald 184 2010 7 HR 1.18 (0.99 1.42)
10 19.9 HR 1.89 (1.28 2.80)
7.0 9.9 HR 1.17 (0.96 1.43)
HF events
Gencer et ale 188 2012 6 10 HR 1.65 (0.84 3.23)
10 HR 1.86 (1.272.72)
Cardiovascular mortality
Razvi et al 182 2008 8 HR 1.09 (0.84 1.41)
Ochs et al 183 2008 8 HR 1.18 (0.98 1.42)
Rodondi et alf 184 2010 10 HR 1.14 (0.9 1.32)
10 19.9 1.58 (1.10 2.27)
7.0 9.9 1.42 (1.031.95)
Total mortality
Singh et al 180 2008 3 RR 1.15 (0.99 1.25)
Haentjens et al 181 2008 9 HR 1.22 (0.96 1.57)
Ochs et al 183 2008 6 RR 1.12 (0.99 1.26)
Rotondi et alg 184 2010 11 HR 1.09 (0.96 1.24)
a
95% CIs are shown in parentheses.
b
The risk of CHD was significantly increased in subjects 65 years of age or younger but not in those 65 or older. The risk was statistically significant only in women in
the meta-analysis by Razvi et al.
c
Risk estimates were lower when higher-quality studies were pooled (RR, 1.021.08).
d
The subgroup analysis revealed no evidence of greater risks of CHD events among pooled participants over 80 years of age.
e
The risk of HF persisted after excluding patients with preexisting HF or AF. Among older participants (80 years old), the risk of HF events was not increased. The risk
of HF was increased after excluding participants using thyroid medications (mainly T4 replacement) at baseline and during periods of follow-up.
f
The subgroup analysis revealed no evidence of greater risks of CHD mortality among pooled participants over 80 years of age.
g
The subgroup analysis revealed no evidence of greater risks of total mortality among pooled participants over 80 years of age.

from 7 cohorts, analyzed with respect to age, sex race, diovascular disease. This analysis demonstrates a signifi-
TSH concentrations, and preexisting cardiovascular dis- cant trend of increased risk of both CHD events and mor-
ease. The severity of SHypo was stratified according to 3 tality at higher serum TSH concentrations, particularly in
categories of TSH concentration: 4.5 to 6.9, 7.0 to 9.9, and participants with a TSH level of 10 mU/L or greater (184).
10.0 to 19.9 mU/L. Separate Cox proportional hazard Other studies indicate that SHypo may worsen cardio-
models were used to assess the relationship between vascular hemodynamics and lead to HF (173, 185), with
SHypo and CHD events and mortality for each cohort. an increased incidence of HF observed in patients with
The risk of CHD events and mortality from CHD in- TSH concentrations of more than 7 to 10 mU/L (104, 186,
creased with higher TSH concentrations. In age- and sex- 187). A recent meta-analysis performed a pooled analysis
adjusted analyses, the HR for CHD events was 1.00 (95% of individual participant data using all the available pro-
CI, 0.86 1.18) for a TSH level of 4.5 to 6.9 mU/L, 1.17 spective cohorts with thyroid function tests and subse-
(95% CI, 0.96 1.43) for a TSH level of 7.0 to 9.9 mU/L, quent follow-up of HF events (188) (Table 8). The pooled
and 1.89 (95% CI, 1.28 2.80) for a TSH level of 10 to analysis was stratified according to age, sex, gender, race,
19.9 mU/L (n 70 events per 235; 38.4 per 1000 person- TSH levels, and preexisting cardiovascular disease and
years; P .001 for trend). The resulting HRs for CHD HF, thereby accounting for the heterogeneity inherent in
mortality were 1.09 (95% CI, 0.911.30), 1.42 (95% CI, the available cohort studies related to the risk of HF
1.031.95), and 1.58 (95% CI, 1.10 2.27, n 28 deaths events. SHypo was defined as a TSH of 4.5 to 19.9 mU/L
per 333; 7.7 per 1000 person-years; P .005 for trend). with normal reference range thyroid hormone levels.
Total mortality was not increased among participants There were 25 390 participants in this individual data
with SHypo, and the results were similar after further ad- analysis that included 2068 subjects with SHypo from 6
justment for traditional cardiovascular risk factors. Risks prospective cohorts from the United States and Europe.
did not significantly differ by age, sex, or preexisting car- The risk of HF events was related to higher TSH levels with

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
452 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

a statistically significantly increased risk among those with thyroid hormone deficiency, the beneficial effects of re-
TSH above 10.0 mU/L (HR, 1.86; 95% CI, 1.272.72). placement therapy with L-T4 on lipid levels were propor-
The increased risk of HF in adults persisted after excluding tional to both the severity of hypothyroidism and the mag-
those with preexisting HF or atrial fibrillation (AF). Fur- nitude of the elevation in lipid levels (112). Patients who
ther adjustments for cardiovascular risk factors and other had higher baseline serum lipid values showed a greater
potentially confounding risk factors for HF did not sig- reduction in serum lipid concentrations after thyroid hor-
nificantly change the association with HF events. Among mone treatment (112).
older participants (80 years old), the risk of HF events No clinical trial has assessed the benefit of replacement
was not increased. The risk of HF was increased after doses of L-T4 on the frequency of CHD and HF events, but
excluding participants using thyroid medications (mainly a survival benefit was associated with L-T4 therapy in some
T4 replacement) at baseline and during periods of fol- studies (190, 191).
low-up (188). In the reanalysis of the Wickham survey on 97 individ-
In our view, these two important meta-analyses (184, uals (mean age, 49 years), during a 20-year follow-up,
188) provide sufficient evidence to justify the treatment of all-cause mortality was significantly lower in L-T4treated
patients with SHypo having a serum TSH level above 10 patients with SHypo than in untreated individuals after
mU/L to reduce the risk of CHD and HF, and L-T4 mono- adjusting for age, gender, and cholesterol levels (190). Fur-
therapy has been recommended for such patients (3) (rec- ther adjustment for other risk factors for ischemic heart
ommendation 15, grade B). Data in the available literature disease did not change the results with an HR of 0.22
results are considered sufficient for this recommendation (0.06 0.81; P .02). Moreover, L-T4 treatment was
with the belief that treatment of this condition may avoid noted to be associated with lower all-cause mortality in
the risk of progression of CHD and cardiovascular risk patients with moderate hypothyroidism (191), and the
(3, 10). risk of HF events was significantly lower in L-T4treated
patients with TSH 10 mU/L in the Cardiovascular Heart
b. Benefit of treatment with replacement doses of L-T4. Various Study (187).
placebo-controlled studies have assessed the effects of L-T4 In summary, the evidence appears substantive for a
replacement therapy on symptoms and signs in patients beneficial effect of L-T4 replacement therapy in patients
with SHypo. The results of these studies have been often with TSH 10 mU/L that could serve to reduce progres-
conflicting (9, 10). A comparative analysis of these studies sion of cardiovascular risk. Despite the lack of definitive
is difficult because they enrolled heterogeneous patients in long-term studies on the outcome of mild to moderate
terms of the causes of hypothyroidism, age, severity of hypothyroidism with and without replacement therapy,
thyroid hormone deficiency, duration of replacement recommendations for treatment of these patients recently
therapy, and L-T4 dose, and in addition, employed differ- have been formulated (3). Prospective randomized clinical
ing scoring systems to assess the symptomatic response to trials will be necessary to assess the beneficial effects of
treatment (9, 10). However, despite these limitations, it is L-T4 in patients with SHypo and establish a guideline for
reasonable to conclude that there are beneficial effects of the optimal TSH cutoff level that would indicate the value
L-T4 treatment on mood, cognition, and symptoms in pa- of replacement therapy that would provide improved
tients with SHypo that are most apparent in patients with symptoms, QOL, and cardiovascular morbidity and mor-
serum TSH greater than 10 mU/L (9, 10). tality. For the longer-term follow-up of these patients, it
The link between SHypo and lipid values is also some- should be mentioned that those patients with serum TSH
what controversial (189). Contradictory results have been above 10 mU/L are more likely to progress to overt hy-
published in regard to whether there are significant dif- pothyroidism with time and need to be monitored with
ferences in levels of serum total cholesterol, LDL-choles- increases in L-T4 dosage as required.
terol, high-density lipoprotein cholesterol, or triglycerides
between euthyroid control subjects and patients with 3. Management of patients with minimally increased
SHypo, especially in patients with a mild increase in TSH serum TSH
(189). Eight placebo-controlled randomized clinical trials a. Evidence in favor of treatment. Whether or not to treat pa-
have examined the effects of replacement L-T4 therapy on tients with mild SHypo remains controversial, in part be-
serum lipids in SHypo (9, 10). L-T4 did not reduce total cause of differing opinions on what constitutes the TSH
cholesterol in 4 studies but had a beneficial effect in the reference range and therefore what constitutes an increase
other 4 studies with the more evident improvement in lipid in serum TSH. However, having considered a given value
profile observed in patients with TSH above 10 mU/L (9, as slightly elevated and instituted T4 treatment, the rate of
10). In a meta-analysis of 13 studies on 200 patients with TSH normalization is faster in patients who have a lesser

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 453

degree of serum TSH elevation (TSH values 4 8 mU/L). On the other hand, treatment of older people with
This is especially true in patients with negative antithyroid SHypo was not associated with similar benefits. After ad-
antibody titers, suggesting that patients with a mild TSH justment for baseline cardiovascular risk factors, age, sex,
increase and nonautoimmune thyroiditis may frequently baseline serum TSH levels, and L-T4 use, the number of
have transient TSH elevation (9, 10). Some justification incident ischemic heart disease events was lower in the
for treatment may be inferred from the presence of ele- L-T4treated younger group (adjusted HR, 0.61; 95% CI,
vated serum lipids, which is more likely to be seen in pa- 0.39 0.95) (193). All-cause mortality was lower in the
tients with SHypo and serum TSH of 10 mU/L, in smok- treated younger group (multivariate-adjusted HR, 0.36;
ers, and in insulin-resistant patients (9, 10). However, an 95% CI, 0.19 0.66), primarily because of a reduction in
argument against treatment may be made based upon 2 circulatory and cancer-related deaths (193). Incident cere-
recent meta-analyses indicating that mild SHypo was not brovascular disease events were unchanged, and incident
associated with an increased risk of CHD and HF events AF was not related to L-T4 exposure (193). This study
(184, 188). clearly demonstrated the benefit of replacement therapy in
Although these 2 important meta-analyses provide ev- young patients with mild TSH increase; however, its major
idence for treating particularly those patients with SHypo limit is the retrospective design.
with serum TSH 10 mU/L, limitations inherent in the Therefore, although some studies have demonstrated
studies reported deserve mention. These include 1) inclu- the potential beneficial effects of L-T4 therapy to improve
sion of a study population that was predominantly white cardiovascular risk patients with mild SHypo, large ran-
with the exception of 2 studies from Japan and Brazil; 2) domized controlled studies will be required to assess the
the lack of follow-up data on thyroid function, which was importance of this treatment in the presence of minimal
assessed only at baseline evaluation, thereby precluding TSH elevation. It has been estimated that an appropriately
assessment of each individuals response to treatment and powered RCT of L-T4 therapy in 40- to 70-year-old people
either progression, stabilization, or regression of the with SHypo will require the recruitment of about 1450
SHypo; and 3) the lack of FT3 assessment in most cohorts, individuals.
which would prove useful to exclude other causes of in- The Thyroid Hormone Replacement for Subclinical
creased TSH levels, eg, as in elderly subjects. Hypothyroidism trial (TRUST), a multicenter double-
blind placebo-controlled randomized trial is recruiting
b. Effects of treatment with replacement doses of L-T4. Reports 3000 older adults with hope of clarifying the effects of
of the effects of L-T4 treatment of SHypo on lipid profile, thyroid hormone replacement therapy in untreated older
specific symptoms of hypothyroidism, or cognitive and adults with SHypo (194). Moreover, the Institute for Ev-
neuropsychiatric symptoms have been conflicting, indi- idence-based Medicine in Old Age (IEMO), started the
cating either benefit or lack of benefit (9, 10). However, 80-plus Thyroid Trial Collaboration, a multicenter trial
there is evidence in some placebo-controlled studies of on participants aged 80 years and older, to evaluate the
potential improvement of cardiovascular hemodynamics effect of treatment with L-T4 on several endpoints (car-
and cardiovascular risk factors with L-T4 therapy in pa- diovascular disease and mortality, cognitive and physical
tients with TSH 10 mU/L (192). Mild SHypo may be function, and QOL) in elderly subjects (195).
associated with a greater cardiovascular risk in young and
middle-aged people (192, 193). c. Recommendations for treatment. Patients with SHypo and
Recently, Razvi and co-workers (193) examined the new onset of symptoms or depression, goiter, or cardio-
outcomes of treated individuals with mild SHypo (serum vascular risk factors (eg, hypertension, hypercholesterol-
TSH of 5.0110.0 mU/L) by analyzing data from the emia, insulin resistance or diabetes, kidney failure, or iso-
United Kingdom General Practitioner Research Database. lated diastolic dysfunction) might benefit from L-T4
All analyses were stratified according to subsequent L-T4 treatment (9, 10). Moreover, treatment of mild SHypo
treatment for younger (40 70 years) vs older ( 70 years) may be indicated in patients with positive antithyroid an-
patients. For a median follow-up period of 7.6 years, tibody tests and a progressively rising TSH level to avoid
52.8% of younger and 49.9% of older patients with mild worsening thyroid function (9, 10). On the other hand, the
SHypo were treated with a median L-T4 dosage of L-T4 of SHypo may be transient in some patients with an initial
75 g/d (range, 12.5175 g/d). This analysis indicated slightly increased TSH level that returns to well within the
that treatment of mild SHypo with L-T4 was associated reference range with further follow-up. In such patients,
with better outcomes in younger (70 years) people with serum TSH concentrations could be assessed at yearly in-
respect to incident fatal and nonfatal ischemic heart dis- tervals to avoid the risk associated with unnecessary treat-
ease events and mortality (193). ment (9, 10). If a therapeutic trial is implemented, replace-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
454 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

ment therapy should be stopped in the absence of a clear significantly affect L-T4 requirement because replacement
beneficial effect or with development of an undetectable doses did not differ between premenopausal and post-
TSH with low doses of L-T4. Young patients with serum menopausal women (199). Moreover, men had a lower
TSH concentrations from 3 to 4.5 mU/L should be mon- requirement than premenopausal women (199).
itored with periodic thyroid function tests, particularly if Gender-based differences in L-T4 dosage became ap-
they have positive TPOAbs (9, 10, 47). parent only when the degree of overweight was included
in the analysis (199).
B. Determinants of L-T4 requirements and starting dose Body weight is directly related to L-T4 dose (200, 201).
of L-T4
Moreover, it has been reported that the most appropriate
Although achieving optimal L-T4 replacement may
replacement therapy dosage correlates best with lean body
seem a rather straightforward task, consistently reaching
mass rather than total body weight, suggesting that adi-
the target TSH may be rendered problematic due to a
pose tissue is less metabolically responsive to L-T4 than the
variety of factors including the cause and severity of hy-
muscle compartment (202204). A large retrospective
pothyroidism, sex, age, gender, menstrual status, body
study in euthyroid patients (without serious chronic con-
weight, lean body mass and body surface area, comor-
ditions or confounding medications) who were receiving
bidities, pathological or physiological conditions, drugs,
L-T4 replacement therapy for primary hypothyroidism
and adherence to therapy (Table 9).
The dose requirement of L-T4 is higher in infants and suggests that age-gender differences in doses of L-T4 may
children (20, 21) and is usually higher in younger com- be secondary to differences in body weight, lean body
pared with elderly patients In some studies, elderly pa- mass, and ideal body weight (199). In this study, multi-
tients required lower doses of L-T4 than younger patients variate analysis indicated that body weight and body com-
with an age-related decline in L-T4 dose in men compared position, but not age, influenced the L-T4 requirement
with women (196 198). The dose requirement of L-T4 (199). These data suggest that the age-related decrease in
also seems to be associated with hormonal status, with L-T4 requirement is mediated by alterations in body weight

lower doses required in postmenopausal women com- and body composition (199).
pared with premenopausal women (198). A study on 50 Most clinicians initiating L-T4 therapy will base dose
patients after thyroidectomy found that the weight-based selection on body weight (200).
L-T4 dosage required to normalize TSH was 1.7 g/kg in Usually body weight is considered a good indicator for
both sexes; however, premenopausal women appeared to calculating an appropriate starting dose of L-T4. Most pa-
need a greater dosage than both postmenopausal women tients are well treated with a narrow dose window that
and men (196). However, in contrast to these studies (196, varies according to body weight from 1.6 to 1.8 g/kg/d
198), in another study, hormonal status in women did not for replacement therapy in young and middle-aged pa-

Table 9. Determinants of L-T4 Requirements


Determinant Dosage Requirements
Etiology of hypothyroidism Larger doses are required in athyreotic patients compared with patients with residual thyroid
function (thyroidectomized patients vs autoimmune hypothyroidism and radioiodine-
ablated patients )
Lower doses are required in patients with CH
Severity of hypothyroidism Larger doses are required in patients with overt hypothyroidism compared with mild and
subclinical disease
Age Decrease in L-T4 requirement with advancing age
Gender Men have a lower requirement than premenopausal women
Adherence to therapy Variable adherence may falsely suggest an increased requirement of L-T4
Body weight Good parameter to start L-T4 therapy
Lean body mass Better parameter to assess L-T4 requirement
Ideal body weight Best parameter to assess L-T4 requirement
Deiodinase activity Reduced activity in elderly patients, obese patients, some chronic and acute illnesses,
important surgery procedures
Increased activity in tumors that express D3 (hemangiomas, vascular tumors)
Deiodinase polymorphisms Increment in dosage
Pregnancy 30%50% increment in dosage
Interferences Increment or decrement in dosage
Food and beverage
Herbal remedies
Drugs

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 455

tients with primary hypothyroidism and 2.0 to 2.5 g/kg/d to significantly reduce the risk of maternal hypothyroid-
when suppressive doses of L-T4 are desired (199). ism during the first trimester and mimic normal physiol-
However, severely obese individuals may need higher ogy (213).
L-T4-suppressive or replacement doses than normal- The increased L-T4 dosage in hypothyroid pregnant
weight individuals due to the impaired L-T4 pharmacoki- women often depends on the etiology of hypothyroidism.
netic parameters (higher plasma volume and/or to delayed In one retrospective study, the average baseline L-T4 dose
GI L-T4 absorption) and their altered T4 to T3 conversion for patients with primary hypothyroidism was 92.5
(205). 32.0, 140.4 62.4, and 153.2 30.3 g daily, respec-
The etiology and severity of hypothyroidism may affect tively, for patients with primary hypothyroidism, in pa-
the L-T4 dosage. The amount of residual functional thy- tients with hypothyroidism resulting from treated Graves
roid tissue may contribute to the T4 production and affect disease or goiter, and in patients with thyroid cancer
the replacement dose of L-T4 (201, 206, 207). Higher (214). The cumulative increase from baseline L-T4 dosage
doses of L-T4 generally are required in thyroidectomized was 13% in the first trimester, 26% in the second trimes-
patients compared with patients with autoimmune hypo- ter, and 26% in the third trimester (P .001) (214).
thyroidism or after radioiodine thyroid ablation for Patients with nephrotic syndrome and other severe ill-
Graves disease (208 210). Patients with SHypo require nesses may have an altered clearance of L-T4 and require
lower doses of L-T4 than patients with overt disease (9, 10). a higher dose (215, 216). GI diseases may reduce L-T4
One randomized control trial evaluated the efficacy of absorption (142). Some drugs or medications may alter
selecting T4 doses on the basis of the initial levels of serum L-T4 absorption, T4 binding, or thyroid hormone metab-
TSH (25 g for TSH 4.0 8.0 mU/L, 50 g for TSH 8 12 olism or affect T4 to T3 conversion (79). Patients with D2
mU/L, and 75 g for TSH 12 mU/L) and reported that polymorphism (threonine 92 alanine) may also need
only minimal adjustments were required after this initial higher doses of L-T4 to restore euthyroidism (217). Rare
dose selection to achieve euthyroidism, according to their tumors that express D3 such as hemangiomas and vascu-
criteria for optimizing TSH on L-T4 replacement therapy lar tumors may increase the L-T4 requirement (64).
(211). Not surprisingly, this study suggested that doses of An RCT has shown that a full dose of L-T4 can be
25 to 75 g daily are usually sufficient for achieving eu- started in adult patients without significant comorbidities
thyroid levels in patients with mild hypothyroidism and (218). In this study, patients were randomized into 2
that larger doses are usually required in patients with more groups who either started L-T4 therapy with a full replace-
severe disease. This study suggests that it might be useful ment dose (1.6 g/kg/d) or with a low L-T4 dose that was
to start treating SHypo with low doses of L-T4 to normalize progressively increased every 4 weeks. After 12 and 24
the serum TSH with the minimal amount of L-T4 to avoid weeks, the authors did not observe any cardiac symptoms
overtreatment. or acute cardiovascular events (218). This approach with
Lower doses tend to be required in patients with CH an initial full replacement dose was deemed safe and more
(81); the estimated starting dosage of L-T4 is 1.3 g/kg/d rapidly effective in reversing symptoms of hypothyroidism
(212). than the traditional approach of gradually increasing L-T4
Optimal management of L-T4 replacement treatment dosage (218).
should take into account conditions increasing the L-T4 However, in some conditions, it is not advisable to start
requirement (eg, pregnancy, glomerular disease, malab- a full L-T4 dosage.
sorption, drugs interfering with L-T4 absorption, in- It is considered prudent in patients older than 50 to 60
creased metabolism, or deiodinase activity) or those con- years without evidence of CHD to start replacement with
ditions that may decrease the L-T4 dose (weight loss, doses of 25 to 50 g/d; this dosage should be titrated
withdrawal of drugs interfering with L-T4, and atrial gradually with a progressive increase every 2 to 3 weeks
arrhythmias). until euthyroidism is reached (173, 175). In patients with
Higher doses of L-T4 are needed in pregnant women. severe ischemic heart disease and in very old patients with
The L-T4 dose often needs to be incremented by 4 to 6 severe hypothyroidism, the recommended starting dose
weeks gestation and may require a 30% to 50% increment should be even lower, with 12.5 g/d with a progressive
in dosage over that taken in the nonpregnant state (17 increase of 12.5 g/d every 4 to 6 weeks (173, 175). Some
19). Overt hypothyroidism discovered during pregnancy patients may need coronary revascularization to tolerate
should be corrected as soon as possible with a full starting the dosage of L-T4 required to achieve euthyroidism be-
dose of L-T4 of about 2.0 to 2.4 g/kg/d (1719). It has cause a full replacement dose of L-T4 may trigger severe
been recommended that a 2-tablet per week increase in angina, myocardial infarction (MI), and arrhythmias in
L-T4 dosage be initiated as soon as pregnancy is confirmed patients with asymptomatic underlying ischemic heart dis-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
456 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

ease (219). Emergency coronary artery bypass grafting in weeks to give sufficient time to reset the pituitary gland
patients with unstable angina or left main coronary artery between dosage changes (3). Changes of about 12.5 to 25
occlusion may be safely performed while the patient is still g/d are initially made, but even smaller changes may be
moderately to severely hypothyroid (173, 175). necessary to achieve the desired TSH levels. In patients
On theoretical grounds, an exacerbation of adrenal in- with primary hypothyroidism, TSH levels should be nor-
sufficiency may occur when starting full replacement ther- malized to a target level within the reference range, con-
apy with L-T4 in patients with central or autoimmune un- sidering the age of patients and concomitant physiological
suspected adrenal disease. Although it rarely occurs, or pathological conditions (3, 9, 10). Periodic follow-up
patients at risk should be treated with clinically appropri- evaluations with repeated TSH testing at 6- and 12-month
ate doses of hydrocortisone until adrenal insufficiency is intervals are appropriate in patients with stable euthyroid-
ruled out (220, 221). ism during L-T4 therapy.
For the newborn, the American Academy of Pediatrics Thyroid hormone requirements are increased in preg-
and the European Society of Pediatric Endocrinology rec- nancy and return to the prepregnancy range after delivery
ommend a dose of 10 to 15 g/kg/d (about 37.550 g/d) when the dose of L-T4 should be reduced (1719). More
(20, 21). This dosage permits prompt FT4 and TSH nor- frequent evaluations of thyroid function are necessary in
malization without adverse effects on growth and skeletal pregnant patients and in patients receiving drugs that can
maturation (20, 21). The discovery of CH in utero is rare; interfere with thyroid function, those with malabsorption,
however, when confirmed or strongly suspected, it can be
and/or those losing or increasing their body weight. Dos-
managed with intra-amniotic injections of L-T4. Alterna-
age adjustments are necessary when medications influenc-
tively, treatment of infants in the first 2 to 4 weeks of life
ing absorption, plasma binding, or metabolism are either
avoids the potential risks associated with treatment in
added or discontinued. In addition to the TSH level, ad-
utero and is associated with an excellent prognosis.
equacy of therapy is mostly reflected by the combination
In conclusion, different conditions may affect the L-T4
of measurements of FT4 and FT3 levels. The assessment of
requirement. The starting doses of L-T4 will vary depend-
serum FT4 targeting the mid reference range should be
ing upon the cause and severity of hypothyroidism, patient
considered when monitoring L-T4 therapy. However, FT3
age and sex, BMI, and any underlying physiological or
levels may remain in the low-normal reference range in a
pathological conditions (pregnancy, CoH, and comor-
significant subset of thyroidectomized patients receiving
bidities) (Table 10).
replacement doses of L-T4 (28 30).
C. Follow-up and target serum TSH in patients receiving In patients with CoH, adequate L-T4 replacement ther-
replacement therapy with L-T4 apy is necessary during the first 2 to 3 years of life to
TSH is the most sensitive indicator of adequacy of L-T4 optimize the likelihood of normal brain development (20
therapy during treatment with replacement doses in hy- 22). It is important to emphasize that the target level of
pothyroid patients (222). TSH levels will decline within a serum TSH in patients with CoH should be 5 mU/L
month after starting L-T4 therapy. Dose adjustments are (optimal values, 0.52.0 mU/L) with a serum FT4 or total
usually guided by serum TSH determinations every 4 to 8 T4 in the upper limit of the normal range. During the first

Table 10. Starting L-T4 Dose According to the Age of the Patients and Physiological and Pathological Conditions
Age/Condition Starting L-T4 Dose
Adults with primary hypothyroidism 1.6 1.8 g/kg/d for replacement doses
2.0 2.5 g/kg/d for TSH suppressive doses of L-T4
Elderly patients 2550 g/d with 2550 g/d incremental dose every 3 4 wk
Very elderly patients 12.525 g/d with 12.525 g/d incremental dose every 3 4 wk
Patients with CHD 12.5 g/d with 12.5 g/d incremental dose every 3 4 wk
Pregnancy 30 50% increment in dosage (2.0 2.4 g/kg/d)
Congenital hypothyroidism 10 15 g/kg/d (37.550 g/d)
Hypothyroidism during the first 6 mo of age 8 10 g/kg/d (2537.5 g/d)
Hypothyroidism after 6 mo of age
6 12 mo 6 8 g/kg/d (50 75 g/d)
15 y 5 6 g/kg/d (75100 g/d)
6 12 y 4 5 g/kg/d (75125 g/d)
12 y to adults 13 g/kg/d (100 200 g/d)
Infantile CH 8 10 g/kg/d (2537.5 g/d)
Juvenile and adult CH 1.3 g /kg/d

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 457

Table 11. Target Serum TSH Values According to the Age of the Patients and Physiological Conditions During
L-T4 Therapy

Age/Condition Target Serum TSH, mU/L


Preconception Lower quartile of the reference range, 1.2
Pregnancy (1st trimester) 22.5
Children with CoH 5 (optimal values (0.52.0) with a serum FT4 or total T4 in the upper limit
of the normal range during the first year of life
Young patients 12.5
Middle-aged patients 1.53
Elderly patients
65 y 4.5
60 70 y 6.0
70 80 y 7.0 8
CH Serum FT4 levels in the upper half of the normal range
Thyroid cancer Risk-related TSH target

year of life, the target value of serum total T4 and FT4 1. Drugs affecting thyroidal synthesis of T4 or T3
should be 130 to 260 nmol/L (10 16 g/dL) and 18 to 30 Some drugs may induce hypothyroidism and may fur-
pmol/L (1.4 2.3 ng/dL), respectively (20 22). Thereafter, ther aggravate thyroid function in patients with a previous
FT4 should be at the upper half of the reference range for diagnosis of thyroid hormone deficiency. This phenome-
age (Table 11). non may serve as a rationale for initiation of L-T4 therapy
In summary, the ideal target serum TSH during L-T4 when hypothyroidism (or perhaps SHypo) is diagnosed. In
replacement therapy should be targeted and achieved with patients receiving drugs known to potentially alter thyroid
function, periodic monitoring of thyroid function tests
consideration of the age of the patients, the cause of
should be performed, and an increase in L-T4 doses is usu-
hypothyroidism, and any underlying physiological or
ally necessary in patients with a previous diagnosis of
pathological conditions (3, 9, 10) (Table 11). The fol-
hypothyroidism.
low-up of patients with primary hypothyroidism is
shown in Table 12. a. Lithium. Lithium carbonate is an effective treatment for
bipolar disorders. Long-term lithium treatment may lead
D. Drugs interfering with L-T4 administration
to goiter in about 50% of cases and may induce subclinical
A large number of drugs may affect thyroid function (34%) or overt (15%) hypothyroidism (224). A review of
and/or may interfere with the L-T4 requirement by inter- 4681 patients treated with lithium reported a prevalence
fering with absorption of L-T4 or by other mechanisms of hypothyroidism between 3.4% and 23% (225). The
(79, 223). Appreciation of the large number of such drugs mechanism involved in lithium-induced hypothyroidism
and their mechanisms of action have been growing in re- is due to the inhibition of both thyroid hormone synthesis
cent years (Table 13). An increased dose of L-T4 is required and release, a cytotoxic effect, and an increased thyroid
in hypothyroid patients receiving these drugs during re- autoimmunity. The risk of hypothyroidism in patients
placement therapy to maintain euthyroidism. treated with lithium is increased in the presence of thyroid

Table 12. Follow-Up of Adult Patients With Primary Hypothyroidism and CH


Evaluation
Baseline Clinical examination, FT4, TSH, TPOAb, anti-Tg antibodies, thyroid ultrasound
ECG and Doppler echocardiography in severe hypothyroidism
Assessment of autoimmune endocrine disease in patients with suspected APS
Follow-up TSH and FT4
2 mo after starting L-T4
4 wk after every dose adjustment
Every 6 mo during the first year
Annually in patients with stable TSH values during replacement L-T4 therapy
Annual thyroid ultrasound in patients with nodular goiter
ECG and Doppler echocardiography every month in patients with CHD, arrhythmias, and pericardial-pleural effusion.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
458 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

Table 13. Drugs Interfering With L-T4 Administration autoantibodies (50%), female gender (female to male,
5:1), older age, and a prolonged duration of treatment
Drug Effect (more than 2 years) (225, 226).
Drugs that influence thyroid hormone synthesis and release and thyroid As a consequence, thyroid function and the presence of
autoimmunity
Reduction of TH synthesis: lithium, flavonoids, resorcinol, aminoglutetimide, thyroid autoantibodies should be evaluated before start-
thionamides, thiocyanate ing lithium therapy and every 6 months thereafter as long
Inhibition of iodine uptake: lithium, ethionamide, amiodarone as treatment is continued. Once patients develop hypo-
Organification of iodine: amiodarone, sulfonamides, lithium, ketoconazole
Cytotoxic effect: lithium, cytokines tyrosine kinase inhibitors thyroidism and L-T4 therapy is initiated, lithium treatment
Alteration of thyroid autoimmunity: lithium, cytokines, amiodarone, may be continued.
tyrosine kinase inhibitors
Drugs altering T4 metabolism
Drugs that increase serum TBG concentration b. Amiodarone and other iodinated drugs. Amiodarone is a
Estrogens benzofuran-derived, iodine-rich antiarrhythmic agent
Tamoxifen used for the treatment of atrial and ventricular arrhyth-
Raloxifene
Heroin mias (227). A daily dose of 200 to 600 mg of amiodarone
Methadone provides approximately 7 to 21 mg iodine. This is an ex-
Mitotane
Fluorouracil
traordinary iodine load relative to the average daily re-
Capecitabine quirement for iodine of 150 g. Because of this iodine
Clofibrate excess, amiodarone may induce hypothyroidism in about
Drugs that decreases serum TBG concentration
Androgens 5% to 15% of patients with the highest risk during the first
Anabolic steroids (danazol) 18 months of treatment, particularly in certain popula-
Slow-release nicotinic acids tions including patients living in iodine-sufficient areas,
Glucocorticoids
Displacement from protein binding sites women, the elderly, and patients with positive thyroid
Salicylates (2 g/d) and salsalate (1.53 g/d) antibodies (227, 228). No clear correlation has been dem-
Drugs and food affecting L-T4 absorption
Cholestyramine
onstrated between the onset of hypothyroidism and the
Colestipol cumulative dose of amiodarone and/or the duration of
Aluminum hydroxide treatment. The pathogenesis of amiodarone-induced hy-
Antacids
Ferrous sulfate pothyroidism is related to the inability of the thyroid gland
Sucralfate to escape from the acute Wolff-Chaikoff effect after an
Laxatives iodine load in patients with preexisting Hashimotos thy-
Calcium carbonate
Proton pump inhibitors and H2 blockers roiditis (227, 228). Chronic amiodarone therapy may also
Lovastatin decrease T4 deiodinase activity, resulting in a 150% in-
Bile acid sequestrants
crease in rT3 and a 15% to 20% reduction in FT3 (227,
Activated charcoal
Anion-exchange resins 228). Amiodarone also inhibits T4 transport in the liver,
Drugs affecting thyroid hormone metabolism (increase L-T4 requirements) decreasing T4 metabolism and resulting in a 20% to 40%
Amiodarone
Phenobarbital
increase in serum T4 and FT4 (227, 228). A transient 20%
Rifampicin to 50% increase in TSH levels is observed during the first
Phenytoin 3 months of treatment with amiodarone. The increases in
Carbamazepine
Tyrosine kinase inhibitors total and FT4 and rT3 usually occur during the first months
Sertraline of administration in euthyroid subjects.
Drugs affecting deiodinase activity Optimal management of the patient treated with ami-
Amiodarone
Glucocorticoids odarone would include a baseline assessment of thyroid
-Blocking drugs function before initiating therapy with repeat follow-up
Propylthiouracil
Drugs affecting TSH secretion
measurements of TSH and FT4 after the first 3 months of
Dopamine and dopamine agonists treatment and every 3 months thereafter. If and when hy-
Somatostatin analogs pothyroidism does develop, initiation of L-T4 therapy may
Dobutamine
Amphetamine also have a beneficial effect on the underlying heart disease
Bexarotene (228). However, it is advisable to titrate the L-T4 replace-
GH ment therapy to maintain the TSH values at the upper limit
Bromocriptine
Peripheral antagonists of thyroid hormone of the normal reference range in view of the likelihood of
L-Carnitine severe underlying heart disease in these patients. Discon-
tinuing L-T4 after 6 to 12 months can be considered to
reassess thyroid function, and spontaneous remission of

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 459

hypothyroidism may be seen in about 60% of the patients anisms have been proposed to explain the development of
who had no antecedent or underlying thyroid disease. hypothyroidism: 1) the inhibition of thyroid iodine uptake
Many other iodine-containing compounds may affect thy- (236), 2) the inhibition of thyroid peroxidase (237), 3) a
roid function such as saturated solution of potassium io- toxic effect with development of painless destructive thy-
dine (amount of iodine, 38 mg/drop), Lugols iodine (6.3 roiditis (238, 239), 4) the inhibition of thyroidal vascu-
mg/drop), povidone iodine (10 mg/mL), iopanoic acid larization leading to thyroid atrophy (240 242), or 5) an
(333 mg/tablet), ipodate sodium (308 mg/capsule), and increase in D3 activity (243). Sunitinib has been associated
various iv radiographic contrast agents (140 380 with hypothyroidism in 14% to 85% of patients, ranging
mg/mL). from transient increases in TSH to persistent hypothyroid-
ism (242).
c. Cytokines. Thyroid dysfunction due to increased thyroid Sorafenib has been associated with TSH elevations in
autoimmunity may develop in patients receiving long- about 18% of patients with metastatic renal cell carci-
term treatment with cytokines for their cytotoxic effects. noma and in 33% of patients with thyroid carcinoma
The administration of interferon ( and ), interleukins, (243255). Moreover, a peripheral effect is suggested by
and granulocyte macrophage colony-stimulating factor the development of thyroid function abnormalities during
has been associated with hypothyroidism and the devel- treatment with second-generation TKIs (imatinib, mote-
opment of thyroid autoimmunity (9). Important risk fac- sanib, nilotinib, and dasatinib) in athyreotic patients
tors for the development of hypothyroidism in patients
(246 248). Imatinib and sorafenib also may increase he-
treated with cytokines include female gender, prolonged
patic metabolism of thyroid hormone.
duration of treatment, advanced age, and preexistent thy-
An increase in dosage of L-T4 may be necessary in athy-
roid autoimmunity (9). Interferon- is a human recom-
reotic patients who are treated with TKIs; therefore, a
binant cytokine used in the treatment of breast cancer,
careful follow-up is necessary during treatment with these
Kaposis sarcoma, chronic hepatitis, and leukemia; its use
drugs (248). Anesthesiologists should be aware of the ef-
is associated with the development of thyroid antibodies in
fects of TKIs on thyroid hormone metabolism that could
20% of patients who were previously antibody negative
prove serious in the perioperative period in view of reports
(229). Persistent hypothyroidism may develop in about
of MC or severe hypothyroidism (249).
30% to 40% of patients with positive antithyroid anti-
bodies before treatment (230). There may be an even
e. Other drugs. Resorcinol may induce hypothyroidism by
greater risk of hypothyroidism with coadministration of
inhibition of TPO and the associated reduction of thyroid
interferon- with ribavirin in patients with hepatitis C
hormone synthesis. Clinical case reports of patients un-
virus-positive hepatitis, or with IL-2 in the treatment of
dergoing resorcinol therapy for dermatological indica-
some tumors (229- 231). The mechanism of onset of thy-
tions reveal thyroid side effects after administration of
roid autoimmunity by interferon- appears related to in-
copious amounts of resorcinol (greater than 34 mg/kg/d)
creases in the expression of cell surface major histocom-
for several months or years (250). Thalidomide may in-
patibility complex class II molecules (232, 233). IL-1 and
crease the risk of hypothyroidism by a mechanism still
TNF- are known to inhibit both iodine organification
and thyroid hormone release. Treatment with IL-2 is as- unknown (251). SHypo developed in about 20% of pa-
sociated with transient painless thyroiditis in about 20% tients with multiple myeloma receiving thalidomide in as-
of patients (233). Granulocyte macrophage colony-stim- sociation with chemotherapy. Some nutraceuticals and di-
ulating factor may induce hypothyroidism in about 10% etary supplements may induce hypothyroidism (252,
of patients (233). 253). In patients taking L-T4, soy protein supplements may
Pretreatment screening is recommended for all patients increase the need for L-T4 by reducing GI absorption
to be treated with cytokines. The development of thyroid (254). Flavonoids are present in plants, beverages, vege-
autoantibodies and the onset of hypothyroidism are not tables, and fruits or in many dietary supplements, and soy
contraindications to continuing therapy with cytokines isoflavinoid has estrogenic activity. The published litera-
(233), but L-T4 therapy should be started as soon as hy- ture provides evidence that these drugs may have antithy-
pothyroidism is documented (233). roid effects by inhibiting thyroid peroxidase and deiodi-
nase activity when administered per os (255, 256). On the
d. Tyrosine kinase inhibitors. TKIs are antineoplastic agents contrary, soy isoflavones have very little significant effect
indicated in the treatment of some metastatic tumors, in- on L-T4 requirement during exposure to vaginal epithe-
cluding thyroid cancer (234, 235). Hypothyroidism is a lium or endometrium (257). Results of a recent study in-
well-recognized side effect of these agents. Several mech- dicated that chronic administration of pure genistein agly-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
460 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

cone (54 mg/d) does not affect thyroid function in 3. Drugs interfering with L-T4 metabolism
postmenopausal women after 3 years of treatment (258). Some drugs may interfere with TSH, FT3, and FT4 eval-
uation in patients receiving replacement doses of L-T4.
2. Drugs interfering with thyroid hormone transport
a. Estrogens, androgens, antiandrogens, and selective estrogen a. Antiepileptic drugs. Phenytoin, carbamazepine, hydan-
receptor modulators. Some drugs, including estrogens, an- toins, phenobarbital, and ritonavir may cause a 20% to
drogens, antiandrogens, selective estrogen receptor mod- 40% decrease in serum total and FT4 concentration and a
ulators, corticosteroids, and 5-fluorouracil, may interfere smaller reduction in serum total and FT3 (223, 263). These
with transport of thyroid hormones by altering the syn-
drugs affect liver cytochrome P450, resulting in an in-
thesis and/or glycidic composition of TBG (223, 259) and
crease in the metabolic clearance and hepatic metabolism
induce changes in circulating levels of total T4 and total T3.
of T4.
Oral estrogens increase TBG concentrations by augment-
Sertraline lowers serum T4 concentrations and raises
ing its sialyation and thereby decreasing its clearance rate
serum TSH. Although the mechanism by which sertraline
and raising its serum concentration. Serum TBG is ele-
vated by about 30% to 50% and serum total T4 by 20% may affect thyroid function is uncertain, some authors
to 35% with use of the commonly employed dosages of 10 have reported that this drug may increase the clearance of
to 100 g of ethinyl estradiol or 0.3 to 2.5 mg of conju- T4. An increased dose of L-T4 is necessary in patients with
gated estrogens (Premarin) (223, 259). As a consequence, hypothyroidism treated with L-T4 who have elevated se-
an adjustment of L-T4 dosage may be necessary in hypo- rum TSH during treatment with sertraline (264).
thyroid patients also taking these agents. The antiestrogen
tamoxifen has only a slight effect on serum TBG concen- b. Other drugs. Rifampicin has similar effects to those of the
tration (223). On the contrary, androgens reduce sialya- latter agents mediated via an important effect on hepatic
tion or synthesis of TBG, thereby decreasing circulating mixed-function oxygenase (79). Aminoglutethimide, lith-
levels of total T4 and total T3 while increasing FT4 and T3 ium, methimazole, propylthiouracil, sulfonamides, tolb-
and perhaps requiring a reduction of L-T4 dosage (223). utamide, propranolol 160 mg/d, and dexamethasone
This effect does not occur in the case of transdermal prep- 4 mg/d may decrease 5-deiodinase activity (79, 223).
arations of estrogens or androgens because of the absence
of the first passage of the drug to the liver. VI. Management of Persistent TSH Elevation in Patients
on High-Dose L-T4 Replacement Therapy
b. Salicylate, clofibrate, furosemide, and heparin. Salicylates in Some patients may have persistently high serum TSH
doses 20 g/d inhibit the binding of thyroid hormones to and low T4 values despite the administration of apparently
TBG and transthyretin, decreasing the total T4 concen- adequate doses of L-T4. In this circumstance, a number of
tration by about 20% to 30% without affecting FT4. Clo- possible interfering factors should be considered and ex-
fibrate, heroin/methadone, and mitotane also may cause a cluded, especially when the prescribed L-T4 doses exceed
decrease in serum TBG concentration. the theoretical daily dose calculated according to body
Large iv doses of furosemide (more than 80 mg) may weight. The following factors should be investigated: 1)
induce a transient increase in serum FT4 and a decrease in poor patient compliance; 2) inadequate or incorrect L-T4
serum total T4 concentrations by inhibiting the T4 binding
dosage and/or administration; 3) increased turnover or
to its carrier proteins (260). Heparin in standard sc doses
excretion of L-T4 related to drugs administered for con-
(5000 U) may inhibit protein binding of T4 because of its
comitant illnesses; 4) heterophile antibody interference
ability to activate lipoprotein lipase. Heparin administra-
with the laboratory test (commonly including antimouse
tion results in a false elevation in the concentration of
antibodies, rheumatoid factor, and autoimmune anti-TSH
unbound (free) T4 in the plasma, particularly when mea-
sured by equilibrium dialysis (261). Some medications antibodies) that may cause a falsely elevated serum TSH;
that increase T4 and T3 in vivo may result in an underes- 5) L-T4 malabsorption due to the coexistence of celiac dis-
timate of free hormone as measured by many hospital or ease, autoimmune gastritis, or administration of drugs
commercial laboratories. that may interfere with L-T4 absorption (eg, calcium, iron,
Total serum T4 corrected for alterations in TBG con- and H2-blockers); 6) coexistence with thyroid hormone
centration will provide a more sensitive estimation of T4 resistance; and 7) coexistence with adrenal insufficiency,
concentration than most popular current methods of free which may induce a TSH elevation reversible with gluco-
hormone estimation (262). corticoid replacement.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 461

