Professional Documents
Culture Documents
in Preterm Infants
Michael P. Sherman, MD, PhD1, David H. Adamkin, MD2, Victoria Niklas, MD3,*, Paula Radmacher, PhD2, Jan Sherman, PhD1,4,
Fiona Wertheimer, DO3, and Karel Petrak, PhD5,
Objective To evaluate the safety and explore the efficacy of recombinant human lactoferrin (talactoferrin [TLf]) to
reduce infection.
Study design We conducted a randomized, double blind, placebo-controlled trial in infants with birth weight of
750-1500 g. Infants received enteral TLf (n = 60) or placebo (n = 60) on days 1 through 28 of life; the TLf dose was
150 mg/kg every 12 hours. Primary outcomes were bacteremia, pneumonia, urinary tract infection, meningitis, and
necrotizing enterocolitis (NEC). Secondary outcomes were sepsis syndrome and suspected NEC. We recorded clinical,
laboratory, and radiologic findings, along with diseases and adverse events, in a database used for statistical analyses.
Results Demographic data were similar in the 2 groups of infants. We attributed no enteral or organ-specific
adverse events to TLf. There were 2 deaths in the TLf group (1 each due to posterior fossa hemorrhage and post-
discharge sudden infant death), and 1 death in the placebo group, due to NEC. The rate of hospital-acquired infec-
tions was 50% lower in the TLf group compared with the placebo group (P < .04), including fewer blood or line
infections, urinary tract infections, and pneumonia. Fourteen infants in the TLf group weighing <1 kg at birth had
no gram-negative infections, compared with only 3 of 14 such infants in the placebo group. Noninfectious outcomes
were not statistically significantly different between the 2 groups, and there were no between-group differences in
growth or neurodevelopment over a 1-year posthospitalization period.
Conclusion We found no clinical or laboratory toxicity and a trend toward less infectious morbidity in the infants
treated with TLf. (J Pediatr 2016;-:---).
Trial registration ClinicalTrials.gov: NCT00854633.
H
ospital-acquired infections represent the majority of diseases affecting preterm infants in neonatal intensive care units
(NICUs).1 Because hospital-acquired infections are associated with increased length of hospital stay and significant in-
creases in the cost of care, the American Academy of Pediatrics has called for strategies to reduce hospital-acquired
infections in NICUs.2,3 Among hospital-acquired infections, bacteria resistant to broad-spectrum antibiotics cause >50% of
patient-associated diseases,4 which has led to an emphasis on antibiotic stewardship.
Modified health care practices have reduced hospital-acquired infections in extremely preterm infants,2,3 but do not address
the underlying immaturity of the mucosal and systemic immune systems.5 Maternal milk is known to reduce the occurrence of
bacteremia and necrotizing enterocolitis (NEC).6,7 Biomolecules in human milk are proposed to synchronously modify the
intestinal microbiome and nascent gut and boost systemic immunity, thereby reducing susceptibility to infection.5,8 Extremely
preterm infants (birth weight <1 kg) are the most vulnerable to infection because maternal colostrum is limited immediately
after birth, or because intestinal dysmotility hinders full enteral feedings for days to weeks.
Human milk proteins enhance the development of intestinal epithelia,
facilitate a healthy intestinal microflora, establish host defenses, and heighten
mucosal defenses. We propose that the human milk protein lactoferrin partly 1
From the Division of Neonatology, Department of Child
2
Health, University of Missouri, Columbia, MO; Division
explains these beneficial effects.9,10 Commercial recombinant human lactoferrin of Neonatal Medicine, Department of Pediatrics,
(talactoferrin [TLf]) TLf became available 20 years ago.11 We found that feeding
3
University of Louisville, Louisville, KY; Division of
Neonatal Medicine, Childrens Hospital Los Angeles,
Keck School of Medicine at the University of Southern
California, Los Angeles, CA; 4Sinclair School of Nursing,
University of Missouri, Columbia, MO; and 5Agennix Inc,
Houston, TX
AE Adverse event *Current address: Prolacta, City of Industry, CA.
bLF bovine lactoferrin Current address: F.J.S. de Oro, 2835 Piso 5-2, 1425
CoNS Coagulase-negative staphylococci Capital Federal Buenos Aires, Argentina.
