Change in Basal Ganglia Volume Over 2 Years

in Patients With Schizophrenia:

Typical Versus Atypical Neuroleptics
Patricia Westmoreland Corson, M.D., Peg Nopoulos, M.D., Del D. Miller, Pharm.D., M.D.,
Stephan Arndt, Ph.D., and Nancy C. Andreasen, M.D., Ph.D.

Objective: For many years, it has been assumed that medications affect brain chemistry and physiology but not structure. Recent reports suggest that neuroleptic medication
changes basal ganglia volume. To explore this possibility, the authors assessed for basal
ganglia volume change in individuals who had their basal ganglia structures delineated
and measured on magnetic resonance scans at the beginning and end of a 2-year period
and who received neuroleptic medication during this time. Method: The basal ganglia volumes of 23 male patients with schizophrenia spectrum disorders were measured from
manual traces delineating the caudate and lenticular nucleus on magnetic resonance images at admission and 2 years later. Patients neuroleptic exposure was calculated over
the 2 years by using a dose-year formula. Results: During the 2-year period, mean basal
ganglia volume of patients receiving predominantly typical neuroleptics increased, while
the opposite was observed for patients receiving mostly atypical neuroleptics. Correlation
analysis for the entire group showed a positive relationship between the 2-year exposure
to typical neuroleptic medication and change in basal ganglia volume and the reverse for
exposure to atypical neuroleptics. Conclusions: In this group, basal ganglia volume increased following exposure to typical neuroleptics and decreased following exposure to
atypical neuroleptics.
(Am J Psychiatry 1999; 156:12001204)

espite discrepancies among the findings of several studies (13), most magnetic resonance imaging
(MRI) and postmortem studies to date have found an
increase in the volume of basal ganglia structures in
patients with schizophrenia compared with age- and
gender-matched comparison subjects (1, 49). It was
originally thought that this increase in volume was
caused by an aberration in neuronal pruning in patients with schizophrenia (4, 6). This theory was presented in early reports as a failure of maturational
synaptic elimination that would normally reduce the
basal ganglia volume during adolescence (4) or a compensatory response to lessened input from the antePresented in part at the International Congress on Schizophrenia Research, Colorado Springs, Colo., April 1216, 1997.
Received April 7, 1998; revision received Jan. 4, 1999; accepted
Feb. 25, 1999. From the Department of Psychiatry, University of
Iowa College of Medicine. Address reprint requests to Dr. Corson,
Mental Health Clinical Research Center, Department of Psychiatry, University of Iowa College of Medicine, Iowa City, IA 52242;
patricia-corson@uiowa.edu (e-mail).
Supported by NIMH grants MH-31593 and MH-40856 and Clinical Research Center grant MH-43271.

1200

rior, temporal, frontal, or thalamic regions (6). More

recent evidence suggests that the effects of neuroleptic
treatment may instead be responsible for the observed
enlargement (812). This hypothesis is based on the
observation of cumulative effects of neuroleptic medication in first-episode, neuroleptic-naive patients who
were scanned at intake and again after several months
of treatment with typical neuroleptics (12). Because
similar observations have been made in patients with
nonschizophrenic psychiatric illness (affective disorder) who were treated with typical neuroleptics (13),
it is probable that the volume increase is caused by
medication, rather than a disease-specific manifestation of schizophrenia.
Two recent studies from our center, using single photon emission computed tomography and positron emission tomography, respectively, reported a relative increase in cerebral blood flow to the basal ganglia
following treatment with neuroleptic medication (14,
15). We hypothesized that neuroleptic-induced receptor
blockade caused an increase in basal ganglia blood flow
and metabolism and that the increase in basal ganglia
Am J Psychiatry 156:8, August 1999

CORSON, NOPOULOS, MILLER, ET AL.