A. Timing of L-T4 administration necessity for a higher L-T4 dosage to maintain euthyroid-
The timing of L-T 4 administration has an important ism during the once-weekly regimen (271). In another
impact on absorption. Awareness of common interfer- small single-blind randomized controlled crossover study,
ences with L-T4 absorption such as food is important to twice-weekly L-T4 administration was able to achieve ac-
achieve the desired stable TSH level. Optimal absorption ceptable biochemical control in elderly patients with pri-
of L-T4 occurs with fasting, whereas a reduction of as much mary hypothyroidism (272).
as 40% to 80% may be observed during food and drink In conclusion, there are conflicting data on the impor-
administration (253, 254). tance of the timing of L-T4 administration. These varied
results may reflect heterogeneity in the causes of hypothy-
1. Morning vs evening administration roidism, differences in the patients studied (newly diag-
Some studies have evaluated the efficacy of bedtime nosed vs patients on a stable dose of L-T4), failure to ac-
dosing of L-T4 (265269). In a small nonrandomized study count for confounders (drugs, etc,), and eating habits
involving 11 hypothyroid patients who had been receiving relative to drug administration. L-T4 is generally taken in
a stable dose of L-T4 each morning, a decrease in mean the morning 30 to 60 minutes before breakfast. The pa-
TSH and an increase in FT4 levels was noted after bedtime tient should be fasting to minimize or avoid interference
administration of L-T4 compared with the same L-T4 dose with L-T4 absorption. Traditionally, most patients are
taken in the morning (267). In a subsequent randomized compliant with taking their daily dose of L-T4 in the morn-
double-blind crossover trial, the same investigators re- ing during the fasting state, which has been demonstrated
ported that bedtime administration was associated with a to be the easiest and the best means of administration.
mean TSH decrease of 1.25 mU/L, a mean FT4 increase of Taking L-T4 1 hour before breakfast results in lower TSH
0.07 ng/dL, and a total T3 increase of 6.5 ng/dL (268). values compared with taking the medication only 30 min-
However, no improvement in QOL, lipid profile, and BP utes before breakfast (266). However, bedtime dosing or
was observed with the bedtime regimen (268). Likewise, a L-T4 administration with breakfast could be useful as an
randomized trial of 77 patients with newly diagnosed hy- alternative strategy in patients who have difficulty taking
pothyroidism found no difference in clinical symptoms, their L-T4 dose regularly in the morning before breakfast.
QOL, and lipid profile between L-T4 administration a half Noncompliant patients who have problems taking
an hour before breakfast or 2 hours after an evening meal once-daily L-T4 may engage in a trial regimen of taking
(269). Bedtime administration did not significantly im- their entire L-T4 dose once weekly or half the dose twice
prove TSH levels in a retrospective study on 15 elderly weekly (273). However, this approach should be avoided
subjects (265). And 65 patients in a crossover design of a in patients with underlying heart disease because of the
3- to 8-week regimen were randomized in an examination potential exacerbation of CHD and arrhythmias due to
of the difference in receiving L-T4 in either a fasting state, the transient supraphysiological hormone concentrations
at bedtime, or with breakfast (266). In this study, serum achieved in the first 1 or 2 days. Crushed L-T4 tablets
TSH levels were lower during L-T4 administration in the suspended in water should be given through a nasogastric
fasting state than during nonfasting conditions. A non- tube to patients receiving enteral feeding. For optimal ab-
fasting regimen of L-T4 administration (at bedtime and sorption, feeding should be interrupted with doses given as
with breakfast) was associated with higher and more vari- long as possible or at least 1 hour before resuming feeding.
able TSH concentrations (266). A weekly im injection of L-T4 may be a useful therapeutic
In a recent prospective, randomized, open-label, cross- approach in obtaining biochemical and clinical euthyroid-
over study of 45 patients with primary hypothyroidism, ism in patients with persistently increased serum TSH de-
L-T4 administration during a fasting state was compared spite high doses of L-T4 replacement therapy (274).
with administration during breakfast. This study con-
firmed previous findings because serum TSH level was 3. Intravenous thyroid hormone therapy
higher during L-T4 administration with breakfast than The option of administering iv L-T4 solution is not uni-
during fasting (2.89 vs 1.9 mU/L, P .028) (270). versally available; it should be considered when oral ad-
ministration cannot be used in patients with severe hypo-
2. Daily vs weekly administration thyroidism (275). Because approximately 70% of an
In a randomized crossover study on 12 hypothyroid orally administered dose of T4 is absorbed, individuals
patients, a once-weekly administration of L-T4 was effec- unable to ingest L-T4 should initially receive 70% or less of
tive and well-tolerated. However, TSH levels were in- their usual dose iv.
creased during the weekly regimen compared with those Patients taking lifelong T4 replacement therapy will po-
obtained during the daily dose regimen, suggesting the tentially suffer an interruption of that treatment when hos-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
462 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

pitalized, especially when they are to undergo a surgical disorders requiring more complex treatment regimens and
procedure and are not able to intake anything by mouth, more frequent or more complex monitoring (276). The
including medications. A delay of 24 to 36 hours in T4 poor compliance rate reached a surprisingly high 82% in
ingestion will not be associated with any significant phys- one study of 100 patients from public hospitals in Campi-
iological alterations, but the patient who is NPO (nilim per nas and Rio de Janeiro, who presented with persistently
os) for several days is best managed by the administration elevated TSH levels irrespective of efforts to increase their
of parenteral L-T4, preferably by the iv route. For example, L-T4 dose (277). Sixty-two percent of these patients had a
prolonged periods of rest or inactivity of the GI tract are poor adherence, including forgetting or neglecting to take
common after intra-abdominal procedures, especially their pills in 62%, complaints of intolerance or side effects
when intestinal resection has been performed. For main- in 12%, and financial stress in 8% (277). Interestingly,
tenance of euthyroidism in these and other surgical pa- about 18% of these patients had serum T4 levels in the
tients, both L-T4 and T3 are widely available in iv hyperthyroid range after 1 month of ostensibly appropri-
preparations. ate L-T4 dosage (277). Improvement in adherence to L-T4
The iv dose is typically administered once daily as a therapy is usually observed with increasing age, and ad-
bolus injection. Adjunctive iv administration of T3 has herence levels of 80% or better are seen in patients aged 60
been advocated by some for the treatment of MC, but years or older (276). On the contrary, the worst medica-
otherwise its use has little application for hospitalized hy- tion compliance is seen in adolescents (276).
pothyroid patients. Because 90% to 100% of orally ad- Irrespective of our awareness of the frequency of non-
ministered T3 is absorbed, the iv dose would not need to compliance, it is often difficult to exclude it as the expla-
be reduced. Adequacy of iv T4 dosage can be monitored in nation for a persistently elevated serum TSH. Some of the
the usual manner with measurements of serum TSH and usual measures for assessing compliance include pill
FT4. counting, counseling methods, interviewing, and labora-
tory tests to estimate L-T4 absorption. In an RCT to assess
B. Poor patient compliance the impact of education on L-T4 adherence, a booklet
One daily dose of L-T4 accounts for 14% of the total about L-T4 medication was distributed to the patients, but
weekly dose. Therefore, a missed dose of L-T4 may affect no improvement in compliance was observed (278).
FT4 and TSH levels over several weeks due to the long Rather than educational materials, it appears that it is the
half-life of L-T4. doctor-patient relationship that plays a key role in achiev-
High serum TSH with normal (or even high) FT4 levels ing adherence to the medication regimen. Adherence to
may be observed in poorly compliant patients, typically treatment may be improved with simple questions to the
after missing several doses over time but then taking the patient about their difficulties in taking L-T4 medication,
L-T4 shortly before their follow-up thyroid function blood by better understanding of psychosocial factors that may
tests. The prevalence of noncompliant hypothyroid pa- impact compliance, or by a discussion of the patients con-
tients has been reported to be between 30% and 80% cerns about potential side effects.
(276 278) despite the simplicity of L-T4 replacement ther- The patients primary-care physician or relatives could
apy and its once-daily oral administration. be enlisted to pay special vigilance to the patients pattern
Some of the following factors may play a role in poor of L-T4 ingestion and help with those noncompliant pa-
adherence to L-T4 therapy: 1) patients could have poor tients who may have psychological or psychiatric disor-
adherence to L-T4 therapy because they may not be satis- ders. Such an approach to surveillance has been recom-
fied with the sense of well-being achieved during the initial mended by psychiatrists for patients presenting the
period of replacement therapy; 2) the cost of the medica- psychopathological features of a Munchhausen syndrome
tion may be an important economic consideration in some or factitious disorder (279) Patients who falsely insist that
countries; and 3) patients may be intolerant of side effects they are being compliant but have thyroid function tests
related to a sudden increase in metabolism induced by L-T4 indicating otherwise (pseudomalabsorption) may be in-
therapy, especially if the initially selected dosage is vestigated further with parenteral infusion of L-T4 (280).
TSH-suppressive. Such assessments may include administration of a single
In one study that assessed compliance in some chronic large dose of T4, ie, 1000 g given in the morning after
illnesses (ie, hypothyroidism, hypertension, diabetes mel- fasting overnight with serial blood sampling for thyroid
litus, hypercholesterolemia, or osteoporosis), only about function tests at 2, 4, and 6 hours after L-T4 administration
68% of hypothyroid patients had good adherence to L-T4 (281). Published data suggest that a serum FT4 peak at 2
therapy (276). In this study, the noncompliance rate in hours rising above the upper limit of the normal range
hypothyroid patients was similar to that in other chronic (more than 25 pmol/L) with an increment of more than 20

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 463

pmol/L suggests poor adherence to treatment, or pseu- to 4 hours after L-T4 is taken. Moreover, the dose of L-T4
domalabsorption in most cases (281285), but unfortu- may need to be increased by 20% to 30% in patients tak-
nately, there are no well-established standards for this test. ing these drugs, and stability of dosage control will be
Although a dose of 1000 g of L-T4 is traditionally used to affected whenever the regimen of the other drugs is altered.
exclude pseudomalabsorption, some authors have pro-
posed that it would be more prudent to consider a 500-g 3. GI diseases
dose of L-T4 to avoid side effects. A radioisotope-labeled Failure to respond appropriately to higher doses of L-T4
L-T4 (double-labeled T4) tracer technique may be used to in compliant patients suggests malabsorption due to a
assess T4 absorption more accurately, but this technique is known or occult GI disorder that may require further in-
not generally available for routine clinical practice (135). vestigation. Celiac disease is a condition resulting in in-
tolerance to dietary gluten and related proteins; it may be
C. Interference with absorption of L-T4 associated with severe symptoms of steatorrhea and
weight loss. Celiac disease affects mainly the jejunum and
1. Food and beverages
proximal ileum, which are both sites involved in L-T4 ab-
The fraction or degree of L-T4 absorption and its con-
sorption (301). Primary hypothyroidism due to autoim-
sistency are important factors that will determine dosage
mune disease and celiac disease-related malabsorption
of replacement L-T4 therapy. Food, juices, milk, coffee, soy
may be closely associated on an autoimmnune basis (301,
products, calcium, iron, multivitamin supplements, higher 302), and coexistent hypothyroidism is commonly de-
fiber intake, and a variety of drugs may influence the ab- tected in patients with newly diagnosed celiac disease. The
sorption of L-T4 (254, 286 289). The ingestion of food is prevalence of celiac disease in autoimmune hypothyroid-
associated with both delayed and reduced L-T4 absorp- ism has been reported to be between 2% and 5% of cases
tion. Coffee, especially espresso coffee, impairs L-T4 ab- (301), whereas the prevalence of autoimmune hypothy-
sorption; it may interact with T4 tablets and keep T4 roidism in patients with celiac disease has been reported to
within the intestinal lumen, making it less available for be as high as 30% (302). The diagnosis of celiac disorder
absorption (287). Dietary fiber and soy products appear to is easily confirmed by the combination of clinical features
have a small but still significant effect on L-T4 absorption and positive serological testing. In the last few years, se-
(254, 288). Grapefruit juice and probably other citric rological screening with IgA, antiendomysial antibodies
fruits may also impair L-T4 absorption (289). A recent that have a high sensitivity (85%98%) and specificity
study by Vita et al (168) suggests that the problem of coffee (97%100%) and tissue transglutaminase (sensitivity of
interference might be overcome by using a soft gel capsule 95%98% and specificity of 94%95%) have substituted
preparation containing T4 dissolved in glycerine to pre- for the lesser sensitive measurement of antigliadin anti-
sumably facilitate more rapid absorption. Patients with bodies. However, the diagnosis of celiac disease can be
impaired acid secretion and potential T4 malabsorption missed in some selected cases of IGA deficiency and /or in
might benefit from this new L-thyroxine preparation. the presence of atypical or subclinical manifestation of this
disorder. Atypical celiac disease accounts for more than
2. Drugs half of all cases; it may be characterized by anemia, os-
Gastric acidity is important for L-T4 absorption (140, teoporosis and subfertility, and may be discovered as a
141). The variability of gastric acid production contrib- result of investigation done after noting L-T4 malabsorp-
utes to the individual variability observed in daily L-T4 tion (302). The diagnosis of celiac disease is confirmed
requirement (140). with observation of the classic findings on endoscopic
Some medications (iron supplements, calcium carbon- small bowel biopsy; after implementation of a gluten-free
ate citrate and acetate, laxatives, cholestyramine, other diet, L-T4 absorption usually will improve (303, 304).
resins like kayexalate or colestipol, lovastatin, sucralfate, Lactose intolerance is another frequent cause of mal-
aluminum, magnesium, orlistat, and raloxifene) may in- absorption (305). Intestinal infections and small intestinal
terfere with L-T4 absorption by altering gastric pH or by bacterial overgrowth may develop in hypothyroid patients
direct sequestration of L-T4 into an insoluble complex due to the decreased GI motility and slower oro-cecal tran-
(290 298). Moreover, antacids such as proton pump in- sit seen before restoration of euthyroidism (306). Lactose
hibitors and H2 receptor antagonists may decrease intes- intolerance may be associated with malabsorption syn-
tinal absorption of L-T4 by increasing gastric pH (142, drome, abdominal pain, flatulence, diarrhea, and weight
299, 300). loss (306).
To counteract these inhibitory effects on absorption, The normal acid environment in the stomach is neces-
calcium and iron supplements should not be taken until 3 sary for L-T4 absorption. Therefore, Helicobacter pylori

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
464 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

infection and atrophic gastritis may affect L-T4 absorption VII. Potential Adverse Effects
due to the severe hypochlorhydria and the increased am- During L-T4 Therapy
monia production (142). Performance of a 14C urea breath
Overtreatment or excessive L-T4 dosage, has been re-
test and serological tests for H. pylori antibodies per-
ported in about 30% to 50% of patients treated with L-T4
formed on blood or saliva (although the sensitivity of sal-
in the United States, England, and Scotland (1214). A
ivary antibody tests is low) and in stool specimens by PCR
higher prevalence of over-replacement therapy has been
should be considered in patients with L-T4 malabsorption
observed in those receiving higher doses of L-T4 (13), in
in whom H. pylori infection is suspected .
elderly individuals, and in patients with low body weight
Autoimmune gastritis is an autoimmune disorder char-
or with diabetes (14).
acterized by the presence of circulating antiparietal cell
antibodies; it is frequently associated with Hashimotos
A. Adverse effects of TSH suppression
thyroiditis (307309). This autoimmune gastric disease
Young patients with endogenous subclinical hyperthy-
may lead to atrophic gastritis with a reduced number and
roidism (EndoSHyper) and those with exogenous subclin-
function of oxyntic glands and a consequent reduction of
ical hyperthyroidism (ExoSHyper) and undetectable se-
gastric acidity with impaired L-T4 absorption. In one study
rum TSH levels during long-term therapy may have
by Centanni et al (142), patients with atrophic gastritis
symptoms and signs of thyroid hormone excess (317
required an increase of 27% in their L-T4 dose, whereas
319). The hypermetabolic state associated with undetect-
patients with H. pylori nonatrophic gastritis required an in-
able serum TSH can impair psychological, social, and
crease of 22% in their dose. And patients with both H. pylori
physical QOL and may be associated with the develop-
and atrophic gastritis needed a 34% higher doses of T4.
ment of some important cardiovascular risk factors (in-
Patients with positive tests for H. pylori should have
creased heart rate and enhanced risk of atrial arrhythmias,
thyroid function retested after treatment of their infection
increased left ventricular [L] mass, and diastolic dysfunc-
to avoid overtreatment with L-T4 when they are cured
tion) in ExoSHyper and EndoSHyper (9, 10, 317320).
(310).
Elderly patients may be asymptomatic for specific symp-
Certain parasitic disorders may be an important cause
toms of thyroid hormone excess compared with younger
of malabsorption. There is a report of a patient with severe
patients. EndoSHyper is not associated with an increased
hypothyroidism caused by intestinal giardiasis (311).
risk of depression and anxiety in patients aged over 65
Bariatric surgery and all operations that reduce the GI
years (321). However, a recent review of 23 studies sup-
absorptive surface may also impair L-T4 absorption (312
ports the evidence that there is an association between
316). Previous GI surgery (jejunostomy, jejuno-ileal by-
EndoSHyper and cognitive impairment, although the
pass, and short bowel syndrome) may be other causes of
mechanism of this association remains unclear (322).
L-T4 malabsorption. L-T4 malabsorption is observed in
patients with cirrhosis, obstructive liver disease, or pan- B. Cardiovascular risk and TSH suppression
creatic insufficiency (16).
ExoSHyper may lead to adverse cardiovascular and
An absorption test with 1000 to 2000 g of L-T4 should
skeletal effects in the elderly. Some prospective studies
be performed in patients suspected of having malabsorp-
have reported a 2 to 3-fold higher risk of AF in patients
tion. The most frequent conditions associated with L-T4
over 60 years of age affected by ExoSHyper (323) or En-
malabsorption are summarized in Table 14.
doSHyper (323325) compared with euthyroid controls.
Two recent meta-analyses have linked SHyper with inci-
dent AF and an increased risk of HF and CHD and mor-
Table 14. Frequent GI Conditions Associated With
tality, mainly in patients with endogenous TSH 0.1
L-T4 Malabsorption
mU/L (188, 326). A pooled analysis of individual partic-
GI Condition ipant data (648 participants) with EndoSHyper from 6
Celiac disease prospective cohorts indicated that the HR for HF events in
Lactose intolerance
Atrophic gastritis
age- and gender-adjusted analyses was 1.46 (95% CI,
H pylori infection 0.94 2.27) compared with euthyroid subjects during a
Intestinal infections and small intestinal bacterial overgrowth median follow-up of 10.4 years. The HR in patients with
Liver diseases: cirrhosis, obstructive liver disease
undetectable serum TSH was 1.94 (95% CI, 1.013.72)
Pancreatic diseases
Pancreatic insufficiency (188). The incidence of HF is increased further in elderly
Previous gastrointestinal surgery, jejunostomy, jejunoileal bypass, short patients (mean age 75.3 years) with a high risk of cardio-
bowel
vascular disease (104). There are conflicting data on the

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 465

risk of stroke in patients with EndoSHyper (184, 327, apy should be avoided in this age group, consistent with
328). The incidence of stroke was increased in patients recent AACE and ATA guidelines (3).
with EndoSHyper in a population-based prospective
study conducted in Denmark on 609 subjects from a gen- C. Fracture risk and TSH suppression
eral practice aged 50 years or above with normal LV func- Overt hyperthyroidism is an important risk factor for
tion (HR was 3.39; 95% CI, 1.1510.00; P .027) after osteoporosis and fractures (331). Four meta-analyses sug-
adjusting for sex, age, and AF (328). Likewise, an in- gest that postmenopausal patients are at risk of bone loss
creased mortality from circulatory diseases (both cardio- during TSH suppression (332335). Several studies have
vascular and cerebrovascular) was associated with a low evaluated whether ExoSHyper increases the risk of frac-
TSH in a 10-year follow-up study of 1191 subjects aged tures. Women with a history of thyroid cancer appeared to
over 60 years with EndoSHyper (327). have their first fracture earlier (P .01) than women with-
There are differing results in studies assessing the risk of out thyroid disease (336). A prospective population-based
total and cardiovascular mortality in patients with ExoSHy- study in patients receiving L-T4 prescriptions reported an
per. A prospective cohort study evaluated the mortality increased risk of fracture in postmenopausal women with
among elderly women (65 years of age), receiving long- a history of hyperthyroidism (337). A 3-fold increased risk
term thyroid hormone therapy (15.8 years) (329). In multi- of hip fracture and a 4-fold increased risk of vertebral
variate analysis, the mortality was similar among users of fracture was found in a prospective study by Bauer et al
thyroid hormone compared with nonusers (HR, 1.11; 95% (338) of 686 women older than 65 years with ExoSHyper
CI, 0.98 1.24; P .09). In this study, only a history of pre- and a serum TSH level 0.1 mU/L who were compared with
vious hyperthyroidism was associated with a small increase women with normal serum TSH levels. The risk of fractures
in all-cause and cardiovascular mortality (329). On the con- was also increased in women receiving L-T4 doses that main-
trary, the Thyroid Epidemiology Audit and Research Study tained serum TSH in the range of 0.1 to 0.5 mU/L (338).
(TEARS), a population-based study performed in Tayside, A suppressed serum TSH (0.03 mU/L) was associated
Scotland, reported that elderly patients (mean age 61.6 years) with a doubled risk of osteoporotic fracture in postmeno-
with suppressed serum TSH (0.03 mU/L) during L- T4 re- pausal women (mean age 60.3 years) in a recent popula-
placement therapy (median follow-up of 4.5 years) had an tion-based study of patients taking L-T4 therapy (330).
increased risk of cardiovascular disease, dysrhythmias, and Patients with a serum TSH below the reference range, but
fractures when compared with patients with a serum TSH not fully suppressed (0.04 0.4 mU/L), had no increased
within the laboratory reference range (330). The risk of these risk of fractures. However, data on FT4 or T3 levels were
outcomes was not increased in patients with a low but not not reported in these studies (330, 338).
suppressed TSH (0.04 0.4 mU/L) (330). For all endpoints, A retrospective cohort study of 213 511 L-T4 users with
there was an increased risk with older age after adjusting for a nested case-control design was performed in Ontario,
sex, history of a previous thyroid condition and cardiovas- Canada, on the available data from the population health
cular disease, socioeconomic status, and presence of diabetes database. Residents 70 to 105 years old receiving at least
(330). 1 prescription for L-T4 were selected (339). This popula-
A recent meta-analysis provided evidence that En- tion-based study found a strong dose-response relation
doSHyper may increase cardiovascular mortality, partic- between current use of L-T4 and the risk of fracture. A
ularly in patients with undetectable serum TSH (326). In significant correlation was found in both sexes between
age- and sex-adjusted analyses, EndoSHyper was associ- current L-T4 use and increased risk of fracture among older
ated with increased total mortality (HR, 1.24; 95% CI, people (70 years) (339). The risk of fractures remained
1.06 1.46) and mortality due to cardiovascular disease higher even among those who discontinued L-T4 within 6
(HR, 1.29; 95% CI, 1.021.62), especially in the presence months (339). Although thyroid function tests were not
of a TSH level 0.10 mU/L (326). performed in this study, a strong correlation was observed
It remains to be established whether or not ExoSHyper between L-T4 dose and the risk of fracture. Higher doses of
and EndoSHyper exert the same adverse effects because T3 L-T4 were associated with a 2 to 3-fold increased risk of
levels are higher in patients with EndoSHyper, whereas fractures compared with lower doses (339). These findings
FT4 concentrations are often elevated in many patients were confirmed in a recent study in which an increase in
undergoing L-T4-suppressive therapy with a greater T4 to risk of femur fracture was found in men aged over 65 years
T3 ratio than in patients with EndoSHyper (28 30) . with EndoSHyper (340). In this prospective cohort of U.S.
However, all of the above referenced data suggest that community-dwelling adults, Lee and co-workers (340)
suppressed serum TSH may have harmful cardiovascular followed 5567 people, 65 years or older, for a median of
effects in elderly patients. Therefore, excessive L-T4 ther- 13 years and showed that the incidence of hip fracture was

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
466 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

Table 15. Consequences of Over- and Undertreatment VIII. Combined Treatment With L-T4 Plus
With L-T4 Liothyronine in Hypothyroid Patients

Potential Consequences of Potential Consequences of A. Persistence of symptoms in hypothyroid patients


Overtreatment With L-T4 Undertreatment with L-T4 with normal serum TSH during L-T4 therapy
Symptoms of hyperthyroidism Symptoms of hypothyroidism About 15% of patients treated with L-T4 do not reach
Accelerated bone loss Elevated total cholesterol and clinical euthyroidism and continue to complain of psy-
LDL-cholesterol
Fractures Impaired cognition
chological impairment, weight gain, depression, and per-
AF Increased diastolic BP sistence of symptoms of thyroid hormone deficiency de-
HF Increased body weight spite biochemical euthyroidism demonstrated by normal
CHD events Increased risk of CHD events
and mortality
serum TSH levels. Three large community-based studies
Increased risk of HF from the United Kingdom (345), The Netherlands (346),
and Norway (347) concluded that successful treatment of
hypothyroidism is associated with only a partial recovery
of neurocognitive improvement and psychological well-
higher in men (but not women) with SHyper compared
being, suggesting that replacement treatment with L-T4
with euthyroid controls (HR, 4.91; 95% CI, 1.1321.27).
might not be fully adequate for optimal brain function.
After excluding those with a baseline use of thyroid-alter-
Vitality and mental heath were significantly lower in pa-
ing medications, men with EndoSHyper had a higher HR
tients receiving L-T4 than in the general population in the
of 4.91 for hip fractures (95% CI, 1.1321.27). No clear study by Wekking et al (346) on 141 patients with com-
association between subclinical dysfunction and fracture pensated hypothyroidism. Similarly, in the large commu-
was observed in women in this study (340). nity-based survey on 381 patients by Saravanan et al
Thus, it appears that current evidence supports the as- (345), 26% of the patients with normal thyroid function
sociation of suppressed serum TSH during L-T4 therapy on L-T4 monotherapy scored significantly worse than did
and an increase risk of fractures, especially in certain pa- euthyroid matched controls on measures of well-being
tients including those receiving excessive therapy after pre- (General Health Questionnaire, 32.3% vs 25.6%) and hy-
vious treatment of hyperthyroidism, the elderly, post- pothyroid symptoms (47% vs 35%). Furthermore, the
menopausal women, and people with risk factors for HUNT 2 study on 1546 women 40 years of age who
osteoporosis. Treatment with L-T4 should be carefully were receiving L- T4 therapy observed higher depression
monitored in patients at a higher risk of bone fracture. and anxiety scores in comparison with 18 137 subjects
Lower doses of replacement L-T4 therapy should be used without thyroid disease and who were not treated with T4
in postmenopausal women and in the presence of risk fac- (depression 18.4% vs 13.0%, P .001; anxiety, 23.4% vs
tors for bone fractures. 18.7%, P .001) (347).
However, traditional subjective procedures that assess
D. Adverse effects of increased serum FT4 during symptoms and QOL are not reliable markers for the eval-
L-T4 therapy uation of the peripheral effects of thyroid hormone, and
Patients receiving replacement L-T4 therapy usually these parameters are not sensitive enough in detecting
have serum FT4 values at the upper limit of the reference slight thyroid hormone deficiency or excess (100, 111,
range compared with euthyroid subjects. Significantly 170) .
higher serum FT4 levels have been reported after total thy- Some cardiovascular risk factors (such as lipid param-
roidectomy in patients receiving L-T4 treatment compared eters, endothelial function, BMI, adiposity, and diastolic
with their prethyroidectomy levels, especially during TSH hypertension) may not be completely normalized after re-
placement therapy (348 355), leading to an increased
suppression as would be expected (28).
morbidity for circulatory diseases and ischemic heart dis-
Some studies have shown adverse effects of high FT4 val-
ease despite apparently adequate treatment with L-T4
ues on prognosis, risk of AF, cardiovascular events, demen-
(356).
tia, and frailty in elderly patients (341343). Moreover, an
increased cardiovascular mortality was found in those 85 B. The effects of low-normal serum TSH during
years of age (101, 344). The data described above suggest L-T4 therapy
that TSH suppression and high FT4 levels should be avoided There are conflicting results on the potential improve-
in elderly patients receiving long-term TSH-suppression ment in QOL when the T4 dosage is titrated until serum
therapy, independent of the serum TSH value (Table 15). TSH is in the lower part of the reference range in patients

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 467

with primary hypothyroidism (155, 170). Small increases latter study, the L-T4 dose needed to normalize thyroid
in L-T4 dosage that approach TSH suppression may im- hormone concentrations was different in each tissue, and
prove resting energy expenditure and reduce fat mass in only combined treatment with L-T4 and L-T3 was able to
hypothyroid patients (357). The effect is due to a signifi- ensure euthyroidism and normalize both plasma and tis-
cantly higher relative increase in resting energy expendi- sue T4 and T3 concentrations (364). Although this exper-
ture as was observed in a prospective 12-month study in imental observation cannot be extrapolated to humans
hypothyroid patients whose TSH level was maintained in without the usual caveats, some studies have reported that
the low-normal TSH range (0.4 2.0 mU/L) during L-T4 low serum FT3 and significantly higher serum FT4 levels
therapy compared with a subgroup maintained within a can be observed in thyroidectomized patients receiving
high-normal TSH range (2.0 4.0 mU/L) (357). However, L-T4 replacement therapy (28 30). For example, in a pro-
in this study, no significant differences were observed be- spective study of 50 patients aged 18 to 65 years who were
tween the 2 groups of patients in other clinical variables scheduled for total thyroidectomy, T3 levels were lower
(lipid profile, body composition, or bone mineral density during replacement doses of L-T4 compared with their pr-
[BMD]) (357). ethyroidectomy levels (28). Similar findings were reported
Chemical or laboratory markers may be more sensitive in a recent retrospective study on 135 patients with thy-
indicators of thyroid hormone effect than the patients roid cancer (30). In this study, higher serum FT4 levels
complaints. For example, changes in T4 dosage did not were observed after total thyroidectomy compared with
produce significant changes in hypothyroid symptoms, those noted before surgery; moreover, serum FT3 levels
well-being, or QOL in a double-blind, RCT with a cross- were significantly lower when serum TSH levels were in
over design of 56 subjects with primary hypothyroidism, the normal reference range (0.33 mU/L) (30). In both of
although the levels of SHBG, alkaline phosphatase, and these studies, post-thyroidectomy serum T3 levels were
deoxypyridinoline were higher and total cholesterol levels equivalent to the pre-thyroidectomy values only when se-
were lower during higher doses of L-T4 compared with rum TSH was suppressed (28, 30).
lower doses (358). Contrary to treatment with a combi- In a recent large cross-sectional study of 1880 athy-
nation of L-T3 plus L-T4 or to baseline levels in control reotic patients on L-T4 treatment, serum T3 levels were
healthy subjects, monotherapy with L-T4 is usually asso- significantly lower, although still within the normal ref-
ciated with higher FT4 levels and a higher T4 to T3 ratio, erence range, compared with the levels observed in 3875
especially in hypothyroid patients receiving TSH replace- matched euthyroid controls (29). TSH was normalized
ment doses of L-T4 (359 361). during L-T4 therapy (0.4 4.0 U/L), but FT4 levels were
Supraphysiological doses of L-T4 are able to suppress above the normal limit of the reference range in 72% of the
serum TSH due to stimulation of D2 activity in the hypo- patients, whereas T3 levels were below the normal limits in
thalamus and pituitary. On the other hand, with a dose of 15.2% and the FT4 to FT3 ratio was higher than the nor-
exogenous L-T4 that suppresses serum TSH, there is inhi- mal range in 29.6% of patients. The FT4 to FT3 ratio
bition of T3 secretion by the thyroid gland and reduced increased with an increasing L-T4 dose, and the widely
peripheral D2 activity with decreased conversion of T4 to variable ratio observed in this study indicates great het-
T3. Therefore, the increased sensitivity of the pituitary/ erogeneity of peripheral T3 production in thyroidecto-
hypothalamic feedback mechanism to serum T4 may ex- mized individuals (29). These data support the concept
plain why suppressive doses of L-T4 can be associated with that the lack of the normal 20% thyroidal T3 secretion in
low T3 levels and high T4 levels (26, 362, 363). Current thyroidectomized subjects might not be compensated by
guidelines and recommendations suggest avoiding low or an increased peripheral deiodination of T4. Higher serum
suppressed serum TSH during replacement L-T4 therapy T4 levels are necessary in thyroidectomized patients to
on the basis of the absence of beneficial effects and the risk achieve normal serum T3 concentrations and compensate
of adverse effects on bone and heart (3). for the absence of T3 secretion by the thyroid gland. There-
fore, in thyroidectomized patients who are persistently
C. L-T4 therapy in thyroidectomized patients symptomatic during L-T4 replacement therapy, it is im-
Intracellular T3 levels do not necessarily reflect similar portant to evaluate whether the normalization of serum
levels in the circulation. Thus, the important moiety is TSH is also associated with both normal FT3 and FT4
primarily intracellular T3, and normal plasma T3 levels do levels (27).
not indicate that T3 is also normal in the peripheral tissues. All of these observations suggest that TSH is not a per-
Administration of T4 (in doses ranging from 0.2 0.8 g/ fect marker of peripheral tissue euthyroidism in thyroid-
100 g body weight per day) was unable to normalize T4 ectomized patients and that in these patients there may be
and T3 in all tissues of thyroidectomized rats (364). In the a wide tissue heterogeneity in T3 production (27).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
468 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

D. Polymorphisms in the genes of the deiodinases E. Effects of combination treatment with T3 and T4
Deiodinase activities are specifically regulated in tis- T3 has a half-life of only 1 day in contrast to the long 6-
sues. This fact may have important consequences for the to 7-day half-life of T4. Wide fluctuations in serum T3
peripheral effects of thyroid hormone and for the set- levels can be observed after its administration (26). Given
points of feedback regulation (27). Polymorphisms in the its short half-life, 3 daily doses of T3 are usually required
genes of deiodinases and in thyroid hormone transporters to obtain physiological and stable circulating T3 levels
may influence tissue T3 availability (365367). The nor- (27). A modest 16% rise in FT4 with no change in FT3
mal brain contains an efficient D2 activity that serves to levels can be observed in the first 4 hours after L-T4 ad-
maintain constant T3 concentrations (368). Defects in the ministration (379). In contrast, a marked rise of 42% in
D2 pathway might be responsible for a reduced T3 in the FT3 levels can be detected within the first 4 hours (T3/
brain or in other tissues and may explain the persistence of T4:T4 6.24:4.63 mU/L, P .001) after L-T3 adminis-
hypothyroid symptoms in some patients receiving L-T4 tration. Moreover, mean exposure to FT3, calculated as
(368). A common Thr92AlaD2 polymorphism has re- AUC is higher (L-T3/L-T4:L-T4 1148:1062, P .0001)
sulted in some studies in tissue T3 insufficiency (368). The during L-T3 treatment (379). In addition, high FT3 levels
Thr92Ala polymorphism has been associated with a num- can persist during chronic L-T3 treatment with potentially
ber of conditions including variation in the hypothalamic- dangerous consequences on the heart and the onset of
pituitary-thyroid axis, altered bone turnover, insulin thyrotoxic symptoms (379).
resistance, increased BMI, mental retardation, osteoar- Several randomized controlled studies have evaluated
thritis, and bipolar disorder (365375). Moreover, a re- the effects of combination therapy with L-T3 and L-T4.
cent meta-analysis including 11 033 individuals showed Several of these studies have had a parallel design, and
that the homozygous form of the D2 Thr92Ala polymor- others were double- or triple-blinded controlled studies
phism is associated with an increased risk of type 2 dia- with an appropriate crossover design (26). Three meta-
betes (376). analyses have evaluated the effects of combination treat-
The presence of the D2 Thr92Ala polymorphism was ment with L-T3 and L-T4 vs L-T4 alone (380 382). The
associated with the need for higher L-T4 doses to normalize L-T4/L-T3 combination therapy had no advantage when
serum TSH in 191 consecutive thyroid cancer patients, compared with standard L-T4 monotherapy in the meta-
previously treated with near-total thyroidectomy and ra- analysis performed by Grozinsky-Glasberg (380) on 11
dioiodine ablation (217). On the contrary, the same poly- RCTs, including a total of 1216 hypothyroid adult pa-
morphism was not associated with a requirement for tients. In this examination, bodily pain, depression, anx-
higher L-T4 doses to normalize serum TSH levels in pa- iety, fatigue, QOL, body weight, total serum cholesterol
tients with autoimmune hypothyroidism (377). Appelhof and triglycerides, serum LDLs, and high-density lipopro-
et al (378) reported that 2 polymorphisms in the D2 gene teins were not different during treatment with L-T4 alone
(the D2 ORFa-Gly3Asp and D2 Thr92Ala polymor- and during combination therapy with L-T4 and L-T3 (380).
phisms) were not determinant for differences in well- Similar results were reported in the meta-analysis by
being, neurocognitive function, or preference for T4/T3 com- Ma et al (381) published in 2009, which included studies
bination therapy in 141 patients with primary autoimmune on a total of 1243 patients. L-T4/L-T3 combination therapy
hypothyroidism. The Weston Area T4/T3 (WATTS) study of was not superior to L-T4 monotherapy in terms of psy-
552 patients on L-T4 replacement therapy suggested that chological and physical well-being with no statistically
there is a small number of patients with the Thr92AlaD2 significant differences in the other variables. The third of
polymorphism with impaired psychological well-being and a the meta-analyses, by Joffe et al (382), on 9 controlled
worse baseline General Health Questionnaire score during studies found no considerable differences of combination
L-T4 replacement therapy (31). This genotype was present in L-T4 plus L-T3 therapy vs L-T4 alone on psychiatric
16% of their study population (31) and was associated with symptoms.
greater clinical improvement on L-T4/L-T3 therapy compared Several meta-analyses have described beneficial thera-
with L-T4 monotherapy (2.3 General Health Questionnaire peutic effects of L-T3 given in combination with tricyclic
points at 3 months and 1.4 points at 12 months). Interest- antidepressants compared with placebo in euthyroid pa-
ingly, this polymorphism had no impact on circulating thy- tients with resistant depression (383, 384). However, al-
roid hormone levels (31). though some studies have investigated the effects of com-
Additional studies will be necessary to confirm the pos- bined L-T4 and L-T3 treatment in depressed hypothyroid
itive neuropsychiatric response to combined L-T4/L-T3 patients (385, 386), only one has reported a beneficial
treatment vs monotherapy with L-T4 in patients with the effect of this treatment in improving depression, mood,
Thr92Ala polymorphism. and cognition (385). Despite the lack of beneficial effects