FDA Food and Drug Administration Funded by the National Institutes of Health (HD057744 to
IND Investigational New Drug Agennix, Inc [PI: K.P. and M.S.]). K.P. served as project
coordinator and Vice-President for Research at Agennix,
NEC Necrotizing enterocolitis Inc until 2012. The other authors declare no conflicts of
NICU Neonatal intensive care unit interest.
RCT Randomized controlled trial Portions of the study were presented at a Mead Johnson
SAE Serious adverse event Pediatric Nutrition Institute, Boston, MA, April 23-24,
2015.
TLf Talactoferrin
VLBW Very low birth weight 0022-3476/$ - see front matter. 2016 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jpeds.2016.04.084
1
THE JOURNAL OF PEDIATRICS www.jpeds.com Volume -
TLf prophylactically to neonatal rats prevented morbidity definitions for hospital-acquired infections.13 The sponsor
and mortality caused by enteroinvasive Escherichia coli.12 established strict criteria for infection, including bloodstream
Our research then became focused on enteral lactoferrin defi- infections, pneumonia, urinary tract infections, meningitis
ciency that occurs in the early life of immature infants. We and NEC.14 A diagnosed infection required antibiotics for
hypothesized that administering TLf would be safe, and con- $72 hours.
ducted a randomized controlled trial (RCT) to assess its Secondary outcomes were mortality, duration of hospi-
safety and efficacy in very low birth weight (VLBW) infants. talization, time to regain birth weight, and the time to reach
full enteral feeds. Disease-related morbidities included
Methods medically or surgically treated patent ductus arteriosus,
intracerebral hemorrhage, periventricular leukomalacia,
In this double-blinded RTC, TLf or placebo was administered retinopathy of prematurity, chronic lung disease (defined
to infants with birth weight between 750 and 1500 g starting as O2 therapy at 36 weeks postconceptual age), suspected
within 24 hours of birth and continuing through 28 days of NEC, clinical sepsis syndrome, and neonatal inflammatory
life (ClinicalTrials.gov: NCT00854633). We excluded infants response syndrome. We defined a clinical sepsis syndrome
who had a major congenital malformation, chromosomal ab- as a negative blood culture, but with clinical and laboratory
normality, documented prenatal or intrapartum neonatal findings necessitating empiric antibiotic therapy. These
infection, or absence of parental consent, or were moribund criteria included elevated inflammatory markers, namely
at birth. We enrolled and randomized subjects between July serial C-reactive protein level $1.5 mg/dL, abnormal serial
1, 2009, to March 17, 2012. Hospital discharge was followed white blood cell count or an elevated immature/total
by a 1-year outpatient follow-up period. Agennix Inc (Hous- neutrophil ratio ($0.3), central thermal instability, apnea
ton, Texas), the trial sponsor, ended the final data analyses in and bradycardia, or respiratory distress. We established sus-
December 2013. pected NEC as a clinical scenario that involved a cessation
Agennix provided TLf to 3 academic health care systems in of enteral feedings and initiation of antibiotics based on
the US. Agennix used good manufacturing practice and sus- gastric residuals, occult or gross blood in the stool, abdom-
pended TLf in sterile, endotoxin-free, phosphate-buffered sa- inal distention, and radiographs showing dilated bowel
line. The excipient served as the placebo with excellent color loops and an abnormal bowel gas pattern, but without a
matching. The Institutional Review Board for each site sentinel loop or pneumatosis intestinalis.