volume previously reported (1, 49) could be attributed

to medication-induced changes in blood flow to the region or changes in receptor proliferation(15).
With the advent of atypical neuroleptic drugs, the effects of typical and atypical neuroleptics, which have
differing affinities for dopamine D2 receptor sites, can
be compared. It has been reported that the volume increase caused by typical neuroleptics is reversed following exposure to the atypical neuroleptic clozapine
(9, 16). Since these studies were initiated, both risperidone and olanzapine have been added to the atypical
neuroleptic armamentarium. Sufficient time has
elapsed since their introduction to gather data on patients scanned at the beginning and end of a 1- to 2year time period, as previously described for clozapine.
To test the proposed relationship between neuroleptic exposure and basal ganglia volume change, we evaluated 23 male patients with schizophrenia spectrum
disorders. The patients exposure to typical and atypical neuroleptics was calculated between times of intake
and 2-year follow-up scans by using a dose-year equation (17). Basal ganglia volumes (caudate and lenticular nucleus) were measured from manual tracings on
the T1-weighted intake and follow-up scans (figure 1).
METHOD
We studied 23 male patients with diagnoses in the schizophrenia
spectrum: schizophrenia (N=19), psychosis not otherwise specified
(N=3), and schizotypal disorder (N=1). Diagnoses were made according to DSM-IV on the basis of a structured interview, the Comprehensive Assessment of Symptoms and History (18). Patients were
chosen for this study according to the stipulation that their initial diagnosis of schizophrenia spectrum disorder remained within that
spectrum by the time of their 2-year follow-up. The mean age of patients at intake was 25.57 years (SD=6.19), and the mean age at illness onset was 22.44 years (SD=6.11). After complete description of
the study to each patient, written informed consent was obtained.
The following neuroleptic drugs were used with these patients:
typical neurolepticshaloperidol, trifluoperazine, thiothixene,
fluphenazine, thioridazine, perphenazine, and chlorpromazine; and
atypical neurolepticsclozapine, risperidone, and olanzapine. The
dose-year formula was used to measure neuroleptic exposure. Neuroleptic exposure was calculated as cumulative dose at the time of
intake scan and further neuroleptic dose accumulated during the 2
years between intake and follow-up scans (17). This required conversion of neuroleptic medication to chlorpromazine equivalents
(19). The older equivalents have recently been extrapolated to atypical neuroleptics, by using clinical efficacy as the criterion for determining equivalency as was done for typical neuroleptics in the work
of Davis (19, cited in reference 20). Exposure was then calculated
over time and weighted for dose. This was a naturalistic study of 23
patients, most of whom had already been medicated to some extent.
At the time of the intake scan, four patients were neuroleptic naive,
14 had been treated with typical neuroleptics only, and five patients
had been treated with both typical and atypical neuroleptics.
Univariate analysis indicated that mean dose years of typical neuroleptics for the entire study group (N=23) at intake was 4.24 (SD=
8.17, range=037.25, median=1.25); similarly, mean dose years of
atypical neuroleptics for the entire study group (N=23) at intake was
0.81 (SD=3.18, range=015.07, median=0).
MRI scans were obtained by using a 1.5-T General Electric Signa
scanner (GE Medical Systems, Milwaukee) at the beginning and end
of the 2-year period. T1-weighted images, using a spoiled grass sequence, were acquired with the following parameters: 1.5-mm coronal slices, 40 flip angle, TR=24 msec, TE=5 msec, number of exci-

Am J Psychiatry 156:8, August 1999

FIGURE 1. Basal Ganglia Structures (Caudate Nuclei and

Lenticular Nuclei) Delineated on T1 Image of a Schizophrenic
Patient

tations=2, field of view=26 cm, and a matrix of 256256192 cm.

Scans were visually assessed for quality and movement artifacts by
using this scale: 4.0=excellent, no obvious movement; 3.0=good,
minimal movement; 2.0=fair, some movement; 1.0=poor, substandard. All scans in this study received quality ratings of 2.5 to 4.0.
The caudate and lenticular nucleus were manually traced on contiguous 1-mm-thick coronal slices by using BRAINS, a software package developed by the Iowa Mental Health Clinical Research Center
(21). The following criteria were established to ensure consistent and
reliable tracing of these two major components of the basal ganglia
(caudate and lenticular nucleus). Each component structure was
traced on coronal slices from its rostral to caudal aspect, first on the
right side of the image, then on the left side. The sequence of tracing
was right caudate, left caudate, right lenticular nucleus, and left lenticular nucleus. Tracing commenced when the structure could be
seen by the naked eye and ended when the structure was no longer
discernible. The tracer was blind to the treatment status of patients.
Taken individually, each major component had several defining
characteristics that were used to guide tracing of the structure. Complete tracing guidelines are available from Dr. Corson on request. Interrater reliabilities (intraclass r) for the traced regions of interest
were as follows: right caudate=0.81, left caudate=0.73, right lenticular nucleus=0.60, and left lenticular nucleus=0.79.

