Original ..............Article Effect of Recombinant Granulocyte Colony-Stimulating Factor on Blood Neutrophil Concentrations among Patients with ‘‘Idiopathic Neonatal Neutropenia’’: A Randomized, Placebo-controlled Trial Sandra E. Juul, MD, PhD INTRODUCTION Robert D. Christensen, MD A severe but self-limited idiopathic variety of neutropenia, termed ‘‘idiopathic neonatal neutropenia’’, has been described among preterm infants.1,2 Neonates with this condition have several features in common, including the fact that no specific variety of OBJECTIVES: neutropenia is diagnosed, it often appears late in the clinical We previously described a severe, prolonged, idiopathic, but self-resolving, course, can be quite prolonged, and generally occurs in otherwise variety of neutropenia among preterm neonates. In the present study, we well infants.1,2 It is not clear whether this condition is a single sought to assess the marrow neutrophil reserves of these patients by entity or represents multiple disorders with the common feature of serially measuring blood neutrophils following the administration of severe but self-limited neutropenia. As neutropenia in preterm recombinant granulocyte colony-stimulating factor (rG-CSF) or placebo. infants can be a sign of incipient severe illness, particularly sepsis, STUDY DESIGN: this variety of neutropenia in the neonatal intensive care unit (NICU) often causes concern, and sometimes prompts evaluations Prospective, randomized trial of rG-CSF vs placebo for infants with for infection and repeated or prolonged courses of antibiotics.1,2 ‘‘idiopathic neonatal neutropenia’’. We previously reported four patients with this variety of À1 RESULTS: neutropenia, all with blood neutrophil counts less than 500 ml . All responded to a 3-day course of recombinant granulocyte colony- During 36 consecutive months, 2407 neonates were admitted to the stimulating factor (rG-CSF) (10 mg/kg/day) with an increase in neonatal intensive care unit; 429 weighed less than 1500 g at birth, 14 of blood neutrophil counts.1 Our investigations suggested that this these were later diagnosed with ‘‘idiopathic neonatal neutropenia’’; 10 variety was the result of reduced neutrophil production, which was were enrolled in this trial. The five rG-CSF recipients had an immediate, not cyclic, not alloimmune, and not associated with recognized marked increase in blood neutrophil concentration, indicating adequate inborn errors, bacterial or viral infections, or medications. rG-CSF-mobilizable marrow neutrophil reserves. This effect persisted to The marrow neutrophil storage pool is a ready reserve of day 5, but counts were not different from those of the five placebo phagocytes that can be released into the circulation upon demand. recipients on days 12 and 15. Patients who have severe and prolonged neutropenia accompanied CONCLUSIONS: by few neutrophils in the marrow reserve have a significant Patients with ‘‘idiopathic neonatal neutropenia’’ have a substantial rG- deficiency in antibacterial defense. Such patients do not display a CSF-mobilizable marrow neutrophil reserve. On that basis, we speculate rapid increase in blood neutrophils following rG-CSF that this variety of neonatal neutropenia does not constitute a significant administration. It is not clear whether patients with ‘‘idiopathic deficiency in antibacterial defense. neonatal neutropenia’’ have a rapid increase in blood neutrophils Journal of Perinatology (2003) 23, 493–497. doi:10.1038/sj.jp.7210961 following rG-CSF administration, signifying an adequate neutrophil reserve, or rather whether such patients lack this response signifying the likelihood of a significant antibacterial defense deficiency. Our four previous cases appeared to have an intact response to rG-CSF, suggesting adequate neutrophil reserves. However, it can be Department of Pediatrics (S.E.J.), University of Washington, WA, USA; and Department of problematical to interpret serial blood neutrophil counts in Pediatrics (R.D.C.), University of South Florida, FL, USA. neonates without a comparison group of placebo recipients. This study was Supported by NIH RR00082. Therefore, we conducted a randomized, placebo-controlled trial of Address correspondence and reprint requests to Sandra Juul, MD, PhD, Department of Pediatrics, rG-CSF administration to patients with ‘‘idiopathic neonatal Division of Neonatology, University of Washington, Box 356320, Seattle, WA 98195, USA. neutropenia’’ to assess the blood neutrophil response and thereby Journal of Perinatology 2003; 23:493–497 r 2003 Nature Publishing Group All rights reserved. 