approved the study. We obtained written consent from the
parents or legal guardians before 24 hours of age. Thereafter, Safety Assessment
each institutional pharmacy randomized a subject via a cen- We used the MedDRA system to report safety outcomes to
tral computer system (inVentiv Clinical Solutions, Houston, the FDA.15 All investigators underwent training in the use
Texas). The inVentiv Web Response system also recorded of this grading and severity scoring system. This system re-
data about participants in this RCT. Agennix sponsored the ports adverse events (AEs) and severe adverse events
research under Food and Drug Administration (FDA) Inves- (SAEs) on a daily basis using an FDA-accepted measurement
tigational New Drug (IND) policies and procedures. The scale.15 The FDA mandates daily recording of clinical infor-
study design included a Data Safety Monitoring Board, a mation during the 28-day prophylactic period, with weekly
centralized serious adverse event reporting system, and peri- recording until discharge. At the 6- and 12-month postdi-
odic onsite monitoring visits that verified clinical, health scharge visits, we collected growth measurements, including
care, laboratory, and radiologic data and pharmacy record head circumference, health outcomes, and developmental
keeping. progress using the Bayley screener III.16 If the subject did
We randomized infants to receive either TLf solution not return, we contacted the primary care physician or par-
(150 mg/mL) at a dose of 300 mg/kg/day or an identical vol- ents by telephone, or parents by US mail, to ascertain the in-
ume of the excipient. Doses were administered every 12 hours fants clinical status.
via nasogastric tube from days of life 1 through 28 or until Agennix did not require any specific collection of clinical
discharge, whichever occurred first. We extrapolated the findings, laboratory tests, or radiologic studies. The FDA
dose of TLf from lactoferrin consumed during enteral breast IND program requires information on daily weight and
milk feeding (150 mL/kg/day) with the content of lactoferrin abdominal circumference, vital signs, physical examination
in human milk estimated at 2 mg/mL. In all subjects, we findings, type and duration of respiratory support, O2 satu-
administered the first dose before 24 hours of age. We ration, volume and type of enteral feeding, intravenous fluids
adjusted the dose at weekly intervals if the weight increased and total parenteral nutrition volume and composition, uri-
by $10% from a previous weight. nary output, gastric residual volumes, number and descrip-
tion of feces passed, and concomitant medications. The
Primary and Secondary Outcomes study protocol collected laboratory and radiographic test
The primary outcome was a significant reduction in hospital- data as ordered by the supervising neonatal attending. Test
acquired infections, including bacteremia, pneumonia, uri- data included complete blood count, C-reactive protein,
nary tract infection, meningitis, and NEC. Our criteria complete metabolic and electrolyte panels, blood gas ana-
were based on Centers for Disease Control and Prevention lyses, gross or occult blood in feces, and results of all
2 Sherman et al
- 2016 ORIGINAL ARTICLES
Sample Size and Statistical Analysis Gestational age, wk, mean SD* 28 6/7 28 6/7
Birth weight, g, mean SD* 1152 206 1143 220
This RCT focused on safety. We based the original sample Birth weight 750-1000 g, n (%)
14 (23) 14 (23)
size for the Phase 1 study on the recommendations of Cohen Birth weight 1001-1500 g, n (%) 46 (75) 46 (75)
(similar to G* Power 3.1) for an ANOVA with 4 groups. Us- Small for gestational age status, n (%) 9 (15) 11 (18)
Male sex, n (%) 33 (55) 36 (60)
ing a power of 0.80, an effect size of 0.5, and an a value of Apgar score, 1 and 5 min, median z
5 and 8 7 and 8
0.05, a minimum of 48 total subjects were needed to com- Multiple births, n (%) 18 (30) 13 (22)
plete the safety phase of the study. We based the Phase 2 sam- Preterm labor, n (%) 42 (70) 41 (68)
Premature rupture of membranes, n (%) 19 (32) 21 (35)
ple size calculations on the same statistical parameters, but >12-hour rupture of membranes, n (%)
8 (13) 15 (25)
with a reduction in effect size to 0.175, resulting in a mini- Maternal antibiotics, n (%) 32 (53) 35 (58)
mum total sample of 360 subjects. Allowing for 10% attri- One or more doses of betamethasone, n (%) 48 (80) 45 (75)
Cesarean/vaginal delivery, n 49/11 48/12
tion, we estimated that 396 subjects were necessary. Race/ethnicity, n
NICU stay (relative risk, 0.52; 95% CI, 0.26-0.99; P < .045).
Table III. Summary of AEs There were no cases of meningitis in either group. Two
Total, TLf Placebo P infants in the TLf group developed NEC, both of whom
Study AEs n (%) group, n (%) group, n (%) value
survived, and 1 formula-fed infant in the placebo group
At least 1 AE, n (%) 118 (98.3) 58 (97) 60 (100) .99 died from NEC.