RESULTS

During the 2 years (mean=2.17 years, SD=0.42,

range=1.183.67, median=2.13) between intake and
follow-up scans, additional neuroleptic dose years
were accumulated by the entire study group (N=23) as
follows: 1) for typical neuroleptics, mean=5.60 dose
years (SD=6.55, range=019.36, median=2.69); 2) for
atypical neuroleptics, mean=5.29 dose years (SD=7.83,
range=027.76, median=2.26). During the 2-year follow-up period, 13 patients were treated almost exclusively with typical neuroleptics: eight of the 13 had
been treated only with typical neuroleptics, whereas
five had also been minimally exposed to atypical neu1201

CHANGE IN BASAL GANGLIA VOLUME

TABLE 1. Change Over 2 Years in Volume of Basal Ganglia

Structures of Schizophrenic Patients Receiving Typical (N=13)
or Atypical (N=10) Neuroleptics
Brain Area
Total basal ganglia
Caudate
Left
Right
Putamen
Left
Right
Globus pallidus
Left
Right

Volume Change (cc)

Typical
Atypical
0.47
0.23
0.18
0.33
0.33
0.16
0.45
0.51
0.58
0.38

0.46
0.39
0.39
0.33
0.20
0.16
0.17
0.44
0.32
0.57

roleptics. For typical neuroleptics, mean=9.05 dose

years (SD=6.89, range=0.6019.36, median=6.74); for
atypical neuroleptics, mean=0.93 dose years (SD=1.53;
range=04.77, median=0). The remaining 10 patients
had been exposed mostly to atypical drugs over the
previous 2 years. Six of these 10 patients were treated
exclusively with atypical neuroleptics (clozapine, olanzapine, or risperidone) during the follow-up period,
while four had minimal exposure to typical neuroleptics (primarily haloperidol); this exposure occurred
during the early part of the follow-up period in all
cases. For atypical neuroleptics, mean=10.96 dose
years (SD=9.14, range=1.2527.76, median=7.18),
and for typical neuroleptics, mean=1.14 dose years
(SD=1.45, range=03.28, median=0).
When viewed as two separate groups in this manner,
exposure to neuroleptics at intake for the two groups
was as follows: group 1: N=13, typical neuroleptic
mean=4.38 dose years (SD=10.11, range=037.25,
median=1.25), atypical neuroleptic mean=0.25 dose
years (SD=0.88, range=03.17, median=0); group 2:
N=10, typical neuroleptic mean=4.08 dose years (SD=
5.17, range=0.0915.00, median=1.61), atypical neuroleptic mean=1.53 dose years (SD=4.76, range=0
15.07, median=0). There was no significant difference
between the mean ages of patients in the two groups.
Age at intake was as follows: group 1: N=13, age at
first scan, mean=25.77 years (SD=5.80, range=19.00
38.00, median=24.00); group 2: N=10, age at first
scan, mean=25.30 years (SD=6.98, range=18.00
41.00, median=23.50). There was also no significant
difference in the length of the time interval between the
time 1 and time 2 scans across groups: group 1: N=13,
time interval between first and follow-up scans, mean=
2.12 years (SD=0.54, range=1.183.67, median=2.12);
group 2: N=10, time interval between first and followup scans, mean=2.21 years (SD=0.18, range=2.00
2.61, median=2.17).
Change in basal ganglia volume over the 2-year interval was compared in these two groups by using a
simple repeated measures analysis of variance. There
was a significant group-by-time interaction (F=12.92,
df=1, 21, p<0.002). This indicates that the changes
over time differ for the two groups (table 1). Follow-up
t tests showed a significant mean decrease in basal gan1202