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 469

of combination therapy vs monotherapy with L-T4 de- iodine for previous hyperthyroidism), their wide age
scribed by the meta-analyses, 3 studies documented that range, and the small sample size in several studies (26).
L-T4 plus L-T3 therapy was preferred by the patients com- Prospective controlled trials will be necessary to eval-
pared with L-T4 monotherapy. Preference rates above uate the potential beneficial effects of combined T3/T4
50% for combined therapy were observed in 66% of sub- treatment vs monotherapy with L-T4 in patients with hy-
jects in the study by Bunevicius et al (385), in 69% of pothyroidism. These trials should enroll homogeneous
patients studied by Escobar-Morreale et al (387), and in groups of patients to ascertain which patients could ben-
52% of those in the study by Appelhof et al (388). Inter- efit from this treatment as well as employ the correct T4 to
estingly, this preference for combined therapy was ob- T3 ratio and monitor the appropriate peripheral param-
served in 2 studies in which serum TSH was suppressed eters. Pharmacogenomic studies would help identify those
(385, 388). In 2 double-blind randomized studies, the patients that might benefit from combined treatment.
mean body weight decreased by 1.7 kg with combination Treatment with currently available formulations of T3 do
treatment compared with monotherapy with L-T4 (388, not result in a physiological or normal serum T3 profile. A
389). In these 2 studies, the patient preference for combi- long-acting slow-release form of T3 would be required to
nation therapy was associated with a significant reduction mimic normal physiological endogenous T3 production in
in body weight and improvement in BMI (388, 389). It hypothyroid patients.
should be underscored that some patients may feel better In the meantime, expert opinion and AACE, ATA, and
just because they are participating in a trial (the Haw- ETA guidelines suggest avoiding treatment with T3 in
thorne effect). pregnant women and in patients with a history of arrhyth-
Few reports have evaluated peripheral parameters of mias or chronic ischemic heart disease (3, 26, 32). The
thyroid hormone action (cardiovascular parameters, lipid addition of T3 to T4 monotherapy should be restricted to
profile, bone metabolism and structure, energy expendi- persistently symptomatic hypothyroid patients despite
ture, or SHBG) during combined therapy; however, the their biochemical euthyroidism during L-T4 replacement
conflicting findings and the relevant methodological lim- therapy with the goal of achieving improved QOL. Be-
itations of these studies do not permit definitive conclu- cause of the risks inherent in T3 treatment, combined ther-
sions on the effects of combination therapy and its poten- apy should be managed only by skilled specialists and
should require careful follow-up to detect and/or avoid
tial beneficial effect (26). Euthyroidism was not reached in
possible adverse effects. Blood samples for evaluation of
most of the available studies that used a fixed L-T3 dose
T3 levels should be obtained before the morning dose. A
(ranging from 7.520 g) (128, 386, 390 395) or a wide
liquid preparation that allows titration of dosage in drops
L-T4 to L-T3 ratio (from 3:1 to 15:1) (387389, 396, 397).
is currently available in Europe and may prove useful for
The physiological L-T4 to L-T3 ratio in humans is about
personalization of L-T4/L-T3 combined treatment.
13:1 to 15:1, and therefore overtreatment or undertreat-
ment (suppressed or elevated TSH or high or low FT3 F. Combination treatment with T3 and T4 in children
levels) was frequently associated with combined treatment with CoH
in these studies (26). The presence of atrial arrhythmias In children with CoH, a high serum TSH (10 20 mU/L)
has been reported in overtreated patients during combi- may persist in about 10% of patients treated with L-T4
nation treatment (128, 396). Once-daily T3 dosage was despite normal serum T4. This has been explained as due
used in most of the studies, and only a few studies divided to either undertreatment or an abnormal maturation of T4
the T3 dosage into 2 daily doses (26). feedback control of TSH secretion. One study reported a
The duration of combination therapy in all of these prevalence of pituitary thyroid hormone resistance in
studies was restricted to only a few weeks, which may be about 43% of infants during the first year of life with an
an insufficient period to evaluate the peripheral effects of improvement toward 10% of cases during adolescence,
combined treatment. Moreover, a potential carryover ef- suggesting that thyroid hormone resistance may improve
fect might have been induced in these studies related to the with age (398). Overtreatment with high L-T4 doses to
long half-life of T4 which could persist for a long time in normalize serum TSH may be contraindicated in this con-
some tissues such as the brain. Finally, other important dition because of risk of induction of deleterious effects on
limitations in establishing the effects of combined replace- growth and school performance. Combined treatment
ment therapy include the heterogeneity of the hypothyroid with L-T3 and L-T4 in a ratio of 4:1 was performed in 10
patients enrolled in the available studies (eg, athyreotic children aged over 5 years in an attempt to normalize T3
patients, patients with autoimmune Hashimotos thyroid- levels in the central nervous system and improve serum
itis, and patients with hypothyroidism induced by radio- TSH. Normalization of serum TSH levels was achieved in

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
470 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

about 7 months in children who were switched to com- in somatotrope, gonadal, or adrenal function or the co-
bination treatment, although they had higher serum T3 existence of cardiac disease.
levels and lower serum T4 levels than patients on treatment
with L-T4 alone (398). C. Interactions with other pituitary hormone deficiencies
and hormone replacement
Alexopoulou et al (405) underscored the importance of
IX. Replacement Therapy in concomitant replacement with other pituitary hormones.
Central Hypothyroidism Contemporary use of other replacement hormones in CH
may require adjustments of T4 treatment dosage. Women
A. Treatment of CH in adults under estrogen treatment and patients under GH treat-
L-T4 therapy is the treatment of choice in patients with ment often will need a higher T4 dose for serum FT4 levels
CH. Several studies have underscored the difficulty in to remain in the euthyroid range (406). With initiation of
achieving optimal substitutive L-T4 therapy in patients thyroid hormone treatment in the presence of concomitant
with CH (399 401). The obvious difference with the ACTH deficiency, there is risk of adrenal crisis due to
management of primary hypothyroidism is that serum adrenal insufficiency, and evaluation of the adrenal axis is
TSH levels cannot guide replacement therapy in CH. indicated before starting L-T4 (407). With documentation
There is no single optimum L-T4 dose for CH, and dos- of adrenal insufficiency, steroid replacement should ac-
age titration is rendered more difficult by the lack of serum company L-T4 replacement with establishment of a eu-
TSH as a guide to therapy. The aim of replacement treat- adrenal state before achieving euthyroidism (407).
ment of CH is to render the patient euthyroid by obtaining Initiation of estrogen therapy may increase L-T4 re-
appropriate serum concentrations of thyroid hormones, quirements in hypothyroid patients due to augmented ca-
and the serum FT4 concentration generally represents the pacity of T4-binding plasma proteins (407), and dosage
most useful marker for this purpose. adjustment will serve to saturate binding sites and restore
Koulouri et al (402) compared FT4 values of patients equilibrium between bound and free thyroid hormone.
with secondary hypothyroidism with those of patients Recombinant human GH (rhGH) treatment may inter-
with primary hypothyroidism who were adequately fere with the activity of the hypothalamus-pituitary-thy-
treated with L-T4. The authors suggested that CH patients roid axis and L-T4 substitutive therapy. In euthyroid in-
tend to be undertreated and that a goal FT4 level of 16 dividuals, the administration of rhGH induces a slight
pmol/L (ie, in the midnormal laboratory reference range of reduction of serum T4 concentrations, an increase in se-
9 25 pmol/L) would be adequate for treatment in most rum T3 concentrations, and a decrease in serum TSH levels
patients. Other authors have suggested that maintenance without changes in rT3 (406, 407). Moreover, rhGH treat-
of FT4 in the upper half of the normal range would be ment augments peripheral deiodination of T4 to T3 and the
appropriate in CH (212, 403). Rather than these some- secretion of somatostatin, which in turn reduces pituitary
what arbitrary FT4 target levels, treatment of CH should TSH secretion (408). A significant reduction in serum T4
be tailored. For example, as in older patients with primary levels without a substantial increase in serum TSH has
hypothyroidism, it is important to consider lower doses of been reported in about 36% of euthyroid adults with GH
L-T4 that result in FT4 levels in the lower end of the ref- deficiency, with a consequent requirement for L-T4 re-
erence range. placement therapy (409). In one study, 16% of patients
with GH deficiency who were receiving L- T4 replacement
B. Limitations in peripheral markers of thyroid therapy required an increase in L-T4 dose (409). GH de-
hormone action ficiency may mask subclinical forms of CH, and patients
The evaluation of peripheral parameters of thyroid hor- on rhGH replacement therapy often require higher re-
mone action (eg, cholesterol, creatinine phosphokinase, placement doses of L-T4 (406, 410).
soluble IL-2 receptor, SHBG, angiotensin-converting en-
zyme, osteocalcin, cross-linked carboxyl-terminal telo- D. Long-term cardiovascular risk in CH
peptide of type I collagen) could be useful in monitoring A recent retrospective study showed that insufficient
L-T4 treatment in CH (212, 404). A recent study employed replacement therapy with thyroid hormone in central hy-
Doppler echocardiography to diagnose subclinical CH in popituitary patients was associated with adverse cardio-
patients with hypothalamic-pituitary disease (404). Un- vascular risk factors independently of other pituitary re-
fortunately, evaluation of tissue parameters of thyroid placement therapy. A decrease in FT4 values during 4 years
hormone action can be frequently ineffective in CH due to were associated with an improvement in BMI, diastolic
confounding effects of concomitant pituitary deficiencies BP, total cholesterol, and LDL-cholesterol after adjust-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 471

Table 16. Management and Follow-Up of CH X. Replacement Therapy for Hypothyroidism


in Specific Conditions
Treatment should be started after the exclusion of adrenal
insufficiency A. Pregnancy
The starting dose should be based on age and co-morbidity
FT4 should be maintained in the middle-normal (in the elderly) or 1. Adverse effects of overt and SHypo, thyroid autoimmunity,
in the upper part of the normal range (in young people) with a and hypothyroxinemia in pregnancy
normal serum TSH Hypothyroidism and autoimmune thyroid disease are
FT4 and TSH levels should be reassessed after replacement
therapy with other pituitary hormones common among women in reproductive age.
Annual assessment (MRI or CT) of hypothalamic pituitary region Pregnant women with overt hypothyroidism have de-
Doppler echocardiography may be useful during the follow-up in
patients without comorbidities
creased fertility and an increased risk of spontaneous
abortion (1719). A positive linear relationship between
pregnancy loss and increased TSH values was reported by
Benhadi et al (415) with the incidence of fetal demise in-
ment for changes in GH and hydrocortisone doses. These creased by 60% for every doubling in serum TSH concen-
results support the importance of optimizing replacement tration. Women with reduced thyroid reserve may develop
therapy with thyroid hormone to reduce the cardiovascu- overt hypothyroidism during pregnancy (416, 417). The
lar risk risk in hypopituitaric patients (411). risk of progression to hypothyroidism is increased in preg-
A summary of aspects of the management and fol- nant women with positive thyroid autoantibodies during
low-up of CH is shown in Table 16. the first trimester and in women with iodine deficiency
(416, 417). Early and late obstetric complications are in-
E. Treatment of CH in pediatric patients creased in women with untreated overt hypothyroidism
Relatively higher L-T4 doses are recommended in hy- with a high risk of miscarriage, anemia, gestational hy-
pothyroid pediatric patients with CH than in adult pa- pertension, placental abruption, postpartum hemor-
tients. As is typically the practice in primary hypothyroid- rhages, and breech presentation (94, 418).
ism, the treatment should be started with full-replacement Additional adverse effects of hypothyroidism in preg-
L-T4 doses to more rapidly achieve adequate circulating nancy include the risk of adverse neonatal outcomes such
FT4 levels and potentially improve neurological develop- as premature birth, low birth weight, and neonatal respi-
ment (21, 413). L-T4 doses should be adjusted every 2 to ratory distress (9, 10).
4 weeks on the basis of results of measurement of FT4 Even SHypo or the presence of autoimmune thyroid
levels (21, 413). disease may have adverse effects on the outcome in preg-
nant women. A recent meta-analysis selected 38 appro-
F. Combination treatment with T3 and T4 in CH priate studies to evaluate the clinical significance of thy-
A double-blind crossover study on 29 patients with CH roid dysfunction and thyroid autoimmunity during
compared 3 regimens of replacement therapy with thyroid pregnancy (419). This examination revealed that SHypo
hormone: 1) empirical L-T4 dosage in patients with normal in early pregnancy may be responsible for an increased risk
FT4 values (1 0.05 g/kg body weight), 2) body weight- of preeclampsia (OR, 1.7; 95% CI, 1.12.6) and perinatal
adapted L-T4 dosage (1.6 k/g body weight), and 3) body mortality (OR, 2.7; 95% CI, 1.6 4.7) compared with that
in patients with normal thyroid function (419). Thyroid
weight-adapted combination of L-T4 and L- T3 therapy (T3
autoimmunity was associated with an increased risk of
to T4 ratio of 1:10) (414). Patients taking the L-T4 therapy
unexplained subfertility (OR, 1.5; 95% CI, 1.12.0), mis-
adapted to body weight were able to achieve a lower BMI
carriage (OR, 3.73; 95% CI, 1.8 7.6), recurrent miscar-
and lower total cholesterol and LDL-cholesterol. Moreover,
riage (OR, 2.3; 95% CI, 1.53.5), preterm birth (OR, 1.9;
patients on the body weight-adapted arm of combined T3/T4
95% CI, 1.13.5), and maternal postpartum thyroiditis
treatment obtained further beneficial effects as indicated by (OR, 11.5; 95% CI, 5.6 24) when compared with the
ankle reflex time and working memory, but at a cost of hav- absence of thyroid antibodies (419). Another meta-anal-
ing supraphysiological serum FT3 levels. These observations ysis on 31 studies (19 cohort and 12 case-control) involv-
suggested that a dose of L-T4 of 1.6 g/kg body weight was ing 12 126 women indicated that the odds of miscarriage
associated with an improved prognosis in patients with CH. were tripled (OR, 3.90; 95% CI, 2.48 6.12; P .001)
Because the 10:1 T4 to T3 ratio differs from the normal ratio and the odds of preterm birth doubled in mothers with
of 14:1, additional studies on combination T3 and T4 treat- positive thyroid autoantibodies (OR, 2.07; 95% CI, 1.17
ment with a correct ratio might better clarify the potential 3.68; P .01) (420). A meta-analysis of 4 prospective
role of combined treatment in patients with CH. studies on 1098 subfertile women undergoing in vitro fer-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
472 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

tilization reported a higher risk for miscarriage in subfer- who were screened for thyroid disease within the first 16
tile women with autoimmune thyroid disease than con- weeks of gestation by means of TSH and FT4 measure-
trols (relative risk [RR], 1.99; 95% CI, 1.42- 2.79; P ments (428). Patients were divided into 2 groups. In the
.001) (421). screening group, thyroid function was evaluated at the
Severe neurodevelopmental delay is associated with hy- time of recruitment, and women with a TSH 97.5th per-
pothyroidism (422, 423). Maternal thyroid deficiency centile and/or FT4 2.5th percentile were treated with
may have detrimental effects on fetal brain development L-T4 at an initial dosage of 150 g/d, starting at 13 weeks
because of the essential role played by maternal thyroid gestation. In the control group, patients had their blood
hormones in early pregnancy because the fetal thyroid tests at the time of recruitment, but their thyroid function
gland is not fully functional until approximately the 29th test results were not known until after delivery. Thyroid
week. In one early study by Man et al (422), children born dysfunction was found in 4.6% of patients in the screening
from mothers with inadequately treated hypothyroidism group and in 5% in the control group. There were no
had significantly reduced intelligence quotients (IQs). The differences in the patients characteristics between the 2
observations in the inadequately treated mothers in the groups (age, smoking, weight, education, or preterm
latter study may relate to those of Haddow et al (178) of birth). The standardized IQ of children in the control
women with untreated SHypo during pregnancy whose group was 100 points (404 children, 73.3% of the children
offspring at 7 9 years of age had an average IQ score that of the women who tested positive), whereas the IQ of
was 7 points below the mean IQ of matched children born children in the screening group (390 children, 78.2% of
from mothers whose serum TSH levels were within the the children of the women who tested positive) was 99.2
reference range (P .005); 19% of these children had IQ points (P .40). (428). The proportion of patients with an
scores less than 85 compared with 5% of the control group IQ 85 points was 12.1% in the screening group vs
(P .07). 14.1% in the control group (P .39). Therefore, no dif-
The incidence of isolated hypothyroxinemia in preg- ferences were found when controls were compared with
nancy ranges from 2% to 25% with an increased risk in different subgroups of patients, namely, women with el-
areas of moderate iodine deficiency. Conflicting results evated TSH only or low FT4 only or women who started
have been reported on the effects of isolated hypothyrox- treatment either before or after 14 weeks gestation (428).
inemia and SHypo on fetal neurodevelopment because In contrast to the study of Haddow et al (178), the
heterogeneous criteria have been used to define normal results of this study indicated that antenatal screening with
levels of maternal FT4 andTSH during early pregnancy treatment of the women discovered to have SHypo did not
(424 427). In a Chinese population, maternal SHypo, result in improved cognitive function in their children at 3
hypothyroxinemia (at 16 20 weeks of gestation), or eu- years of age. A possible explanation for the negative find-
thyroidism with elevated TPOAb titers were significant ings of this RCT might be that L-T4 treatment started too
predictors of lower motor and intellectual development in late in pregnancy to exert a beneficial impact on brain
the offspring evaluated at 25 to 30 months of age (426). development.
Isolated hypothyroxinemia during the second trimester A recent study reported a consistent association be-
was not associated with significantly lower cognitive, lan- tween severe maternal hypothyroxinemia in early gesta-
guage, and motor scores at age 2 years compared with tion (gestational weeks 6 18) and autistic symptoms in
scores for offspring of matched euthyroxinemic women in offspring. A 4-fold increase in the odds of having a prob-
another study (427). The results of a recent review have able autistic child (adjusted OR, 3.89; 95% CI, 1.83
suggested that the effects of maternal hypothyroxinemia 8.20; P .001) was associated in this study with hypo-
on the offsprings neurocognitive development depends thyroxinemia (429).
on the timing of this condition during pregnancy (424). A Iodine deficiency is the most common cause of maternal
lower neurocognitive function has been observed in in- hypothyroxinemia. Some studies have evaluated the po-
fancy and early childhood in the studies where maternal tential adverse effects of hypothyroxinemia during preg-
hypothyroxinemia have been investigated during early nancy in iodine-deficient areas. Irreversible fetal brain
pregnancy; on the contrary, conflicting results have been damage has been reported in the presence of severe iodine
observed when this condition has been assessed in mid- deficiency (65). A meta-analysis of 37 studies performed in
pregnancy. No association has been found during late China comparing IQs in children from iodine-sufficient
pregnancy (424). and -deficient areas reported that the IQ of children ex-
The Antenatal Thyroid Screening and Childhood Cog- posed to severe iodine deficiency was reduced by 12.45
nitive Function study (CATS) was a prospective, random- points and was recovered by 8.7 IQ points with iodine
ized, multicenter trial that involved almost 22 000 women supplementation (430).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 473

Children born from mothers with prolonged hypothy- included members of the Asia and Oceania Thyroid Associ-
roxinemia living in iodine-deficient areas may have a def- ation and the Latin American Thyroid Society. The Ameri-
icit of motor and mental development and an increased can College of Obstetrics and Gynecology published guide-
incidence of attention deficit disorder, hyperactivity dis- lines for diagnosis and treatment of hypothyroidism in 2007
order, and reduced IQ with a higher risk of expressive that were reaffirmed in 2010 (443).
language delay and nonverbal cognitive delay compared
with controls (431 436). The World Health Organization a. Treatment of overt hypothyroidism. The most recent ES
recommends that iodine intake during pregnancy should guidelines recommend treatment of overt hypothyroidism
be between 200 and 250 g/d and suggests that a median in pregnant women (recommendation level A) because
urinary iodine concentration of 150 to 249 g/L indicates there is evidence that overt maternal hypothyroidism may
adequate iodine intake in pregnant women (437). Data lead to serious adverse effects on the fetus and mother (19).
from the NHANES study reported that the median values Therefore, L-T4 treatment clearly should be considered in
for urinary iodine excretion of childbearing women were all pregnant women with a TSH concentration above the
below the median values established by these World trimester-specific reference interval with a decreased FT4
Health Organization recommendations (438). Therefore, and in all women with a TSH concentration above 10.0
a deficiency in iodine intake during pregnancy was found mU/L irrespective of the level of FT4 (18, 19).
in the United States, and similar data have been reported
in Europe (439). The 2012 Endocrine Society guidelines b. Treatment of SHypo and TPO women. The 2011 AACE-
recommend that vitamins containing 150 to 200 g iodine ATA guidelines recommended the treatment of SHypo in
(potassium iodide or iodate) should be administered be- all women positive for TPOAb (18). On the contrary, the
fore conception to ensure that all pregnant women are 2012 ES guidelines recommended L-T4 replacement in
protected from iodine deficiency during pregnancy (19). women with SHypo who are TPOAb or TPOAb (19).
Breastfeeding women should maintain a daily intake of A prospective trial demonstrated a 69% increase in the
250 g of iodine to ensure a supply of 100 g iodine per rate of spontaneous miscarriage in TPOAb women with
day to the infants. TSH between 2.5 and 5 mU/L during the first trimester of
pregnancy compared with women with TSH below 2.5
2. Beneficial effects of replacement L-T4 therapy in mI/L (444). However, there is no evidence of an improve-
hypothyroid pregnant women ment of neurological development in the offspring, and no
Adequate treatment with replacement doses of L-T4 in prospective studies have demonstrated a benefit of treat-
early pregnancy improved the outcome of pregnancy in ing TPOAb women. Consequently, the ES guidelines rec-
hypothyroid women in terms of maternal and neonatal ommendations for treatment of SHypo in TPOAb pa-
morbidity (440, 441). tients are only based on expert opinion that the benefits of
A significant 52% RR reduction in miscarriages (RR, treatment outweigh the potential risk of not treating (rec-
0.48; 95% CI, 0.25 0.92; P .03) was reported after ommendation level C; evidence fair for obstetrical out-
initiation of adequate L-T4 therapy (440). Early fetal loss come) (19). In contrast, the American College of Obstet-
and preterm delivery were shown to be significantly re- rics and Gynecology recommends against treatment of
duced in adequately treated hypothyroid women com- SHypo (445).
pared with inadequately treated (4% vs 31% and 1.6% vs The 2012 ES guidelines do not recommend treatment
12.5% respectively) (441). In a large prospective study, with L-T4 in euthyroid women with positive thyroid an-
women with SHypo treated with L-T4 had a significantly tibodies because only 1 randomized trial has demon-
lower rate of obstetrical and neonatal complications com- strated a decrease in the miscarriage rate in euthyroid an-
pared with untreated women (442). tibody-positive women during the first trimester (446). In
this prospective, randomized trial, 984 unselected women
3. Treatment of hypothyroidism in pregnancy (differences were screened for TPOAb positivity and thyroid function
among guidelines) tests during their first obstetrical appointment and were
Guidelines have been published with recommendations divided into 3 groups. Group A included TPOAb women
for optimal treatment of thyroid hormone deficiency during treated with L-T4; group B included TPOAb women who
pregnancy. The AACE and ATA guidelines on the diagnosis did not receive L-T4 and Group C consisted of all TPOAb
and treatment of thyroid disease during pregnancy and post- untreated women. The results demonstrated that the first-
partum were published in 2011 (18). The first ES guidelines trimester miscarriage rate was significantly lower in
were first published on 2007 (17) and were revised by the groups A (3.5%) and C (2.4%) than in group B (13.8%)
inclusion of 27 new articles in 2012 (19). The ES task force (P .05) (446). Given the increased risk of progression to

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
474 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

hypothyroidism, the 2012 ES guidelines recommended out residual functional thyroid tissue (eg, after radioiodine
that women with TPOAb antibodies should be moni- ablation or total thyroidectomy) than in those with resid-
tored only every 4 to 6 weeks for the potential risk of serum ual thyroid tissue (eg, Hashimotos thyroiditis) (214).
TSH rising above the normal range during pregnancy as an About 30% to 50% of hypothyroid women being
indication of progression into hypothyroidism (19). An treated during pregnancy have a TSH concentration out-
ongoing study from the United Kingdom, the TABLET side of the desired reference range (449). In a prospective
(Thyroid Antibodies and L-Thyroxine) study, a multi- study on 119 consective pregnancies, about 49% of
center placebo double-blind controlled trial, hopes to clar-
women on L-T4 replacement therapy had serum TSH out-
ify the impact of L-T4 therapy in euthyroid women with
side the reference range (30% had elevated TSH concen-
positive thyroid autoantibodies planning pregnancy.
trations, and 20% had low TSH concentrations) during
There has been increased interest in the role of selenium
the first trimester of pregnancy (449). The risk of fetal loss
deficiency in thyroid disease. Treatment with selenome-
thionine decreased TPOAb levels in euthyroid women was significantly increased in these women compared with
with recurrent pregnancy loss and reduced the risk of post- that in euthyroid pregnant women (449). Ideally, women
partum thyroiditis (447, 448). However, large random- with hypothyroidism should have their thyroid function
ized studies are necessary to evaluate the potential benefits evaluated before pregnancy to avoid the risks that are as-
of selenium supplementation in pregnancy. At present, sociated with maternal subclinical and overt hypothyroid-
given the potential risk of developing diabetes, the guide- ism during pregnancy. Lower preconception TSH values
lines suggest that selenium supplementation should not be (within the nonpregnant reference range) reduce the risk
recommended for TPOAb women during pregnancy. of TSH elevation during the first trimester. In the study by
Rotondi et al (450), preconception TSH levels in the lower
c. Treatment of isolated hypothyroxinemia. The AACE-ATA quartile of the reference range during L-T4 therapy were
guidelines do not recommend the treatment of isolated associated with an adequate TSH level at the first evalu-
hypothyroxinemia (18). ation during pregnancy. Similar findings were reported by
Although there is no formal recommendation in the
Abalovich et al (451) and indicated that only 17.2% of
2012 ES guidelines, the task force suggests that in patients
women with preconception TSH below 1.2 mU/L needed
with isolated hypothyroxinemia, a partial replacement
an increment in L-T4 dosage during pregnancy, whereas
therapy may be initiated at the discretion of the caregiver,
women with preconception TSH levels between 1.2 and
with continued monitoring (19). Therefore, one may in-
fer from the ES guidelines that initiation of treatment of 2.4 mU/L required a 50% increase in L-T4 dosage.
this condition falls under the judgment of the clinician There is evidence that high serum FT4 levels during L-T4
until new data might clarify the consequences of isolated therapy should be avoided. In one study, high-normal FT4
hypothyroxinemia (19). concentrations in early pregnancy (between 17.01 and
22.00 pmol/L) were associated with reduced fetal growth,
d. Recommendations on L-T4 treatment and follow-up in preg- resulting in 116 g lower birth weight as well as with an
nancy. The ES, AACE, ATA, and ETA guidelines recom- increased risk of small-for-gestational-age newborns
mend that oral L-T4 should be used for treatment of ma- (452). Similar results were reported by Shields et al (453)
ternal hypothyroidism (1719). All guidelines recommend who found a significant positive associations between FT4
against the use of other thyroid preparations such as T3 or in the cord blood at birth and birth weight.
desiccated thyroid (3, 1719, 32). Rapid correction of hy- Thyroid function tests should be repeated approxi-
pothyroidism is advised as soon as diagnosed in preg-
mately 40 days after the first adjustment of L-T4 dosage.
nancy. The goal of L-T4 treatment is to normalize maternal
Thereafter, maternal serum TSH should be monitored ev-
serum TSH values to within the trimester-specific preg-
ery 4 to 6 weeks because further L-T4 dose adjustments are
nancy reference range (serum TSH 2.5 mU/L during the
often required (1719). In the postpartum period, L-T4
first trimester and 3.0 mU/L during the second and third
trimester). Women who are already being treated with should be reduced to the patients preconception dose (17
L-T4 should increase their dosage to 30% to 50% above
19), with TSH testing performed at approximately 6
the preconception dosage by 4 to 6 weeks of gestation weeks postpartum. According to the 2012 ES guidelines,
(1719). The etiology of maternal hypothyroidism and the women at high risk for postpartum thyroiditis (eg, women
degree of elevation of the preconception level of TSH may with TPO, type 1 diabetes and/or a history of postpartum
guide clinicians on the magnitude of the increase in L-T4 thyroiditis) should be screened with serum TSH at 6 to 12
dosage. The increment should be greater in women with- weeks postpartum (19).

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 475

4. Screening for thyroid dysfunction in pregnancy evaluate the impact of prepregnant treatment of thyroid
Different guidelines from the various professional so- dysfunction on fetal brain development.
cieties for clinical endocrinologists, thyroidologists, and
gynecologists have assessed the necessity for universal B. SHypo in the elderly
screening of healthy women for thyroid dysfunction. The The diagnosis of SHypo in the elderly often may be
American College of Obstetrics and Gynecologists does more difficult than in younger patients. An increased se-
not recommend routine screening of thyroid function in rum TSH value in older individuals may frequently reflect
pregnancy (443). The 2007 ES guidelines recommended the response to recovery from systemic illness or the pre-
case finding targeted to specific groups of patients who are vious use of drugs that had interfered with thyroid func-
at an increased risk of developing hypothyroidism such as tion. Moreover, in healthy individuals, serum TSH con-
women with 1) a personal or family history of thyroid centrations are higher in the elderly than in younger people
disease, 2) goiter, 3) positive thyroid antibodies, 4) symp- because of a shift in the TSH distribution with age. Re-
toms or clinical signs suggesting thyroid dysfunction, 5) analysis of NHANES III data indicates that the 97.5th
history of autoimmune disease or type 1diabetes mellitus, percentile of the TSH upper limit is approximately 5.9
6) infertility, 7) history of miscarriage or preterm delivery, mU/L in subjects aged 70 to 79 years and 7.5 mU/L in those
and 8) a history of head and neck irradiation (17). Similar 80 years and older (49).
recommendations appear in the AACE-ATA guidelines, Elderly people with SHypo seem to be less symptomatic
than younger patients. In a large cross-sectional commu-
which also included an age of 30 years or greater as an
nity-based study carried out in primary care practices in
additional criterion (18).
England, SHypo (defined as a TSH 5.5 mU/L) was not
However, one report (454) indicated that screening
associated with cognitive dysfunction, anxiety, or depres-
only high-risk women will miss about one-third of patients
sion in subjects 65 years of age, supporting the concept
who remain undiagnosed with overt hypothyroidism or
that a mild TSH increase need not be treated in the elderly
SHypo. Another study supported this contention as
in an attempt to improve the QOL (321). Four randomized
screening of women at low risk identified sufficient cases
placebo-controlled studies have reported marginal differ-
of hypothyroidism to subsequently improve rates of ob-
ences in cognitive function in elderly patients treated with
stetrical and fetal complications (442).
L-T4 compared with controls (456 459).
Screening pregnant women in the first trimester of preg-
The largest RCT of L-T4 replacement in an elderly co-
nancy proved cost-effective for detecting and treating
hort was performed by Parle (460) on 94 patients (mean
overt hypothyroidism in the study by Dosiou et al (455) by
age 73.8 years) with SHypo who were treated with L-T4 for
employing testing for either TSH or antiperoxidase anti-
53 weeks. No significant improvement in cognitive func-
body at the initial evaluation. The 2012 ES guidelines de- tion tests was noted in the L-T4treated subjects compared
scribed conflicting opinions on the screening of hypothy- with patients receiving placebo. In another community-
roidism in women among the panel members Some based study of people aged 70 to 79 years, SHypo was
members recommended screening of all pregnant women associated with increased walking speed and retention of
for serum TSH abnormalities by the ninth week or at the physical function across a 2-year follow up period com-
time of their first visit (19). Others recommended neither pared with subjects with normal thyroid function (461). In
for nor against universal screening of all pregnant women this study, a TSH level of 7 mU/L was the cutoff point
at the time of their first visit or supported aggressive case separating adults 70 years into 2 groups: those with mild
findings to identify and test high-risk women by the ninth SHypo (TSH level, 4.57.0 mU/L) and those with mod-
week or at the time of their first visit before and during erate SHypo (TSH level, 720 mU/L). No negative influ-
pregnancy (19). Clarification of the more optimal ap- ence of SHypo was found after using functional mobility
proach to screening of pregnant women may have to await as a marker of health status in this elderly population
the results of some currently ongoing and future trials. (461).
One such trial is underway by the National Institutes of A diagnosis of SHypo is of some concern because of the
Health Maternal Fetal Medicine TSH study on 120 000 risk of progression to overt hypothyroidism, and this con-
pregnant women and is assessing the outcomes of ran- cern pertains in the elderly as well as in younger popula-
domizing treatment or nontreatment of women with tions and has been examined in a limited number of stud-
SHypo and isolated hypothyroxinemia. Another trial in ies. In the United States Cardiovascular Health Study on
China, the Subclinical Hypothyroid and Iodine Deficiency 3996 individuals aged at least 65 years, a serum TSH of 10
in Early Pregnancy and Women Planning for Pregnancy mU/L or higher and TPOAb positivity were independently
(SHEP) trial, will screen 21 500 and treat 4800 women to associated with a significant risk of progression to overt

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
476 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

hypothyroidism (462). Persistence of SHypo without pro- although TSH may increase over time in older individuals,
gression was observed in 56% of cases, whereas reversion these changes are not correlated with increased or de-
to euthyroidism was significantly more common in pa- creased mortality. On the contrary, higher FT4 levels are
tients with a TSH of 4.5 to 6.9 mU/L and with TPOAb associated with AF and higher death rates in the elderly
negativity (462). (341, 463).
Of note in this regard are the results of 2 recent longi- Treatment of elderly subjects with mild thyroid hor-
tudinal studies that assessed the changes in TSH and thy- mone deficiency was not associated with benefit in the
roid hormone levels with aging and concluded that a mild study by Razvi et al (193). Incident ischemic heart disease
progressive increase in TSH may be a marker of progres- events, all-cause mortality, cause-specific mortality, and
sion toward thyroid hormone deficiency only in younger incident cerebrovascular events were comparable in older
patients (463, 464). In fact, a progressive increase in FT4 untreated and L-T4treated groups (193).
levels and a decrease in FT3 levels were associated with These findings raise concern in regard to possible over-
increases in serum TSH levels during the years of fol- treatment of mild TSH elevations in subjects of advanced
low-up in elderly subjects. These data have been inter- age. Replacement therapy should be individualized in el-
preted to suggest that higher FT4 levels might be necessary derly and very elderly patients with serum TSH concen-
in the elderly to achieve comparable TSH values as in trations of more than 10 mU/L (9, 10). Only low doses of
younger subjects. The mild progressive TSH increase in L-T4 (2550 g/d) are often required to normalize serum
elderly subjects might simply reflect the natural history of TSH concentrations in elderly patients because of their
thyroid function in advanced age, which may relate to a decreased T4 metabolism. The target or goal TSH serum
decline in 5-deiodinase activity with aging (59). These concentration should be higher in individuals older than
data have been cited to support recommendations for an 70 years than in younger patients to mimic physiological
increase in the upper limit of the reference range in the values (eg, 4 7 mU/L). Although large randomized trials
elderly and the exercise of caution in diagnosing SHypo in still are needed, existing evidence suggests that treatment
elderly subjects. of mild SHypo should probably be avoided in patients
In regard to the risks of not treating SHypo in the el- older than 60 years of age because there is no definitive
derly, 2 meta-analyses reported that an association be- evidence that these patients are symptomatic or that L-T4
tween SHypo and CHD existed only in patients younger treatment will improve their QOL and/or reduce their car-
than 60 years (182, 183), and subgroup analysis in a third diovascular mortality. As in younger patients, overtreat-
recent meta-analysis (184) revealed no evidence of greater ment with L-T4 should be avoided because of the adverse
risks of CHD events, CHD mortality, and total mortality cardiovascular and skeletal consequences of iatrogenic hy-
among pooled participants over 80 years of age. HF may perthyroidism in elderly people.
develop in healthy elderly subjects with SHypo and a TSH
level above 10 mU/L (188). On the other hand, modestly C. Congenital hypothyroidism
increased serum TSH levels have been associated with lon- Treatment with replacement doses of L-T4 should be
gevity in several cross-sectional studies in elderly patients promptly started when the diagnosis of CoH is confirmed
vs younger controls and in nonagenarians with reported by means of an initial T4 level 10th percentile and a TSH
familial longevity (49). A pattern of decreased mortality in level 9 mU/L.
SHypo was observed in the Leiden 85-Plus Study of 558
individuals aged 85 years who had been monitored for 4 1. Time of TSH normalization
years (101). There was no link between persistent SHypo The goal of treatment in patients with CoH is to obtain
and cardiovascular mortality (HR, 1.07; 95% CI, 0.87 prompt control of thyroid function after having made an
1.31) in 679 patients with SHypo of at least 65 years of early diagnosis. L-T4 is the treatment of choice. The timing
age, enrolled in the recent Cardiovascular Health Study of L- therapy is important because an inverse relationship
and not taking thyroid preparations (465). Moreover, no has been reported between IQ and the age at diagnosis of
association was found between SHypo, TSH level, or per- CoH (466). Infants with CoH treated in the first few weeks
sistent TPOAbs and death in elderly subjects in the latter of life usually have a normal or slightly reduced IQ,
study (465). Conversely, FT4 levels were positively asso- whereas mean IQ was reduced to 89 when the diagnosis of
ciated with death (HR per ng/dL, 2.57; 95% CI, 1.32 CoH was delayed to 3 months of age, and it further de-
5.02) (465). clined to 71 with diagnosis between 3 and 6 months and
The data of this important large epidemiological study, to 34 after 6 months of age (467). A serum FT4 level 129
which measured TSH during the follow-up and included nmol /L (below 10 g/dL) should be reached as rapidly as
only subjects with persistent TSH elevations, suggest that possible, usually within 1 or 2 weeks after starting L-T4

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 477

therapy to obtain a normal TSH after 1 month of treat- with target values of 130 to 206 nmol/L (10 16 g/dL) for
ment. Selva et al (468) reported that IQ was 10 points serum total T4 and 18 to 30 pmol/L (1.4 2.3 ng/dL) for
lower in children in whom the serum TSH and FT4 levels serum FT4. Serum TSH should be kept below 5 mU/L (20,
were normalized over 2 weeks after starting L-T4 21). The New England Collaborative Study reported that
treatment. children with persistent serum T4 below 10 g/dL in the
first year of life had an IQ that was 18 points lower com-
2. L-T4 formulations for CoH pared with patients with T4 above 10 g/dL (474). Some
L-T4 tablets are approved for treatment of CoH in the
studies have reported an increased risk of craniosynosto-
United States; administration is facilitated by crushing the sis, hyperactivity, delinquency, aggressiveness, and poor
tablets, mixing with water, and giving with a spoon. In attention in children with CoH who were overtreated with
Europe, a liquid L-T4 formulation is available with admin- L-T4 (475).
istration of the desired dosage by a dropper. In one study,
a liquid formulation of L-T4 was administered to 28 con- 4. Follow-up in patients with CoH
secutive newborns with primary CoH (469) in a median The AAP suggests that the measurement of serum T4 or
starting dose of 12.3 g L-T4/kg/d that was reduced to FT4 and TSH should be performed 2 or 4 weeks after
about 5 g L-T4/kg/d after 9 months. The median time of initial treatment with L-T4 and then every 1 to 2 months
normalization of TSH to 6 mU/L was 2 weeks. TSH during the first 6 months of life, every 3 to 4 months be-
levels normalized within a median of 1 week in 21 patients tween 6 months and 3 years of age and subsequently every
who received a median starting dosage of 12.7 g L-T4/kg 6 to 12 months until growth is completed (21). Routine
(range 9.8 17.1 g/kg), and after a median of 2 months in thyroid function tests should be repeated 4 weeks after a
7 patients receiving only 10.1 g /kg (469). These results change in L-T4 dosage (21). The European Society of Pe-
demonstrated comparable efficacy of the liquid formula- diatric Endocrinology recommends monitoring T4 or FT4
tion to that of L-T4 tablets but in a formulation that is every 1 or 2 weeks after the starting dose of L-T4 until TSH
easier to administer. and FT4 are completely normalized (20). The AAP rec-
However, a recent study by Cassio et al (470) of 42
ommends performing a 30-day withdrawal of L-T4 treat-
consecutive infants with CH demonstrated that drops and
ment to assess whether the diagnosis of CoH will be per-
tablets are not bioequivalent, especially in infants with
manent at the age of 2 or 3 years in children whose initial
severe hypothyroidism .
thyroid scan did not show an ectopic or absent thyroid
3. L-T4 dosage in CoH
gland, with a serum TSH 50 mU/L at screening and who
The American Academy of Pediatrics (AAP) (21) and had no increase in TSH after the newborn period (21).
the European Society for Pediatric Endocrinology (20) Treatment with L-T4 should be restarted if an elevated
recommend an L-T4 starting dose of 10 15 g/kg by TSH develops after L-T4 withdrawal.
mouth once daily (about 37.550 g/d). Conversely, if the TSH remains normal after L-T4 with-
Higher performance scores for behavior, reading, spell- drawal, the diagnosis of CoH can be considered to have
ing, and math were reported in newborns who received 50 been only transient. One study suggested the use of rhTSH
g/d L-T4 (1217 g/kg/d) vs 37.5 g (10 15 g/kg/d) to perform thyroid scintigraphy to confirm the diagnosis
(471). In one study, 83 infants were assigned to receive 3 of CoH so as to avoid the adverse effects of L-T4 with-
different starting doses of thyroid hormone at birth (6.0 drawal (476). Another means of assessing whether the
8.0 vs 8.110.0 and 10.115.0 g/kg/d) (472). Evaluation hypothyroidism will be permanent is to reduce the L-T4
at 4 years of age indicated that infants with severe CoH dosage by 50% for 30 days; if the serum TSH remains
who had started L-T4 treatment at the highest dose had the normal after this reduction, L-T4 should be discontinued
highest intellectual assessment scores (472). Some studies tai- and thyroid function tests should be reevaluated subse-
lored the starting L-T4 dose according to the severity of hy- quently to confirm maintenance of euthyroidism. Rab-
pothyroidism. An L-T4 dose of 15 g/kg/d was started in biosi et al (477) reported that major risk factors for per-
infants with athyreosis, whereas 12 g/kg/d was adminis- manent CH were prematurity, first-degree familial history
tered to infants with an ectopic gland and 10 g/kg/d in of goiter/nodules, thyroid hypoplasia at diagnosis, and
infants with dyshormonogenesis. The outcome of this study high L-T4 requirements at follow-up.
indicated that FT4 was normalized within 7 days in 78% of The management and follow-up of CoH are summa-
the infants and in 100% after 14 days of treatment (473). rized in Table 17. No specific guidelines have been devel-
The serum FT4 or total T4 should be maintained in the oped for monitoring genetic conditions associated with
upper limit of the normal range during the first year of life hypothyroidism.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
478 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