0743-8346/03 $25 www.nature.com/jp 493 Juul and Christensen Idiopathic Neonatal Neutropenia gain insight into the adequacy or inadequacy of their rG-CSF- Table 1 Clinical Characteristics of Patients <1500 g mobilizable marrow neutrophil reserves. Infants <1500 g Sex (% male) 240/429 (56%) MATERIALS AND METHODS Ventilator support 317/429 (73.9%) From June 1, 1998 through July 1, 2001, patients in the NICU of Bronchopulmonary dysplasia 115/429 (26.8%) the Shands Children’s Hospital at the University of Florida Intraventricular hemorrhage underwent evaluation for neutropenia in accordance with our All grades 65/429 (15.2%) consensus document, ‘‘A consistent approach to evaluating Grade III 29/429 (6.7%) 3 neutropenia in the NICU’’. Patients with a blood neutrophil count Grade IV 8/429 (1.9%) <500 mlÀ1 for more than two days, or <1000 mlÀ1 for more than 4,5 5 days, had maternal blood testing for antineutrophil antibodies. Retinopathy of prematurity If those tests were negative, yet the neutropenia persisted with All stages 48/429 (11.2%) no diagnosis for the neutropenia, a marrow biopsy was performed Stage III 9/429 (2.1%) using a micro-method we reported.6 If no specific diagnosis for Stage IV 4/429 (0.9%) the neutropenia was still evident, the condition was termed ‘‘idiopathic neonatal neutropenia’’1 and the patient was eligible for performed. After this evaluation, 14 infants were identified with this study. ‘‘idiopathic neonatal neutropenia’’. In total, 10 were enrolled in After parental informed consent was obtained, eligible patients this study, six of whom were male. None of the 10 were clinically were randomized by the Shand’s Hospital Research Pharmacist, ill at the time of study entry. Blood neutrophil counts on the day of using a random-number table, to receive either rG-CSF 10 mg/kg/ study entry were 691±202 mlÀ1. The gestational age of these day i.v. over 30 minutes or placebo for 3 consecutive days. Study infants (mean±SE) was 28.7±0.7 weeks, median 29, range 25 to subjects had blood obtained for neutrophil concentrations at preset 32 weeks, with birth weights of 1067±80 g, median 1136, range intervals: 24 h, 48 h, 72 hours, 5, 12, and 15 days after the first 741 to 1474 g. Apgar scores averaged 6 at 1 minute, and 7 at 5 injection. Nosocomial infections between study entry and hospital minutes. Postnatal age of presentation with neutropenia ranged discharge were diagnosed by the clinicians caring for the patients. from 5 to 13 weeks (7.1±0.8 weeks, median 6.3 weeks). All had the No special infectious surveillance or cultures were part of this diagnosis of ‘‘idiopathic neonatal neutropenia’’, in that none had evaluation. Rather the hospital records were reviewed at discharge positive neutrophil antibody tests, none were delivered to women searching specifically for the diagnosis of nosocomial infection. with pregnancy-induced hypertension, and none had a recognized The study was approved by the Institutional Review Board of the infectious disease at the time of study. University of Florida and by the University of Florida Clinical As part of our ‘‘Consistent approach to neonatal neutropenia’’,3 Research Center Advisory Committee. Informed consent included all of these subjects with severe or prolonged neutropenia had a permission to use relevant clinical data recorded in the chart. marrow biopsy.3 In order to differentiate individuals with Data were analyzed by ANOVA, w2 test, or Pearson’s r correlation pseudoneutropenia due to excessive marginization, lidocaine with as appropriate. Results are reported as mean±SE, range and epinephrine was used as the topical anesthetic for these biopsies. median. Significance was considered at p<0.05. Only one infant had a significant increase in circulating neutrophils after the procedure; the prebiopsy neutrophil count was 913 mlÀ1, rising to 1596 mlÀ1 after the epinephrine exposure. RESULTS Excessive neutrophil marginization (pseudo-neutropenia) was not During the 3-year study period, 2407 infants were admitted to the thought to be a significant issue in these patients. NICU; 429 of these were less than 1500 g birth weight. Marrow findings and the specifics of the neutrophil counts from Demographics of the infants less than 1500 g are shown in Table 1. all 10 subjects are shown in Table 2. No specific diagnoses were Screening CBCs were not routinely performed; however, if an infant evident from any marrow study. Eight had varying degrees of left- was noted to be neutropenic based on a clinically indicated CBC, shifted myeloid maturation consistent with a consumptive process, they were evaluated according to our algorithm, ‘‘A consistent and two had decreased neutrophil precursors consistent with approach to evaluating neutropenia in the NICU’’.3 This decreased neutrophil production. Four of the marrow aspirates had assessment included an evaluation for sepsis if clinically increased B lymphocytes with decreased T lymphocytes. appropriate, and an appraisal