At least 1 AE of grade 52 (43) 27 (45) 25 (42) .91
3/4/5, n (%) Table VI lists the types of bacteria that caused the
At least 1 SAE, n (%) 15 (13) 7 (12) 8 (13) .95 infections. In the TLf group, there was a reduction in
At least 1 TLf or placebo 0 0 0 - gram-positive bacterial isolates, with coagulase-negative
AE, n (%)
At least 1 AE causing drug 22 (18) 14 (23) 8 (13) .33 staphylococci (CoNS) accounting for most of these isolates.
discontinuation, n (%) In 14 infants with birth weight <1 kg, we identified no
AEs related to study cases of gram-negative bacterial infection and 2 cases of
drug, n (%)
Not related 106 (88) 53 (88) 53 (88) .94 CoNS-related bacteremia (14%). In the placebo group, 5 of
Possibly related 12 (10) 5 (8) 7 (12) .83 14 infants (36%) with birth weight <1 kg had bacteremia
AEs by degree of caused by CoNS (n = 2) or Klebsiella pneumoniae (n = 1),
severity, n (%)
Grade 1 29 (24) 14 (23) 15 (25) .94 or pneumonia caused by Klebsiella oxytoca (n = 2).
Grade 2 37 (31) 17 (28) 20 (33) .84
Grade 3 46 (38) 25 (42) 21 (35) .83 Secondary and Other Efficacy Outcomes
Grade 4 4 (3) 1 (2) 3 (5) .65
Grade 5 2 (2) 1 (2) 1 (2) .48 Tables VII and VIII (available at www.jpeds.com)
Deaths, n 3 2 1 1.0 summarize Secondary Outcomes in breast-fed and
Treatment-related 0 0 0 - formula-fed infants, respectively. Secondary outcomes,
deaths, n
including cumulative weight gain from birth and duration
MedDRA Version 14.1 defined classes, organ systems, and preferred terms. Scoring criteria are of hospitalization, did not differ by study arm or feeding
available at http://www.meddra.org/.
type. Follow-up records in the SAS database after hospital
discharge identified no abnormalities in growth or
(72%), and respiratory disorders (72%) were the most development between the TLf and placebo groups. We
commonly reported treatment-emergent AEs. These AEs proposed that TLf might reduce inflammation in treated
were often the reason for study drug or placebo infants9; however, the peak C-reactive protein level was not
discontinuation. The rate of at least 1 SAE was also similar in significantly different between the TLf group (n = 30) and
the TLf and placebo arms. All SAEs were associated with the placebo group (n = 38) (1.6 1.6 mg/dL vs
complications of very preterm birth rather than with 2.8 6.0 mg/dL, respectively).
administration of TLf or placebo. The Data Safety
Monitoring Board never halted the progression of the RCT. Discussion
Primary Outcomes To date, 4 investigative groups have published clinical trial
Table IV summarizes primary outcome data. Table V information using enteral administration of bovine lactofer-
(available at www.jpeds.com) shows the demographics of rin (bLF) to prevent late-onset sepsis and NEC during hospi-
infant having the first episode of a hospital-acquired talization of infants with birth weight #2000 g.17-20 The
infection. Compared with the placebo group, the TLf group current RCT is a Phase 1/2 trial that provides safety and pre-
had a lower risk of hospital-acquired infections during the liminary efficacy data associated with enteral administration
NNT, number needed to treat; NS, not significant; RR, relative risk.
*The number of hospital-acquired infections/total number of infants per group.
P value from c2 test for infectious comparisons; hospital-acquired infection/1000 hospital days from t test; significance P # .05.
zFirst identified episode of hospital-acquired infection in bloodstream, spinal fluid, urine, and lung fluid.