Table 17. Management and Follow-Up of CoH were multicenter studies. In this analysis, children with
SHypo had a lower risk of progression toward overt dis-
Treatment should be started at L-T4 dose of 10 15 g/kg/d ease compared with the adult population, with the rate of
Target ranges for TT4, FT4, and TSH values during the first year
TT4 130 206 nmol/L (10 16 g/dL) progression ranging between 0% and 28.8% (484). How-
FT4 18 30 pmol/L (1.4 2.3 ng/dL). ever, children with initial TSH levels greater than 7.5
Serum TSH should be kept under 5 mU/L
Follow-up should be performed by TSH, TT4, or FT4 evaluation
mU/L, particularly girls, were at a greater risk of persistent
At 2 and 4 wk after starting L-T4 treatment hypothyroidism (484).
Every 12 mo during the first 6 mo of life Moreover, a progressive increase in TPOAbs and
Every 3 4 mo between 6 mo and three y of age
Every 6 12 mo thereafter until growth is complete plasma TSH values were predictive factors for the devel-
4 wk after any change in L-T4 dose opment of persistent hypothyroidism (484). The presence
Abbreviation: TT4, total T4. of autoimmune thyroiditis, celiac disease, elevated TSH,
and TPOAbs increased the risk of developing hypothy-
D. Treatment of SHypo in children and adolescents roidism by 4-, 3.4-, and 3.5-fold, respectively, in a retro-
L-T4 is the treatment of choice in children and adoles- spective study of 87 children with autoimmune thyroiditis
cents. The L-T4 dosage requirement in CoH progressively who were compared with 59 children with isolated hy-
declines from 10 to 15 /kg/d in infants to 4 to 5 g/kg/d perthyrotropinemia (485). On the contrary, no predictive
by the age of 5 years due to a progressive decrease in the factors for progression were identified in patients with
rate of T4 turnover (20, 21) . isolated hyperthyrotropinemia after a 3-year follow-up
The most common cause of acquired SHypo in children (486).
is autoimmune Hashimotos thyroiditis. The risk of Few studies assessed the effects of replacement therapy
SHypo is particularly common in children with Downs in children with SHypo. Two longitudinal studies evalu-
syndrome and Turner syndrome (478, 479). Persistent ated the effects of replacement doses of L-T4 on growth
SHypo may develop in children who are classified as false- velocity (487, 488). Cetinkaya et al (487) studied 24 pre-
positive at neonatal screening for the presence of short- pubertal and 15 pubertal Turkish children with short stat-
term neonatal hyperthyrotropinemia. It has been sug- ure and SHypo. They observed a significant improvement
gested that the mild persistent TSH increase in these in growth velocity after 6 months and 1 year of treatment
children may reflect congenital anatomic abnormalities with L-T4; this improvement was more significant in the
such as hypoplasia and hemiagenesis and thyroid function pubertal group (487). On the contrary, no effects of re-
alterations (480). placement therapy with L-T4 were reported on height and
Some SHypo patients without features of autoimmune BMI in a longitudinal study of 69 patients with a mild
thyroiditis may have increased TSH due to thyroid resis- increase in TSH (510 mU/L) treated for 2 years; however,
tance to the effect of TSH. Defined mutations in the TSH growth velocity was not reported in this study (488). Sim-
receptor gene have been reported in studies evaluating the ilar results were noted in a retrospective study in which no
etiology of SHypo in pediatric populations (481, 482). significant difference in growth velocity was reported be-
Full replacement doses of L-T4 should be administered tween treated and untreated SHypo children compared
when the diagnosis of hypothyroidism is formulated in with a control group of short children with normal TSH
children and adolescents. Transient psychosis and intra- (489). A significant improvement in growth velocity after
cranial hypertension have been reported during the early L-T4 treatment was reported in prepubertal hypothyroid
phase of L-T4 therapy as well as in subjects undergoing children with diabetes (TSH 50 mU/L) compared with
progressive increase of L-T4 doses (483). In prepubertal less severe hypothyroid patients (TSH level 10.150
children, bone maturation and pubertal development mU/L) and to matched diabetic controls, with a more sig-
should be monitored. Currently, no recommendations or nificant effect in children with higher TSH levels (490).
guidelines have been formulated on the management of The NHANES III from the United States analyzed the
SHypo in the pediatric population. cognitive function in 1327 adolescents aged between 13
Few studies have assessed the risk of progression and and 16 years, of whom 1.7% had SHypo (491). The results
the potential adverse effects of untreated SHypo in chil- of this study indicated no differences in neuropsycholog-
dren and adolescents. Only 27 relevant articles (with over- ical test scores; SHypo was associated with better perfor-
all data from a total of 4018 children) were identified in a mance in some areas of cognitive function largely related
recent review article that assessed the natural history of to higher mean reading and block design scores in the
SHypo and the potential effects of replacement therapy SHypo children compared with the scores of 1275 euthy-
with L-T4 in children and adolescents (484). Among these roid subjects (491). Similarly, no effects of L-T4 on neu-
selected studies, only 6 were longitudinal trials and only 4 ropsychological functions were reported in children with

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 479

SHypo in the study by Aijaz et al (492). Conflicting results (CKD) is characterized by a low-T3 syndrome that has
have been reported in 2 studies that assessed the effect on been correlated with higher levels of markers of inflam-
L-T4 on thyroid volume in children (493, 494). A signifi- mation (highly sensitive C-reactive protein, IL-6, etc,),
cant reduction in thyroid volume was found with auto- malnutrition (lower prealbumin and IGF-1), endothelial
immune SHypo in the study by Svensson et al (493); this dysfunction, impaired cardiac function, and poor survival
was the only study that assessed thyroid volume by (497, 498). CKD patients also have an increased incidence
ultrasonograpy. of both primary overt hypothyroidism and SHypo.
No large prospective trials have evaluated cardiovas- In NHANES III, the prevalence of hypothyroidism was
cular risk of SHypo in a pediatric population. Elevated reported to be between 10.9% and 23.1% in patients with
TSH values in obese adolescents with nonalcoholic fatty stage 2 to 5 CKD (499). Moreover, the prevalence of pri-
liver disease were positively correlated with important mary hypothyroidism increased with progressively lower
metabolic and cardiovascular risk parameters such as total levels of kidney function; more than 20% of those with an
cholesterol, triglycerides, LDL-cholesterol, insulin, ca- estimated glomerular filtration rate (eGFR) of 60 mL/
rotid intima-media thickness, and left ventricular mass min/1.73 m2 had clinical or primary SHypo after control-
(495). These findings could suggest that obese adolescents ling for age, gender, and race/ethnicity (499). The preva-
with nonalcoholic fatty liver disease and increased serum lence of SHypo was reported to be about 18% in patients
TSH may have an increased cardiovascular risk (495). with CKD not requiring chronic dialysis and was inde-
However, caution should be exercised in making a diag- pendently associated with progressively lower eGFR
nosis of SHypo in children with high TSH levels. This is (500). Patients with eGFR of 60 ml/min/1.73 m2 had an
because the overweight state and obesity are not infre- increased OR of primary SHypo after adjusting for age,
quently associated with increased serum TSH in children; gender, fasting plasma glucose, total cholesterol, and tri-
however, TSH normalization generally occurs after glyceride concentrations (500).
weight loss. Although treatment of adults with SHypo Furthermore, a TSH level within the normal range was
negatively associated with eGFR (501). In a cross-sec-
often has been discouraged on the basis of risk of over-
tional, population-based study of 29 480 adults without
treatment, oversuppression of TSH has been noted only
previously known thyroid disease, the prevalence of CKD
infrequently in a pediatric population (484).
was higher in people with TSH in the middle and highest
In summary, the literature suggests that treatment of
thirds of the reference range and CKD was more common
SHypo may be useful in children with TSH 10 mU/L and
in those with SHypo (501). Making a diagnosis of hypo-
clinical signs or symptoms of thyroid hormone deficiency
thyroidism may be difficult in patients with CKD because
and when SHypo is associated with short stature and im-
several of the signs and symptoms of thyroid hormone
paired growth velocity or increased thyroid volume. On
deficiency may be attributed to uremia. On the other hand,
the other hand, replacement therapy with L-T4 is more
overt hypothyroidism may worsen renal hemodynamics
difficult to justify when serum TSH is 10 mU/L and in the
by decreasing cardiac output, leading to a progressive de-
absence of goiter and/or positive antithyroid antibodies,
cline in GFR. Therefore, the normalization of thyroid
because of the low risk to progression to overt hypothy-
function with L-T4 replacement therapy in hypothyroid
roidism in these children. Randomized controlled pro-
patients may improve kidney function (502507).
spective studies will be necessary to clarify the necessity of In a recent study that evaluated the effects of L-T4 ther-
treating milder degrees of SHypo in the pediatric age apy on renal function in 309 patients with SHypo and
group. stage 2 to 4 CKD (507), the overall rate of decline in eGFR
was significantly greater in SHypo patients who did not
E. Replacement therapy with L-T4 in hypothyroid
receive L-T4 treatment compared with the L-T4 treatment
patients with comorbidities
group (507). Kaplan-Meier analysis indicated that renal
1. Acute and chronic kidney disease event-free survival was significantly lower in the nontreat-
Thyroid hormones affect renal development and phys- ment group (P .01). In a multivariate Cox regression
iology, with significant changes in renal function, electro- analysis, thyroid hormone replacement therapy was found
lytes, and water homeostasis noted in patients with thy- to be an independent predictor of renal outcome (HR,
roid dysfunction (496 498). 0.28; 95% CI, 0.12 0.68; P .01) (507).
Hypothyroidism induces a decrease in glomerular fil- SHypo has been identified as a strong predictor of all-
tration, hyponatremia, and reduced facilitation of water cause mortality in chronic dialysis patients. In the study by
excretion (496). Renal disease, in turn, leads to significant Rhee et al (508), patients who were euthyroid during ex-
changes in thyroid function. Chronic kidney disease ogenous thyroid replacement were not at higher mortality

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
480 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

risk than spontaneous euthyroid controls, suggesting that (519). Given this high frequency, screening for thyroid
L-T4 replacement therapy could improve the prognosis of dysfunction is recommended in children and adolescents
CKD patients on chronic dialysis. However, an issue of with T1DM (520). Consistent with this recommendation
safety was raised by one study that evaluated the effects of are current guidelines that suggest that baseline thyroid
L-T4 in patients with acute renal failure and the euthyroid function tests should be measured in newly diagnosed pa-
sick syndrome. In this prospective, randomized, placebo- tients with T1DM (518). The British Thyroid Association
controlled, double-blind trial, mortality was increased suggests TPOAb testing at baseline and TSH monitoring
3.3-fold with 300 g/70 kg/d of iv L-T4 for 2 days in pa- at yearly intervals in patients with T1DM (518). On the
tients with acute renal failure (509). The increased mor- contrary, thyroid function tests are recommended in
tality was correlated with suppression of TSH, suggesting T2DM only when there is a suspicion of an autoimmune
that euthyroid sick patients should not be replaced with thyroid disease (518).
thyroid hormone (509). The coexistence of thyroid hormone deficiency may
Hypothyroidism has also been described as a conse- worsen the long-term morbidity in diabetic patients (521
quence, rather than the cause, of renal dysfunction. T4 cir- 523). SHypo may increase the risk of retinopathy and ne-
culates in the bloodstream largely protein bound to the 3 phropathy in patients with diabetes (521523). In a cross-
major T4 binding proteins, and the T4-protein complex can sectional analysis on 588 patients with T2DM, SHypo was
be lost in the urine in the nephrotic syndrome. Thus, an undi- associated with a greater prevalence of diabetic nephropathy
agnosed nephrotic syndrome should be considered in some in- (OR, 3.15; 95% CI, 1.48 6.69) compared with euthyroid
stances of persistent hypothyroidism. Similarly, in patients with persons with diabetes (521). Moreover, the risk of cardio-
known nephritic syndrome, it is essential to periodically mon- vascular events was significantly increased in T2DM persons
itor thyroid function and make any indicated adjustments in with SHypo after adjustment for age, sex, cardiovascular risk
L-T4 dosage as may be warranted by serum TSH levels. factors, and medication (521). There are conflicting results
In conclusion, the clinical experience reported in the on cardiovascular mortality in patients with diabetes and
literature suggests that thyroid hormone deficiency should thyroid hormone deficiency (181, 524, 525).
be treated in patients with CKD and may also improve the A significant improvement of insulin sensitivity has
kidney function. Because it may be more difficult to make been reported after replacement therapy with L-T4 (526).
a clinical diagnosis of hypothyroidism in patients with On the basis of the latter studies, there may be a potential
CKD, it may be worthwhile to consider screening for thy- benefit in treating thyroid hormone deficiency in insulin-
roid dysfunction in these patients. resistant diabetic patients, but prospective studies will be
required to support this assumption.
2. Diabetes
Thyroid hormone has important effects on glucose ho- 3. Heart disease
meostasis; it influences insulin levels and counterregula- Hypothyroidism is a cause of reversible HF (185188).
tory hormones, intestinal glucose absorption, hepatic glu- Moreover, hypothyroidism may frequently coexist in pa-
cose production, and fat and muscle uptake of glucose tients with chronic HF, with a prevalence of about 18% in
(510, 511). Insulin resistance has been reported in patients all patients with HF and a higher prevalence (23%) among
with overt hypothyroidism and SHypo and when present men and Hispanic subjects (527). The onset of hypothy-
may further increase cardiovascular risk in patients with roidism may exacerbate progression of HF in cardiac pa-
thyroid hormone deficiency, especially when associated tients. A recent study assessed the prognostic role of thy-
with hyperlipidemia and hypertension (510 512). On the roid function deficiency in 338 consecutive outpatients
other hand, the prevalence of overt hypothyroidism and with chronic HF (528). The results of this study indicated
SHypo is increased in type 1 diabetes mellitus (T1DM), in that TSH, considered as a continuous variable, was sig-
T2DM patients with autoantibodies against glutamic acid nificantly associated with HF events in multivariate anal-
decarboxilase 65, and in patients with metabolic syn- ysis. Even mild or SHypo was independently associated
drome (513517). with a greater likelihood of HF progression in patients
A meta-analysis of 10 920 patients with diabetes indi- with chronic HF (528).
cated a prevalence of thyroid disease of 11% with an in- These results may explain why the American College of
creased prevalence in women and no significant difference Cardiology guidelines for HF published in 2010 recom-
in frequency between T1DM and T2DM (518). A study of mended screening with serum TSH of all newly diagnosed
pediatric patients with T1DM found a rising prevalence of cases of HF (529).
thyroid antibodies with age, from 3.5% in patients less Hypothyroidism is associated with an increased risk of
than 5 years old up to 25% between 15 and 20 years of age mortality in patients with acute and chronic cardiac dis-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 481

ease. Iervasi et al (530) assessed the cardiac outcome in Cardiac function is depressed with low cardiac output,
3121 patients with chronic cardiac disease and subclinical reduced inotropism and chronotropism, increased SVR,
thyroid dysfunction during a mean follow-up of 32 vasoconstriction, increased diastolic BP, reduced blood
months. After adjustment for several risk factors, the HRs volume, and reduced glomerular filtration with water re-
for cardiac and overall deaths were higher in patients with tention and nonpitting edema. Hyponatremia is often
SHypo and low serum T3 than in euthyroid patients (530). present as a result of decreased free water clearance and is
The same authors evaluated the long-term prognostic usually associated with a low serum osmolality and high
role of subclinical thyroid disease in 1026 patients with levels of creatinine and antidiuretic hormone.
acute cardiac disease. Similar to the results of the previous The mortality rate of MC has been reported to be be-
study, the HR for cardiac death was higher in patients with tween 30% and 60%. Factors associated with a poor
SHypo and the low-T3 syndrome, after adjustment for prognosis include advanced age, hypotension, and brady-
several risk factors (531). Recently, the Sudden Cardiac cardia at presentation, persistent hypothermia not respon-
Death in Heart Failure trial showed that patients with sive to treatment, sepsis, and intake of sedatives (537).
moderately symptomatic HF and abnormal thyroid func- Sequential Organ Failure Assessment (SOFA) scores
tion, either at baseline or during the 5-year follow-up, 6 and a high Acute Physiology and Chronic Health Eval-
have had a 60% increase in the RR of death compared uation (APACHE) II score may help clinicians identify
with those with normal thyroid function (532). patients at risk of mortality at an earlier stage (536, 537),
In summary, the cited results suggest that the onset of and such patients should be hospitalized in an intensive
thyroid hormone deficiency may worsen the prognosis of care unit. Hypothermia, hypovolemia, and electrolyte ab-
cardiac disease and that replacement therapy with L-T4 normalities should be corrected. Mechanical ventilation
could be justified. However, adverse effects may accom- may be necessary (538), and cardiovascular function
pany the long-term use of thyroid hormone, and prospec- should be monitored, especially after iv administration of
tive studies will be necessary to clarify the most appropri- thyroid hormone replacement therapy. The possibility of
ate therapeutic approach to improve the cardiovascular underlying infectious diseases should be investigated by
mortality in patients with thyroid hormone deficiency. means of blood and urine cultures and chest radiograph.
Empiric treatment with broad-spectrum iv antibiotics can
4. Treatment of MC be considered. Most importantly, steroid supplementa-
MC is a rare and under-recognized condition that usu- tion and thyroid hormone replacement therapy should be
ally occurs in elderly patients with longstanding hypothy- promptly started. Hydrocortisone should be administered
roidism (40, 533, 534). About 300 cases of MC have been iv at a dosage of 50 to 100 mg every 6 to 8 hours to avoid
reported in the literature (535). Common precipitating the possible precipitation of an adrenal crisis in patients
factors of MC are hypothermia, infections (particularly with central or autoimmune hypothyroidism who may
influenza, pneumonia, and urosepsis), trauma, and certain have coincident marginal adrenal function due to under-
medications (amiodarone, -blocking drugs, narcotics, lying Addisons disease. Severe hyponatremia (120
lithium, barbiturates, and sedatives). Anesthesia, discon- mEq/L) should be corrected with the administration of
tinuation of L-T4 replacement therapy, cerebrovascular hypertonic saline solution (50 100 mL of 3% sodium
events, congestive HF (CHF), or GI bleeding may play an chloride) followed by an iv bolus of 40 to 120 mg furo-
important role in the onset of MC (536). Sinus bradycar- semide. Conivaptan and tolvaptan, nonpeptide inhibitors
dia, low-voltage complexes on an electrocardiogram of antidiuretic hormone, could be useful in treating hypo-
(ECG), complete heart block, nonspecific ST-T changes, natremia in MC where elevated vasopressin levels are
prolongation of the QT interval, and increased QT dis- present. Conivaptan has been successfully used to treat
persion, represent frequent findings at the ECG evaluation euvolemic hyponatremia in a starting dose of 20 mg by iv
in patients with severe hypothyroidism. Polymorphic ven- infusion followed by continued infusion at a rate of 20
tricular tachycardia (torsades de pointes), pericardial ef- mg/d for 2 to 4 days (539). Unfortunately, the rare routine
fusion, and MI are alarming complications of MC. An availability of the vasopressin assay may limit the use of
elevated creatine phosphokinase levels in association with this helpful drug in MC.
nonspecific ECG findings may generate an incorrect di- Severe hypotension should be treated with the cautious
agnosis of MI. administration of iv fluids (5% to 10% glucose in isotonic
Depression of cerebral function, hyponatremia, hypo- sodium chloride solution resuscitation plus hydrocorti-
glycemia, respiratory acidosis, hypoxia, hypercapnia, and sone). Caution also should be exercised when using vaso-
reduced cerebral blood flow are frequent in the clinical pressor drugs because these drugs might exacerbate car-
presentation and may lead to respiratory failure and coma. diac arrhythmias with the iv administration of thyroid

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
482 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

hormone. About 200 g of T4 iv followed by 100 g T4 iv iv followed by 100 g T4 iv was both effective and safe
per day can be administered to restore vital functions in when compared with higher doses of L-T4 (537).
patients with severe coma. Alternatively, a 500-g dose of Combined therapy with T4 and T3 may also be useful.
oral L-T4 should be promptly started and followed by 125 T4 may be started with an initial dose of 4 g/kg lean body
to 150 g L-T4 as a daily maintenance dose orally in pa- weight and may be followed by 100 g 24 hours later with
tients who are able to take oral medication. Patients who a further progressive reduction to 50 g daily iv or orally.
received oral L-T4 had no difference in the outcome from T3 may also be started simultaneously with T4 at a dose of
those receiving iv T4 (537). 10 g iv, and the same dose is given every 8 to 12 hours
Because the onset of T4 action may be delayed due to a until the patient can take maintenance oral doses of T4.
reduced rate of activating conversion of T4 to T3, iv L-T3 The treatment of MC is summarized in Table 18.
administration may be more effective than T4. T3 would be
administered in a starting doses of 10 to 20 g, followed
by 10 g every 4 hours for the first 24 hours and then 10 XI. L-T4 Therapy in Benign and Malignant
g every 6 hours for 1 or 2 days until the patient suffi- Thyroid Nodular Disease
ciently improves their cerebral function to take oral med-
ication (540). Adverse cardiac events may develop during A. L-T4 therapy in benign hypofunctioning thyroid nodules
the early treatment of MC (arrhythmias or MI), especially The incidental detection of unsuspected thyroid nod-
with doses above 500 g/d of L-T4 and 75 g/d of L-T3, ules or incidentalomas has dramatically increased in the
leading to increased mortality (541). As a consequence, last several years due in large part to the frequent use of
elderly patients and patients with comorbidities are more ultrasound for the evaluation of the thyroid gland and
safely treated with lower doses of L-T4, usually in the range lesions of the neck (10). Discovery of a thyroid nodule and
of 100 to 150 g oral L-T4 daily. In one report, 200 g T4 presentation to an endocrinologist is relatively common.

Table 18. Treatment and Prognosis of MC


Monitoring in ICU setting
Treat precipitating factors
Hypoventilation: O2 and mechanical ventilation
Hypotension: administration of iv fluids (5%10% glucose in isotonic NaCl solution resuscitation plus hydrocortisone; caution with vasopressor
drugs because they may exacerbate cardiac arrhythmias with iv thyroid replacement therapy)
Hypothermia treated with warm room temperature/blankets; avoid rapid rewarming because of peripheral vascular dilatation, which may
precipitate hypotension and shock
Hypoglycemia: managed with dextrose infusion
Hyponatremia may be treated with saline solution and loop diuretics
Treat and then exclude coexisting adrenal insufficiency: hydrocortisone 50 100 iv every 6 8 h
Thyroid replacement therapy
Controversial: L-T4 alone, L-T3 alone, L-T4/L-T3 combination
L-T4 administration
Parenteral replacement can be preferred initially due to bowel wall edema and gastric atony with unpredictable gastrointestinal absorption
L-T4: 200 500 g iv followed by 50 100 g iv per day in severe coma
Advantages: iv formulation, smaller risk of cardiac complications
Disadvantages: L-T4 less metabolically active than L-T3
L-T4 500 g of oral dose followed by 150 g of L-T4 as maintenance dose in patients who are able to take oral medication
Liothyronine
L-T3 starting doses of 10 20 g, followed by 10 g every 4 h for the first 24 h and then 10 g every 6 h for 1 or 2 d
Advantages: biologically active, faster therapeutic
Disadvantages: higher risk of cardiac complications (ischemia, lethal arrhythmias)
Combined therapy with T4 and T3
T4 stated at a dose of 4 g/kg of lean body weight followed by 100 g 24 h later with a further progressive reduction to 50 g daily iv or
orally; T3 started simultaneously with T4 at a dose of 10 g iv, and the same dose is given every 8 12 h until the patient can take maintenance
oral doses of T4
Prognosis
Mortality rate 30% 60%
Poor prognosis associated with
Initial temperature 93F or no response to therapy within 3 d
Advanced age
Bradycardia
Hypotension
Cardiac complications (including MI, CHF)
Sequential Organ Failure Assessment (SOFA) scores more than 6
High APACHE II score

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 483

The prevalence of thyroid nodules is higher in women in Some meta-analyses have failed to show a clear benefit
areas of iodine deficiency and increases with advanced of TSH suppression in the management of nontoxic thy-
age, being up to 50% in individuals over the age of 50 years roid nodules, and this treatment is debated in terms of
as may be found on ultrasound evaluation. Iodine defi- efficacy, the optimal level of TSH suppression, the optimal
ciency is the most frequent contributing factor to the de- duration of treatment, and potential side effects. A critical
velopment of multinodular goiter. The pathogenesis be- examination of trials on TSH-suppressive therapy was
gins with iodine deficiency leading to reduced thyroid performed in an early cumulative meta-analysis that eval-
hormone production, which in turn results in increased uated the capacity of L-T4-suppressive therapy to decrease
serum TSH, and TSH serves as a growth factor for thyroid the volume of a nontoxic solid thyroid nodule to 50% of
epithelial cells. its baseline value (548). The analysis indicated that L-T4
A diffuse goiter is frequently observed in young subjects treatment was associated with decreased nodule volume in
living in iodine-deficient areas whereas MNG is com- 17% of patients and was able to inhibit growth in another
monly observed in the elderly (65). The natural history of 10% (548). In the meta-analysis by Castro et al (549) that
untreated isolated nodules and MNG is characterized by included 6 randomized trials for a total of 346 patients,
a progressive worsening; thyroid nodule volume increases only 22% of patients treated with TSH-suppressive doses
over time, and the size of the surrounding thyroid tissue of L-T4, compared with 10% in the control group, had a
enlarges presumably due to TSH stimulation, increasing decrease in nodule volume by more than 50%. However,
the risk of the onset of new nodules. Two retrospective this degree of improvement during thyroid hormone ther-
studies from Japan and England have examined the nat- apy for longer than 6 months did not achieve statistical
ural history of untreated hypofunctioning thyroid nodules significance (549).
followed for an average of 15 and 5 years, respectively Nine studies were included in the meta-analysis of
(542, 543). These studies have demonstrated that un- Richter et al (550), which indicated that T4-suppressive
treated solid thyroid nodules increased in size in about therapy led to a nonsignificant improvement in the rate of
13% to 15% of cases, whereas nodules with greater cystic response to therapy (defined as 50% nodule volume re-
content were less likely to grow (542, 543). duction by ultrasound) (pooled RR, 1.83; 95% CI, 0.9
Thyroid autonomy may develop in one or more of the 3.73). The mean nodule volume or diameter reduction was
nodules in euthyroid elderly patients with MNG, leading not statistically significant between the L-T4 treatment
to development of endogenous subclinical hyperthyroid- group and the control group (550). In the meta-analysis of
ism. This condition may then exist for several years, pre- Sdano et al (551), a thyroid hormone TSH suppression
ceding the onset of overt disease and may be responsible therapy (THST) appeared more likely than placebo or no
for severe cardiovascular complications (544). Obstruc- treatment to significantly reduce benign thyroid nodule
tive symptoms with airway compression, dysphagia, and volume. Finally, a meta-analysis by Yousef et al (552) as-
paralysis of a recurrent laryngeal nerve may occur in the sessed the effects of THST on the basis of ultrasonography
presence of a large MNG. of the nodules rather than physical examination. A total of
Approximately 5% to 10% of thyroid nodules may be 417 patients were treated with L-T4, and 326 patients re-
malignant, and the risk of thyroid cancer may increase in ceived placebo. Of the 11 included studies, 8 were RCTs
people who were exposed to ionizing radiation in infancy (6 single-center studies and 2 multicenter trials) and 3 were
and childhood . cohort studies. The analysis indicated that L-T4treated
Surgical treatment is recommended in the presence of patients were up to twice as likely to have nodule reduction
large goiter with compressive symptoms, progressive at the end of follow-up (minimum follow-up of 6 months)
growth of thyroid nodules, and documented or suspected (552).
thyroid malignancy (2325, 545, 546). Given the results of these various studies and meta-
Murray in 1891 (125) reported that the sc injection of analyses, it is understandable why the issue of THST for
sheep thyroid extract for the treatment of myxedema re- nodules remained controversial. It is relevant to consider
duced the size of goiter. In 1953, Greer and Astwood (547) some of the limits or confounding factors inherent in the
described goiter regression in two-thirds of patients treated studies or these meta-analyses including: 1) the heteroge-
with thyroid extract in an uncontrolled trial. Treatment with neity of the patients evaluated, 2) the variability in L-T4
L-T4 in doses sufficient to suppress TSH has long been used treatment in terms of doses and duration of therapy, 3) the
with the aim of preventing or reducing the growth of thyroid different definitions of response to treatment, 4) the dif-
nodules or the formation of new nodules. However, the ef- ferent levels of TSH suppression obtained, 5) the various
fectiveness of TSH suppression of thyroid nodules with L-T4 methods used to assess nodule volume, and 6) the inclu-
has been controversial for decades. sion of nodules with different characteristics (solid, mixed,

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
484 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

or cystic nodules). Moreover, the follow-up of some stud- therapy was superior to T4, iodine, or placebo. L-T4 ther-
ies was relatively short, and the side effects were poorly apy was administered in doses able to obtain a TSH target
documented. range of 0.2 to 0.8 mU/L. This study revealed that the
A critical analysis of these results suggests that the re- volume of benign nodules greater than 1 cm in diameter
duction of nodule volume was more effective in those nod- was significantly reduced by TSH-suppressive L-T4 doses
ules of certain characteristics such as: 1) small nodules in combination with iodine compared with T4 alone, io-
(2.4 mL in volume or 1.7 cm in diameter); 2) recently dine alone, or placebo in an iodine-deficient German pop-
diagnosed thyroid nodules, ie, nodules likely of shorter ulation (557). The nodule volume reduction was, respec-
duration; 3) lesions with colloid features on cytological tively, 17.3% in the T4 plus iodine group, 7.3% in the
examination; 4) patients from geographic regions with T4 group, and 4.0% in the iodine group as compared
borderline or frank iodine deficiency; and 5) nodules with with placebo. Thyroid volume reduction (7.9%) was
cytological diagnosis of Hashimotos thyroiditis. On the superior during combined L-T4-iodine therapy compared
contrary, a reduction in size or volume was rarely observed with iodine (5.2%); however, this reduction was com-
in fibrotic, hyperplastic nodules. parable to L-T4 alone (557).
Some degree of regrowth of nodules sometimes occurs Thus, in a region with an insufficient iodine supply,
after the cessation of L-T4 therapy. Moreover, THST in treatment with a combination of iodine and T4 of patients
doses that reduced TSH levels to 0.1 mU/L were more with normal to modestly elevated TSH levels could serve
effective than those doses associated with TSH levels 0.1 to reduce thyroid nodule volume. However, we should
mU/L (553). A few studies assessed the ability of L-T4 to caution that high doses of iodine may induce development
reduce or even prevent the development of additional nod- of hyperthyroidism especially in elderly subjects. It should
ules (554, 555). One 5-year follow-up study indicated that be noted that ATA management guidelines state that there
L-T4 treatment was ineffective in achieving nodule shrink- is no benefit of THST of thyroid nodules proven benign by
age (TSH 0.1 mU/L) but was effective in significantly cytological examination (23, 24).
reducing the frequency of new nodules (554). Although Another interesting aspect of the relationship between
still unsettled due to the heterogeneous results of these serum TSH and thyroid neoplasia is the growing body of
studies, it remains conceivable that L-T4 therapy might be literature indicating that a high serum TSH level may be an
more efficient than placebo in preventing the development independent predictor for the development of thyroid can-
of additional nodules and the increase in the thyroid vol- cer in patients with thyroid nodules (558, 559). Thus, sev-
ume under certain specific clinical circumstances. This eral studies report a direct relationship between higher
possibility is supported by a prospective multicenter, ran- TSH levels and the risk of papillary thyroid cancer (PTC)
domized, double-blind placebo-controlled French trial, in in patients with nodular thyroid disease (558, 559). This
which THST was effective in reducing the growth of sol- may constitute another rationale for L-T4 treatment of
itary thyroid nodules and in preventing the development patients with nodules in hope of reducing the potential
of new nodules (555). Pharmacological doses of iodine development of thyroid cancer by reducing serum TSH
have been used in some studies for the treatment of pa- levels (560).
tients with MNG in areas of iodine deficiency (556). The This approach is supported by the results of a recent
rationale for iodine treatment is based on the known re- study that investigated the effect of L-T4 treatment on the
lationship of goiter development to iodine deficiency, and frequency of PTC diagnosed by cytology in a large series
the possibility that part of the effect of L-T4 therapy lies in of patients with nodular goiter. In this study, L-T4treated
the adjunctive provision of the iodine content of the L-T4. patients had a significantly lower serum TSH and enjoyed
Indeed, La Rosa et al (556) reported that patients receiving a lower prevalence of PTC (3.2% vs 5.1%; P .0001)
potassium iodide (1.5 mg every 2 weeks) had a decrease of compared with untreated patients (560). The frequency of
their basal nodule volume by 23%. PTC was closely related to the serum TSH level, being
Recently, a multicenter, double-blind, randomized, lowest in patients with TSH below the reference range and
placebo-controlled trial (LISA trial) was performed in highest in patients with TSH above the reference range and
Germany to assess the effects of iodine and L-T4 treatment showing a progressive increase from the lower to the upper
in patients with nodular goiters, during a 12-month treat- quartile of the TSH reference range (560).
ment period (557). A total of 1020 euthyroid patients aged Although the latter studies indicating a relationship be-
18 to 65 years with 1 or more thyroid nodules (minimal tween TSH and thyroid neoplasia have been very recent,
diameter 10 mm) participated in the study. They were the data should not have been surprising given the known
recruited from 60 German centers and were randomized mitogenic potential of TSH. Thus, TSH is known to stim-
to evaluate whether the effect of combined T4 plus iodine ulate thyroid cell proliferation and the growth of experi-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 485

mental tumors with its removal shown to provoke accel- Open questions remain on THST which should be clar-
erated tumor growth. Clinically, TSH levels can play a role ified in the future. In particular, it is unknown whether
in the progression of preexisting PTC (561563). Further- subnormal TSH may induce the same beneficial and ad-
more, high TSH values have been associated with a more verse effects associated with an undetectable serum TSH.
advanced stage of thyroid cancer (564 566). Conceiv- Therefore, it remains to be established what level of TSH
ably, the link between Hashimotos thyroiditis and a suppression could be effective in improving thyroid nod-
higher risk of PTC might be mediated as well by the like- ule growth and malignancy without increasing the risk of
lihood of higher TSH levels in these patients. adverse effects. Moreover, it will be necessary to under-
All of these results clearly indicate that TSH suppres- stand how long patients should be treated with L-T4 to
sion could affect the natural history of thyroid nodules. improve the nodular disease and the risk of PTC, consid-
A recent meta-analysis assessed the association be- ering that the beneficial effect of this treatment might be
tween serum TSH and thyroid cancer by the evaluation of lost when therapy is discontinued.
28 studies, including a total of 42 032 subjects and 5786 RCTs with a long period of follow-up will be required
thyroid cancer cases (567). This analysis confirmed that a to evaluate the risk to benefit ratio of treatment with L-T4
higher serum TSH concentration is associated with a with the aim of reducing the risk of nodule or MNG
higher risk of thyroid cancer. The dose-response model growth, the risk of thyroid cancer, and the risk of adverse
OR was 1.72 (95% CI, 1.422.07) per mU/ L TSH below effects. The ability of L-T4 treatment to prevent the devel-
1 mU/L and changed to an OR of 1.16 (95% CI, 1.12 opment of functional autonomy in patients with nontoxic
1.21) per mU/ L TSH at levels of 1 mU/L and greater. In a MNG has not been demonstrated. This important issue
prospective model, the odds of thyroid cancer were 3 times should be another objective of future controlled prospec-
greater in patients with a serum TSH level of 4 mU/L com- tive clinical trials. It may be possible to identify a subgroup
pared with those with a serum TSH of 0 mU/L and doubled of patients with nodular disease in whom THST could be
between a serum TSH level of 2.2 and 7 mU/L. shown to be clearly beneficial.
Future longitudinal studies are necessary to further as-
sess this important issue before including serum TSH in a B. L-T4 therapy in DTC
diagnostic nomogram for thyroid cancer prevention Prolonged thyroid hormone-suppressive therapy has
(567). Currently, L-T4 treatment of thyroid nodules to pre- long been used to prevent the recurrence or progression of
vent thyroid cancer is not recommended. well-differentiated thyroid cancer. As alluded to above,
Of course, there are risks attendant to THST that per- experimental studies have shown that TSH stimulates thy-
tain to their use for benign nodular disease and that have roid cell proliferation, radioiodine uptake, and Tg pro-
added to the debate and principally relate to the cardiac duction (568, 569). Moreover, clinical data have demon-
and skeletal side effects of this therapy (9). Specifically, strated that TSH suppression with L-T4 reduces the
THST can lead to an unintentional exogenous subclinical likelihood of progression and/or recurrence of thyroid
hyperthyroidism that may increase the risk of AF and may cancer (570 572). Several older studies have shown fewer
reduce BMD, especially in the elderly and postmenopausal cancer recurrences and cancer-related deaths in patients
women with even more harmful effects in patients with who have received TSH-suppressive L-T4 doses than in
comorbidities (9). Although recent ATA guidelines do not those who have not received TSH-suppressive therapy
recommend routine suppressive therapy with L-T4 of be- (573, 574). These studies have supported the survival ben-
nign thyroid nodule (23, 24), this treatment is still used efit of a high level of TSH suppression (ranging from 0.05
among members of the ETA, especially in areas of bor- 0.1 mU/L) especially in patients at high risk of recurrence
derline-low iodine intake (25). after total thyroidectomy and thyroid remnant ablation,
According to AACE/AME/ETA guidelines, TSH sup- suggesting that a constant TSH suppression (0.1 mU/L)
pression should be particularly avoided in postmeno- could improve relapse-free survival. Pujol et al reported
pausal women with evidence of low bone mass, in the that persistent TSH suppression 0.05 mU/mL was asso-
elderly, and in those with cardiac disease; in those nodules ciated with longer relapse-free survival compared with se-
with suspicious cytological lesions; and in general, in pa- rum TSH levels of 1 mU/mL or higher (575). In this study,
tients in whom the risk of this therapy outweighs its un- the degree of TSH suppression was an independent pre-
certain benefits (25). A THST can be considered in patients dictor of recurrence in the multivariate analysis (575).
living in iodine-deficient areas, in young patients with These findings were confirmed by a meta-analysis in-
small thyroid nodules, and in patients with nodular goiters cluding 10 studies that indicated the benefit of TSH sup-
with no evidence of functional autonomy. Raised FT3 and pression (576). However, these studies did not differenti-
FT4 levels should be avoided during THST. ate between the beneficial effects of TSH suppression

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
486 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

according to the aggressiveness of DTC. Moreover, all of tween those patients who were receiving TSH-suppressive
the studies were limited by the lack of randomization and doses and those in whom TSH remained within the ref-
appropriate controls (576). erence range (580).
Recommendations for the degree of TSH-suppression A recent retrospective study examined the survival of
therapy required have been modified in recent years be- 157 DTC patients with distant metastases (581). The re-
cause of appreciation of the risk of adverse effects of ex- sults indicated that DTC-specific survival was signifi-
ogenous TSH suppression and the evidence that suppres- cantly improved in patients with a median TSH level 0.1
sion may be necessary only in high-risk patients, whereas mU/L (median survival 15.8 years) than in those with a
those patients who appear to be disease-free will enjoy nonsuppressed TSH level (median survival 7.1 years; P
excellent survival without suppression. Such recommen- .001). However, no further improvement in survival was
dations have evolved related to studies such as that re- found when TSH suppression was lower than 0.1 mU/ L
ported by Cooper et al (577) who performed a multicenter, (581). These results suggest that TSH-suppression therapy
prospective study using the database of the National Thy- decreases the rate of recurrences and mortality only in
roid Cancer Cooperative Registry. This study assessed the patients with high-risk DTC. As a result of these studies,
results of 617 patients with PTC followed for a median of clinicians target aggressive TSH suppression with L-T4 in
4.5 years and showed that TSH suppression was an inde- patients with metastatic DTC, whereas less aggressive
pendent predictor of disease progression in high-risk pa- TSH suppression is deemed appropriate in disease-free
tients. However, a greater degree of TSH suppression was patients.
not useful to prevent disease progression in patients at low Recent management guidelines allowed clinicians to
risk for recurrence (577). stratify the degree of TSH suppression according to the
A second study from the National Thyroid Cancer Co- degree of aggressiveness of thyroid cancer (24, 582585).
operative Registry included 2936 patients with DTC The ATA guidelines have defined risk of recurrence as
(578). This study confirmed that subnormal to undetect- either low, intermediate, or high based upon the patients
able serum TSH levels were necessary to improve overall status after the initial therapy (thyroidectomy with or
survival and disease-specific survival in stage III and IV without radioiodine ablation) and recommends a risk-
patients. On the contrary, this study proved that subnor- adapted titration of TSH suppression. The guidelines of
mal serum TSH levels were sufficient to improve the over- both the ATA and ETA recommend maintaining unde-
all survival in stage II patients compared with normal or tectable serum TSH levels in high-risk patients beginning
elevated serum TSH levels (578). A retrospective study at the time of their initial management (24, 582). On the
assessed recurrence risk in 366 patients with DTC treated contrary, in low-risk patients who represent about 80% of
with total thyroidectomy and radioiodine ablation (579). all thyroid cancer patients, the ATA and ETA guidelines
In this study, serum TSH values 4.5 mU/L were an in- do not support any benefit from TSH suppression. Fur-
dependent predictor of death. Furthermore, serum TSH thermore, ATA and ETA guidelines recommend that high-
levels 2 mU/L were associated with an increased risk of risk patients with evidence of persistent disease should
thyroid cancer-specific death and recurrences. However, have their serum TSH level maintained at 0.1 mU/L. And
no differences in deaths or recurrences were observed in in those high-risk patients who have achieved a disease-
patients who had their serum TSH levels maintained be- free status after treatment, the ATA guidelines recommend
tween 0.1 and 0.4 mU/L (580). maintaining the TSH level between 0.1 and 0.5 mU/L for
A recent RCT assessed the efficacy of TSH-suppression 5 to 10 years of follow-up, whereas the ETA guidelines
therapy on thyroid cancer recurrences (580). Low- and recommend a TSH level of 0.1 mU/L for 3 to 5 years. For
high-risk patients with PTC were stratified according to low-risk patients, the ATA guidelines suggest that initial
the AMES (age, metastasis, extension, and size) risk-group serum TSH levels may be between 0.1 and 0.5 mU/L,
classification. Disease-free survival patients were ran- whereas the ETA guidelines suggest a suppressed TSH of
domly assigned to receive postoperative TSH-suppression 0.1 mU/L. In low-risk patients who are shown to be
therapy (group A) or not (group B). The 218 patients in- disease-free on follow-up testing, both guidelines recom-
cluded in group A received doses of L-T4 to keep serum mend that serum TSH should be kept at the low end of the
TSH levels below 0.01 mU/L, and the 215 enrolled in reference range (24, 582).
group B were treated with L-T4 doses to obtain a serum In addition to the potential for adverse effects on heart
TSH level within the reference range (0.4 5.0 mU/L). and bone, TSH-suppressive therapy is not without possi-
Neck ultrasonography and chest computed tomography ble other risks in patients with DTC (586, 587). For ex-
were used to detect recurrences. The results of this study ample, a recent review noted an increased incidence of
indicated that disease-free survival was not different be- kidney, pancreas, ovarian, and breast cancers (586). A

