The safety and ability of GM-CSF to influence the postnatal immune development of preterm (<32 weeks g.a.) neonates is being tested in a randomised controlled trial. 100 neonates are being randomised to receive GM-CSF (“Leucomax”, Sandoz) 10μg/kg/day by s.c. injection for 5 consecutive days, commencing within 72 hours of birth. Laboratory parameters of neutrophil proliferation and function are being tested pre-, immediately post and 1 week post GM-CSF therapy. Interim analysis of the first 50 babies is reported.

Gestational age and birth weight in the 2 groups are comparable. Control: median 27 (range 24-30) weeks; median 0.935 (range 0.480-1.486)kg. Treatment: 27 (23-31) weeks; 0.995 (0.570-1.476)kg. 19 babies were growth retarded(≤10th centile). There has been no evidence of toxicity in treated babies compared with controls, as assessed by irritability or weight gain while receiving GM-CSF, oxygen requirement during the first 28 postnatal days, development of chronic lung disease or GI toxicity. Babies receiving GM-CSF had higher neutrophil counts during treatment: mean 19.0 (range 2.3-53.4)×109/1 vs. controls: 7.6 (0.9-21.6) ×109/1. In 4 babies GM-CSF was stopped before the 5th dose because the neutrophil count exceeded 50 ×109/1. Among the 8 treated growth retarded (GR) babies there were no neutrophil counts below the normal range during the 28 study period, compared with 9 episodes of neutropenia in 11 control GR babies. GM-CSF caused in vivo neutrophil activation during treatment as indicated by increased CR3 (CD11b) expression: 2010 vs 1664 mean flourescence; and reduced FcRIII (CD16): 590 vs 1344. While not a primary endpoint in this pilot study, evidence to date suggests a trend to less morbidity from sepsis in babies who received GM-CSF.

GM-CSF therapy is safe when given prospectively to preterm neonates for 5 days immediately after birth. Interim data analysis shows GM-CSF activates neutrophils, prevents growth retardation related neutropenia and may reduce sepsis-related morbidity.