4 Sherman et al
- 2016 ORIGINAL ARTICLES
of TLf. The study used a recombinant human lactoferrin pro- Finally, studies in progress should evaluate the safety of
duced under good manufacturing practices and with FDA lactoferrin by using an internationally accepted method like
approval as an IND. This study also used the MedDRA sys- MedDRA. Based on the suggested mechanisms of action
tem to measure safety during and after administration of for lactoferrin,9,19 we suggest that future RCTs should also
TLf, an instrument used by the International Conference report on differences in inflammatory biomarkers between
on Harmonization of Technical Requirements for Registra- lactoferrin and placebo control subjects. One attractive strat-
tion of Pharmaceuticals for Human Use.12 The results egy may be to examine inflammation in twins who are
demonstrate the safety of the TLf molecule and provide an randomized to enteral lactoferrin vs placebo. Thoughtful ad-
initial report of efficacy related to reducing hospital- aptations of the traditional RCT design may provide oppor-
acquired infections and possibly other noninfectious- tunities to test whether lactoferrin supplementation will
related outcomes. reduce infectious and other morbidities in VLBW infants. n
The 4 previous studies using bLF in preterm infants re-
ported a reduced rate of infection of variable magnitude, The authors appreciate the care provided by our fellow neonatologists,
and 1 report using prophylaxis with bLF noted a reduction neonatal-perinatal medicine fellows, neonatal intensive care nurses,
in late-onset sepsis among infants with birth weight respiratory care practitioners, and pharmacists during this RCT.
<1 kg.17 In the current TLf trial, we found a 14% rate of infec-
Submitted for publication Jan 14, 2016; last revision received Mar 22, 2016;
tion in infants with birth weight <1 kg given enteral TLf vs accepted Apr 25, 2016.
36% in babies given placebo. Our results are in agreement Reprint requests: Michael P. Sherman, MD, PhD, Womens & Childrens
with those reported by Manzoni et al,17 with a comparable Hospital, University of Missouri Healthcare, Suite 206 Neonatology, 404 Keene
reduction in infection rates in extremely preterm infants. St, Columbia, MO 65201. E-mail: shermanmp@missouri.edu
In all studies using bLF, the biological agent was generally re-
garded as a safe food supplement. In contrast, the source of References
TLf was biotechnology rather than a food supplement iso-
lated from bovine milk. Furthermore, compared with the 1. Polin RA, Denson S, Brady MT. Epidemiology and diagnosis of health
aforementioned RCT using bLF,17 TLf showed similar thera- care-associated infections in the NICU. Pediatrics 2012;129:e1104-9.
2. Bizzarro MJ, Shabanova V, Baltimore RS, Dembry LM, Ehrenkranz RA,
peutic efficacy as bLf in reducing gram-positive bacterial in-
Gallagher PG. Neonatal sepsis 2004-2013: the rise and fall of coagulase-
fections (Table VI). We propose that this decline caused the negative staphylococci. J Pediatr 2015;166:1193-9.
overall reduction in hospital-acquired infections in infants 3. Polin RA, Denson S, Brady MT. Strategies for prevention of health care-
treated with TLf compared with those treated with placebo associated infections in the NICU. Pediatrics 2012;129:e1085-93.
(Table IV). 4. Cantey JB, Milstone AM. Bloodstream infections: epidemiology and
resistance. Clin Perinatol 2015;42:1-16.
There were no urinary tract infections in the TLf group
5. Collado MC, Cernada M, Neu J, Perez-Martnez G, Gormaz M,
but some in the placebo group; however, the difference was Vento M. Factors influencing gastrointestinal tract and microbiota im-
not statistically significant. An association between urinary mune interaction in preterm infants. Pediatr Res 2015;77:726-31.
infections and NEC was reported recently.21 Because the 6. Patel AL, Johnson TJ, Engstrom JL, Fogg LF, Jegier BJ, Bigger HR, et al.
fecal microbiome is a microbial reservoir for urogenital Impact of early human milk on sepsis and health-care costs in very low
birth weight infants. J Perinatol 2013;33:514-9.
colonization, we suggest a possible relationship between
7. Sullivan S, Schanler RJ, Kim JH, Patel AL, Traw oger R, Kiechl-
enteric prophylaxis with TLf and reduced urinary infec- Kohlendorfer U, et al. An exclusively human milk-based diet is associ-
tions, and would encourage examination of this association ated with a lower rate of necrotizing enterocolitis than a diet of human
in upcoming clinical trials of bLF. Future studies of lacto- milk and bovine milk-based products. J Pediatr 2010;156:562-7.e1.