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 487

Table 19. ATA Risk Assessment for DTC at the Time of Initial Surgery
Low Risk Intermediate Risk High Risk
No local or distant metastases Microscopic invasion into the Increased age (4550 y)
Complete surgery perithyroidal soft tissues Increased size (4 cm)
No locoregional or vascular invasion Aggressive histology Macroscopic tumor invasion
No aggressive histology (eg, tall cell, insular, Vascular invasion Incomplete tumor resection
columnar cell carcinoma) Distant metastases
No uptake outside the thyroid bed on initial Radioiodine uptake outside the thyroid bed after a
posttreatment scan (if performed) posttreatment radioiodine scan performed after
ablation of remnant thyroid

large epidemiological study in Norway including 29 000 of ExoSHyper. Stratification is based upon those data in-
patients with suppressed TSH levels to less than 0.5 mU/L dicating that more aggressive TSH suppression is war-
who had been monitored for 9 years confirmed an in- ranted in patients with high risk disease or recurrent tu-
creased cancer incidence (HR, 1.34) (587). It has been mor, whereas less aggressive TSH suppression is targeted
postulated that integrin activation by TSH could be re- for low-risk patients. The underlying concept holds that
sponsible for promoting angiogenesis through the activa- long-term treatment with L-T4 should be individualized
tion of MAPK (or MAPK pathway) (588). Finally, an in- and balanced against the potential risk for adverse effects
creased thyrotoxic periodic paralysis with profound during the follow-up in patients who are both at high risk
hypokalemia and muscle paralysis due to overtreatment of recurrence and at high risk of adverse effects. Elderly
with L-T4 was reported in Caucasians (589). patients with DTC, especially those with comorbidities,
Therefore, the beneficial role or necessity for TSH-sup- are usually at higher risk for both cancer progression and
pressive therapy in all patients with DTC is questioned adverse effects from L-T4 therapy. Therefore, these pa-
today, given the above-mentioned recent randomized tients should undergo periodic cardiological evaluation
study that failed to show any benefit for TSH-suppressive and bone density assessment, whereas the degree of TSH
therapy in patients at low risk levels and the overall in- suppression is individualized and adapted to the clinical
creased risk of adverse cardiac and bone effects and the less situation, which may change with time. In high-risk pa-
likely but possible development of other cancers. tients who are clinically and biochemically free of disease
Stratification into a risk-adapted management scheme during 5 to 10 years of follow-up, the degree of TSH-
for L-T4 therapy in patients with DTC was summarized in suppressive dosage can be lightened to achieve measurable
a recent review that considered the aggressiveness of the serum TSH levels between 0.1 and 0.5 mU/L, whereas
DTC as well as the potential adverse effects induced by serum FT4 should be maintained within its reference
iatrogenic subclinical hyperthyroidism (569). To this end, range.
patient age and the presence of preexisting cardiovascular Until the presence of residual disease can be ruled out,
and skeletal risk factors were considered. Tables 19 and 20 an undetectable TSH value (0.1 mU/L) should be main-
stratify different potential patient categories according to tained in the first 1 to 3 years after initial diagnosis and
the aggressiveness of DTC and the risk of adverse effects treatment of patients at intermediate risk of thyroid can-

Table 20. Assessment of Risk of Adverse Effects From TSH-Suppressive Doses of L-T4 in DTC Patients
Mild Risk Moderate Risk Higher Risk
Young and middle-aged patients Elderly subjects Arrhythmias
Asymptomatic patients Hypertension Atrial fibrillation
No cardiovascular disease Symptoms and signs of Comorbidities (eg, heart disease,
No alterations of cardiac rhythm adrenergic overactivity diabetes, renal failure)
No symptoms of adrenergic overactivity Cigarette smoking Previous stroke or TIA
No cardiovascular risk factors Cardiovascular risk factors Left atrial dilatation
No comorbiditiesPremenopausal Postmenopausal women Increased risk factors for stroke
women Osteopenia HF
Normal BMD Men and women with risk CHD
No risk factors for osteoporosis factors for osteoporosis Valvular heart disease
Vascular disease (coronary or peripheral
arterial disease)
Osteoporosis
Modified from Ref. 569.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
488 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

cer. However, in those at high risk of adverse effects of who claim to feel better after taking T3. The frequency of
therapy, the degree of TSH suppression should be reeval- complaints attributed to Wilsons syndrome were re-
uated during the follow-up to obtain a normalization of viewed some years ago in an interesting study published in
serum TSH in disease-free patients, especially in the pres- the Annals of Internal Medicine called Functional so-
ence of significant comorbidities. matic syndromes (591).
A TSH value between 0.1 and 0.5 mU/L is suggested for In 2005, the ATA published a position statement on
initial therapy in patients at low risk of cancer progression Wilsons syndrome after reviewing the material presented
and a low risk of adverse effects. However, once cure or on the Wilsons syndrome website (592). This statement
complete remission from cancer has been established, al- concluded that the evidence supporting existence of this
lowing serum TSH levels to drift up into the normal ref- syndrome is inconclusive and without scientific merit.
erence range is desirable, with a TSH target within the Moreover, the diagnostic criteria for Wilsons syndrome
low-normal reference range (0.32 mU/L). were considered vague or imprecise with potential risk of
In patients at low risk of cancer progression and a high leading patients to mistaken or delayed diagnoses of other
and intermediate risk of adverse effects, a TSH value be- conditions. Finally, the prescribed dosage for T3 for Wil-
tween 0.5 and 1 mU/L would be reasonable for initial sons syndrome was found to be inconsistent with normal
therapy, and a TSH value of 1 to 2 mU/L would be ap- physiology with clear potential for harmful effects on the
propriate during the follow-up in disease-free patients. heart and skeleton.
In the future, drugs that specifically decrease pituitary
TSH secretion without adverse effects on the cardiovas- B. Replacement therapy with thyroid hormone in
cular system and the skeleton could be useful in high-risk obese patients
DTC patients. This would be feasible by an integrated Thyroid hormones are inappropriately and frequently
molecular and genetic approach to identify patients with used in a misguided attempt to induce weight loss in obese
DTC at high risk of recurrences who would benefit from euthyroid subjects. However, reduced caloric intake with
TSH suppression. weight loss may induce a significant reduction in the serum
FT3 level with an increased rT3 due to reduced 5-deiodi-
nation. Thereafter, continuing caloric deprivation further
reduces energy expenditure and the decreased T3 levels
XII. Appropriate and Inappropriate Use of L-T4
Replacement Therapy in Other Conditions may be responsible for the difficulties in maintaining or in
achieving further weight loss (52)
A. Fatigue (Wilsons syndrome) Consequently, some workers have offered this altera-
Wilsons syndrome refers to a cluster of debilitating tion during caloric deprivation as a potential rationale for
symptoms in euthyroid patients. Dr E. Dennis Wilson, the use of thyroid hormone in obese subjects to improve
MD, of Florida suggests that this condition represents their weight loss (52). Ideally, the best treatment of obesity
functional hypothyroidism, or a form of presumed thyroid should increase fat loss without decreasing skeletal muscle
hormone deficiency in euthyroid patients. Dr Wilson mass and without inducing cardiac arrhythmias. A recent
and/or his protogees consider that patients with this syn- meta-analysis assessed the results of T3 therapy in 14 stud-
drome could be responsive to treatment with a special ies in obese subjects during caloric deprivation (6 random-
preparation of L-T3. He maintains that patients with this ized controlled studies and 8 prospective observational
syndrome may have abnormally low body temperatures studies). Weight loss was significantly increased in only 5
(90% with temperatures 97.8F or lower) together with studies with T3 doses ranging from 38 to 117 g/70 kg/d
symptoms indicative of hypothyroidism such as fatigue, (593). T3 therapy decreased serum TSH and T4 concen-
fluid retention, inappropriate weight gain, unhealthy nails trations. It is difficult to ascertain the effects of T3 on
and skin, irritability, depression, memory loss, poor mo- resting metabolic rate, protein breakdown, fat loss, and
tivation, anxiety, panic attacks, joint aches, muscle aches, heart rate due to the small number of obese subjects en-
hypoglycemia, constipation, irritable bowel syndrome, rolled in each study and the poor quality of most of these
and chronic fatigue. The evidence for beneficial effects of studies. Therefore, the authors concluded that data avail-
the T3 preparation recommended for Wilsons syndrome able in the literature are inconclusive to evaluate the ef-
is available on Dr Wilsons website on which are listed fectiveness of thyroid hormone therapy as a treatment for
about 37 symptoms that are considered part of this con- obesity (593). Prospective large controlled studies will be
dition (590). On the same website, the evidence justifying necessary to assess the potential beneficial effect of treat-
T3 treatment is described (590). The evidence appears ment of obese patients with thyroid hormones or their
based on anecdotal reports and testimonials from people analogs.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 489

C. Effects of replacement therapy with thyroid hormone levels were significantly reduced, and SHBG was signifi-
in severe cardiac diseases and dyslipidemia cantly increased in the tiratricol group compared with the
The physiological action of thyroid hormone is medi- L-T4 group. However, adverse effects of TRIAC were seen
ated by its genomic nuclear effects. These effects result in regard to increased bone resorption and turnover with
from the binding of T3 to specific nuclear TRs. marked elevation of osteocalcin, pyridinium cross-links,
In the human heart, 2 TR genes are expressed and each and urinary calcium. The study investigators concluded
gene generates 2 isoforms of receptors (TR1, TR2, that TRIAC had increased hepatic and skeletal actions but
TR1, and TR2) (129). TR transcripts are abundant in did not have pituitary-specific selectivity (603). However,
the heart, whereas TR2 acts as a negative regulator be- because TRIAC selectively increases the function of TR1
cause it functions by suppressing TR1 transcription without affecting TR1, some utility became apparent for
(129). TR1 is widely expressed and has a high expression the treatment of symptomatic patients with selective pi-
in cardiac and skeletal muscles. TR1 is predominantly tuitary resistance to thyroid hormone (604) and TSH-de-
expressed in the brain, liver, and kidney. TR2 expression pendent hyperthyroidism (605). In euthyroid patients,
is limited to the hypothalamus and pituitary (129). TRIAC may significantly increase metabolic activity with
induction of symptoms of hyperthyroidism, and it has
1. Analogs of thyroid hormone for treatment of obesity been used inappropriately as a dietary supplement for
and dyslipidemia weight loss. In the United States, the FDA has not ap-
The selective activation of different TR-mediated path- proved use of tiratricol as a dietary supplement due to the
ways is a promising strategy for treating lipid disorders potential risk of serious health consequences. It could have
and obesity (594 598). Preclinical studies have suggested a role in the therapy of hypothyroid patients with accom-
that thyromimetics might be useful for the treatment of panying hypercholesterolemia, but the negative effects on
obesity and dyslipidemia. Selective thyromimetics are syn- bone are a drawback.
thetic analogs of thyroid hormones that can selectively
stimulate TR, avoiding harmful effects on the heart and b. Sobetirome. Development of sobetirome, or GC-1 (3,5-
bone. Animal data have suggested that thyromimetics dimethyl-4 4-hydroxy-3-isopropylbenzyl phenoxy ace-
might be useful in the treatment of obesity, hepatic ste- tic acid), was derived from the first-generation TR-se-
atosis, and atherosclerosis (599, 600). In the past, the Cor- lective agonists as a selective thyromimetic with 10-fold
onary Drug Project (1966 1975) performed a clinical preferential action on TR 1. It binds TR with an affinity
experiment using dextro-T4 for hyperlipidemia (601). similar to T3 but binds TR with a 10-fold lower affinity
However, overall mortality increased with dextro-T4 de- (606). Sobetirome preferentially accumulates in the liver,
spite reductions in serum cholesterol, possibly due to the reduces LDL-cholesterol, and initially was thought to
high rates of contamination of dextro-T4 preparations have a promising role as an antiobesity agent because it is
with the L-enantiomer (601). able to induce a 20% decrease in fat mass and improve
lipid profile without reduction in food intake and without
a. 3,5,3-Triiodothyroacetic acid. The decarboxylation and affecting the heart or BMD (607). Treatment of rats with
deamination of thyroid hormone leads to 3,5,3-triiodo- 3 g T3/100 g body weight and equimolar amounts of
thyroacetic acid (TRIAC; tiratricol), a naturally occurring GC-1 (3 g GC-1/100 g body weight) both resulted in a
metabolite of T4 with thyromimetic activity. TRIAC has loss of fat mass. However, T3 caused widespread loss of
affinity for TR some 10 to 20 times that of T3. The first muscle mass, whereas GC-1 had no effect (608). There was
reports in rats postulated that TRIAC could serve as a optimism that a second generation of highly selective TR
favorable alternative agent for TSH suppression without agonists or compounds would have additional tissue-spe-
adverse effects on the heart (602). It has been administered cific effects (ie, at the level of adipose tissue) and demon-
to patients with isolated pituitary resistance to thyroid strate promise for the treatment of obesity and type 2
hormone and was thought to have TSH-suppressive ef- diabetes.
fects greater than T4. In a trial to assess the specific thy-
romimetic effects of tiratricol, 24 athyreotic patients were c. KB141. KB141 is another TR agonist that is approxi-
randomized to blinded treatment with either TRIAC (24 mately 15 times more selective for TR than for TR in
g/kg twice daily) or L-T4 (1.9 g/kg daily) to obtain a vitro and is able to induce weight loss and reduce choles-
TSH level less than 0.1 mU/L (603). No significant differ- terol and lipoprotein (a) with no effect on heart rate (609).
ences were found between the 2 groups in terms of weight,
resting metabolic rate, urea excretion, symptoms, and car- d. Eprotirome (KB2115). KB2115 (eprotirome; 3(-3,5-di-
diovascular function. Plasma total and LDL-cholesterol bromo-4-(4-hydroxy-3-(1-methylethyl)- phenoxy)-phe-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
490 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

nyl)-amino-3-oxopropanoic acid) is a TR-selective li- modify this interpretation (612, 613). Studies performed
gand that is preferentially taken up by the liver. in patients after acute MI or cardiopulmonary bypass and
A randomized, placebo-controlled, double-blind, mul- in those with chronic HF have demonstrated the negative
ticenter trial was performed to assess the safety and effi- prognostic impact of low-T3 syndrome and the potential
cacy of KB2115 in lowering the level of serum LDL-cho- improvement of cardiac dysfunction after thyroid hor-
lesterol in patients with hypercholesterolemia who were mone administration (614 621).
already receiving simvastatin or atorvastatin but had per- Experimental and clinical studies in HF have confirmed
sistent LDL levels above 116 mg/dL. Eprotirome induced that a low-T3 state is associated with a return to a fetal gene
a 23% to 29% decrease in LDL-cholesterol and a 22% to program with the development of a hypothyroid-like car-
38% lowering in triglycerides with a 37% to 45% de- diac condition (a shift from -myosin heavy chain to
crease in apolipoprotein A1 and apolipoprotein B (610). -myosin heavy chain and a decreased sarcoplasmic re-
Serum T4 was decreased but remained within the reference ticulum calcium ATPase), leading to a subsequent wors-
range without significant changes in serum TSH or T3. ening in cardiac remodeling and a progressive decline in
There were no observed changes in body weight in patients cardiac function (619 625). These observations have led
receiving eprotirome (610). No potentially deleterious to studies examining the administration of L-T4, L-T3, or
cardiac or bone effects were found using this drug, and thyroid hormone analogs to patients with HF to poten-
only a minor transient increase in liver enzymes was ob- tially improve their prognosis (35, 622 635). The first
served, but unfortunately, cartilage damage in long-term study was performed by Hamilton et al (626) and was a
dog models led to the withdrawal of eprotirome from clin- small nonrandomized trial with administration of an iv
ical trials (611). Sternal cartilage maturation was impaired bolus of L-T3 followed by L-T3 infusion in patients with
with chondrocyte loss and structural changes noted in advanced HF and low-T3 syndrome. Cardiac output im-
multiple areas. The mechanism may relate to overactivity proved significantly 2 hours after T3 administration with
of normal T3 effects on regulation of chrondrocyte a significant decrease in SVR. Interestingly, heart rate did
proliferation. not change and there were no side effects attributable to
the intervention (626).
2. Thyroid hormone and thyroid hormone analogs for
treatment of HF a. Treatment with thyroid hormone. In 2 randomized place-
Recent interest in the potential role of thyroid hor- bo-controlled studies, Moruzzi et al (627, 628) assessed
mones in heart disease is due to the following consider- the cardiovascular short-term and long-term effects of L
ations: 1) evidence of positive effects of thyroid hormone -T4 administered orally at a dose of 0.1 mg/d in patients
on the heart and vascular function (172), 2) proof that with dilated cardiomyopathy. In both of these studies,
even mild thyroid hormone deficiency is associated with a L-T4 significantly improved cardiac function and en-
worse prognosis in cardiac patients (528, 529), and 3) the hanced resting left ventricular ejection fraction, cardiac
fact that important changes in thyroid hormone metabo- output, and exercise capacity (627, 628).
lism have been reported in patients with severe cardiac Malik et al (629) administered iv L-T4 (20 g/h) in 10
disease, including HF, MI, and patients with ischemic consecutive patients with severe systolic HF that had pro-
heart disease undergoing coronary artery bypass (612, gressed to cardiogenic shock. The authors observed a sig-
613). nificant improvement in cardiac index, pulmonary capil-
A low-T3 syndrome occurs in approximately 20% to lary wedge pressure, and mean arterial pressure 24 and 36
30% of patients with HF with a significantly higher inci- hours after L-T4 administration.
dence in patients with New York Heart Association class Iervasi et al (630) administered physiological doses of
L-T3 ( 20 g/d/m body surface) for a period of 96 hours
2
IIII-IV than in those with New York Heart Association
class I-II. The basic pathophysiological mechanisms un- in 6 patients with advanced HF and low-T3 syndrome.
derlying the low-T3 syndrome are related to the reduced These authors observed a progressive reduction in SVR, an
activity of 5-monodeiodinase, resulting in a significant increase in LV ejection fraction, and an improvement of
fall in circulating T3 levels with a consequent increase in cardiac output after T3 administration (630).
rT3 (612, 613). In a more recent placebo-controlled study, the same
This condition has long been interpreted as reflecting a authors administered a 3-day L-T3 infusion in patients
beneficial state in patients with HF because it was consid- with chronic and stable dilated cardiomyopathy and
ered an adaptive process to reduce energy expenditure, low-T3 syndrome (631). They reported an improved car-
negative nitrogen balance, and metabolic demand. How- diac performance that was not associated with an increase
ever, some experimental and clinical evidence has begun to in myocardial O2 consumption and an increase in total

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 491

cardiac work. Moreover, a concomitant improvement of pass surgery (621). Low T3 levels usually develop after
the neuroendocrine pattern with decreased noradrenaline surgery due to the effects on T4 conversion to T3 of cyto-
plasma levels and N-terminal prohormone of brain natri- kines and the administration of corticosteroids. T3 levels
uretic peptide was seen (631). No randomized trials have spontaneously improve 6 to 7 days after surgery. Patients
compared the effects of L-T4 vs L-T3 in similar clinical with lower T3 values in the postoperative period have an
settings. increased risk of death and complications (621).
A recent meta-analysis assessed the results of 13 RCTs
b. 3,5-Diiodothyropropionic acid. 3,5-Diiodothyropropi- of euthyroid adult patients that evaluated the effect of
onic acid (DITPA), a thyroid hormone analog, has cardiac thyroid hormone after cardiac surgery in patients with
inotropic selectivity compared with thyroid hormone with nonthyroidal illnesses (636). T3 therapy was administered
minimal effects on heart rate and metabolic activity (632). iv in 10 studies and orally in 3 studies. High T3 iv doses
This thyromimetic drug has been used to treat CHF with (0.175 0.333 g/kg /h for 6 9 hours were employed in 7
promising preliminary results (35, 633, 634). studies, whereas low doses (0.0275 0.0333 g/kg/ hr for
Long-term administration of DITPA also stimulates 14 to 24 hours), were used in three studies. Both high- and
coronary arteriolar growth without inducing cardiac hy- low-doses of both iv and oral T3 therapy significantly in-
pertrophy by upregulating key angiogenic growth factors creased cardiac index at 4 to 6 hours in pooled analysis in
(635). In a recent RCT in patients with stable CHF, DITPA patients after coronary artery bypass grafting surgery
had no effect on CHF symptoms and was generally poorly (636). The effects on SVR, heart rate, pulmonary capillary
tolerated (35); however, DITPA was able to improve car- wedge pressure, new-onset AF, inotropic function, and
diac index and diastolic function and to decrease SVR. serum TSH and T4 were considered inconclusive (636).
Total and LDL-cholesterol values and triglycerides im- The data were considered insufficient to evaluate the du-
proved and weight loss of an average of 11 pounds was ration of either the stay in the intensive care unit or in the
noted (35). The weight loss may have been related to GI hospital (636). Mortality was not significantly altered by
side effects that were common in the patients treated with high-dose iv T3 therapy and was not able to be assessed for
DITPA (35). DITPA induced suppression of the hypotha- low-dose iv or oral T3 (636).
lamic-pituitary-thyroid axis and had a negative effect on The quality of all of the studies included in this meta-
bone due to increased bone turnover (35). Unfortunately, analysis was considered high because all of the trials were
several adverse events led to its withdrawal from clinical double-blinded and placebo-controlled. However, the
trials. small number of studies and the small number of enrolled
Although all of these studies have documented the po- subjects was responsible for the inconclusive findings of
tential beneficial effects of thyroid hormone and its ana- this otherwise accurate meta-analysis.
logs, large prospective studies are needed to assess the Replacement therapy with thyroid hormone is not rec-
potential therapeutic use of thyroid hormones in treating ommended in postoperative cardiac patients because clinical
and/or preventing HF. The risks of such therapy should be benefits and potential adverse effects are not adequately ad-
underscored because high doses of thyroid hormone, es- dressed. Further prospective studies are required to assess
pecially during infusion or over a prolonged period, may potential beneficial effects of thyroid hormone on cardiac
have detrimental effects on the heart after an initial im- work, myocardial oxygen consumption, ischemia, infarc-
provement in cardiac performance. Future research will be tion, and arrhythmias.
important to establish which drug (T3, T4, or analogs)
could be useful in treating cardiac patients, including the E. Thyroid hormone therapy for the surgical or
best schedule (dosage of thyroid hormone and route of perioperative patient
administration) for therapy. A total of 1500 patients Surgical patients demonstrate the spectrum of altered
would be needed to detect enhanced survival by 20% with thyroid function tests associated with the euthyroid sick
95% confidence, with mortality due to nonthyroidal ill- syndrome or nonthyroidal systemic illness.
ness of 20%. Notwithstanding the very low serum T3 levels that may
be seen in sick patients, we do not believe that there is a role
D. Replacement therapy with thyroid hormone for for T3 treatment for the overwhelming majority of these
postoperative nonthyroidal illnesses patients. An exception to this practice may be the patient
The low-T3 syndrome is also often associated with ma- with head injury and prolonged coma in whom pituitary
jor surgical procedures (612, 613, 621). dysfunction may lead to secondary hypothyroidism. Even
The profile of thyroid hormone measurements show a only relative starvation or nutritional deficiency for a few
progressive decline in T4 and T3 after coronary artery by- days will elicit the euthyroid sick (or low-T3) syndrome.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
492 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

Indeed, a fall in free and total serum T3 with a rise in rT3 F. Treatment of MCT8 deficiency
may be seen after only an overnight fast in preparation for Deficiency of MCT8 causes Allan-Herndon-Dudley
surgery. syndrome of childhood psychomotor retardation marked
It is likely that both a patients sense of well-being and by high serum T3, low T4, and normal or high TSH due to
the general process of wound healing and postoperative impaired access of T4 to the brain. Access of DITPA to the
brain is independent of MCT8, and in a multicenter trial
recovery may be negatively impacted by thyroid hormone
on 4 children affected by MCT8 deficiency, the adminis-
deficiency. Given that thyroid hormone action affects vir-
tration of DITPA (12 mg/kg/d) for 26 to 40 months com-
tually all body systems, it is important in the perioperative
pletely normalized thyroid function tests; reduced SHBG,
patient to restore therapy, either iv or as soon as oral dos-
heart rate, and ferritin; increased cholesterol levels; and
age may be given, to avoid cardiovascular, GI, or renal reduced the hypermetabolism and the tendency for weight
dysfunction (637). The older patient with hypothyroidism loss (412).
and underlying cardiovascular disease may be particularly
at risk, because a worsened hypothyroid state can cause XIV. The Future of Replacement Therapy With
Thyroid Hormone
impaired cardiac contractility with reduced cardiac out-
TSH level is not an optimal marker of adequate thyroid
put, increased angina, reduced blood volume, and in-
hormone replacement therapy in all hypothyroid patients.
creased peripheral vascular resistance. The surgical pa-
In the future, the use of other more sensitive peripheral
tient most at risk is one with preexisting coronary artery markers of thyroid hormone action at tissue levels might
disease or CHF. In general, it is best to undergo general help clinicians to personalize the treatment of thyroid hor-
anesthesia and surgery while euthyroid (637). Hence, thy- mone deficiency.
roid hormone deficiency should be corrected before elec- Large prospective studies studies are necessary to clar-
tive surgical procedures. However, in the case of emergent ify the potential adverse effects of mild subclinical hypo-
procedures including cardiac bypass surgery for impend- thyroidism to improve the treatment of this mild form of
ing MI, it has been proposed that hypothyroid patients can thyroid hormone deficiency, especially in children and el-
safely undergo the surgery and have postoperative resto- derly patients.
ration of euthyroidism. In the latter patient population, Prospective randomized controlled studies are neces-
rigorous attention to intraoperative anesthesia and post- sary to evaluate whether replacement therapy with L-T4
operative ventilation and recovery will be essential for an may improve or completely counteract the adverse effects
of even mild hypothyroidism and its attendant risks.
optimal outcome. Pulmonary function may be impaired
In the future, appropriate RCTs will clarify the poten-
by hypothyroidism in such patients due to reduced lung
tial beneficial effects of combination treatment with T3
volumes or reduced ventilatory excursions from muscle
and T4 vs L-T4 monotherapy especially in thyroidecto-
weakness, reduced exercise capacity, or obesity when
mized patients and in patients with certain polymorphisms
present (637). The slowed metabolism of drugs often ad- in deiodinase activity.
ministered in the perioperative period can also play a role, A long-acting slow-release form of T3 will be required
including analgesics, narcotics, and hypnotics. Impaired to obtain physiological and stable TSH levels with a cir-
renal function in hypothyroidism may be a factor for those cadian T3 rhythm over 24 hours.
drugs largely eliminated via the kidneys. Moreover, the Randomized controlled studies are necessary to evalu-
physician should be alert to the possibility of a coagulopa- ate the potential therapeutic use of thyroid hormones and
thy complicating surgery in the hypothyroid patient (637). their analogs in treating low-T3 syndrome in patients with
Bleeding thought secondary to relative factor VIII defi- HF and nonthyroidal illnesses .
ciency with prolonged partial thromboplastin time has Future studies will clarify whether TSH-suppressive
been described as well as enhanced coagulation due to a therapy of thyroid nodules may prevent the development
prolonged half-life of factors II, VII, and X. In the post- of thyroid cancer. The discovery of thyroid hormone an-
alogs that suppress pituitary TSH secretion with less effect
operative setting, the latter patients may be at increased
on the cardiovascular system and the skeleton might im-
risk for deep venous thrombosis and pulmonary embo-
prove the treatment of patients with aggressive DTC.
lism. Finally, unexplained hypotension in the periopera-
tive patient with a history of hypothyroidism may be a clue
to the presence of partial or subclinical adrenal insuffi- Acknowledgments
ciency which should be considered, ruled out, and/or Address requests for reprints to: Bernadette Biondi, Department of Clin-
treated if confirmed or strongly suspected. ical Medicine and Surgery. University of Naples Federico II, Via S.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 493

Pansini 5, 80131 Naples, Italy. E-mail: bebiondi@unina.it, bebiondi@ 14. Somwaru LL, Arnold AM, Joshi N, Fried LP, Cappola AR.
libero.it. High frequency of and factors associated with thyroid hor-
Disclosure Summary: B.B. has been a speaker at symposia organized mone over-replacement and under-replacement in men and
by Merck Serono and IBSA, is part of the Editorial Board of the European women aged 65 and over. J Clin Endocrinol Metab. 2009;
Journal of Endocrinology, Thyroid, European Thyroid Journal, and
94:13421345.
Frontiers in Thyroid Endocrinology. L.W. serves as a consultant for
15. Okosieme OE, Belludi G, Spittle K, Kadiyala R, Richards
Asurogen, Inc; Genzyme, Inc; and IBSA, Inc, and as Editor-in-Chief of the
Journal of Clinical Endocrinology and Metabolism. J. Adequacy of thyroid hormone replacement in a general
population. QJM. 2011;104:395 401.
16. Liwanpo L, Hershman JM. Conditions and drugs interfer-
ing with thyroxine absorption. Best Pract Res Clin Endo-
References crinol Metab. 2009;23:781792.
17. Abalovich M, Amino N, Barbour LA, et al. Management
1. Singer PA, Cooper DS, Levy EG, et al. Treatment guidelines
of thyroid dysfunction during pregnancy and postpartum:
for patients with hyperthyroidism and hypothyroidism.
an Endocrine Society clinical practice guideline. J Clin En-
Standards of Care Committee, American Thyroid Associ-
docrinol Metab. 2007;92:S1S47.
ation. JAMA. 1995;273:808 812.
18. Stagnaro-Green A, Abalovich M, Alexander E, et al.
2. Surks MI, Ortiz E, Daniels GH, Sawin CT, Col NF, Cobin
Guidelines of the American Thyroid Association for the
RH, Franklyn JA, Hershman JM, Burman KD, Denke MA,
diagnosis and management of thyroid disease during preg-
Gorman C, Cooper RS, Weissman NJ. Subclinical thyroid
nancy and postpartum. Thyroid. 2011;21:10811125.
disease: scientific review and guidelines for diagnosis and
19. De Groot L, Abalovich M, Alexander EK, et al. Manage-
management. JAMA. 2004;291:228 238.
3. Garber JR, Cobin RH, Gharib H, et al; American Associ- ment of thyroid dysfunction during pregnancy and post-
ation of Clinical Endocrinologists and American Thyroid partum: an Endocrine Society clinical practice guideline.
Association Taskforce on Hypothyroidism In Adults. Clin- J Clin Endocrinol Metab. 2012;97:25432565.
ical practice guidelines for hypothyroidism in adults: co- 20. Revised guidelines for neonatal screening programmes for
sponsored by the American Association of Clinical Endo- primary congenital hypothyroidism. Working Group on
crinologists and the American Thyroid Association. Neonatal Screening of the European Society for Paediatric
Thyroid. 2012;22:1200 1235. Endocrinology. Horm Res. 1999;52:49 52.
4. Roti E, Minelli R, Gardini E, Braverman LE. The use and 21. Rose SR, Brown RS, Foley T, et al; American Academy of
misuse of thyroid hormone. Endocr Rev. 1993;14:401 Pediatrics; Section on Endocrinology and Committee on
423. Genetics, American Thyroid Association; Public Health
5. Cerqueira C, Knudsen N, Ovesen L, et al. Doubling in the Committee, Lawson Wilkins Pediatric Endocrine Society.
use of thyroid hormone replacement therapy in Denmark: Update of newborn screening and therapy for congenital
association to iodization of salt? Eur J Epidemiol. 2011; hypothyroidism. Pediatrics. 2006;117:2290 2303.
26:629 635. 22. Balhara B, Misra M, Levitsky LL. Clinical monitoring
6. Mitchell AL, Hickey B, Hickey JL, Pearce SH. Trends in guidelines for congenital hypothyroidism: laboratory out-
thyroid hormone prescribing and consumption in the UK. come data in the first year of life. J Pediatr. 2011;158:532
BMC Public Health. 2009;9:132. 537.
7. IMS Institute for Healthcare Informatics. The use of med- 23. Cooper DS, Doherty GM, Haugen BR, et al; American
icines in the United States: review of 2011. April 2012. IMS Thyroid Association Guidelines Taskforce. Management
Health website. http://www.imshealth.com/ims/Global/ guidelines for patients with thyroid nodules and differen-
Content/Insights/IMS%20Institute%20for%20Healthcare% tiated thyroid cancer. Thyroid. 2006;16:109 1042.
20Informatics/IHII_Medicines. 24. Cooper DS, Doherty GM, Haugen BR, et al; American
8. McDermott MT, Ridgway EC. Subclinical hypothyroid- Thyroid Association (ATA) Guidelines Taskforce on Thy-
ism is mild thyroid failure and should be treated. J Clin roid Nodules and Differentiated Thyroid Cancer. Revised
Endocrinol Metab. 2001;86:4585 4590. American Thyroid Association management guidelines for
9. Biondi B, Cooper DS. The clinical significance of subclin- patients with thyroid nodules and differentiated thyroid
ical thyroid dysfunction. Endocr Rev. 2008;29:76 131. cancer. Thyroid. 2009;19:11671214.
10. Cooper DS, Biondi B. Subclinical thyroid disease. Lancet. 25. Gharib H, Papini E, Paschke R, et al; AACE/AME/ETA
2012;379:11421154. Task Force on Thyroid Nodules. American Association of
11. Klubo-Gwiezdzinska J, Wartofsky L. Thyrotropin blood Clinical Endocrinologists, Associazione Medici Endocri-
levels, subclinical hypothyroidism, and the elderly patient. nologi, and European Thyroid Association medical guide-
Arch Intern Med 2009;169:1949 1951. lines for clinical practice for the diagnosis and management
12. Canaris GJ, Manowitz NR, Mayor G, Ridgway EC. The of thyroid nodules. J Endocrinol Invest. 2010;33:150.
Colorado thyroid disease prevalence study. Arch Intern 26. Biondi B, Wartofsky L. Combination treatment with T4
Med. 2000;160:526 534. and T3: toward personalized replacement therapy in hy-
13. Parle JV, Franklyn JA, Cross KW, Jones SR, Sheppard MC. pothyroidism? J Clin Endocrinol Metab. 2012;97:2256
Thyroxine prescription in the community: serum thyroid 2271.
stimulating hormone level assays as an indicator of under- 27. Celi FS, Zemskova M, Linderman JD, et al. The pharma-
treatment or overtreatment. Br J Gen Pract. 1993;43:107 codynamic equivalence of levothyroxine and liothyronine:
109. a randomized, double blind, cross-over study in thyroid-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
494 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

ectomized patients. Clin Endocrinol (Oxf). 2010;72:709 45. Surks MI, Goswami G, Daniels GH. The thyrotropin ref-
715. erence range should remain unchanged. J Clin Endocrinol
28. Jonklaas J, Davidson B, Bhagat S, Soldin SJ. Triiodothy- Metab. 2005;90:5489 5496.
ronine levels in athyreotic individuals during levothyroxine 46. Baloch Z, Carayon P, Conte-Devolx B, et al; Guidelines
therapy. JAMA. 2008;299:769 777. Committee, National Academy of Clinical Biochemistry.
29. Gullo D, Latina A, Frasca F, Le Moli R, Pellegriti G, Vigneri Laboratory medicine practice guidelines. Laboratory sup-
R. Levothyroxine monotherapy cannot guarantee euthy- port for the diagnosis and monitoring of thyroid disease.
roidism in all athyreotic patients. PLoS One. 2011;6: Thyroid. 2003;13:3126.
e22552. 47. Biondi B. The normal TSH reference range: what has
30. Ito M, Miyauchi A, Morita S, et al. TSH-suppressive doses changed in the last decade? J Clin Endocrinol Metab. 2013;
of levothyroxine are required to achieve preoperative na- 98:3584 3587.
tive serum triiodothyronine levels in patients who have 48. Taylor PN, Razvi S, Pearce SH, Dayan C. A review of the
undergone total thyroidectomy. Eur J Endocrinol. 2012; clinical consequences of variation in thyroid function
167:373378. within the reference range. J Clin Endocrinol Metab. 2013;
31. Panicker V, Saravanan P, Vaidya B, et al. Common vari- 98:35623571.
ation in the DIO2 gene predicts baseline psychological 49. Surks MI, Boucai L. Age- and race-based serum thyrotro-
well-being and response to combina-tion thyroxine plus pin reference limits. J Clin Endocrinol Metab. 2010;95:
triiodothyronine therapy in hypothyroid patients. J Clin 496 502.
Endocrinol Metab. 2009;94:16231629. 50. Boucai L, Hollowell JG, Surks MI. An approach for de-
32. Wiersinga WM, Duntas L, Fadeyev V, Nygaard B, Vander- velopment of age-, gender-, and ethnicity-specific thyro-
pump MP. ETA guidelines: the use of L-T4 L-T3 in the tropin reference limits. Thyroid. 2011;21:511.
treatment of hypothyroidism. Eur Thyroid J. 2012;1:55 51. Surks MI, Hollowell JG. Age-specific distribution of serum
71. thyrotropin and antithyroid antibodies in the U.S. popu-
33. Bryzgalova G, Effendic S, Khan A, et al. Anti-obesity, anti- lation: implications for the prevalence of subclinical hypo-
diabetic, and lipid lowering effects of the thyroid receptor thyroidism. J Clin Endocrinol Metab. 2007;92:4575
-subtype selective agonist KB-141. J Steroid Biochem Mol 4582.
Biol. 2008;111:262267. 52. Biondi B. Thyroid and obesity: an intriguing relationship.
34. Ladenson PW, Kristensen JD, Ridgway EC, et al. Use of the J Clin Endocrinol Metab. 2010;95:3614 3617.
thyroid hormone analogue eprotirome in statin-treated 53. Teng W, Shan Z, Teng X, et al. Effect of iodine intake on
dyslipidemia. N Engl J Med. 2010;362:906 916. thyroid diseases in China. N Engl J Med. 2006;354:2783
35. Goldman S, McCarren M, Morkin E, et al. DITPA (3,5- 2793.
diiodothyropropionic acid), a thyroid hormone analog to 54. Krassas GE, Poppe K, Glinoer D. Thyroid function and
treat heart failure: phase II trial veterans affairs cooperative human reproductive health. Endocr Rev. 2010;31:702
study. Circulation. 2009;119:30933100. 755.
36. Roberts CG, Ladenson PW. Hypothyroidism. Lancet. 55. Adler SM, Wartofsky L. The nonthyroidal illness syn-
2004;363:793 803. drome. Endocrinol Metab Clin North Am. 2007;36:657
37. Samuels MH, Ridgway EC. Central hypothyroidism. En- 672, vi.
docrinol Metab Clin North Am. 1992;21:903919. 56. Arnaud-Lopez L, Usala G, Ceresini G, et al. Phosphodies-
38. Lania A, Persani L, Beck-Peccoz P. 2008 Central hypothy- terase 8B gene variants are associated with serum TSH
roidism. Pituitary. 2008;11:181186. levels and thyroid function. Am J Hum Genet. 2008;82:
39. Ord WM. On myxedema, a term proposed to be applied to 1270 1280.
an essential condition in the cretinoid affection occa- 57. Medici M, van der Deure WM, Verbiest M, et al. A large-
sionally observed in middle-aged women. Med Chir Trans. scale association analysis of 68 thyroid hormone pathway
1878;61:5778. genes with serum TSH and FT4 levels. Eur J Endocrinol.
40. Wartofsky L. Myxedema coma. Endocrinol Metab Clin 2011;164:781788.
North Am. 2006;35:687 698. viiviii. 58. Taylor PN, Panicker V, Sayers A, et al. A meta-analysis of
41. Rastogi MV, LaFranchi SH. Congenital hypothyroidism. the associations between common variation in the PDE8B
Orphanet J Rare Dis. 2010;5:17. gene and thyroid hormone parameters, including assess-
42. Hollowell JG, Staehling NW, Flanders WD, et al. Serum ment of longitudinal stability of associations over time and
TSH, T4, and thyroid antibodies in the United States pop- effect of thyroid hormone replacement. Eur J Endocrinol.
ulation (1988 to 1994): National Health and Nutrition 2011;164:773780.
Examination Survey (NHANES III). J Clin Endocrinol 59. Donda A, Lemarchand-Braud T. Aging alters the activity
Metab. 2002;87:489 499. of 5-deiodinase in the adenohypophysis, thyroid gland,
43. Hamilton TE, Davis S, Onstad L, Kopecky KJ. Thyrotro- and liver of the male rat. Endocrinology. 1989;124:1305
pin levels in a population with no clinical, autoantibody, or 1309.
ultrasonographic evidence of thyroid disease: implications 60. Duntas LH, Biondi B. The interconnections between obe-
for the diagnosis of subclinical hypothyroidism. J Clin En- sity, thyroid function, and autoimmunity: the multifold
docrinol Metab. 2008;93:1224 1230. role of leptin. Thyroid. 2013;23(6):646 653.
44. Wartofsky L, Dickey RA. The evidence for a narrower thy- 61. Glinoer D. The regulation of thyroid function in pregnan-
rotropin reference range is compelling. J Clin Endocrinol cy: pathways of endocrine adaptation from physiology to
Metab. 2005;90:54835488. pathology. Endocr Rev. 1997;18:404 433.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 495

