Apparently Severe Late-Onset Neutropenia in Two Very Low Birth Weight Infants

To the Editor. In 2000, Omar et al1 reported for the first time late-onset neutropenia in very low birth weight (VLBW) infants. The authors showed an incidence rate of late-onset neutropenia (defined as absolute neutrophil count lower than 1500/mm3 after 3 weeks of life) as high as 22%. The average nadir of the neutrophil in the neutropenic infants was 1067 ± 44/mm3. Severe late-onset neutropenia has to our knowledge not yet been described.We report 2 cases of premature infants who developed apparently severe late-onset neutropenia.Case 1 was born at 29 weeks of a bichorial biamniotic pregnancy (birth weight = 1.285 kg). The child required artificial ventilation for 2 days and had by then a normal evolution. She received rHuEpo 3 times a week (250 IU/kg) from day 7 to day 40. However, the child required transfusion on day 69. On day 69, a complete blood count performed before the transfusion showed an absolute neutrophil blood count of 423/mm3. White blood cells were 4700/mm3, hemoglobin 8.5 mg/dL, and blood platelets 333 000/mm3. The child was asymptomatic and, as no specific cause was found, no specific treatment was performed. A blood count performed on day 76 showed an absolute neutrophil blood count of 1211/mm3. The child was discharged on day 77.Case 2 was born at 28 weeks (birth weight = 1.280 kg) after having developed chronic intrauterine transfusion syndrome that caused the intrauterine death of her twin. She required artificial ventilation for 4 days and had by then a normal evolution. She received rHuEpo 3 times a week (250 IU/kg) from day 7 to day 39. She developed a progressive neutropenia with a nadir of 532/mm3 on day 69. Hemoglobin was then 9.2 g/dL and blood platelets 353 000/mm3. The absolute neutrophil blood count slowly increased and the child was discharged on day 75. On day 82, the neutrophil blood count was 1350/mm3. By then, she had a normal evolution.Our patients developed late-onset neutropenia as defined by Omar et al, but we think that the decreasing absolute neutrophil blood count could have been worsened by the rHuEpo administration, as reported by Latini et al for non-neutropenic infants.2 However, the dose of rHuEpo received by our patient was lower than the ones who developed neutropenia in the report of Latini et al (750 IU/kg/week vs 1200 IU/kg/week) and was given 3 times a week rather than 1 time a week.Omar et al1 recommend avoiding institution of agressive therapy such as granulocyte colony-stimulating factor (G-CSF) for late-onset neutropenia. However, Christensen et al3recommend G-CSF when the absolute neutrophil count is below 500/mm3 for 2 or 3 days, regardless of the cause.We believe that additional reports and studies are required before recommending agressive and expensive therapy of late-onset neutropenia, even when the absolute neutrophil blood count is around 500/mm3.In Reply. I appreciate the opportunity to respond to the letter from Servais et al regarding apparently severe late-onset neutropenia in VLBW infants. The incidence of late-onset neutropenia in our study was 22% (51/225 infants),1 and the nadir absolute neutrophil count (ANC) in the neutropenic infants was 1067 ± 44/mm3. Apparently severe late-onset neutropenia, defined as ANC <750/mm3, was detected in 10 of the infants in our study (4.4%). The mean gestational age of infants with apparently severe late-onset neutropenia was 28 ± 1 weeks (range: 25–32 weeks), and mean birth weight was 1056 ± 96 g (range: 726–1430 g). The nadir ANC was 577 ± 43/mm3 (range: 279–720/mm3) with a concomitant hemoglobin of 9.8 ± 0.4 g/dL (range: 6.7–11.1 g/dL), reticulocyte count of 7.5% ± 1.6% (range: 4.2%–14.6%), and platelet count of 339 ± 38 ×103/mm3 (range: 139–497 ×103/mm3). The mean ANC subsequent to the nadir ANC was 1901 ± 351/mm3 (range: 1044–3180/mm3). The mean postnatal age at onset was 6 ± 1 weeks (range: 4–10 weeks). This apparently severe late-onset neutropenia was transient and lasted for 1 week in 9 infants and 2 weeks in 1 infant. The lowest ANC of 294 and 444/mm3 were detected in 2 infants with subsequent increase of ANC to 1125 and 1776/mm3, respectively. All 10 infants with apparently severe late-onset neutropenia were stable, growing in full oral feeding. In contrast to the 2 infants reported by Sarvais et al, none of the infants in our study received erythropoietin. Late-onset neutropenia is a phenomenon that occurs in VLBW infants with anemia of prematurity and marked reticulocytosis.1 It seems that neutropenia detected in premature infants treated with erythropoietin2,3 has a similar mechanism and is secondary to marked reticulocytosis rather than to erythropoietin therapy. We agree with Servais et al and continue to recommend avoiding aggressive and expensive therapy such as G-CSF for late-onset neutropenia, even when ANC is around 500/mm3. Additional studies are needed to examine the incidence of nosocomial infection in stable, growing VLBW infants with late-onset neutropenia. In contrast, G-CSF may be a reasonable therapy for VLBW infants with persistent early-onset neutropenia during the first 3 weeks of life. VLBW infants with early-onset neutropenia are usually sick with multiple foreign bodies, such as central lines, peripheral intravenous lines, and endotracheal tubes, which increase their susceptibility to sepsis. Multiple studies have shown a possible beneficial effect of G-CSF in premature infants with early-onset neutropenia.4–8