glia volume for the group of patients who received

mostly atypical neuroleptic medication (mean=0.99
cc, SD=1.06) (paired t test=2.93, df=9, p<0.02, twotailed). By contrast, the group medicated mostly with
typical neuroleptics showed a significant increase in
basal ganglia volume (mean=0.52 cc, SD=0.94) (paired
t test=1.98, df=12, p<0.04, one-tailed, based on a priori prediction of increase in size; this was the only onetailed t test used).
Correlation analysis (two-tailed) for the entire study
group showed a significant inverse relationship between exposure to atypical neuroleptics and basal ganglia volume change (rs=0.46, df=21, p<0.03) and a
significant positive relationship with typical neuroleptics (rs=0.47, df=21, p<0.03). Using multiple regression
analysis, we found that the overall basal ganglia volume change was not significantly accounted for by volume change of any one component structure, although
direction of volume change (i.e., greater caudate and
lenticular nucleus volume with typical neuroleptic exposure and the reverse with exposure to atypical neuroleptics) was in each case consistent with the model
anticipated for that particular class of drug.
DISCUSSION

The basal ganglia structures mediate many of the

cognitive and behavioral processes disrupted in schizophrenia (4, 68, 22). Because of their high density of
dopamine D2 receptors, the basal ganglia structures
are a major target to which dopaminergic pathways
project. For many years, the clinical pharmacology of
schizophrenia has been based on the use of dopamine
receptor antagonists as a treatment modality. Typical
neuroleptics show a markedly higher dopamine D2 receptor affinity and occupancy than do atypical neuroleptics (2327), whose efficacy may also be related to
their high degree of 5-HT2 occupancy (25).
Studies of the relationship between striatal volume
and neuroleptic exposure have steadily evolved since
Chakos et al. (10) examined caudate nuclei in patients
with first-episode schizophrenia who had minimal previous exposure to neuroleptics. Following 18 months
of treatment with typical neuroleptics, a significant
volume increase was observed in the caudate nuclei of
patients versus comparison subjects, and this finding
was attributed to the effects of neuroleptic drugs on
the dopaminergic system. In this study, all drugs were
in the typical class (10). This finding has been replicated (11, 13), and Chakos et al. (16) extended their
original study to observe patients switched from typical neuroleptics to the atypical neuroleptic clozapine,
finding a decrement in caudate volumes of as much as
10% after a 1-year exposure. This result has been replicated in a small group of patients with childhood-onset schizophrenia (9).
Functional imaging studies from our center suggest
that neuroleptic treatment is associated with changes
in basal ganglia perfusion and that perfusion changes
Am J Psychiatry 156:8, August 1999

CORSON, NOPOULOS, MILLER, ET AL.

differentially depending on the class of medication

typical or atypical. Vascular engorgement resulting
from the greater perfusion could lead to a noticeable
volume change (14, 15). It is also possible that these
differential effects could result in differences in receptor size or receptor proliferation. On an ultrastructural
level, the greater number of dopamine D2 receptors in
the basal ganglia of patients with schizophrenia has
been a replicable finding in postmortem neurochemical
studies of schizophrenia (28). Such an increase in receptor density is considered by many to be a consequence of neuroleptic treatment and has been shown
to occur in animals (8, 29) as well as humans (25, 30,
31) following chronic exposure to typical neuroleptic
medication.
In this group of 23 male patients, we confirm in a
within-subjects study design that basal ganglia volumes of patients with schizophrenia, as measured over
a 2-year period, change in response to neuroleptic exposure and that the direction of the change differs for
typical and atypical neuroleptics. The direction of the
volume change correlates with the overall class of neuroleptic used. Basal ganglia volume across the entire
study group of 23 patients increased following treatment with typical neuroleptics and decreased following treatment with atypical neuroleptics.
When we examined the major basal ganglia substructures separately, none was found to contribute
significantly in itself to the overall basal ganglia volume change. However, component structures of the
basal ganglia are now often highlighted as being different with regard to dopamine D2 receptor structure and
function (22, 23, 32), and various studies reporting
basal ganglia volume increase have reported a differential increase for the major substructures (1, 4, 5, 9, 32).
Finally, another aspect that remains unclear is
whether the volume decrement observed in the basal
ganglia once medication is changed from typical to
atypical neuroleptics is a direct effect of treatment with
atypical neuroleptics or whether withdrawal of typical
neuroleptics in and of itself results in the same shrinkage pattern (9, 16). This question has so far been difficult to answer, because virtually no patient in any of
the studies reported to date has been treated solely
with atypical neuroleptic medication without first
having had some previous exposure to the typical
class. A case-by-case inspection of our study group
gives modest preliminary evidence that atypical neuroleptics could produce basal ganglia volume decrement in and of themselves and not merely as part of a
reversal of the mechanism instituted by typical neuroleptics. Four of our patients who had received relatively minor doses of typical neuroleptics at intake
(0.433.33 dose years) were then switched to atypical
neuroleptics, receiving large doses of atypical medication (between 4.73 and 21.67 dose years). These were
the four patients who experienced the largest basal
ganglia volume decrements (1.41 to 2.69 cc). Further studies of a larger number of patients exposed
Am J Psychiatry 156:8, August 1999