62. LaFranchi S. Approach to the diagnosis and treatment of 80. Goldner WS, Sandler DP, Yu F, Hoppin JA, Kamel F, Levan
neonatal hypothyroidism. J Clin Endocrinol Metab. 2011; TD. Pesticide use and thyroid disease among women in the
96:2959 2967. Agricultural Health Study. Am J Epidemiol. 2010;171:
63. Mandel SJ, Hermos RJ, Larson CA, Prigozhin AB, Rojas 455 464.
DA, Mitchell ML. Atypical hypothyroidism and the very 81. Persani L. Central hypothyroidism: pathogenic, diagnos-
low birthweight infant. Thyroid. 2000;10:693 695. tic, and therapeutic challenges. J Clin Endocrinol Metab.
64. Jassam N, Visser TJ, Brisco T, Bathia D, McClean P, Barth 2012;97:3068 3078.
JH. Consumptive hypothyroidism: a case report and re- 82. Carl A, Blow Pedersen I, et al. Smoking cessation is fol-
view of the literature. Ann Clin Biochem. 2011;48:186 lowed by a sharp but transient rise in the incidence of overt
189. autoimmune hypothyroidisma population-based, case-
65. Zimmermann MB. Iodine deficiency. Endocr Rev. 2009; control study. Clin Endocrinol (Oxf). 2012;77:764 772.
30:376 408. 83. Visser TJ. Thyroid hormone transporters and resistance.
66. Teng X, Shan Z, Chen Y, et al. More than adequate iodine Endocr Dev. 2013;24:110.
intake may increase subclinical hypothyroidism and auto- 84. Refetoff S, Weiss RE, Usala SJ. The syndromes of resistance
immune thyroiditis: a cross-sectional study based on two to thyroid hormone. Endocr Rev. 1993;14:348 399.
Chinese communities with different iodine intake levels. 85. Bochukova E, Schoenmakers N, Agostini M, et al. A mu-
Eur J Endocrinol. 2011;164:943950. tation in the thyroid hormone receptor gene. N Engl
67. Michels AW, Gottlieb PA. Autoimmune polyglandular J Med. 2012;366:243249.
syndromes. Nat Rev Endocrinol. 2010;6:270 277. 86. van Mullem A, van Heerebeek R, Chrysis D, et al. Clinical
68. De Moraes AV, Pedro AB, Romaldini JH. Spontaneous phenotype and mutant TR1. N Engl J Med. 2012;366:
hypothyroidism in the follow up of Graves hyperthyroid 14511453.
patients treated with antithyroid drugs. South Med J. 2006; 87. van Mullem AA, Chrysis D, Eythimiadou A, et al. Clinical
99:1068 1072. phenotype of a new type of thyroid hormone resistance
69. Metso S, Jaatinen P, Huhtala H, Luukkaala T, Oksala H, caused by a mutation of the TR1 receptor: consequences
Salmi J. Long-term follow-up study of radioiodine treat- of L-T4 treatment. J Clin Endocrinol Metab. 2013;98:
ment of hyperthyroidism. Clin Endocrinol (Oxf). 2004; 3029 3038.
61:641 648. 88. Foley TP Jr. Acquired hypothyroidism in infants, children
70. Ross DS. Radioiodine therapy for hyperthyroidism. and adolescents. In: Braverman LE, Utiger RD, eds. Wer-
N Engl J Med. 2011;364:542550. ner, Ingbars The Thyroid. 8th ed. Philadelphia, PA: Lip-
71. Ceccarelli C, Bencivelli W, Vitti P, Grasso L, Pinchera A. pincott Williams, Wilkins; 2000:983988.
Outcome of radioiodine-131 therapy in hyperfunctioning 89. Foley TP Jr. Hypothyroidism. In: Hoekelman RA, Adam
thyroid nodules: a 20 years retrospective study. Clin En- HM, Nelson NM, Weitzman ML, Wilson MH, eds. Pri-
docrinol (Oxf). 2005;62:331335. mary Pediatric Care. 4th ed. St. Louis, MO: Mosby; 2001:
72. Holm LE, Lundell G, Israelsson A, Dahlqvist I. Incidence of 15611565.
hypothyroidism occurring long after iodine-131 therapy 90. Tunbridge WM, Evered DC, Hall R, et al. The spectrum of
for hyperthyroidism. J Nucl Med. 1982;23:103107. thyroid disease in a community: the Whickham survey.
73. Vaiman M, Nagibin A, Hagag P, Kessler A, Gavriel H. Clin Endocrinol (Oxf). 1977;7:481 493.
Hypothyroidism following partial thyroidectomy. Otolar- 91. Vanderpump MP, Tunbridge WM, French JM, et al. The
yngol Head Neck Surg. 2008;138:98 100. incidence of thyroid disorders in the community: a twenty-
74. Johner A, Griffith OL, Walker B, et al. Detection and man- year follow-up of the Whickham Survey. Clin Endocrinol
agement of hypothyroidism following thyroid lobectomy: (Oxf). 1995;43:55.
evaluation of a clinical algorithm. Ann Surg Oncol. 2011; 92. Mariotti S, Franceschi C, Cossarizza A, Pinchera A. The
18:2548 2554. aging thyroid. Endocr Rev. 1995;16:686 715.
75. Vogelius IR, Bentzen SM, Maraldo MV, Petersen PM, 93. Dez JJ, Iglesias P, Burman KD. Spontaneous normaliza-
Specht L. Risk factors for radiation-induced hypothyroid- tion of thyrotropin concentrations in patients with sub-
ism: a literature-based meta-analysis. Cancer. 2011;117: clinical hypothyroidism. J Clin Endocrinol Metab. 2005;
5250 5260. 90:4124 4127.
76. Boice JD Jr. Thyroid disease 60 years after Hiroshima and 94. Li Y, Teng D, Shan Z, et al. Antithyroperoxidase and an-
20 years after Chernobyl. JAMA. 2006;295:1060 1102. tithyroglobulin antibodies in a five-year follow-up survey
77. Quach A, Ji L, Mishra V, et al. Thyroid and hepatic func- of populations with different iodine intakes. J Clin Endo-
tion after high-dose 131 I-metaiodobenzylguanidine (131 crinol Metab. 2008;93:17511757.
I-MIBG) therapy for neuroblastoma. Pediatr Blood Can- 95. Walsh JP, Bremner AP, Feddema P, Leedman PJ, Brown SJ,
cer. 2011;56:191201. OLeary P. Thyrotropin and thyroid antibodies as predic-
78. Bishton MJ, Leahy MF, Hicks RJ, Turner JH, McQuillan tors of hypothyroidism: a 13-year, longitudinal study of a
AD, Seymour JF. Repeat treatment with iodine-131-ritux- community-based cohort using current immunoassay tech-
imab is safe and effective in patients with relapsed indolent niques. J Clin Endocrinol Metab. 2010;95:10951104.
B-cell non-Hodgkins lymphoma who had previously re- 96. svold BO, Vatten LJ, Midthjell K, Bjro T. Serum TSH
sponded to iodine-131-rituximab. Ann Oncol. 2008;19: within the reference range as a predictor of future hypo-
1629 1633. thyroidism and hyperthyroidism: 11-year follow-up of the
79. Surks MI, Sievert R. Drugs and thyroid function. N Engl HUNT Study in Norway. J Clin Endocrinol Metab. 2012;
J Med. 1995;333:1688 1694. 97:9399.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
496 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

97. Klubo-Gwiezdzinska J, Burman KD, Van Nostrand D, hypothyroidism. Best Pract Res Clin Endocrinol Metab.
Wartofsky L. Levothyroxine treatment in pregnancy: in- 2009;23:793 800.
dications, efficacy, and therapeutic regimen. J Thyroid Res. 116. Cappelli C, Rotondi M, Pirola I, et al. 2009 TSH-lowering
2011;2011:843591. effect of metformin in type 2 diabetic patients: differences
98. Allan WC, Haddow JE, Palomaki GE, et al. Maternal thy- between euthyroid, untreated hypothyroid, and euthyroid
roid deficiency and pregnancy complications: implications on L-T4 therapy patients .Diabetes Care. 2009;32:1589
for population screening. J Med Screen. 2000;7:127130. 1590.
99. Canaris GJ, Steiner JF, Ridgway EC. Do traditional symp- 117. Cappelli C, Rotondi M, Pirola I, et al. Thyreotropin levels
toms of hypothyroidism correlate with biochemical dis- in diabetic patients on metformin treatment. Eur J Endo-
ease? J Gen Intern Med. 1997;12:544 550. crinol. 2012;167:261265.
100. Roberts LM, Pattison H, Roalfe A, et al. Is subclinical 118. Spencer C, Eigen A, Shen D, et al. Specificity of sensitive
thyroid dysfunction in the elderly associated with depres- assays of thyrotropin (TSH) used to screen for thyroid dis-
sion or cognitive dysfunction? Ann Intern Med. 2006;145: ease in hospitalized patients. Clin Chem. 1987;33:1391
573581. 1396.
101. Gussekloo J, van Exel E, de Craen AJ, Meinders AE, 119. Stricker R, Echenard M, Eberhart R, et al. Evaluation of
Frlich M, Westendorp RG. Thyroid status, disability and maternal thyroid function during pregnancy: the impor-
cognitive function, and survival in old age. JAMA. 2004; tance of using gestational age-specific reference intervals.
292:25912599. Eur J Endocrinol. 2007;157:509 514.
102. Patel KV. Epidemiology of anemia in older adults. Semin 120. Kahric-Janicic N, Soldin SJ, Soldin OP, West T, Gu J, Jonk-
Hematol. 2008;45:210 217. laas J. Tandem mass spectrometry improves the accuracy
103. Samuels MH. Subclinical thyroid disease in the elderly. of free thyroxine measurements during pregnancy. Thy-
Thyroid. 1998;8:803 813. roid. 2007;17:303311.
104. Nanchen D, Gussekloo J, Westendorp RG, et al; PROSPER 121. Corbetta C, Weber G, Cortinovis F, et al. A 7-year expe-
Group. Subclinical thyroid dysfunction and the risk of rience with low blood TSH cutoff levels for neonatal
heart failure in older persons at high cardiovascular risk.
screening reveals an unsuspected frequency of congenital
J Clin Endocrinol Metab. 2012;97:852 861.
hypothyroidism (CH). Clin Endocrinol (Oxf). 2009;71:
105. Levy EG. Thyroid disease in the elderly. Med Clin North
739 745.
Am. 1991;75:151167.
122. Hinton CF, Harris KB, Borgfeld L, et al. Trends in inci-
106. Tachman ML, Guthrie GP. Hypothyroidism: diversity of
dence rates of congenital hypothyroidism related to select
presentation. Endocr Rev. 1984;5:456 465.
demographic factors: data from the United States, Califor-
107. Bemben DA, Hamm RM, Morgan L, Winn P, Davis A,
nia, Massachusetts, New York, and Texas. Pediatrics.
Barton E. Thyroid disease in the elderly. Part 2. Predict-
2010;125:S37S47.
ability of subclinical hypothyroidism. J Fam Pract. 1994;
123. Woo HC, Lizarda A, Tucker R, et al. Congenital hypo-
38:583588.
thyroidism with a delayed thyroid-stimulating hormone
108. Billewicz WZ, Chapman RS, Crooks J, et al. Statistical
elevation in very premature infants: incidence and growth
methods applied to the diagnosis of hypothyroidism. QJM.
1969;38:255266. and developmental outcomes. J Pediatr. 2011;158:538
109. Seshadri MS, Samuel BU, Kanagasabapathy AS, Cherian 542.
AM. Clinical scoring system for hypothyroidism: is it use- 124. LaFranchi SH 2010 Newborn screening strategies for con-
ful? J Gen Intern Med. 1989;4:490 492. genital hypothyroidism: an update. J Inherit Metab Dis
110. Zulewski H, Mller B, Exer P, Miserez AR, Staub JJ. Es- 33(Suppl 2):S225S233.
timation of tissue hypothyroidism by a new clinical score: 125. Murray GR. Note on the treatment of myxoedema by hy-
evaluation of patients with various grades of hypothyroid- podermic injections of an extract of the thyroid gland of a
ism and controls. J Clin Endocrinol Metab. 1997;82:771 sheep. BMJ. 1891;2:796 797.
776. 126. Lindholm J, Laurberg P. Hypothyroidism and thyroid sub-
111. van de Ven AC, Netea-Maier RT, de Vegt F, et al. Is there stitution: historical aspects. J Thyroid Res. 2011;2011:
a relationship between fatigue perception and the serum 809341.
levels of thyrotropin and free thyroxine in euthyroid sub- 127. Braverman LE, Ingbar SH, Sterling K. Conversion of thy-
jects? Thyroid. 2012;22:1236 1243. roxine (T4) to triiodothyronine (T3) in athyreotic human
112. Danese MD, Ladenson PW, Meinert CL, Powe NR. Effect subjects. J Clin Invest. 1970;49:855 864.
of thyroxine therapy on serum lipoproteins in patients with 128. Smith RN, Taylor SA, Massey JC. Controlled clinical trial
mild thyroid failure: a quantitative review of the literature. of combined triiodothyronine and thyroxine in the treat-
J Clin Endocrinol Metab. 2000;85:29933001. ment of hypothyroidism. BMJ. 1970;4:145148.
113. Ladenson PW, Singer PA, Ain KB, et al. American Thyroid 129. Brent G. Mechanisms of thyroid hormone action. J Clin
Association guidelines for detection of thyroid dysfunc- Invest. 2012;122:30353043.
tion. Arch Intern Med. 2000;160:15731575. 130. Andersen S, Pedersen KM, Bruun NH, Laurberg P. Nar-
114. Karmisholt J, Andersen S, Laurberg P. Variation in thyroid row individual variations in serum T4 and T3 in normal
function in subclinical hypothyroidism: importance of subjects: a clue to the understanding of subclinical thyroid
clinical follow-up and therapy. Eur J Endocrinol. 2011; disease. J Clin Endocrinol Metab. 2002;87:1068 1072.
164:317323. 131. Pilo A, Iervasi G, Vitek F, Ferdeghini M, Cazzuola F, Bi-
115. Haugen BR. Drugs that suppress TSH or cause central anchi R. Thyroidal and peripheral production of 3,5,3-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 497

triiodothyronine in humans by multicompartmental anal- 148. US Food and Drug Administration. FDA orange book:
ysis. Am J Physiol. 1990;258:E715E726. approved drug products with therapeutic equivalence eval-
132. Bianco AC, Salvatore D, Gereben B, Berry MJ, Larsen PR. uations for thyroxine. U.S. FDA website. http://www.
Biochemistry, cellular and molecular biology, and physi- accessdata.fda.gov/scripts/cder/ob/docs/tempai.cfm. Accessed
ological roles of the iodothyronine selenodeiodinases. En- January 12, 2013.
docr Rev. 2002;23:38 89. 149. Hennessey JV. Levothyroxine a new drug? Since when?
133. Maia AL, Kim BW, Huang SA, Harney JW, Larsen PR. How could that be? Thyroid. 2003;13:279 282.
Type 2 iodothyronine deiodinase is the major source of 150. American Thyroid Association, Endocrine Society, Amer-
plasma T3 in euthyroid humans. J Clin Invest. 2005;115: ican Association of Clinical Endocrinologists. Joint state-
2524 2533. ment on the U. S. Food and Drug Administrations decision
134. Hoang TD, Olsen CH, Mai VQ, Clyde PW, Shakir MK. regarding bioequivalence of levothyroxine sodium. Thy-
Desiccated thyroid extract compared with levothyroxine roid. 2004;14:486.
in the treatment of hypothyroidism: a randomized, double- 151. Hennessey JV, Malabanan AO, Haugen BR, Levy EG. Ad-
blind, crossover study. J Clin Endocrinol Metab. 2013;98: verse event reporting in patients treated with levothyrox-
19821990. ine: results of the Pharmacovigilance Task Force survey of
135. Penny R, Frasier SD. Elevated serum concentrations of tri- the American Thyroid Association, American Association
iodothyronine in hypothyroid patients. Values for patients of Clinical Endocrinologists, and The Endocrine Society.
receiving USP thyroid. Am J Dis Child. 1980;134:16 18. Endocr Pract. 2010;16:357370.
136. Jha S, Waghdhare S, Reddi R, Bhattacharya P. Thyroid 152. Wartofsky L. Levothyroxine: therapeutic use and regula-
storm due to inappropriate administration of a com- tory issues related to bioequivalence. Expert Opin Phar-
pounded thyroid hormone preparation successfully macother. 2002;3:727732.
treated with plasmapheresis. Thyroid. 2012;22:1283 153. Eisenberg M, Distefano JJ. TSH-based protocol, tablet in-
1286. stability, and absorption effects on L-T4 bioequivalence.
137. Hays MT. Localization of human thyroxine absorption. Thyroid. 2009;19:103110.
Thyroid. 1991;1:241248.
154. Blakesley V, Awni W, Locke C, Ludden T, Granneman
138. Wenzel KW, Kirschsieper HE. Aspects of the absorption of
GR, Braverman LE. Are bioequivalence studies of levo-
oral L-thyroxine in normal man. Metabolism. 1977;26:
thyroxine sodium formulations in euthyroid volunteers re-
1 8.
liable? Thyroid. 2004;14:191200.
139. Read DG, Hays MT, Hershman JM. Absorption of oral
155. Carr D, McLeod DT, Parry G, Thornes HM. Fine adjust-
thyroxine in hypothyroid and normal man. J Clin Endo-
ment of thyroxine replacement dosage: comparison of the
crinol Metab. 1970;30:798 799.
thyrotrophin releasing hormone test using a sensitive thy-
140. Yao X, Forte JG. Cell biology of acid secretion by the pa-
rotrophin assay with measurement of free thyroid hor-
rietal cell. Annu Rev Physiol. 2003;65:103131.
mones and clinical assessment. Clin Endocrinol (Oxf).
141. Lahner E, Annibale B, Delle Fave G. Systematic review:
1988;28:325333.
Helicobacter pylori infection and impaired drug absorp-
156. Dong BJ, Hauck WW, Gambertoglio JG, et al. Bioequiva-
tion. Aliment Pharmacol Ther. 2009;29:379 386.
142. Centanni M, Gargano L, Canettieri G, et al. Thyroxine in lence of generic and brand-name levothyroxine products in
goiter, Helicobacter pylori infection, and chronic gastritis. the treatment of hypothyroidism. JAMA. 1997;277:1205
N Engl J Med. 2006;354:17871795. 1213.
143. Benvenga S, Bartolone L, Squadrito S, Lo Giudice F, Tri- 157. Mayor GH, Orlando T, Kurtz NM. Limitations of levo-
marchi F. Delayed intestinal absorption of levothyroxine. thyroxine bioequivalence evaluation: analysis of an at-
Thyroid. 1995;5:249 253. tempted study. Am J Ther. 1995;2:417 432.
144. Liel Y, Harman-Boehm I, Shany S. Evidence for a clinically 158. Escalante DA, Arem N, Arem R. Assessment of inter-
important adverse effect of fiber-enriched diet on the bio- changeability of two brands of levothyroxine preparations
availability of levothyroxine in adult hypothyroid patients. with a third-generation TSH assay. Am J Med. 1995;98:
J Clin Endocrinol Metab. 1996;81:857 859. 374 378.
145. Food and Drug Administration Center for Drug Evalua- 159. Grant TM, Zhu T, Brandquist C, Teuscher NS, Lamson
tion and Research. Guidance for industry: levothyroxine MJ. Randomized, open-label, single-dose, two-way cross-
sodium products enforcement of August 14, 2001: com- over study to determine the relative bioavailability of a
pliance date and submission of new applications. Wash- reformulated levothyroxine tablet compared with the com-
ington, DC: US Department of Health and Human Services mercial tablet (Levoxyl) in healthy volunteers. Clin Phar-
Food and Drug Administration; 2001. macol Ther. 2009;85:PIII-85 S95.
146. US Food and Drug Administration. Bioequivalence AB2 160. Lomenick JP, Wang L, Ampah SB, Saville BR, Greenwald
ratings of L-thyroxine products. U.S. FDA website. www. FI. Generic levothyroxine compared with synthroid in
accessdata.fda.gov/scripts/cder/drugsatfda/index.cfm. young children with congenital hypothyroidism. J Clin En-
Accessed May 17, 2005. docrinol Metab. 2013;98:653 658.
147. US Food and Drug Administration. Guidance on narrow- 161. Carswell JM, Gordon JH, Popovsky E, Hale A, Brown RS.
ing (95% rule) permissible variance in L-thyroxine tablet Generic and brand-name L-thyroxine are not bioequivalent
content. U.S. FDA website. http://www.fda.gov/cder/drug/ for children with severe congenital hypothyroidism. J Clin
infopage/levothyroxine/qa.htm. Accessed October 22, Endocrinol Metab. 2013;98:610 617.
2007. 162. Hennessey JV. Generic vs name brand L-thyroxine prod-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
498 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

ucts interchangeable or still not? J Clin Endocrinol Metab. 178. Haddow JE, Palomaki GE, Allan WC, et al. Maternal thy-
2013;98:511514. roid deficiency during pregnancy and subsequent neuro-
163. Yavuz S, Linderman JD, Smith S, Zhao X, Pucino F, Celi psychological development of the child. N Engl J Med.
FS. The dynamic pituitary response to escalating-dose 1999;341:549 555.
TRH stimulation test in hypothyroid patients treated with 179. Vlzke H, Schwahn C, Wallaschofski H, Drr M. The as-
liothyronine or levothyroxine replacement therapy. J Clin sociation of thyroid dysfunction with all-cause and circu-
Endocrinol Metab. 2013;98:E862E826. latory mortality: is there a causal relationship? J Clin En-
164. Celi FS, Zemskova M, Linderman JD, et al. Metabolic ef- docrinol Metab. 2007;92:24212429.
fects of liothyronine therapy in hypothyroidism: a random- 180. Singh S, Duggal J, Molnar J, Maldonado F, Barsano CP,
ized, double-blind, crossover trial of liothyronine versus Arora R. Impact of subclinical thyroid disorders on coro-
levothyroxine. J Clin Endocrinol Metab. 2011;96:3466 nary heart disease, cardiovascular and all-cause mortality:
3474. a meta-analysis. Int J Cardiol. 2008;125:41 48.
165. Colucci P, DAngelo P, Mautone G, Scarsi C, Ducharme 181. Haentjens P, Van Meerhaeghe A, Poppe K, Velkeniers B.
MP. Pharmacokinetic equivalence of a levothyroxine so- Subclinical thyroid dysfunction and mortality: an estimate
dium soft capsule manufactured using the new Food and of relative and absolute excess all-cause mortality based on
Drug Administration potency guidelines in healthy volun- time-to-event data from cohort studies. Eur J Endocrinol.
teers under fasting conditions. Ther Drug Monit. 2011; 2008;159:329 341.
33:355361. 182. Razvi S, Shakoor A, Vanderpump M, Weaver JU, Pearce
166. Leggio GM, Incognito T, Privitera G, Marano MR, Drago SH. The influence of age on the relationship between sub-
F. Comparative bioavailability of different formulations of clinical hypothyroidism and ischemic heart disease: a
levothyroxine and liothyronine in healthy volunteers. J En- metaanalysis. J Clin Endocrinol Metab. 2008;93:2998
docrinol Invest. 2006;29:RC35RC38. 3007.
167. Pabla D, Akhlaghi F, Zia H. A comparative pH-dissolution 183. Ochs N, Auer R, Bauer DC, et al. Meta-analysis: subclin-
profile study of selected commercial levothyroxine prod- ical thyroid dysfunction and the risk for coronary heart
ucts using inductively coupled plasma mass spectrometry. disease and mortality. Ann Intern Med. 2008;148:832
Eur J Pharm Biopharm. 2009;72:105110. 845.
168. Vita R, Saraceno G, Trimarchi F, Benvenga S. A novel 184. Rodondi N, den Elzen WP, Bauer DC, et al; Thyroid Stud-
formulation of L-thyroxine (L-T4) reduces the problem of ies Collaboration. Subclinical hypothyroidism and the risk
L-T4 malabsorption by coffee observed with traditional of coronary heart disease and mortality. JAMA. 2010;304:
tablet formulations. Endocrine. 2013;43:154 160. 13651374.
169. Virili C, Santaguida MG, Cellini M, Del Duca SC, Gargano 185. Biondi B. Mechanisms in endocrinology: heart failure and
L, Centanni M. 2012 Pilot study with softgel thyroxine thyroid dysfunction. Eur J Endocrinol. 2012;167:609
preparation in the treatment of patients with T4 malab- 618.
sorption due to gastric disorders. Endocr Rev. 2013;34: 186. Rodondi N, Newman AB, Vittinghoff E, de Rekeneire N,
OR50 OR54. Satterfield S, Harris TB, Bauer DC. Subclinical hypothy-
170. Walsh JP, Ward LC, Burke V, et al. Small changes in thy- roidism and the risk of heart failure, other cardiovascular
roxine dosage do not produce measurable changes in hy- events, and death. Arch Intern Med. 2005;165:2460
pothyroid symptoms, well-being, or quality of life: results 2466.
of a double-blind, randomized clinical trial. J Clin Endo- 187. Rodondi N, Bauer DC, Cappola AR, et al. Subclinical thy-
crinol Metab. 2006;91:2624 2630. roid dysfunction, cardiac function, and the risk of heart
171. Biondi B, Palmieri EA, Lombardi G, Fazio S. Effects of failure. The Cardiovascular Health Study. J Am Coll Car-
subclinical thyroid dysfunction on the heart. Ann Intern diol. 2008;52:11521159.
Med. 2002;137:904 914. 188. Gencer B, Collet TH, Virgini V, et al; Thyroid Studies
172. Fazio S, Palmieri EA, Lombardi G, Biondi B. Effects of Collaboration. Subclinical thyroid dysfunction and the
thyroid hormone on the cardiovascular system. Recent risk of heart failure events: an individual participant data
Prog Horm Res. 2004;59:3150. analysis from 6 prospective cohorts. Circulation. 2012;
173. Biondi B, Klein I. Hypothyroidism as a risk factor for car- 126:1040 1049.
diovascular disease. Endocrine. 2004;24:113. 189. Pearce EN. Update in lipid alterations in subclinical hypo-
174. Becker C. Hypothyroidism and atherosclerotic heart dis- thyroidism. J Clin Endocrinol Metab. 2012;97:326 333.
ease: pathogenesis, medical management, and the role of 190. Razvi S, Weaver JU, Vanderpump MP, Pearce SH. The
coronary artery bypass surgery. Endocr Rev. 1985;6:432 incidence of ischemic heart disease and mortality in people
440. with subclinical hypothyroidism: reanalysis of the Whick-
175. Cappola AR, Ladenson PW. Hypothyroidism and athero- ham Survey cohort. J Clin Endocrinol Metab. 2010;95:
sclerosis. J Clin Endocrinol Metab. 2003;88:2438 2444. 1734 1740.
176. Duntas LH, Wartofsky L. Cardiovascular risk and sub- 191. McQuade C, Skugor M, Brennan DM, Hoar B, Stevenson
clinical hypothyroidism: focus on lipids and new emerging C, Hoogwerf BJ. Hypothyroidism and moderate subclin-
risk factors. What is the evidence? Thyroid. 2007;17: ical hypothyroidism are associated with increased all-cause
10751084. mortality independent of coronary heart disease risk fac-
177. Thvilum M, Brandt F, Brix TH, Hegeds L. A review of the tors: a PreCIS database study. Thyroid. 2011;21:837 843.
evidence for and against increased mortality in hypothy- 192. Biondi B. Cardiovascular effects of mild hypothyroidism.
roidism. Nat Rev Endocrinol. 2012;8:417 424. Thyroid. 2007;17:625 630.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 499

193. Razvi S, Weaver JU, Butler TJ, Pearce SH. Levothyroxine 210. Bearcroft CP, Toms GC, Williams SJ, Noonan K, Monson
treatment of subclinical hypothyroidism, fatal and nonfa- JP. Thyroxine replacement in post-radioiodine hypothy-
tal cardiovascular events, and mortality. Arch Intern Med. roidism. Clin Endocrinol (Oxf). 1991;34:115118.
2012;172;811 817. 211. Teixeira PF, Reuters VS, Ferreira MM, et al. Treatment of
194. Rodondi N. Spotlight: investigating the relationships be- subclinical hypothyroidism reduces atherogenic lipid lev-
tween thyroid dysfunction and cardiovascular disease, and els in a placebo-controlled double-blind clinical trial.
testing the impact of treatment of thyroid dysfunction on Horm Metab Res. 2008;40:50 55.
clinical outcomes. Circulation. 2012;126:f127f129. 212. Ferretti E, Persani L, Jaffrain-Rea ML, Giambona S, Tam-
195. Mooijaart SP, IEMO 80-plus Thyroid Trial Collabora- burrano G, Beck-Peccoz P. Evaluation of the adequacy of
tion. Subclinical thyroid disorders. Lancet. 2012;380:335; levothyroxine replacement therapy in patients with central
author reply 336 337. hypothyroidism. J Clin Endocrinol Metab. 1999;84:924
196. Jonklaas J. Sex and age differences in levothyroxine dosage 929.
requirement. Endocr Pract. 2010;16:7179. 213. Yassa L, Marqusee E, Fawcett R, Alexander EK. Thyroid
197. Sawin CT, Herman T, Molitch ME, London MH, Kramer hormone early adjustment in pregnancy (the THERAPY)
SM. Aging and the thyroid. Decreased requirement for thy- trial. J Clin Endocrinol Metab. 2010;95:3234 3241.
roid hormone in older hypothyroid patients. Am J Med. 214. Loh JA, Wartofsky L, Jonklaas J, Burman KD. The mag-
1983;75:206 209. nitude of increased levothyroxine requirements in hypo-
198. Arafah BM. Increased need for thyroxine in women with thyroid pregnant women depends upon the etiology of the
hypothyroidismduring estrogen therapy. N Engl J Med. hypothyroidism. Thyroid. 2009;19:269 275.
2001;344:17431749. 215. Trouillier S, Delvaux I, Ranc N, Andr M, Voinchet H,
199. Devdhar M, Drooger R, Pehlivanova M, Singh G, Jonklaas Aumatre O. Nephrotic syndrome: dont forget to search
J. Levothyroxine replacement doses are affected by gender for hypothyroidism [in French]. Rev Med Interne. 2008;
and weight, but not age. Thyroid. 2011;21:821 827. 29:139 144.
200. Mandel SJ, Brent GA, Larsen PR. Levothyroxine therapy 216. Junglee NA, Scanlon MF, Rees DA. Increasing thyroxine
in patients with thyroid disease. Ann Intern Med. 1993; requirements in primary hypothyroidism: dont forget the
119:492502. urinalysis! J Postgrad Med. 2006;52:201203.
201. Sukumar R, Agarwal A, Gupta S, et al. Prediction of LT4 217. Torlontano M, Durante C, Torrente I, et al. Type 2 deio-
replacement dose to achieve euthyroidism in subjects un- dinase polymorphism (threonine 92 alanine) predicts L-
dergoing total thyroidectomy for benign thyroid disorders. thyroxine dose to achieve target thyrotropin levels in thy-
World J Surg. 2010;34:527531. roidectomized patients. J Clin Endocrinol Metab. 2008;
202. Santini F, Pinchera A, Marsili A, et al. Lean body mass is 93:910 913.
a major determinant of levothyroxine dosage in the treat- 218. Roos A, Linn-Rasker SP, van Domburg RT, Tijssen JP,
ment of thyroid diseases. J Clin Endocrinol Metab. 2005; Berghout A. The starting dose of levothyroxine in primary
90:124 127. hypothyroidism treatment: a prospective, randomized,
203. Cunningham JJ, Barzel US. Lean body mass is a predictor double-blind trial. Arch Intern Med. 2005;165:1714
of the daily requirement for thyroid hormone in older men 1720.
and women. J Am Geriatr Soc. 1984;32:204 207. 219. Keating FR Jr, Parkin TW, Selby JB, Dickinson LS. Treat-
204. Sartorio A, Ferrero S, Trecate L, Bedogni G. Thyroid func- ment of heart disease associated with myxedema. Prog
tion is more strongly associated with body impedance than Cardiovasc Dis. 1961;3:364 381.
anthropometry in healthy subjects. J Endocrinol Invest. 220. Peterson RE. The infuence of the thyroid on adrenal cor-
2002;25:620 623. tical function. J Clin Invest. 1958;37:736 743.
205. Michalaki MA, Gkotsina MI, Mamali I, et al. Impaired 221. Boelaert K, Newby PR, Simmonds MJ, et al. Prevalence
pharmacokineticsof levothyroxine in severely obese vol- and relative risk of other autoimmune diseases in subjects
unteers. Thyroid. 2011;21:477 481. with autoimmune thyroid disease. Am J Med. 2010;123:
206. Gordon MB, Gordon MS. Variations in adequate levothy- 183.
roxine replacement therapy in patients with different 222. Kabadi UM, Jackson T. Serum thyrotropin in primary hy-
causes of hypothyroidism. Endocr Pract. 1999;5:233 pothyroidism. A possible predictor of optimal daily levo-
238. thyroxine dose in primary hypothyroidism. Arch Intern
207. Burmeister LA, Goumaz MO, Mariash CN, Oppenheimer Med. 1995;155:1046 1048.
JH. Levothyroxine dose requirements for thyrotropin sup- 223. Barbesino G. Drugs affecting thyroid function. Thyroid.
pression in the treatment of differentiated thyroid cancer. 2010;20:763770.
J Clin Endocrinol Metab. 1992;75:344 350. 224. Lazarus JH. The effects of lithium therapy on thyroid and
208. Bartalena L, Martino E, Pacchiarotti A, et al. Factors af- thyrotropin-releasing hormone. Thyroid. 1998;8:909
fecting suppression of endogenous thyrotropin secretion 913.
by thyroxine treatment: retrospective analysis in athyreotic 225. Lazarus JH. Lithium and thyroid. Best Pract Res Clin En-
and goitrous patients. J Clin Endocrinol Metab. 1987;64: docrinol Metab. 2009;23:723733.
849 855. 226. Bocchetta A, Bernardi F, Pedditzi M, et al. Thyroid ab-
209. Olubowale O, Chadwick DR. Optimization of thyroxine normalities during lithium treatment. Acta Psychiatr
replacement therapy after total or near-total thyroidec- Scand. 1991;83:193198.
tomy for benign thyroid disease. Br J Surg. 2006;93:57 227. Basaria S, Cooper DS. Amiodarone and the thyroid. Am J
60. Med. 2005;118:706 714.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
500 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

228. Martino E, Bartalena L, Bogazzi F, Braverman LE. The cinoma treated with sorafenib. Ann Oncol. 2008;19:265
Effects of amiodarone on the thyroid. Endocr Rev. 2001; 268.
22:240 254. 245. Gupta-Abramson V, Troxel AB, Nellore A, et al. Phase II
229. Tran HA, Malcolm Reeves GE, Gibson R, Attia JR. De- trial of sorafenib in advanced thyroid cancer. J Clin Oncol.
velopment of thyroid diseases in the treatment of chronic 2008;26:4714 4719.
hepatitis C with -interferon may be a good prognosticator 246. de Groot JW, Zonnenberg BA, Plukker JT, van Der Graaf
in achieving a sustained virological response: a meta-anal- WT, Links TP. Imatinib induces hypothyroidism in pa-
ysis. J Gastroenterol Hepatol. 2009;24:11631168. tients receiving levothyroxine. Clin Pharmacol Ther. 2005;
230. Vezali E, Elefsiniotis I, Mihas C, Konstantinou E, Saroglou 78:433 438.
G. Thyroid dysfunction in patients with chronic hepatitis 247. de Groot JW, Links TP, van der Graaf WT. Tyrosine kinase
C: virus- or therapy-related? J Gastroenterol Hepatol. inhibitors causing hypothyroidism in a patient on levothy-
2009;24:1024 1029. roxine. Ann Oncol. 2006;17:1719 1720.
231. Hamnvik OP, Larsen PR, Marqusee E. Thyroid dysfunc- 248. Sherman SI, Wirth LJ, Droz JP, et al; Motesanib Thyroid
tion from antineoplastic agents. J Natl Cancer Inst. 2011; Cancer Study Group. Motesanib diphosphate in progres-
103:15721587. sive differentiated thyroid cancer. N Engl J Med. 2008;
232. Kung AW, Jones BM, Lai CL. Effects of interferon- ther- 359:31 42.
apy on thyroid function, T-lymphocyte subpopulations 249. Lele AV, Clutter S, Price E, De Ruyter ML. Severe hypo-
and induction of autoantibodies. J Clin Endocrinol Metab. thyroidism presenting as myxedema coma in the postop-
1990;71:1230 1234. erative period in a patient taking sunitinib: case report and
233. Ward DL, Bing-You RG. Autoimmune thyroid dysfunc- review of literature. J Clin Anesth. 2013;25:4751.
tion induced by interferon-alpha treatment for chronic 250. Lynch BS, Delzell ES, Bechtel DH. Toxicology review and
hepatitis C: screening and monitoring recommendations. risk assessment of resorcinol: thyroid effects. Regul Toxi-
Endocr Pract. 2001;7:5258. col Pharmacol. 2002;36:198 210.
234. Schlumberger M, Sherman SI. Clinical trials for progres- 251. Badros AZ, Siegel E, Bodenner D, et al. Hypothyroidism in
sive differentiated thyroid cancer: patient selection, study patients with multiple myeloma following treatment with
design, and recent advances. Thyroid. 2009;19:1393 thalidomide. Am J Med. 2002;112:412 413.
1400. 252. Mechanick JI, Chausmer AB, Kelman AS, Spitz AF, Wal-
235. Carhill AA, Cabanillas ME, Jimenez C, et al. The nonin- lach S. 2001 Nutraceuticals. In: Palumbo PJ, Bower BF,
vestigational use of tyrosine kinase inhibitors in thyroid Braithwaite SS (AACE Self-Assessment Profile Commit-
cancer: establishing a standard for patient safety and mon- tee). eds. Self-Assessment Profile in Endocrinology and
itoring. J Clin Endocrinol Metab. 2013;98:31 42. Metabolism. Jacksonville, FL: The American Association
236. Mannavola D, Coco P, Vannucchi G, et al. A novel ty- of Clinical Endocrinologists and the American College of
rosine-kinase selective inhibitor, sunitinib, induces tran- Endocrinology; 2001:271280.
sient hypothyroidism by blocking iodine uptake. J Clin 253. Mechanick JI, Brett EM, Chausmer AB, Dickey RA, Wal-
Endocrinol Metab. 2007;92:35313534. lach S; American Association of Clinical Endocrinologists.
237. Wong E, Rosen LS, Mulay M, et al. Sunitinib induces hy- American Association of Clinical Endocrinologists medi-
pothyroidism in advanced cancer patients and may inhibit cal guidelines for the clinical use of dietary supplements
thyroid peroxidase activity. Thyroid. 2007;17:351355. and nutraceuticals. Endocr Pract. 2003;9:417 470.
238. Faris JE, Moore AF, Daniels GH. Sunitinib (Sutent)-in- 254. Bell DS, Ovalle F. Use of soy protein supplement and re-
duced thyrotoxicosis due to destructive thyroiditis: a case sultant need for increased dose of levothyroxine. Endocr
report. Thyroid. 2007;17:11471149. Pract. 2001;7:193194.
239. Alexandrescu DT, Popoveniuc G, Farzanmehr H, Dasanu 255. Cassidy A, Albertazzi P, Lise Nielsen I, et al. Critical review
CA, Dawson N, Wartofsky L. Sunitinib-associated lym- of health effects of soyabean phyto-oestrogens in post-
phocytic thyroiditis without circulating antithyroid anti- menopausal women. Proc Nutr Soc. 2006;65:76 92.
bodies. Thyroid. 2008;18:809 812. 256. Messina M, Redmond G. Effects of soy protein and soy-
240. Kamba T, McDonald DM. Mechanisms of adverse effects bean isoflavones on thyroid function in healthy adults and
of anti-VEGF therapy for cancer. Br J Cancer. 2007;96: hypothyroid patients: a review of the relevant literature.
1788 1795. Thyroid. 2006;16:249 258.
241. Kamba T, Tam BY, Hashizume H, et al. VEGF-dependent 257. Duncan AM, Underhill KE, Xu X, Lavalleur J, Phipps WR,
plasticity of fenestrated capillaries in the normal adult mi- Kurzer MS. Modest hormonal effects of soy isoflavones in
crovasculature. Am J Physiol Heart Circ Physiol. 2006; postmenopausal women. J Clin Endocrinol Metab. 1999;
290:H560 H576. 84:3479 3484.
242. Makita N, Miyakawa M, Fujita T, Iiri T. Sunitinib induces 258. Bitto A, Polito F, Atteritano M, et al. Genistein aglycone
hypothyroidism with a markedly reduced vascularity. Thy- does not affect thyroid function: results from a three-year,
roid. 2010;20:323326. randomized, double-blind, placebo-controlled trial. J Clin
243. Abdulrahman RM, Verloop H, Hoftijzer H, et al. Endocrinol Metab. 2010;95:30673072.
Sorafenib-induced hypothyroidism is associated with in- 259. Knopp RH, Bergelin RO, Wahl PW, Walden CE, Chap-
creased type 3 deiodination. J Clin Endocrinol Metab. man MB. Clinical chemistry alterations in pregnancy and
2010;95:3758 3762. oral contraceptive use. Obstet Gynecol. 1985;66:682
244. Tamaskar I, Bukowski R, Elson P, et al. Thyroid function 690.
test abnormalities in patients with metastatic renal cell car- 260. Stockigt JR, Topliss DJ. Assessment of thyroid function