ferrin prophylaxis should study the fecal microbiome, 8. Neu J, Mihatsch WA, Zegarra J, Supapannachart S, Ding ZY, Murgua-
Peniche T. Intestinal mucosal defense system, Part 1. Consensus recom-
because the microbiota present may be transferred to other
mendations for immunonutrients. J Pediatr 2013;162(3 Suppl):S56-63.
body sites. Metagenomic technology can accurately classify 9. Sherman MP, Adamkin DH, Radmacher PG, Sherman J, Niklas V. Pro-
the fecal translocation of pathogenic bacteria originating in tective proteins in mammalian milks: lactoferrin steps forward. Neore-
the feces.22 views 2012;13:e293-300.
10. Jiang R, Du X, L onnerdal B. Comparison of bioactivities of talactoferrin 17. Manzoni P, Rinaldi M, Cattani S, Pugni L, Romeo MG, Messner H, et al.
and lactoferrins from human and bovine milk. J Pediatr Gastroenterol Bovine lactoferrin supplementation for prevention of late-onset sepsis in
Nutr 2014;59:642-52. very low-birth-weight neonates: a randomized trial. JAMA 2009;302:
11. Ward PP, Piddington CS, Cunningham GA, Zhou X, Wyatt RD, 1421-8.
Conneely OM. A system for production of commercial quantities of hu- 18. Manzoni P, Meyer M, Stolfi I, Rinaldi M, Cattani S, Pugni L, et al. Bovine
man lactoferrin: a broad spectrum natural antibiotic. Biotechnology (N lactoferrin supplementation for prevention of necrotizing enterocolitis
Y) 1995;13:498-503. in very-low-birth-weight neonates: a randomized clinical trial. Early
12. Edde L, Hipolito RB, Hwang FF, Headon DR, Shalwitz RA, Sherman MP. Hum Dev 2014;90(Suppl 1):S60-5.
Lactoferrin protects neonatal rats from gut-related systemic infection. 19. Akin IM, Atasay B, Dogu F, Okulu E, Arsan S, Karatas HD, et al. Oral
Am J Physiol Gastrointest Liver Physiol 2001;281:G1140-50. lactoferrin to prevent nosocomial sepsis and necrotizing enterocolitis
13. McKibben L, Horan TC, Tokars JI, Fowler G, Cardo DM, Pearson ML, of premature neonates and effect on T-regulatory cells. Am J Perinatol
et al. Guidance on public reporting of healthcare-associated infections: 2014;31:1111-20.
recommendations of the Healthcare Infection Control Practices Advi- 20. Ochoa TJ, Zegarra J, Cam L, Llanos R, Pezo A, Cruz K, et al. Randomized
sory Committee. Infect Control Hosp Epidemiol 2005;26:580-7. controlled trial of lactoferrin for prevention of sepsis in Peruvian neo-
14. Walsh MC, Kliegman RM. Necrotizing enterocolitis: treatment based on nates less than 2500 g. Pediatr Infect Dis J 2015;34:571-6.
staging criteria. Pediatr Clin North Am 1986;33:179-201. 21. Pineda LC, Hornik CP, Seed PC, Cotten CM, Laughon MM, Bidegain M,
15. Brajovic S. MedDRA use at FDA. http://www.ich.org/fileadmin/Public_ et al. Association between positive urine cultures and necrotizing entero-
Web_Site/Training/GCG-_Endorsed_Training_Events/ASEAN_MedDRA_ colitis in a large cohort of hospitalized infants. Early Hum Dev 2015;91:
March_2010/Day_3/Regulatory_Perspective_SBrajovic.pdf. Accessed April 583-6.
26, 2016. 22. Sherman MP, Minnerly J, Curtiss W, Rangwala S, Kelley ST. Research on
16. Weiss LG, Oakland T, Aylward GP, Eds. Bayley III. Clinical use and neonatal microbiomes: what neonatologists need to know. Neonatology
interpretation. Burlington, MA: Elsevier Inc.; 2010. 2014;105:14-24.
6 Sherman et al
- 2016 ORIGINAL ARTICLES
6.e2 Sherman et al
- 2016 ORIGINAL ARTICLES