solely to atypical medication are needed to elucidate

whether this occurs consistently and, if so, what type
of mechanism could be responsible for basal ganglia
volume shrinkage as an ab initio response to atypical
neuroleptic medication.
REFERENCES
1. Heckers S, Heinsen H, Heinsen Y, Beckmann H: Cortex, white
matter and basal ganglia in schizophrenia: a volumetric postmortem study. Biol Psychiatry 1991; 29:556566
2. Gur RE, Pearlson GD: Neuroimaging in schizophrenia research. Schizophr Bull 1993; 19:337353
3. De Lisi LE, Hoff AL, Tew W, Sakuma M, Grimson R: Brain
structure change over time in first episode schizophrenia (abstract). Schizophr Res 1997; 24:141142
4. Jernigan TL, Zisook S, Heaton RK, Moranville JT, Hesselink
JR, Braff DL: Magnetic resonance imaging abnormalities in
lenticular nuclei and cerebral cortex in schizophrenia. Arch
Gen Psychiatry 1991; 48:881890
5. Breier A, Buchanan RW, Elkashef A, Munson RC, Kirkpatrick
B, Gellad F: Brain morphology in schizophrenia: a magnetic
resonance imaging study of limbic, prefrontal cortex and caudate structures. Arch Gen Psychiatry 1992; 49:921926
6. Swayze VW II, Andreasen NC, Alliger RJ, Yuh WT, Ehrhardt
JC: Subcortical and temporal structures in affective disorder
and schizophrenia: a magnetic resonance imaging study. Biol
Psychiatry 1992; 31:221240
7. Nasrallah HA, Chu O, Olsen SC, Martin R: Increased caudate
volume in schizophrenia: a controlled MRI study (abstract).
Schizophr Res 1993; 9:204
8. Elkashef AM, Buchanan RW, Gellad F, Munson RC, Breier A:
Basal ganglia pathology in schizophrenia and tardive dyskinesia: an MRI quantitative study. Am J Psychiatry 1994; 151:
752755
9. Frazier JA, Gied JN, Kaysen D, Albus K, Hamburger S, Alaghband-Rad J, Lenane MC, McKenna K, Breier A, Rapoport JL:
Childhood-onset schizophrenia: brain MRI rescan after 2
years of clozapine maintenance treatment. Am J Psychiatry
1996; 153:564566
10. Chakos MH, Lieberman JA, Bilder RM, Borenstein M, Lerner
G, Bogerts B, Wu H, Kinon B, Ashtari M: Increase in caudate
nuclei volumes of first-episode schizophrenic patients taking
antipsychotic drugs. Am J Psychiatry 1994; 151:14301436
11. Keshavan MS, Bagwell WW, Haas GL, Sweeney JA, Schooler
N, Pettegrew JW: Changes in caudate volume with neuroleptic treatment (letter). Lancet 1994; 344:1434
12. Rodriguez R, Flaum M, Nopoulos P: The effect of cumulative
neuroleptic exposure on caudate and putamen volume (abstract). Schizophr Res 1997; 24:154155
13. Doraiswamy PM, Tupler LA, Krishnan KRR: Neuroleptic treatment and caudate plasticity (letter). Lancet 1995; 345:734
735
14. Miller DD, Rezai K, Alliger R, Andreasen NC: The effect of antipsychotic medication on relative cerebral blood perfusion in
schizophrenia: assessment with technetium-99m hexamethylpropyleneamine oxime single photon emission computed tomography. Biol Psychiatry 1997; 41:550559
15. Miller DD, Andreasen NC, OLeary DS, Rezai K, Watkins GL,
Ponto LL, Hichwa RD: Effect of antipsychotics on regional cerebral blood flow measured with positron emission tomography. Neuropsychopharmacology 1997; 17:230240; correction, 1998; 18:323324
16. Chakos MH, Lieberman JA, Alvir J, Bilder R, Ashtari M: Caudate nuclei volumes in patients treated with typical antipsychotics or clozapine (letter). Lancet 1995; 345:456457
17. Miller DD, Flaum MA, Nopoulos P, Arndt S, Andreasen NC:
The concept of dose years: a reliable method for calculating
lifetime psychotropic drug exposure (abstract). Schizophr Res
1995; 15:159
18. Andreasen NC, Flaum M, Arndt S: The Comprehensive Assessment of Symptoms and History (CASH): an instrument