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 501

during high-dosage furosemide therapy. Arch Intern Med. 278. Crilly M. Thyroxine adherence in primary hypothyroid-
1989;149:973. ism. Lancet. 2004;363:1558.
261. Jaume JC, Mendel CM, Frost PH, Greenspan FS, Laughton 279. Gattaz WF, Dressing H, Hewer W. Munchhausen syn-
CW. Extremely low doses of heparin release lipase activity drome: psychopathology and management. Psychopathol-
into the plasma and can thereby cause artifactual eleva- ogy. 1990;23:3339.
tions in the serum-free thyroxine concentration as mea- 280. Jauk B, Mikosch P, Gallowitsch HJ, et al. Unusual mal-
sured by equilibrium dialysis. Thyroid. 1996;6:79 83. absorption of levothyroxine. Thyroid. 2000;10:9395.
262. Stockigt JR, Lim CF. Medications that distort in vitro tests 281. Lips DJ, van Reisen MT, Voigt V, Venekamp W. Diagnosis
of thyroid function, with particular reference to estimates and treatment of levothyroxine pseudomalabsorption.
of serum free thyroxine. Best Pract Res Clin Endocrinol Neth J Med. 2004;62:114 118.
Metab. 2009;23:753767. 282. Ogawa D, Otsuka F, Mimura U, et al. Pseudomalabsorp-
263. Kundra P, Burman KD. The effect of medications on thy- tion of levothyroxine: a case report. Endocr J. 2000;47:
roid function tests. Med Clin North Am. 2012;96:283 4550.
295. 283. Eledrisi MS, Szymajda A, Alshanti M, Urban RJ. Non-
264. McCowen KC, Garber JR, Spark R. Elevated serum thy- compliance with medical treatment: pseudomalabsorption
rotropin in thyroxine-treated patients with hypothyroid- of levothyroxine. South Med J. 2001;94:833 836.
ism given sertraline. N Engl J Med. 1997;337:1010 1011. 284. Morris JC. How do you approach the problem of TSH
265. Elliott DP. Effect of levothyroxine administration time on elevation in a patient on high-dose thyroid hormone re-
serum TSH in elderly patients. Ann Pharmacother. 2001; placement? Clin Endocrinol (Oxf). 2009;70:671 673.
35:529 532. 285. Ain KB, Refetoff S, Fein HG, Weintraub BD. Pseudoma-
266. Bach-Huynh TG, Nayak B, Loh J, Soldin S, Jonklaas J. labsorption of levothyroxine. JAMA. 1991;266:2118
Timing of levothyroxine administration affects serum thy- 2120.
rotropin concentration. J Clin Endocrinol Metab. 2009; 286. Simonian HP, Vo L, Doma S, Fisher RS, Parkman HP.
94:39053912. Regional postprandial differences in pH within the stom-
267. Bolk N, Visser TJ, Kalsbeek A, van Domburg RT, Bergh-
ach and gastroesophageal junction. Dig Dis Sci. 2005;50:
out A. Effects of evening vs morning thyroxine ingestion on
2276 2285.
serum thyroid hormone profiles in hypothyroid patients.
287. Benvenga S, Bartolone L, Pappalardo MA, et al. Altered
Clin Endocrinol (Oxf). 2007;66:43 48.
intestinal absorption of L-thyroxine caused by coffee. Thy-
268. Bolk N, Visser TJ, Nijman J, Jongste IJ, Tijssen JG, Bergh-
roid. 2008;18:293301.
out A. Effects of evening vs morning levothyroxine intake:
288. Conrad SC, Chiu H, Silverman BL. Soy formula compli-
a randomized double-blind crossover trial. Arch Intern
cates management of congenital hypothyroidism. Arch Dis
Med. 2010;170:1996 2003.
Child. 2004;89:37 40.
269. Rajput R, Chatterjee S, Rajput M. Can levothyroxine be
289. Lilja JJ, Laitinen K, Neuvonen PJ. Effects of grapefruit juice
taken as evening dose? Comparative evaluation of morn-
on the absorption of levothyroxine. Br J Clin Pharmacol.
ing versus evening dose of levothyroxine in treatment of
2005;60:337341.
hypothyroidism. J Thyroid Res. 2011;2011:505239.
270. Perez CL, Araki FS, Graf H, de Carvalho GA. Serum thy- 290. Sherman SI, Tielens ET, Ladenson PW. Sucralfate causes
rotropin levels following levothyroxine administration at malabsorption of L-thyroxine. Am J Med. 1994;96:531
breakfast. Thyroid. 2013;23:779 784. 535.
271. Grebe SK, Cooke RR, Ford HC, et al. Treatment of hypo- 291. Singh N, Singh PN, Hershman JM. Effect of calcium car-
thyroidism with once weekly thyroxine. J Clin Endocrinol bonate on the absorption of levothyroxine. JAMA. 2000;
Metab. 1997;82:870 875. 283:28222825.
272. Taylor J, Williams BO, Frater J, Stott DJ, Connell J. Twice- 292. Liel Y, Sperber AD, Shany S. Nonspecific intestinal ad-
weekly dosing for thyroxine replacement in elderly pa- sorption of levothyroxine by aluminum hydroxide. Am J
tients with primary hypothyroidism. J Int Med Res. 1994; Med. 1994;97:363365.
22:273277. 293. Campbell NR, Hasinoff BB, Stalts H, Rao B, Wong NC.
273. Rangan S, Tahrani AA, Macleod AF, Moulik PK. Once Ferrous sulfate reduces thyroxine efficacy in patients with
weekly thyroxine treatment as a strategy to treat non-com- hypothyroidism. Ann Intern Med. 1992;117:1010 1013.
pliance. Postgrad Med J. 2007;83:e3. 294. Shakir KM, Michaels RD, Hays JH, Potter BB. The use of
274. Anderson L, Joseph F, Goenka N, Patel V. Isolated thy- bile acid sequestrants to lower serum thyroid hormones in
roxine malabsorption treated with intramuscular thyrox- iatrogenic hyperthyroidism. Ann Intern Med. 1993;118:
ine injections. Am J Med Sci. 2009;337:150 152. 112113.
275. Hays MT. Parenteral thyroxine administration. Thyroid. 295. Demke DM. Drug interaction between thyroxine and lo-
2007;17:127129. vastatin. N Engl J Med. 1989;321:13411342.
276. Briesacher BA, Andrade SE, Fouayzi H, Chan KA. Com- 296. Madhava K, Hartley A. Hypothyroidism in thyroid carci-
parison of drug adherence rates among patients with seven noma follow-up: Orlistat may inhibit the absorption of
different medical conditions. Pharmacotherapy. 2008;28: thyroxine. Clin Oncol (R Coll Radiol). 2005;17:492.
437 443. 297. Siraj ES, Gupta MK, Reddy SS. Raloxifene causing mal-
277. Bagattoli RM, Vaisman M, Lima JS, Ward LS. Estudo de absorption of levothyroxine. Arch Intern Med. 2003;163:
adeso ao tratamento do hipotiroidismo. Arq Bras Endo- 13671370.
crinol Metabol. 2000;44:483 487. 298. John-Kalarickal J, Pearlman G, Carlson HE. New medi-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
502 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

cations which decrease levothyroxine absorption. Thy- patients dependent on prolonged thyroxine infusion
roid. 2007;17:763765. through a jejunostomy. Clin Nutr. 2000;19:65 67.
299. Sachmechi I, Reich DM, Aninyei M, Wibowo F, Gupta G, 316. Stone E, Leiter LA, Lambert JR, Silverberg JD, Jeejeebhoy
Kim PJ. Effect of proton pump inhibitors on serum thyroid- KN, Burrow GN. L-Thyroxine absorption in patients with
stimulating hormone level in euthyroid patients treated short bowel. J Clin Endocrinol Metab. 1984;59:139 141.
with levothyroxine for hypothyroidism. Endocr Pract. 317. Biondi B, Fazio S, Cuocolo A, et al. Impaired cardiac re-
2007;13:345349. serve and exercise capacity in patients receiving long-term
300. Ananthakrishnan S, Braverman LE, Levin RM, Magnani thyrotropin suppressive therapy with levothyroxine. J Clin
B, Pearce EN. The effect of famotidine, esomeprazole, and Endocrinol Metab. 1996;81:4224 4228.
ezetimibe on levothyroxine absorption. Thyroid. 2008;18: 318. Biondi B, Palmieri EA, Fazio S, et al. Endogenous subclin-
493 498. ical hyperthyroidism affects quality of life and cardiac mor-
301. Valentino R, Savastano S, Tommaselli AP, et al. Prevalence phology and function in young and middle-aged patients.
of coeliac disease in patients with thyroid autoimmunity. J Clin Endocrinol Metab. 2000;85:4701 4705.
Horm Res. 1999;51:124 127. 319. Biondi B, Fazio S, Carella C, et al. Cardiac effects of long
302. Virili C, Bassotti G, Santaguida MG, et al. Atypical celiac term thyrotropin-suppressive therapy with levothyroxine.
disease as cause of increased need for thyroxine: a system- J Clin Endocrinol Metab. 1993;77:334 338.
atic study. J Clin Endocrinol Metab. 2012;97:E419 E422. 320. Biondi B. Should we treat all subjects with subclinical thy-
303. McDermott JH, Coss A, Walsh CH. Celiac disease pre- roid disease the same way? Eur J Endocrinol. 2008;159:
senting as resistant hypothyroidism. Thyroid. 2005;15: 343345.
386 388. 321. Roberts LM, Pattison H, Roalfe A, et al. Is subclinical
304. Collin P, Kaukinen K, Vlimki M, Salmi J. Endocrino- thyroid dysfunction in the elderly associated with depres-
logical disorders and celiac disease. Endocr Rev. 2002;23: sion or cognitive dysfunction? Ann Intern Med. 2006;145:
464 483. 573581.
305. Muoz-Torres M, Varsavsky M, Alonso G. Lactose intol- 322. Gan EH, Pearce SH. The thyroid in mind: cognitive func-
tion and low thyrotropin in older people. J Clin Endocrinol
erance revealed by severe resistance to treatment with levo-
Metab. 2012;97:3438 3449.
thyroxine. Thyroid. 2006;16:11711173.
323. Sawin CT, Geller A, Wolf P, et al. Low serum thyrotropin
306. Lauritano EC, Bilotta AL, Gabrielli M, et al. Association
concentrations as a risk factor for atrial fibrillation in older
between hypothyroidism and small intestinal bacterial
persons. N Engl J Med 1994;331:1249 1252.
overgrowth. J Clin Endocrinol Metab. 2007;92:4180
324. Auer J, Scheibner P, Mische T, Langsteger W, Eber O, Eber
4184.
B. Subclinical hyperthyroidism as a risk factor for atrial
307. Lahner E, Centanni M, Agnello G, et al. Occurrence and
fibrillation. Am Heart J. 2001;142:838 842.
risk factors for autoimmune thyroid disease in patients
325. Cappola AR, Fried LP, Arnold AM, et al. Thyroid status,
with atrophic body gastritis. Am J Med. 2008;121:136
cardiovascular risk, and mortality in older adults. JAMA.
141.
2006;295:10331041.
308. Checchi S, Montanaro A, Pasqui L, et al. L-Thyroxine re-
326. Collet TH, Gussekloo J, Bauer DC, et al; Thyroid Studies
quirement in patients with autoimmune hypothyroidism Collaboration. Subclinical hyperthyroidism and the risk of
and parietal cell antibodies. J Clin Endocrinol Metab. coronary heart disease and mortality. Arch Intern Med.
2008;93:465 469. 2012;172:799 809.
309. Silva CM, Souza MV. Autoimmune hypothyroidism non 327. Parle JV, Maisonneuve P, Sheppard MC, Boyle P, Franklyn
responsive to high doses of levothyroxine and severe hy- JA. Prediction of all-cause and cardiovascular mortality in
pocalcemia [in Portuguese]. Arq Bras Endocrinol Metabol. elderly people from one low serum thyrotropin result: a
2005;49:599 603. 10-year cohort study. Lancet. 2001;358:861 865.
310. Sachs G, Shin JM, Munson K, et al. The control of gastric 328. Schultz M, Kistorp C, Raymond I, et al. Cardiovascular
acid and Helicobacter pylori eradication. Aliment Phar- events in thyroid disease: a population based, prospective
macol Ther. 2000;14:13831401. study. Horm Metab Res. 2011;43:653 659.
311. Seppel T, Rose F, Schlaghecke R. Chronic intestinal giar- 329. Bauer DC, Rodondi N, Stone KL, Hillier TA. Thyroid hor-
diasis with isolated levothyroxine malabsorption as reason mone use, hyperthyroidism and mortality in older women.
for severe hypothyroidismimplications for localization Am J Med. 2007;120:343349.
of thyroid hormone absorption in the gut. Exp Clin En- 330. Flynn RW, Bonellie SR, Jung RT, MacDonald TM, Morris
docrinol Diabetes. 1996;104:180 182. AD, Leese GP. Serum thyroid-stimulating hormone con-
312. Padwal R, Brocks D, Sharma AM. A systematic review of centration and morbidity from cardiovascular disease and
drug absorption following bariatric surgery and its theo- fractures in patients on long-term thyroxine therapy. J Clin
retical implications. Obes Rev. 2010;11:4150. Endocrinol Metab. 2010;95:186 193.
313. Bevan JS, Munro JF. Thyroxine malabsorption following 331. Vestergaard P, Weeke J, Hoeck HC, et al. Fractures in
intestinal bypass surgery. Int J Obes. 1986;10:245246. patients with primary idiopathic hypothyroidism. Thy-
314. Azizi F, Belur R, Albano J. Malabsorption of thyroid hor- roid. 2000;10:335340.
mones after jejunoileal bypass for obesity. Ann Intern Med. 332. Faber J, Galle AM. Changes in bone mass during pro-
1979;90:941942. longed subclinical hyperthyroidism due to L-thyroxine
315. Smyrniotis V, Vaos N, Arkadopoulos N, Kostopanagiotou treatment: a meta-analysis. Eur J Endocrinol. 1994;130:
G, Theodoraki K, Lambrou A. Severe hypothyroidism in 350 356.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 503

333. Uzzan B, Campos J, Cucherat M, Nony P, Boissel JP, Perret 349. Bauer M, Silverman DH, Schlagenhauf F, et al. Brain glu-
GY. Effects on bone mass of long term treatment with cose metabolism in hypothyroidism: a positron emission
thyroid hormones: a meta-analysis. J Clin Endocrinol tomography study before and after thyroid hormone re-
Metab. 1996;81:4278 4289. placement therapy. J Clin Endocrinol Metab. 2009;94:
334. Quan ML, Pasieka JL, Rorstad O. Bone mineral density in 29222929.
well-differentiated thyroid cancer patients treated with 350. Lomenick JP, El-Sayyid M, Smith WJ. Effect of levo-thy-
suppressive thyroxine: a systematic overview of the liter- roxine treatment on weight and body mass index in chil-
ature. J Surg Oncol. 2002;79:62 69; discussion 69 70. dren with acquired hypothyroidism. J Pediatr. 2008;152:
335. Heemstra KA, Hamdy NA, Romijn JA, Smit JW. The ef- 96 100.
fects of thyrotropin-suppressive therapy on bone metabo- 351. Hoogwerf BJ, Nuttall FQ. Long-term weight regulation in
lism in patients with well-differentiated thyroid carci- treated hyperthyroid and hypothyroid subjects. Am J Med.
noma. Thyroid. 2006;16:583591. 1984;76:963970.
336. Solomon BL, Wartofsky L, Burman KD. Prevalence of 352. Franklyn JA, Daykin J, Betteridge J, et al. Thyroxine re-
fractures in postmenopausal women with thyroid disease. placement therapy and circulating lipid concentrations.
Thyroid. 1993;3:1723. Clin Endocrinol (Oxf). 1993;38:453 459.
337. Sheppard MC, Holder R, Franklyn JA. Levothyroxine 353. Clausen P, Mersebach H, Nielsen B, Feldt-Rasmussen B,
treatment and occurrence of fracture of the hip. Arch In- Feldt-Rasmussen U. Hypothyroidism is associated with
tern Med. 2002;162:338 343. signs of endothelial dysfunction despite 1-year replace-
338. Bauer DC, Ettinger B, Nevitt MC, Stone KL; Study of Os- ment therapy with levothyroxine. Clin Endocrinol (Oxf).
teoporotic Fractures Research Group. Risk for fracture in 2009;70:932937.
women with low serum levels of thyroid-stimulating hor- 354. Ruhla S, Arafat AM, Osterhoff M, et al. Levothyroxine
mone. Ann Intern Med. 2001;134:561568. medication is associated with adiposity independent of
339. Turner MR, Camacho X, Fischer HD, et al. Levothyroxine TSH. Exp Clin Endocrinol Diabetes. 2012;120:351354.
dose and risk of fractures in older adults: nested case-con- 355. Jonklaas J, Nsouli-Maktabi H. Weight changes in euthy-
trol study. BMJ. 2011;342:d2238. roid patients undergoing thyroidectomy. Thyroid. 2011;
340. Lee JS, Buzkov P, Fink HA, et al. Subclinical thyroid dys- 21:13431351.
function and incident hip fracture in older adults. Arch 356. Flynn RW, Macdonald TM, Jung RT, Morris AD, Leese
Intern Med. 2010;170:1876 1883. GP. Mortality and vascular outcomes in patients treated
341. Gammage MD, Parle JV, Holder RL, et al. Association for thyroid dysfunction. J Clin Endocrinol Metab. 2006;
between serum free thyroxine concentration and atrial fi- 91:2159 2164.
brillation. Arch Intern Med. 2007;167:928 934. 357. al-Adsani H, Hoffer LJ, Silva JE. Resting energy expendi-
342. de Jong FJ, Masaki K, Chen H, et al. Thyroid function, the ture is sensitive to small dose changes in patients on chronic
risk of dementia and neuropathologic changes: the Hono- thyroid hormone replacement. J Clin Endocrinol Metab.
lulu-Asia aging study. Neurobiol Aging. 2009;30:600 1997;82:1118 1125.
606. 358. Boeving A, Paz-Filho G, Radominski RB, Graf H, Amaral
343. Yeap BB, Alfonso H, Chubb SA, et al. Higher free thyrox- de Carvalho G. Low-normal or high-normal thyrotropin
ine levels are associated with frailty in older men: the target levels during treatment of hypothyroidism: a pro-
Health In Men Study. Clin Endocrinol (Oxf). 2012;76: spective, comparative study. Thyroid. 2011;21:355360.
741748. 359. Walsh JP, Ward LC, Burke V, et al. Small changes in thy-
344. van den Beld AW, Visser TJ, Feelders RA, Grobbee DE, roxine dosage do not produce measurable changes in hy-
Lamberts SW. Thyroid hormone concentrations, disease, pothyroid symptoms, well-being, or quality of life: results
physical function, and mortality in elderly men. J Clin En- of a double-blind, randomized clinical trial. J Clin Endo-
docrinol Metab. 2005;90:6403 6409. crinol Metab. 2006;91:2624 2630.
345. Saravanan P, Chau WF, Roberts N, Vedhara K, Green- 360. Fish LH, Schwartz HL, Cavanaugh J, Steffes MW, Bantle
wood R, Dayan CM. Psychological well-being in patients JP, Oppenheimer JH. Replacement dose, metabolism, and
on adequate doses of L-thyroxine: results of a large, con- bioavailability of levothyroxine in the treatment of hypo-
trolled community-based questionnaire study. Clin Endo- thyroidism. Role of triiodothyronine in pituitary feedback
crinol (Oxf). 2002;57:577585. in humans. N Engl J Med. 1987;316:764 770.
346. Wekking EM, Appelhof BC, Fliers E, et al. Cognitive func- 361. Salmon D, Rendell M, Williams J, et al. Chemical hyper-
tioning and well-being in euthyroid patients on thyroxine thyroidism: serum triiodothyronine levels in clinically eu-
replacement therapy for primary hypothyroidism. Eur J thyroid individuals treated with levothyroxine. Arch In-
Endocrinol. 2005;153:747753. tern Med. 1982;142:571573.
347. Panicker V, Evans J, Bjro T, Asvold BO, Dayan CM, 362. Woeber KA. Levothyroxine therapy and serum free thy-
Bjerkeset O. A paradoxical difference in relationship be- roxine and free triiodothyronine concentrations. J Endo-
tween anxiety, depression and thyroid function in subjects crinol Invest. 2002;25:106 109.
on and not on T4: findings from the HUNT study. Clin 363. Hoermann R, Midgley JE, Larisch R, Dietrich JW. Is pi-
Endocrinol (Oxf). 2009;71:574 580. tuitary TSH an adequate measure of thyroid hormone-
348. Samuels MH, Schuff KG, Carlson NE, Carello P, Janowsky controlled homoeostasis during thyroxine treatment? Eur
JS. Health status, psychological symptoms, mood, and J Endocrinol. 2013;168:271280.
cognition in L-thyroxine-treated hypothyroid subjects. 364. Escobar-Morreale HF, del Rey FE, Obregn MJ, de Esco-
Thyroid. 2007;17:249 258. bar GM. Only the combined treatment with thyroxine and

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
504 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

triiodothyronine ensures euthyroidism in all tissues of the bined thyroxine/3,5,3-triiodothyronine therapy. J Clin
thyroidectomized rat. Endocrinology. 1996;137:2490 Endocrinol Metab. 2005;90:6296 6299.
2502. 379. Saravanan P, Siddique H, Simmons DJ, Greenwood R,
365. Butler PW, Smith SM, Linderman JD, et al. The Thr92Ala Dayan CM. Twenty-four hour hormone profiles of TSH,
5 type 2 deiodinase gene polymorphism is associated with free T3 and free T4 in hypothyroid patients on combined
a delayed triiodothyronine secretion in response to the thy- T3/T4 therapy. Exp Clin Endocrinol Diabetes. 2007;115:
rotropin-releasing hormone stimulation test: a pharma- 261267.
cogenomic study. Thyroid. 2010;20:14071412. 380. Grozinsky-Glasberg S, Fraser A, Nahshoni E, Weizman A,
366. Peeters RP, van den Beld AW, van Toor H, et al. A poly- Leibovici L. Thyroxine-triiodothyronine combination
morphism in type I deiodinase is associated with circulat- therapy versus thyroxine monotherapy for clinical hypo-
ing free insulin-like growth factor I levels and body com- thyroidism: meta-analysis of randomized controlled trials.
position in humans. J Clin Endocrinol Metab. 2005;90: J Clin Endocrinol Metab. 2006;91:25922599.
256 263. 381. Ma C, Xie J, Huang X, et al. Thyroxine alone or thyroxine
367. Canani LH, Capp C, Dora JM, et al. The type 2 deiodinase plus triiodothyronine replacement therapy for hypothy-
A/G (Thr92Ala) polymorphism is associated with de- roidism. Nucl Med Commun. 2009;30:586 593.
creased enzyme velocity and increased insulin resistance in 382. Joffe RT, Brimacombe M, Levitt AJ, Stagnaro-Green A.
patients with type 2 diabetes mellitus. J Clin Endocrinol Treatment of clinical hypothyroidism with thyroxine and
Metab. 2005;90:34723478. triiodothyronine: a literature review and metaanalysis.
368. Bianco AC, Casula S. Thyroid hormone replacement ther- Psychosomatics. 2007;48:379 384.
apy: three simple questions, complex answers. Eur Thy- 383. Aronson R, Offman HJ, Joffe RT, Naylor CD. Triiodo-
roid J. 2012;1:88 98. thyronine augmentation in the treatment of refractory de-
369. Guo TW, Zhang FC, Yang MS, et al. Positive association pression. A meta-analysis. Arch Gen Psychiatry. 1996;53:
of the DIO2 (deiodinase type 2) gene with mental retar- 842 848.
dation in the iodine-deficient areas of China. J Med Genet. 384. Altshuler LL, Bauer M, Frye MA, et al. Does thyroid sup-
2004;41:585590. plementation accelerate tricyclic antidepressant response?
370. Gumieniak O, Perlstein TS, Williams JS, et al. Ala92 type A review and meta-analysis of the literature. Am J Psychi-
2 deiodinase allele increases risk for the development of atry. 2001;158:16171622.
hypertension. Hypertension. 2007;49:461 466. 385. Bunevicius R, Kazanavicius G, Zalinkevicius R, Prange AJ
371. Meulenbelt I, Min JL, Bos S, et al. Identification of DIO2 Jr. Effects of thyroxine as compared with thyroxine plus
as a new susceptibility locus for symptomatic osteoarthri- triiodothyronine in patients with hypothyroidism. N Engl
tis. Hum Mol Genet. 2008;17:18671875. J Med. 1999;340:424 429.
372. He B, Li J, Wang G, et al. Association of genetic polymor- 386. Sawka AM, Gerstein HC, Marriott MJ, MacQueen GM,
phisms in the type II deiodinase gene with bipolar disorder Joffe RT. Does a combination regimen of thyroxine (T4)
in a subset of Chinese population. Prog Neuropsychop- and 3,5,3-triiodothyronine improve depressive symptoms
harmacol Biol Psychiatry. 2009;33:986 990. better than T4 alone in patients with hypothyroidism? Re-
373. Grineva E, Babenko A, Vahrameeva N, et al. Type 2 de- sults of a double-blind, randomized, controlled trial. J Clin
iodinase Thr92Ala polymorphism impact on clinical Endocrinol Metab. 2003;88:4551 4555.
course and myocardial remodeling in patients with Graves 387. Escobar-Morreale HF, Botella-Carretero JI, Escobar del
disease. Cell Cycle. 2009;8:25652569. Rey F, Morreale de Escobar G. Treatment of hypothyroid-
374. Heemstra KA, Hoftijzer H, van der Deure WM, et al. The ism with combinations of levothyroxine plus liothyronine.
type 2 deiodinase Thr92Ala polymorphism is associated J Clin Endocrinol Metab. 2005;90:4946 4954.
with increased bone turnover and decreased femoral neck 388. Appelhof BC, Fliers E, Wekking EM, et al. Combined ther-
bone mineral density. J Bone Miner Res. 2010;25:1385 apy with levothyroxine and liothyronine in two ratios,
1391. compared with levothyroxine monotherapy in primary hy-
375. Mentuccia D, Proietti-Pannunzi L, Tanner K, et al. Asso- pothyroidism: a double-blind, randomized, controlled
ciation between a novel variant of the human type 2 deio- clinical trial. J Clin Endocrinol Metab. 2005;90:2666
dinase gene Thr92Ala and insulin resistance: evidence of 2674.
interaction with the Trp64Arg variant of the -3-adrener- 389. Nygaard B, Jensen EW, Kvetny J, Jarlv A, Faber J. Effect
gic receptor. Diabetes. 2002;51:880 883. of combination therapy with thyroxine (T4) and 3,5,3-
376. Dora JM, Machado WE, Rheinheimer J, Crispim D, Maia triiodothyronine versus T4 monotherapy in patients with
AL. Association of the type 2 deiodinase Thr92Ala poly- hypothyroidism, a double-blind, randomised cross-over
morphism with type 2 diabetes: case-control study and study. Eur J Endocrinol. 2009;161:895902.
meta-analysis. Eur J Endocrinol. 2010;163:427 434. 390. Fadeyev VV, Morgunova TB, Sytch JP, Melnichenko GA.
377. Heemstra KA, Hoftijzer HC, van der Deure WM, et al. TSH and thyroid hormones concentrations in patients with
Thr92Ala polymorphism in the type 2 deiodinase is not hypothyroidism receiving replacement therapy with L-thy-
associated with T4 dose in athyroid patients or patients roxine alone or in combination with L-triiodothyronine.
with Hashimoto thyroiditis. Clin Endocrinol (Oxf). 2009; Hormones. 2005;4:101107.
71:279 283. 391. Walsh JP, Shiels L, Lim EM, et al. Combined thyroxine/
378. Appelhof BC, Peeters RP, Wiersinga WM, et al. Polymor- liothyronine treatment does not improve well-being, qual-
phisms in type 2 deiodinase are not associated with well- ity of life, or cognitive function compared to thyroxine
being, neurocognitive functioning, and preference for com- alone: a randomized controlled trial in patients with pri-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 505

mary hypothyroidism. J Clin Endocrinol Metab. 2003;88: apy become mandatory? J Clin Endocrinol Metab. 2002;
4543 4550. 87:20422045.
392. Clyde PW, Harari AE, Getka EJ, Shakir KM. Combined 407. Jrgensen JO, Pedersen SA, Laurberg P, Weeke J, Skakke-
levothyroxine plus liothyronine compared with levothy- baek NE, Christiansen JS. Effects of growth hormone ther-
roxine alone in primary hypothyroidism: a randomized apy on thyroid function of growth hormone-deficient
controlled trial. JAMA. 2003;290:29522958. adults with and without concomitant thyroxine-substi-
393. Rodriguez T, Lavis VR, Meininger JC, Kapadia AS, Staf- tuted central hypothyroidism. J Clin Endocrinol Metab.
ford LF. Substitution of liothyronine at a 1:5 ratio for a 1989;69:11271132.
portion of levothyroxine: effect on fatigue, symptoms of 408. Jrgensen JO, Mller J, Laursen T, Orskov H, Christian-
depression, and working memory versus treatment with sen JS, Weeke J. Growth hormone administration stimu-
levothyroxine alone. Endocr Pract. 2005;11:223233. lates energy expenditure and extrathyroidal conversion of
394. Bunevicius R, Prange AJ. Mental improvement after re- thyroxine to triiodothyronine in a dose-dependent manner
placement therapy with thyroxine plus triiodothyronine: and suppresses circadian thyrotrophin levels: studies in
relationship to cause of hypothyroidism. Int J Neuropsy- GH-deficient adults. Clin Endocrinol (Oxf). 1994;41:
chopharmacol. 2000;3:167174. 609 614.
395. Bunevicius R, Jakubonien N, Jurkevicius R, Cernicat J, 409. Agha A, Walker D, Perry L, et al. Unmasking of central
Lasas L, Prange AJ Jr. Thyroxine vs thyroxine plus triio- hypothyroidism following growth hormone replacement
dothyronine in treatment of hypothyroidism after thyroid- in adult hypopituitary patients. Clin Endocrinol (Oxf).
ectomy for Graves disease. Endocrine. 2002;18:129 133. 2007;66:7277.
396. Siegmund W, Spieker K, Weike AI, et al. Replacement ther- 410. Portes ES, Oliveira JH, Maccagnan P, Abucham J. Changes
apy with levothyroxine plus triiodothyronine (bioavail- in serum thyroid hormones levels and their mechanisms
able molar ratio 14:1) is not superior to thyroxine alone to during long-term growth hormone (GH) replacement ther-
improve well-being and cognitive performance in hypo- apy in GH deficient children. Clin Endocrinol (Oxf). 2000;
thyroidism. Clin Endocrinol (Oxf). 2004;60:750 757. 53:183189.
397. Regalbuto C, Maiorana R, Alagona C, et al. Effects of 411. Klose M, Marina D, Hartoft-Nielsen ML, et al. 2013 Cen-
either LT4 monotherapy or LT4/LT3 combined therapy in tral hypothyroidism and its replacement have a significant
patients totally thyroidectomized for thyroid cancer. Thy- influence on cardiovascular risk factors in adult hypopi-
roid. 2007;17:323331. tuitary patients. J Clin Endocrinol Metab. [Epub ahead of
398. Cassio A, Cacciari E, Cicognani A, et al. Treatment for print]
congenital hypothyroidism: thyroxine alone or thyroxine 412. Verge CF, Konrad D, Cohen M, et al. Diiodothyropropi-
plus triiodothyronine? Pediatrics. 2003;111:10551060. onic acid (DITPA) in the treatment of MCT8 deficiency.
399. Wardle CA, Fraser WD, Squire CR. Pitfalls in the use of J Clin Endocrinol Metab. 2012;97:4515 4523.
thyrotropin concentration as a first-line thyroid-function 413. Koch CA, Sarlis NJ. The spectrum of thyroid diseases in
test. Lancet. 2001;357:10131014. childhood and its evolution during transition to adult-
400. Oppenheimer JH, Braverman LE, Toft A, Jackson IM, hood: natural history, diagnosis, differential diagnosis and
Ladenson PW. A therapeutic controversy. Thyroid hor- management. J Endocrinol Invest. 2001;24:659 675.
mone treatment: when and what? J Clin Endocrinol 414. Slawik M, Klawitter B, Meiser E, et al. Thyroid hormone
Metab. 1995;80:28732883. replacement for central hypothyroidism: a randomized
401. Carrozza V, Csako G, Yanovski JA, et al. Levothyroxine controlled trial comparing two doses of thyroxine (T4)
replacement therapy in central hypothyroidism: a practice with a combination of T4 and triiodothyronine. J Clin En-
report. Pharmacotherapy. 1999;19:349 355. docrinol Metab. 2007;92:4115 4122.
402. Koulouri O, Auldin MA, Agarwal R, et al. Diagnosis and 415. Benhadi N, Wiersinga WM, Reitsma JB, Vrijkotte TG,
treatment of hypothyroidism in TSH deficiency compared Bonsel GJ. Higher maternal TSH levels in pregnancy are
to primary thyroid disease: pituitary patients are at risk of associated with increased risk for miscarriage, fetal or neo-
underreplacement with levothyroxine. Clin Endocrinol natal death. Eur J Endocrinol. 2009;160:985991.
(Oxf). 2011;74:744 749. 416. Glinoer D, Riahi M, Grn JP, Kinthaert J. Risk of subclin-
403. Iverson JF, Mariash CN. Optimal free thyroxine levels for ical hypothyroidism in pregnant women with asymptom-
thyroid hormone replacement in hypothyroidism. Endocr atic autoimmune thyroid disorders. J Clin Endocrinol
Pract. 2008;14:550 555. Metab. 1994;79:197204.
404. Doin FC, Rosa-Borges M, Martins MR, Moiss VA, Abu- 417. Vermiglio F, Lo Presti VP, Castagna MG, et al. Increased
cham J. Diagnosis of subclinical central hypothyroidism in risk of maternal thyroid failure with pregnancy progres-
patients with hypothalamic-pituitary disease by Doppler sion in an iodine deficient area with major iodine deficiency
echocardiography. Eur J Endocrinol. 2012;166:631 640. disorders. Thyroid. 1999;9:19 24.
405. Alexopoulou O, Beguin C, De Nayer P, Maiter D. Clinical 418. Idris I, Srinivasan R, Simm A, Page RC. Maternal hypo-
and hormonal characteristics of central hypothyroidism at thyroidism in early and late gestation: effects on neonatal
diagnosis and during follow-up in adult patients. Eur J and obstetric outcome. Clin Endocrinol (Oxf). 2005;63:
Endocrinol. 2004;150:1 8. 560 565.
406. Porretti S, Giavoli C, Ronchi C, et al. Recombinant human 419. van den Boogaard E, Vissenberg R, Land JA, et al. Signif-
GH replacement therapy and thyroid function in a large icance of (sub)clinical thyroid dysfunction and thyroid au-
group of adult GH-deficient patients: when does L-T4 ther- toimmunity before conception and in early pregnancy: a

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
506 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

systematic review. Hum Reprod Update. 2011;17:605 and hyperactivity disorders in the offspring of mothers ex-
619. posed to mild-moderate iodine deficiency: a possible novel
420. Thangaratinam S, Tan A, Knox E, Kilby MD, Franklyn J, iodine deficiency disorder in developed countries. J Clin
Coomarasamy A. Association between thyroid autoanti- Endocrinol Metab. 2004;89:6054 6060.
bodies and miscarriage and preterm birth:meta-analysis of 437. Zimmermann MB. Iodine deficiency in pregnancy and the
evidence. BMJ. 2011;342:d2616. effects of maternal iodine supplementation on the off-
421. Toulis KA, Goulis DG, Venetis CA, et al. Risk of sponta- spring: a review. Am J Clin Nutr. 2009;89:668S 672S.
neous miscarriage in euthyroid women with thyroid auto- 438. Caldwell KL, Miller GA, Wang RY, Jain RB, Jones RL.
immunity undergoing IVF: a meta-analysis. Eur J Endo- Iodine status of the U.S. population, National Health and
crinol. 2010;162:643 652. Nutrition Examination Survey 20032004. Thyroid.
422. Man EB, Jones WS, Holden RH, Mellits ED. Thyroid func- 2008;18:12071214.
tion in human pregnancy. 8. Retardation of progeny aged 439. Tan TO, Cheng YW, Caughey AB. Are women who are
7 years; relationships to maternal age and maternal thyroid treated for hypothyroidism at risk for pregnancy compli-
function. Am J Obstet Gynecol. 1971;111:905916. cations? Am J Obstet Gynecol. 2006;194:e1 e3.
423. Glinoer D. Maternal and fetal impact of chronic iodine 440. Abalovich M, Gutierrez S, Alcaraz G, Maccallini G, Garcia
deficiency. Clin Obstet Gynecol. 1997;40:102116. A, Levalle O. Overt and subclinical hypothyroidism com-
424. Henrichs J, Ghassabian A, Peeters RP, Tiemeier H. Ma- plicating pregnancy. Thyroid. 2002;12:63 68.
ternal hypothyroxinemia and effects on cognitive function- 441. Stagnaro-Green A. Maternal thyroid disease and preterm
ing in childhood: how and why? Clin Endocrinol (Oxf). delivery. J Clin Endocrinol Metab. 2009;94:2125.
2013;79:152162. 442. Negro R, Schwartz A, Gismondi R, Tinelli A, Mangieri T,
425. Casey BM, Dashe JS, Spong CY, McIntire DD, Leveno KJ, Stagnaro-Green A. Universal screening versus case finding
Cunningham GF. Perinatal significance of isolated mater- for detection and treatment of thyroid hormonal dysfunc-
nal hypothyroxinemia identified in the first half of preg- tion during pregnancy. J Clin Endocrinol Metab. 2010;95:
nancy. Obstet Gynecol. 2007;109:1129 1135. 1699 1707.
426. Li Y, Shan Z, Teng W, et al. Abnormalities of maternal 443. American College of Obstetricians and Gynecologists.
thyroid function during pregnancy affect neuropsycholog- Thyroid disease in pregnancy. Technical Bulletin no. 37
ical development of their children at 2530 months. Clin (reaffirmed 2010). Washington, DC: American College of
Endocrinol (Oxf). 2010;72:825 829. Obstetricians and Gynecologists; 2002.
427. Craig WY, Allan WC, Kloza EM, et al. Mid-gestational 444. Negro R, Schwartz A, Gismondi R, Tinelli A, Mangieri T,
maternal free thyroxine concentration and offspring neu- Stagnaro-Green A. Increased pregnancy loss rate in thyroid
rocognitive development at age two years. J Clin Endocri- antibody negative women with TSH levels between 2.5 and
nol Metab. 2012;97:E22E28. 5.0 in the first trimester of pregnancy. J Clin Endocrinol
428. Lazarus JH, Bestwick JP, Channon S, et al. Antenatal thy- Metab. 2010;95:E44 E48.
roid screening and childhood cognitive function. N Engl 445. American College of Obstetricians and Gynecologists.
J Med. 2012;366:493501. Committee opinion. Subclinical hypothyroidism in preg-
429. Romn GC, Ghassabian A, Bongers-Schokking JJ, et al. nancy. Number 381 (reaffirmed 2010). Washington, DC:
Association of gestational maternal hypothyroxinemia American College of Obstetricians and Gynecologists;
and increased autism risk. Ann Neurol. 2013;74:733742. 2007.
430. Qian M, Wang D, Watkins WE, et al. The effects of iodine 446. Negro R, Formoso G, Mangieri T, Pezzarossa A, Dazzi D,
on intelligence in children: a meta-analysis of studies con- Hassan H. Levothyroxine treatment in euthyroid pregnant
ducted in China. Asia Pac J Clin Nutr. 2005;14:32 42. women with autoimmune thyroid disease: effects on ob-
431. Pop VJ, Kuijpens JL, van Baar AL, et al. Low maternal free stetrical complications. J Clin Endocrinol Metab. 2006;
thyroxine concentrations during early pregnancy are as- 91:25872591.
sociated with impaired psychomotor development in in- 447. Al-Kunani AS, Knight R, Haswell SJ, Thompson JW, Lin-
fancy. Clin Endocrinol (Oxf). 1999;50:147155. dow SW. The selenium status of women with a history of
432. Pop VJ, Brouwers EP, Vader HL, Vulsma T, van Baar AL, recurrent miscarriage. BJOG. 2001;108:1094 1097.
de Vijlder JJ. Maternal hypothyroxinaemia during early 448. Negro R, Greco G, Mangieri T, Pezzarossa A, Dazzi D,
pregnancy and subsequent child development: a 3-year fol- Hassan H. The influence of selenium supplementation on
low-up study. Clin Endocrinol (Oxf). 2003;59:282288. postpartum thyroid status in pregnant women with thy-
433. Kooistra L, Crawford S, van Baar AL, Brouwers EP, Pop roid peroxidase autoantibodies. J Clin Endocrinol Metab.
VJ. Neonatal effects of maternal hypothyroxinemia during 2007;92:12631268.
early pregnancy. Pediatrics. 2006;117:161167. 449. Hallengren B, Lantz M, Andreasson B, Grennert L. Preg-
434. Li Y, Shan Z, Teng W, et al. Abnormalities of maternal nant women on thyroxine substitution are often dysregu-
thyroid function during pregnancy affect neuropsycholog- lated in early pregnancy. Thyroid. 2009;19:391394.
ical development of their children at 2530 months. Clin 450. Rotondi M, Mazziotti G, Sorvillo F, et al. Effects of in-
Endocrinol (Oxf). 2010;72:825 829. creased thyroxine dosage pre-conception on thyroid func-
435. Henrichs J, Bongers-Schokking JJ, Schenk JJ, et al. Mater- tion during early pregnancy. Eur J Endocrinol. 2004;151:
nal thyroid function during early pregnancy and cognitive 695700.
functioning in early childhood: the Generation R Study. 451. Abalovich M, Alcaraz G, Kleiman-Rubinsztein J, et al. The
J Clin Endocrinol Metab. 2010;95:4227 4234. relationship of preconception thyrotropin levels to require-
436. Vermiglio F, Lo Presti VP, Moleti M, et al. Attention deficit ments for increasing the levothyroxine dose during preg-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 507

