Incidence, significance, and kinetic mechanism responsible for leukemoid reactions in patients in the neonatal intensive care unit: a prospective evaluation.
OBJECTIVE: To prospectively investigate the incidence, significance, and kinetic mechanism responsible for leukemoid reactions in patients in the neonatal intensive care unit (NICU). DESIGN: We prospectively studied all infants admitted to the NICU at the University of Florida who, during a period of 12 consecutive months, had a leukemoid reaction. All those identified had a standardized evaluation consisting of (1) karyotype analysis, (2) bacterial cultures, (3) evaluations for toxoplasmosis, other (congenital syphilis and viruses), rubella, cytomegalovirus, and herpes simplex virus) (TORCH), (4) determination of blood viscosity, (5) use of marrow aspirates for morphology, clonogenic progenitor cell assays, and cell-cycle analysis of progenitors, (6) determination of serum concentrations of granulocyte and granulocyte-macrophage colony-stimulating factors, and (7) serial complete blood cell counts until the leukemoid reaction remitted. RESULTS: During 12 months, 707 patients were admitted to the NICU and 4262 complete blood cell counts were performed on samples from these patients. A leukemoid reaction was identified in nine patients, all of whom were preterm (born at 24 to 38 weeks' gestation). Peak blood leukocyte concentrations were 51.7 +/- 15.6 x 10(3)/microl (mean +/- SD). The leukemoid reactions were detected during the first 4 days of life in seven patients, on day 9 in one, and on day 25 in one. An abnormal karyotype (47, XY, +21) was present in one infant. Mothers of four infants had received betamethasone antenatally. None had elevated whole blood viscosity or positive findings on bacterial or TORCH evaluations. None of the bone marrow findings were consistent with steroid-induced leukocytosis; all studies indicated accelerated neutrophil production. Serum concentrations of granulocyte-macrophage colony-stimulating factor were either negligible or nondetectable. Serum granulocyte colony-stimulating factor was elevated in three patients, low in two, and nondetectable in four. The leukemoid reactions persisted for 5 to 32 days, the longest being in the patient with trisomy 21. CONCLUSIONS: Leukemoid reactions were not particularly rare in our NICU (1.3% of patients). The reactions were not associated with hyperviscosity and, except in one patient with a karyotype abnormality, were transient. The responsible kinetic mechanism was increased neutrophil production, not steroid-induced leukocytosis.