1203

CHANGE IN BASAL GANGLIA VOLUME

19.
20.
21.

22.
23.
24.

25.

26.

for assessing diagnosis and psychopathology. Arch Gen Psychiatry 1992; 49:615623
Davis JM: Dose equivalence of the antipsychotic drugs. J
Psychiatry Res 1974; 11:6569
American Psychiatric Association: Practice Guideline for the
Treatment of Patients With Schizophrenia. Am J Psychiatry
1997; 154(April suppl)
Andreasen NC, Cohen G, Harris G, Cizadlo T, Parkkinen J,
Rezai K, Swayze VW II: Image processing for the study of
brain structure and function: problems and programs. J Neuropsychiatry Clin Neurosci 1992; 4:125133
Graybiel AM: The basal ganglia and cognitive pattern generators. Schizophr Bull 1997; 23:459469
Arnt J, Skarsfeldt T, Hyttel J: Differentiation of classical and
novel antipsychotics using animal models. Int Clin Psychopharmacol 1997; 12(suppl 1):S9S17
Pehek EA, Yamamoto BK: Differential effects of locally administered clozapine and haloperidol on dopamine efflux in the
rat prefrontal cortex and caudate-putamen. J Neurochem
1994; 63:21182124
Sedvall G, Pauli S, Farde L, Karlsson P, Nyberg S, Nordstrom
A-L: Recent developments in PET scan imaging of neuroreceptors in schizophrenia. Isr J Psychiatry Relat Sci 1995; 32:
2229
Pilowsky LS, Mulligan RS, Acton PD, Ell P, Costa DC, Kerwin
RW: Limbic selectivity of clozapine. Lancet 1997; 350:490
491

1204

27. Pilowsky LS, Busatto GF, Taylor M, Costa DC, Sharma TC,
Sigmundsson T, Ell PJ, Nohria V, Kerwin RW: Dopamine D2
receptor occupancy in vivo by the novel antipsychotic olanzapinea 123I IBZM single proton emission tomography (SPET)
study. Psychopharmacology (Berl) 1996; 124:148153
28. Clardy JA, Hyde TM, Kleinman JE: Postmortem neurochemical and neuropathological studies in schizophrenia, in Schizophrenia: From Mind to Molecule. Edited by Andreasen NC.
Washington, DC, American Psychiatric Press, 1994, pp 123
145
29. Clow A, Theodorou A, Jenner P, Marsden CD: A comparison
of striatal and mesolimbic dopamine function in the rat during
6-month trifluoperazine administration. Psychopharmacology
(Berl) 1980; 69:227233
30. Seeman P, Bzowej NH, Guan HC: Human brain D1 and D2
dopamine receptors in schizophrenia, Alzheimers, Parkinsons and Huntingtons diseases. Neuropsychopharmacology
1987; 1:515
31. Joyce JN, Lexow N, Bird E: Organization of dopamine D1 and
D2 receptors in the human striatum: receptor autoradiographic studies in Huntingtons disease and schizophrenia.
Synapse 1988; 2:546557
32. Hokama H, Shenton ME, Nestor PG, Kikinis R, Levitt JJ, Metcalf D, Wible C, ODonnell B, Jolesz FA, McCarley RW: Caudate, putamen and globus pallidus volume in schizophrenia: a
quantitative MRI study. Psychiatry Res Neuroimaging 1995;
61:209229

Am J Psychiatry 156:8, August 1999