nancy in women with primary hypothyroidism. Thyroid. tal laboratory screening versus clinical symptoms as indi-
2010;20:11751178. cators leading to diagnosis. Br Med J (Clin Res Ed). 1984;
452. Medici M, Timmermans S, Visser W, et al. Maternal thy- 289:11711175.
roid hormone parameters during early pregnancy and birth 467. Klein AH, Meltzer S, Kenny FM. Improved prognosis in
weight: the Generation R Study. J Clin Endocrinol Metab. congenital hypothyroidism treated before age three
2013;98:59 66. months. J Pediatr. 1972;81:912915.
453. Shields BM, Knight BA, Hill A, Hattersley AT, Vaidya B. 468. Selva KA, Harper A, Downs A, Blasco PA, Lafranchi SH.
Fetal thyroid hormone level at birth is associated with fetal Neurodevelopmental outcomes in congenital hypothy-
growth. J Clin Endocrinol Metab. 2011;96:E934 E938. roidism: comparison of initial T4 dose and time to reach
454. Vaidya B, Anthony S, Bilous M, et al. Detection of thyroid target T4 and TSH. J Pediatr. 2005;147:775780.
dysfunction in early pregnancy: universal screening or tar- 469. von Heppe JH, Krude H, LAllemand D, Schnabel D, Grters
geted high-risk case finding? J Clin Endocrinol Metab. A. The use of L-T4 as liquid solution improves the practica-
2007;92:203207. bility and individualized dosage in newborns and infants with
455. Dosiou C, Sanders GD, Araki SS, Crapo LM. Screening congenital hypothyroidism. J Pediatr Endocrinol Metab.
pregnant women for autoimmune thyroid disease: a cost- 2004;17:967974.
effectiveness analysis. Eur J Endocrinol. 2008;158:841 470. Cassio A, Monti S, Rizzello A, et al. Comparison between
851. liquid and tablet formulations of levothyroxine in the ini-
456. Nystrom E, Caidahl K, Fager G, Wikkelso P, Lundberg A, tial treatment of congenital hypothyroidism. J Pediatr.
Linndstedt G. A double-blind cross-over 12-month study 2013;162:1264 1269, 1269.e12.
of L-thyroxine treatment of women with subclinical hy- 471. Selva KA, Harper A, Downs A, Blasco PA, Lafranchi SH.
pothyroidism. Clin Endocrinol (Oxf). 1988;29:6376. Neurodevelopmental outcomes in congenital hypothy-
457. Jaeschke R, Guyatt G, Gerstein H, et al. Does treatment roidism: comparison of initial T4 dose and time to reach
with L-thyroxine influence health status in middle-aged target T4 and TSH. J Pediatr. 2005;147:775780.
and older adults with subclinical hypothyroidism? J Gen 472. Salerno M, Militerni R, Bravaccio C, et al. Effect of dif-
Intern Med. 1996;11:744 749. ferent starting doses of levothyroxine on growth and in-
458. Samuels MH, Schuff KG, Carlson NE, Carello P, Janowsky tellectual outcome at four years of age in congenital hy-
JS. Health status, mood, and cognition in experimentally pothyroidism. Thyroid. 2002;12:4552.
induced subclinical hypothyroidism. J Clin Endocrinol 473. Mathai S, Cutfield WS, Gunn AJ, et al. A novel therapeutic
Metab. 2007;92:25452551. paradigm to treat congenital hypothyroidism. Clin Endo-
459. Jorde R, Waterloo K, Storhaug H, Nyrnes A, Sundsfjord J, crinol (Oxf). 2008;69:142151.
Jenssen TG. Neuropsychological function and symptoms 474. Characteristics of infantile hypothyroidism discovered on
in subjects with subclinical hypothyroidism and the effect neonatal screening. J Pediatr. 1984;104:539 544.
of thyroxine treatment. J Clin Endocrinol Metab. 2006; 475. Rovet J, Alvarez M. Thyroid hormone and attention in
91:145153. congenital hypothyroidism. J Pediatr Endocrinol Metab.
460. Parle J, Roberts L, Wilson S, et al. A randomized controlled 1996;9:63 66.
trial of the effect of thyroxine replacement on cognitive 476. Fugazzola L, Persani L, Vannucchi G, et al. Thyroid scin-
function in community-living elderly subjects with sub- tigraphy and perchlorate test after recombinant human
clinical hypothyroidism: the Birmingham Elderly Thyroid TSH: a new tool for the differential diagnosis of congenital
Study. J Clin Endocrinol Metab. 2010;95:36233632. hypothyroidism during infancy. Eur J Nucl Med Mol Im-
461. Simonsick EM, Newman AB, Ferrucci L, et al; Health ABC aging. 2007;34:1498 1503.
Study. Subclinical hypothyroidism and functional mobility 477. Rabbiosi S, Vigone MC, Cortinovis F, et al. Congenital
in older adults. Arch Intern Med. 2009;169:20112017. hypothyroidism with eutopic thyroid gland: analysis of
462. Somwaru LL, Rariy CM, Arnold AM, Cappola AR. The clinical and biochemical features at diagnosis and after
natural history of subclinical hypothyroidism in the elder- re-evaluation. J Clin Endocrinol Metab. 2013;98:1395
ly: the cardiovascular health study. J Clin Endocrinol 1402.
Metab. 2012;97:19621969. 478. Grueiro de Papendieck L, Chiesa A, Bastida MG, Alonso
463. Waring AC, Arnold AM, Newman AB, Bzkov P, Hirsch G, Finkielstain G, Heinrich JJ. Thyroid dysfunction and
C, Cappola AR. Longitudinal changes in thyroid function high thyroid stimulating hormone levels in children with
in the oldest old and survival: the Cardiovascular Health Downs syndrome. J Pediatr Endocrinol Metab. 2002;15:
Study All-Stars Study. J Clin Endocrinol Metab. 2012;97: 15431548.
3944 3950. 479. Gawlik A, Gawlik T, Januszek-Trzciakowska A, Patel H,
464. Bremner AP, Feddema P, Leedman PJ, et al. Age-related Malecka-Tendera E. Incidence and dynamics of thyroid
changes in thyroid function: a longitudinal study of a com- dysfunction and thyroid autoimmunity in girls with Turn-
munity-based cohort. J Clin Endocrinol Metab. 2012;97: ers syndrome: a long-term follow-up study. Horm Res
1554 1562. Paediatr. 2011;76:314 3120.
465. Hyland KA, Arnold AM, Lee JS, Cappola AR. Persistent 480. Leonardi D, Polizzotti N, Carta A, et al. Longitudinal study
subclinical hypothyroidism and cardiovascular risk in the of thyroid function in children with mild hyperthyro-
elderly: the cardiovascular health study. J Clin Endocrinol tropinemia at neonatal screening for congenital hypothy-
Metab. 2013;98:533540. roidism. J Clin Endocrinol Metab. 2008;93:2679 2685.
466. Alm J, Hagenfeldt L, Larsson A, Lundberg K. Incidence of 481. Narumi S, Muroya K, Abe Y, et al. TSHR mutations as a
congenital hypothyroidism: retrospective study of neona- cause of congenital hypothyroidism in Japan: a popula-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
508 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

tion-based genetic epidemiology study. J Clin Endocrinol 496. Iglesias P, Dez JJ. Thyroid dysfunction and kidney disease.
Metab. 2009;94:13171323. Eur J Endocrinol. 2009;160:503515.
482. Nicoletti A, Bal M, De Marco G, et al. Thyrotropin-stim- 497. Carrero JJ, Qureshi AR, Axelsson J, et al. Clinical and
ulating hormone receptor gene analysis in pediatric pa- biochemical implications of low thyroid hormone levels
tients with non-autoimmune subclinical hypothyroidism. (total and free forms) in euthyroid patients with chronic
J Clin Endocrinol Metab. 2009;94:4187 4194. kidney disease. J Intern Med. 2007;262:690 701.
483. Van Dop C, Conte FA, Koch TK, Clark SJ, Wilson-Davis 498. Zoccali C, Mallamaci F, Tripepi G, Cutrupi S, Pizzini P.
SL, Grumbach MM. Pseudotumor cerebri associated with Low triiodothyronine and survival in end-stage renal dis-
initiation of levothyroxine therapy for juvenile hypothy- ease. Kidney Int. 2006;70:523528.
roidism. N Engl J Med. 1983;308:1076 1080. 499. Lo JC, Chertow GM, Go AS, Hsu CY. Increased preva-
484. Monzani A, Prodam F, Rapa A, et al. Endocrine disorders lence of subclinical and clinical hypothyroidism in persons
in childhood and adolescence. Natural history of subclin- with chronic kidney disease. Kidney Int. 2005;67:1047
ical hypothyroidism in children and adolescents and po- 1052.
tential effects of replacement therapy: a review. Eur J En- 500. Chonchol M, Lippi G, Salvagno G, Zoppini G, Muggeo M,
docrinol. 2013;168:R1R11. Targher G. Prevalence of subclinical hypothyroidism in
485. Lazar L, Frumkin RB, Battat E, Lebenthal Y, Phillip M, patients with chronic kidney disease. Clin J Am Soc Neph-
Meyerovitch J. Natural history of thyroid function tests rol. 2008;3:1296 1300.
over 5 years in a large pediatric cohort. J Clin Endocrinol 501. Asvold BO, Bjro T, Vatten LJ. Association of thyroid
Metab. 2009;94:1678 1682. function with estimated glomerular filtration rate in a pop-
486. Radetti G, Maselli M, Buzi F, et al. The natural history of ulation-based study: the HUNT study. Eur J Endocrinol.
the normal/mild elevated TSH serum levels in children and 2011;164:101105.
adolescents with Hashimotos thyroiditis and isolated hy- 502. Capasso G, De Tommaso G, Pica A, et al. Effects of thyroid
perthyrotropinaemia: a 3-year follow-up. Clin Endocrinol hormones on heart and kidney functions. Miner Electro-
(Oxf). 2012;76:394 398. lyte Metab. 1999;25:56 64.
487. Cetinkaya E, Aslan A, Vidinlisan S, Ocal G. Height im- 503. den Hollander JG, Wulkan RW, Mantel MJ, Berghout A.
provement by L-thyroxine treatment in subclinical hypo- Correlation between severity of thyroid dysfunction and
thyroidism. Pediatr Int. 2003;45:534 537. renal function. Clin Endocrinol (Oxf). 2005;62:423 427.
488. Wasniewska M, Corrias A, Aversa T, et al. Comparative 504. Montenegro J, Gonzlez O, Saracho R, Aguirre R,
evaluation of therapy with L-thyroxine versus no treatment Gonzlez O, Martinez I. Changes in renal function in pri-
in children with idiopathic and mild subclinical hypothy- mary hypothyroidism. Am J Kidney Dis. 1996;27:195
roidism. Horm Res Paediatr. 2012;77:376 381. 198.
489. Zen VL, Czepielewski MA, de Paula LP, Schwerz JC. Sub- 505. Kreisman SH, Hennessey JV. Consistent reversible eleva-
clinical hypothyroidism prevalence in children evaluated tions of serum creatinine levels in severe hypothyroidism.
for short stature. In: Proceedings from the 91st Annual Arch Intern Med. 1999;159:79 82.
Meeting of The Endocrine Society; June 2009; Washing- 506. Mooraki A, Broumand B, Neekdoost F, Amirmokri P,
ton, DC. Bastani B. Reversible acute renal failure associated with
490. Chase HP, Garg SK, Cockerham RS, Wilcox WD, Walra- hypothyroidism: report of four cases with a brief review of
vens PA. Thyroid hormone replacement and growth of literature. Nephrology. 2003;8:57 60.
children with subclinical hypothyroidism and diabetes. 507. Shin DH, Lee MJ, Kim SJ, et al. Preservation of renal func-
Diabet Med. 1990;7:299 303. tion by thyroid hormone replacement therapy in chronic
491. Wu T, Flowers JW, Tudiver F, Wilson JL, Punyasavatsut kidney disease patients with subclinical hypothyroidism.
N. Subclinical thyroid disorders and cognitive perfor- J Clin Endocrinol Metab. 2012;97:27322740.
mance among adolescents in the United States. BMC Pe- 508. Rhee CM, Alexander EK, Bhan I, Brunelli SM. Hypothy-
diatr. 2006;6:12. roidism and mortality among dialysis patients. Clin J Am
492. Aijaz NJ, Flaherty EM, Preston T, Bracken SS, Lane AH, Soc Nephrol. 2013;8:593 601.
Wilson TA. Neurocognitive function in children with com- 509. Acker CG, Singh AR, Flick RP, Bernardini J, Greenberg A,
pensated hypothyroidism: lack of short term effects on or Johnson JP. A trial of thyroxine in acute renal failure. Kid-
off thyroxin. BMC Endocr Disord. 2006;20;6:2. ney Int. 2000;57:293298.
493. Svensson J, Ericsson UB, Nilsson P, et al. Levothyroxine 510. Brenta G. Diabetes and thyroid disorders. Br J Diabetes
treatment reduces thyroid size in children and adolescents Vasc Dis. 2010;10:172177.
with chronic autoimmune thyroiditis. J Clin Endocrinol 511. Duntas L, Orgiazzi J, Brabant G. The interface between
Metab. 2006;91:1729 1734. thyroid and diabetes mellitus. Clin Endocrinol (Oxf).
494. Rother KI, Zimmerman D, Schwenk WF. Effect of thyroid 2011;75:19.
hormone treatment on thyromegaly in children and ado- 512. Duntas LH, Biondi B. New insights into subclinical hypo-
lescents with Hashimoto disease. J Pediatr. 1994;124: thyroidism and cardiovascular risk. Semin Thromb He-
599 601. most. 2011;37:2734.
495. Sert A, Pirgon O, Aypar E, Yilmaz H, Odabas D. Subclin- 513. Perros P, McCrimmon RJ, Shaw G, Frier BM. Frequency
ical hypothyroidism as a risk factor for the development of of thyroid dysfunction in diabetic patients: value of annual
cardiovascular disease in obese adolescents with nonalco- screening. Diabet Med. 1995;12:622 627.
holic fatty liver disease. Pediatr Cardiol. 2013;34:1166 514. Chubb SA, Davis WA, Inman Z. Prevalence and progres-
1174. sion of subclinical hypothyroidism in women with type 2

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 509

diabetes: the Fremantle Diabetes Study. Clin Endocrinol increased mortality and mild thyroid dysfunction in car-
(Oxf). 2005;62:480 486. diac patients. Arch Intern Med. 2007;167:1526 1532.
515. Gambelunghe G, Forini F, Laureti S, et al. Increased risk for 531. Molinaro S, Iervasi G, Lorenzoni V, et al. Persistence of
endocrine autoimmunity in Italian type 2 diabetic patients mortality risk in patients with acute cardiac diseases and
with GAD65 autoantibodies. Clin Endocrinol (Oxf). mild thyroid dysfunction. Am J Med Sci. 2012;343:6570.
2000;52:565573. 532. Mitchell JE, Hellkamp AS, Mark DB, et al. Thyroid func-
516. Libman IM, Sun K, Foley TP, Becker DJ. Thyroid auto- tion in heart failure and impact on mortality. JACC Heart
immunity in children with features of both type 1 and type Fail. 2013;1:48 55.
2 diabetes. Pediatr Diabetes. 2008;9:266 271. 533. Ridgway EC, McCammon JA, Benotti J. Maloof F. Acute
517. Uzunlulu M, Yorulmaz E, Oguz A. Prevalence of subclin- metabolic responses in myxedema to large doses of intra-
ical hypothyroidism in patients with metabolic syndrome. venous L-thyroxine. Ann Intern Med. 1972;77:549 555.
Endocr J. 2007;54:7176. 534. Sheu CC, Cheng MH, Tsai JR, Hwang JJ. Myxedema co-
518. Kadiyala R, Peter R, Okosieme OE. Thyroid dysfunction in ma: a well-known but unfamiliar medical emergency. Thy-
patients with diabetes: clinical implications and screening roid. 2007;17:371372.
strategies. Int J Clin Pract. 2010;64:1130 1139. 535. Wall CR. Myxedema coma: diagnosis and treatment. Am
519. Kordonouri O, Klinghammer A, Lang EB, Grters-Kies- Fam Physician. 2000;62:24852490.
lich A, Grabert M, Holl RW. Thyroid autoimmunity in 536. Rodrguez I, Fluiters E, Prez-Mndez LF, Luna R, Pramo
children and adolescents with type 1 diabetes: a multi- C, Garca-Mayor RV. Factors associated with mortality of
center survey. Diabetes Care. 2002;25:1346 1350. patients with myxoedema coma: prospective study in 11
520. American Diabetes Association. Standards of medical care cases treated in a single institution. J Endocrinol. 2004;
in diabetes2008. Diabetes Care. 2008;31:S12S54. 180:347350.
521. Chen HS, Wu TE, Jap TS, et al. Subclinical hypothyroid- 537. Dutta P, Bhansali A, Masoodi S, Bhadada S, Sharma N,
ism is a risk factor for nephropathy and cardiovascular Rajput R. Predictors of outcome in myxoedema coma: a
diseases in type 2 diabetic patients. Diabet Med. 2007;24: study from a tertiary care centre. Crit Care. 2008;12:R1.
1336 1344. 538. Ladenson PW, Goldenheim PD, Ridgway EC. Prediction
522. Yang JK, Liu W, Shi J, Li YB. An association between and reversal of blunted ventilatory responsiveness in pa-
subclinical hypothyroidism and sight-threatening diabetic tients with hypothyroidism. Am J Med. 1988;84:877 883.
retinopathy in type 2 diabetic patients. Diabetes Care. 539. Ferguson-Myrthil N. Novel agents for the treatment of
2010;33:1018 1020. hyponatremia: a review of conivaptan and tolvaptan. Car-
523. Kim BY, Kim CH, Jung CH, Mok JO, Suh KI, Kang SK. diol Rev. 2010;18:313321.
Association between subclinical hypothyroidism and se- 540. Pereira VG, Haron ES, Lima-Neto N, Medeiros-Neto GA.
vere diabetic retinopathy in Korean patients with type 2 Management of myxedema coma: report on three success-
diabetes. Endocr J. 2011;58:10651070. fully treated cases with nasogastric or intravenous admin-
524. Sathyapalan T, Manuchehri AM, Rigby AS, Atkin SL. Sub- istration of triiodothyronine. J Endocrinol Invest. 1982;
clinical hypothyroidism is associated with reduced all- 5:331334.
cause mortality in patients with type 2 diabetes. Diabetes 541. Yamamoto T, Fukuyama J, Fujiyoshi A. Factors associated
Care. 2010;33:e37. with mortality of myxedema coma: report of eight cases
525. Chubb SA, Davis WA, Davis TM. Subclinical hypothy- and literature survey. Thyroid. 1999;9:11671174.
roidism and mortality in women with type 2 diabetes. Clin 542. Kuma K, Matsuzuka F, Kobayashi A, et al. Outcome of
Endocrinol (Oxf). 2006;64:476 477. long standing solitary thyroid nodules. World J Surg.
526. Handisurya A, Pacini G, Tura A, Gessl A, Kautzky-Willer 1992;16:583587; discussion 587588.
A. Effects of thyroxine replacement therapy on glucose 543. Alexander EK, Hurwitz S, Heering JP, et al. Natural his-
metabolism in subjects with subclinical and overt hypo- tory of benign solid and cystic thyroid nodules. Ann Intern
thyroidism. Clin Endocrinol (Oxf). 2008;69:963969. Med. 2003;138:315318.
527. Marzouka G, Cortazar F, Alvarez JA, Dias A, Hebert K. 544. Biondi B, Kahaly GJ. Cardiovascular involvement in pa-
Racial and sex differences in prevalence of hypothyroidism tients with different causes of hyperthyroidism. Nat Rev
in patients with cardiomyopathies enrolled into a heart Endocrinol. 2010;6:431 443.
failure disease management program. Congest Heart Fail. 545. Paschke R, Hegeds L, Alexander E, Valcavi R, Papini E,
2011;17:133139. Gharib H. Thyroid nodule guidelines: agreement, disagree-
528. Iacoviello M, Guida P, Guastamacchia E, et al. Prognostic ment and need for future research. Nat Rev Endocrinol.
role of sub-clinical hypothyroidism in chronic heart failure 2011;7:354 361.
outpatients. Curr Pharm Des. 2008;14:2686 2692. 546. Gharib H, Papini E, Paschke R. Thyroid nodules: a review
529. Jessup M, Abraham WT, Casey DE, et al. 2009 focused of current guidelines, practices, and prospects. Eur J En-
update: ACCF/AHA Guidelines for the Diagnosis and docrinol. 2008;159:493505.
Management of Heart Failure in Adults: a report of the 547. Greer MA, Astwood EB. Treatment of simple goiter with
American College of Cardiology Foundation/American thyroid. J Clin Endocrinol Metab. 1953;13:13121331.
Heart Association Task Force on Practice Guidelines: de- 548. Zelmanovitz F, Genro S, Gross JL. Suppressive therapy
veloped in collaboration with the International Society for with levothyroxine for solitary thyroid nodules: a double-
Heart and Lung Transplantation. Circulation. 2009;119: blind controlled clinical study and cumulative meta-anal-
19772016. yses. J Clin Endocrinol Metab. 1998;83:38813885.
530. Iervasi G, Molinaro S, Landi P, et al. Association between 549. Castro MR, Caraballo PJ, Morris JC. Effectiveness of thy-

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
510 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

roid hormone suppressive therapy in benign solitary thy- patients: a single-center experience. Endocrine. 2011;39:
roid nodules: a meta-analysis. J Clin Endocrinol Metab. 259 265.
2002;87:4154 4159. 563. Kim SS, Lee BJ, Lee JC, et al. Preoperative serum thyroid
550. Richter B, Neises G, Clar C. Pharmacotherapy for thyroid stimulating hormone levels in well-differentiated thyroid
nodules. A systematic review and meta-analysis. Endocri- carcinoma is a predictive factor for lateral lymph node
nol Metab Clin North Am. 2002;31:699 722. metastasis as well as extrathyroidal extension in Korean
551. Sdano MT, Falciglia M, Welge JA, Steward DL. Efficacy of patients: a single-center experience. Endocrine. 2011;39:
thyroid hormone suppression for benign thyroid nodules: 259 265.
meta-analysis of randomized trials. Otolaryngol Head 564. Holm LE, Blomgren H, Lwhagen T. Cancer risks in pa-
Neck Surg. 2005;133:391396. tients with chronic lymphocytic thyroiditis. N Engl J Med.
552. Yousef A, Clark J, Doi SA. Thyroxine suppression therapy 1985;312:601 604.
for benign, non-functioning solitary thyroid nodules: a 565. Boi F, Lai ML, Marziani B, Minerba L, Faa G, Mariotti S.
quality-effects meta-analysis. Clin Med Res. 2010;8:150 High prevalence of suspicious cytology in thyroid nodules
158. associated with positive thyroid autoantibodies. Eur J En-
553. Koc M, Ersoz HO, Akpinar I, Gogas-Yavuz D, Deyneli O, docrinol. 2005;153:637 642.
Akalin S. Effect of low- and high-dose levothyroxine on 566. Fiore E, Rago T, Latrofa F, et al. Hashimotos thyroiditis
thyroid nodule volume: a crossover placebo-controlled is associated with papillary thyroid carcinoma: role of TSH
trial. Clin Endocrinol (Oxf). 2002;57:621 628. and of treatment with L-thyroxine. Endocr Relat Cancer.
554. Papini E, Petrucci L, Guglielmi R, et al. Long-term changes 2011;18:429 437.
in nodular goiter: a 5-year prospective randomized trial of 567. McLeod DS, Watters KF, Carpenter AD, Ladenson PW,
levothyroxine suppressive therapy for benign cold thyroid Cooper DS, Ding EL. Thyrotropin and thyroid cancer di-
nodules. J Clin Endocrinol Metab. 1998;83:780 783. agnosis: a systematic review and dose-response meta-anal-
555. Wmeau JL, Caron P, Schvartz C, et al. Effects of thyroid- ysis. J Clin Endocrinol Metab. 2012;97:26822692.
stimulating hormone suppression with levothyroxine in 568. Biondi B, Filetti S, Schlumberger M. Thyroid-hormone
reducing the volume of solitary thyroid nodules and im- therapy and thyroid cancer: a reassessment. Nat Clin Pract
proving extranodular nonpalpable changes: a randomized, Endocrinol Metab. 2005;1:32 40.
double-blind, placebo-controlled trial by the French Thy- 569. Biondi B, Cooper DS. Benefits of thyrotropin suppression
roid Research Group. J Clin Endocrinol Metab. 2002;87: versus the risks of adverse effects in differentiated thyroid
4928 4934. cancer. Thyroid. 2010;20:135146.
556. Hegedus L, Paschke R, Krohn K, Bonnema SJ. Multinod- 570. Brabant G. Thyrotropin suppressive therapy in thyroid
ular goiter. In: Jameson JL, DeGroot L, eds. Endocrinol- carcinoma: what are the targets? J Clin Endocrinol Metab.
ogy. 6th ed. Philadelphia, PA: Saunders Elsevier; 2010: 2008;93:11671169.
1636 1649. 571. Balme HW. Metastatic carcinoma of the thyroid success-
557. Grussendorf M, Reiners C, Paschke R, Wegscheider K; fully treated with thyroxine. Lancet. 1954;266:812 813.
LISA Investigators. Reduction of thyroid nodule volume 572. Crile G Jr. Endocrine dependency of papillary carcinomas
by levothyroxine and iodine alone and in combination: a of the thyroid. JAMA. 1966;195:721724.
randomized, placebo-controlled trial. J Clin Endocrinol 573. Mazzaferri EL, Jhiang SM. Long-term impact of initial
Metab. 2011;96:2786 2795. surgical and medical therapy on papillary and follicular
558. Boelaert K, Horacek J, Holder RL, Watkinson JC, Shep- thyroid cancer. Am J Med. 1994;97:418 428.
pard MC, Franklyn JA. Serum thyrotropin concentration 574. Cady B, Cohn K, Rossi RL, et al. The effect of thyroid
as a novel predictor of malignancy in thyroid nodules in- hormone administration upon survival in patients with dif-
vestigated by fine-needle aspiration. J Clin Endocrinol ferentiated thyroid carcinoma. Surgery. 1983;94:978
Metab. 2006;91:4295 4301. 983.
559. Haymart MR, Repplinger DJ, Leverson GE, et al. Higher 575. Pujol P, Daures JP, Nsakala N, Baldet L, Bringer J, Jaffiol
serum thyroid stimulating hormone level in thyroid nodule C. Degree of thyrotropin suppression as a prognostic de-
patients is associated with greater risks of differentiated terminant in differentiated thyroid cancer. J Clin Endocri-
thyroid cancer and advanced tumor stage. J Clin Endocri- nol Metab. 1996;81:4318 4323.
nol Metab. 2008;93:809 814. 576. McGriff NJ, Csako G, Gourgiotis L, Lori CG, Pucino F,
560. Fiore E, Rago T, Provenzale MA, et al. Lower levels of TSH Sarlis NJ. Effects of thyroid hormone suppression therapy
are associated with a lower risk of papillary thyroid cancer on adverse clinical outcomes in thyroid cancer. Ann Med.
in patients with thyroid nodular disease: thyroid autonomy 2002;34:554 564.
may play a protective role. Endocr Relat Cancer. 2009; 577. Cooper DS, Specker B, Ho M, et al. Thyrotropin suppres-
16:12511260. sion and disease progression in patients with differentiated
561. Haymart MR, Glinberg SL, Liu J, Sippel RS, Jaume JC, thyroid cancer: results from the National Thyroid Cancer
Chen H. Higher serum TSH in thyroid cancer patients oc- Treatment Cooperative Registry. Thyroid. 1998;8:737
curs independent of age and correlates with extrathyroidal 744.
extension. Clin Endocrinol (Oxf). 2009;71:434 439. 578. Jonklaas J, Sarlis NJ, Litofsky D, et al. Outcomes of pa-
562. Kim SS, Lee BJ, Lee JC, et al. Preoperative serum thyroid tients with differentiated thyroid carcinoma following ini-
stimulating hormone levels in well-differentiated thyroid tial therapy. Thyroid. 2006;16:1229 1242.
carcinoma is a predictive factor for lateral lymph node 579. Hovens GC, Stokkel MP, Kievit J, et al. Associations of
metastasis as well as extrathyroidal extension in Korean serum thyrotropin concentrations with recurrence and

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
doi: 10.1210/er.2013-1083 edrv.endojournals.org 511

death in differentiated thyroid cancer. J Clin Endocrinol 596. Cheng SY, Leonard JL, Davis PJ. Molecular aspects of thy-
Metab. 2007;92:2610 2615. roid hormone actions. Endocr Rev. 2010;31:139 170.
580. Sugitani I, Fujimoto Y. Does postoperative thyrotropin 597. Baxter JD, Dillmann WH, West BL, et al. Selective mod-
suppression therapy truly decrease recurrence in papillary ulation of thyroid hormone receptor action. J Steroid
thyroid carcinoma? A randomized controlled trial. J Clin Biochem Mol Biol. 2001;76:31 42.
Endocrinol Metab. 2010;95:4576 4583. 598. Moreno M, de Lange P, Lombardi A, Silvestri E, Lanni A,
581. Diessl S, Holzberger B, Mder U, Grelle I, Smit JW, Buck Goglia F. Metabolic effects of thyroid hormone deriva-
AK, Reiners C, Verburg FA. Impact of moderate vs strin- tives. Thyroid. 2008;18:239 253.
gent TSH suppression on survival in advanced differenti- 599. Silvestri E, Glinni D, Cioffi F, et al. Metabolic effects of the
ated thyroid carcinoma. Clin Endocrinol (Oxf). 2012;76: iodothyronine functional analogue TRC150094 on the
586 592. liver and skeletal muscle of high-fat diet fed overweight
582. Pacini F, Schlumberger M, Dralle H, et al; European Thy- rats: an integrated proteomic study. Mol Biosyst. 2012;8:
roid Cancer Taskforce. European consensus for the man- 19872000.
agement of patients with differentiated thyroid carcinoma 600. Cioffi F, Lanni A, Goglia F. Thyroid hormones, mitochon-
of the follicular epithelium. Eur J Endocrinol. 2006;154: drial bioenergetics and lipid handling. Curr Opin Endo-
787 803. crinol Diabetes Obes. 2010;17:402 407.
583. Sundram F, Robinson BG, Kung A, et al. Well-differenti- 601. The coronary drug project. Findings leading to further
ated epithelial thyroid cancer management in the Asia Pa- modifications of its protocol with respect to dextrothyrox-
cific region: a report and clinical practice guideline. Thy- ine. The coronary drug project research group. JAMA.
roid. 2006;16:461 469. 1972;220:996 1008.
584. Pitoia, F, Ward L, Wohllk N, et al. Recommendations of 602. Liang H, Juge-Aubry CE, OConnell M, Burger A G. Or-
the Latin American Thyroid Society on diagnosis and man- gan-specific effects of 3,5,3-triiodothyroacetic acid in
agement of differentiated thyroid cancer. Arq Bras Endo- rats. Eur J Endocrinol. 1997;137:537544.
crinol Metabol. 2009;53:884 897. 603. Sherman SI, Ringel MD, Smith MJ, Kopelen HA, Zoghbi
585. British Thyroid Association, Royal College of Physicians.
WA, Ladenson PW. Augmented hepatic and skeletal thy-
Guidelines for management of thyroid cancer. British Thy-
romimetic effects of tiratricol in comparison with levothy-
roid Foundation website. http://www.rcplondon.ac.uk/
roxine. J Clin Endocrinol Metab. 1997;82:21532158.
publications/guidelines-management-thyroid-cancer. Ac-
604. Takeda T, Suzuki S, Liu RT, DeGroot LJ. Triiodothy-
cessed December 1, 2007.
roacetic acid has unique potential for therapy of resistance
586. Zafn C. TSH-suppressive treatment in differentiated thy-
to thyroid hormone. J Clin Endocrinol Metab. 1995;80:
roid cancer. A dogma under review. Endocrinol Nutr.
20332040.
2012;59:125130.
605. Beck-Peccoz P, Piscitelli G, Cattaneo MG, Faglia G. Suc-
587. Hellevik AI, Asvold BO, Bjro T, Romundstad PR, Nilsen
cessful treatment of hyperthyroidism due to nonneoplastic
TI, Vatten LJ. Thyroid function and cancer risk: a pro-
pituitary TSH hypersecretion with 3,5,3-triiodothy-
spective population study. Cancer Epidemiol Biomarkers
roacetic acid (TRIAC). J Endocrinol Invest. 1983;6:217
Prev. 2009;18:570 574.
588. Bergh JJ, Lin HY, Lansing L, et al. Integrin V3 contains 223.
a cell surface receptor site for thyroid hormone that is 606. Chiellini G, Apriletti JW, Yoshihara HA, Baxter JD, Ri-
linked to activation of mitogen-activated protein kinase beiro RC, Scanlan TS. A high-affinity subtype-selective ag-
and induction of angiogenesis. Endocrinology. 2005;146: onist ligand for the thyroid hormone receptor. Chem Biol.
2864 2871. 1998;5:299 306.
589. Hannon MJ, Behan LA, Agha A. Thyrotoxic periodic pa- 607. Lin V, Klepp H, Hanley R. Sobetirome is a thyroid hor-
ralysis due to excessive L-thyroxine replacement in a Cau- mone receptor - and liver selective thyromimetic that can
casian man. Ann Clin Biochem. 2009;46:423 425. effect substantial LDL-C lowering without significant
590. Wilsons Temperature Syndrome website. www.wilson- changes in heart rate or the thyroid axis in euthyroid men.
syndrome.com. In: Proceedings from the 90th Annual Meeting of The En-
591. Barsky AJ, Borus JF. Functional somatic syndromes. Ann docrine Society, June 1518, 2008; San Francisco, CA.
Intern Med. 1999;130:910 921. OR36 OR33.
592. American Thyroid Association. American Thyroid Asso- 608. Villicev CM, Freitas FR, Aoki MS, et al. Thyroid hormone
ciation Statement on Wilsons Syndrome 2005. Posted receptor -specific agonist GC-1 increases energy expen-
on May 24, 2005. http://www.thyroid.org/american- diture and prevents fat-mass accumulation in rats. J En-
thyroid-association-statement-on-wilsons-syndrome/. docrinol. 2007;193:2129.
593. Kaptein EM, Beale E, Chan LS. Thyroid hormone therapy 609. Bryzgalova G, Effendic S, Khan A, et al. Anti-obesity, anti-
for obesity and nonthyroidal illnesses: a systematic review. diabetic, and lipid lowering effects of the thyroid receptor
J Clin Endocrinol Metab. 2009;94:36633675. subtype selective agonist KB-141. J Steroid Biochem Mol
594. Utiger RD. Altered thyroid function in nonthyroidal illness Biol. 2008;111:262267.
and surgery. To treat or not to treat? N Engl J Med. 1995; 610. Ladenson PW, McCarren M, Morkin E, et al. Effects of the
333:15621563. thyromimetic agent diiodothyropropionic acid on body
595. Baxter JD, Webb P. Thyroid hormone mimetics: potential weight, body mass index, and serum lipoproteins: a pilot
applications in atherosclerosis, obesity and type 2 diabetes. prospective, randomized, controlled study. J Clin Endo-
Nat Rev Drug Discov. 2009;8:308 320. crinol Metab. 2010;95:1349 1354.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.
512 Biondi and Wartofsky Thyroid Hormone Treatment Endocrine Reviews, June 2014, 35(3):433512

611. Sweetlove M. Phase III trial of eprotirome: discontinued 626. Hamilton MA, Stevenson LW, Fonarow GC, Steimle A,
for safety concerns. Pharm Med. 2012;26:3. Goldhaber JI, Child JS, Chopra IJ, Moriguchi JD, Hage A.
612. Wartofsky L, Burman K. Alterations in thyroid function in Safety and hemodynamic effects of intravenous triiodo-
patients with systemic illness: the euthyroid sick syn- thyronine in advanced congestive heart failure. Am J Car-
drome. Endocr Rev. 1982;3:164 217. diol. 1998;81:443 447.
613. Opasich C, Pacini F, Ambrosino N, et al. Sick euthyroid 627. Moruzzi P, Doria E, Agostoni PG, Capacchione V, Sgan-
syndrome in patients with moderate-to-severe chronic zerla P. Usefulness of L-thyroxine to improve cardiac and
heart failure. Eur Heart J. 1996;17:1860 1866. exercise performance in idiopathic dilated cardiomyopa-
614. Pingitore A, Landi P, Taddei MC, Ripoli A, LAbbate A, thy. Am J Cardiol. 1994;73:374 378.
Iervasi G. Triiodothyronine levels for risk stratification of 628. Moruzzi P, Doria E, Agostoni PG. Medium-term effective-
patients with chronic heart failure. Am J Med. 2005;118: ness of L-thyroxine treatment in idiopathic dilated cardio-
132136. myopathy. Am J Med. 1996;101:461 467.
615. Gerdes AM, Iervasi G Thyroid replacement therapy and 629. Malik FS, Mehra MR, Uber PA, Park MH, Scott RL, Van
heart failure. Circulation. 2010;122:385393. Meter CH. Intravenous thyroid hormone supplementation
616. Friberg L, Drvota V, Bjelak AH, Eggertsen G, Ahnve S. in heart failure with cardiogenic shock. J Card Fail. 1999;
Association between increased levels of reverse triiodothy- 5:3137.
ronine and mortality after acute myocardial infarction. 630. Iervasi G, Emdin M, Colzani RM, et al. Beneficial effects of
Am J Med. 2001;111:699 703. long-term triiodothyronine (T3) infusion in patients with
617. Hamilton MA, Stevenson LW, Luu M, Walden JA. Altered advanced heart failure and low T3 syndrome. In: Kimchi A,
thyroid hormone metabolism in advanced heart failure. ed. Second International Congress on Heart Disease: New
Trends in Research, Diagnosis and Treatment. Engle-
J Am Coll Cardiol. 1990;16:9195.
wood, NJ: Medimond Medical Publications; 2001:549
618. Coceani M, Iervasi G, Pingitore A, Carpeggiani C,
553.
LAbbate A. Thyroid hormone and coronary artery dis-
631. Pingitore A, Galli E, Barison A, et al. Acute effects of tri-
ease: from clinical correlations to prognostic implications.
iodothyronine (T3) replacement therapy in patients with
Clin Cardiol. 2009;32:380 385.
chronic heart failure and low-T3 syndrome: a randomized,
619. Kozdag G, Ural D, Vural A, et al. Relation between free
placebo-controlled study. J Clin Endocrinol Metab. 2008;
triiodothyronine/free thyroxine ratio, echocardiographic
93:13511358.
parameters and mortality in dilated cardiomyopathy. Eur
632. Pennock GD, Raya TE, Bahl JJ, Goldman S, Morkin E.
J Heart Fail. 2005;7:113118.
Cardiac effects of 3,5-diiodothyropropionic acid, a thyroid
620. Iervasi G, Pingitore A, Landi P, et al. Low-T3 syndrome: a hormone analog with inotropic selectivity. J Pharmacol
strong prognostic predictor of death in patients with heart Exp Ther. 1992;263:163169.
disease. Circulation. 2003;107:708 713. 633. Morkin E, Pennock GD, Raya TE, Bahl JJ, Goldman S.
621. Klemperer JD, Klein I, Gomez M, et al. Thyroid hormone Development of a thyroid hormone analogue for the treat-
treatment after coronary-artery bypass surgery. N Engl ment of congestive heart failure. Thyroid. 1996;6:521
J Med. 1995;333:15221527. 526.
622. Ladenson PW, Sherman SI, Baughman KL, Ray PE, Feld- 634. Morkin E, Pennock GD, Spooner PH, et al. Clinical and
man AM. Reversible alterations in myocardial gene ex- experimental studies on the use of 3,5-diiodothyropropi-
pression in a young man with dilated cardiomyopathy and onic acid, a thyroid hormone analogue, in heart failure.
hypothyroidism. Proc Natl Acad Sci U S A. 1992;89:5251 Thyroid. 2002;12:527533.
5255. 635. Wang X, Zheng W, Christensen LP, Tomanek RJ. DITPA
623. Katzeff HL, Powell SR, Ojamaa K. Alterations in cardiac stimulates bFGF, VEGF, angiopoietin, and Tie-2 and fa-
contractility and gene expression during low-T3 syndrome: cilitates coronary arteriolar growth. Am J Physiol Heart
prevention with T3. Am J Physiol. 1997;273:E951E956. Circ Physiol. 2003;284:H613H618.
624. Kinugawa K, Minobe WA, Wood, et al. Signaling path- 636. Kaptein EM, Sanchez A, Beale E, Chan LS. Thyroid hor-
ways responsible for fetal gene induction in the failing hu- mone therapy for postoperative nonthyroidal illnesses: a
man heart: evidence for altered thyroid hormone receptor systematic review and synthesis. J Clin Endocrinol Metab.
gene expression. Circulation. 2001;103:1089 1094. 2010;95:4526 4534.
625. Taegtmeyer H, Sen S, Vela D. Return to the fetal gene 637. Stathatos N, Wartofsky L. Perioperative management of
program: a suggested metabolic link to gene expression in patients with hypothyroidism. Endocrinol Metab Clin
the heart. Ann N Y Acad Sci. 2010;1188:191198. North Am. 2003;32:503518.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 06 July 2016. at 00:03 For personal use only. No other uses without permission. . All rights reserved.

You might also like