Neonatal nucleated red blood cells and the prediction of cerebral white matter injury in preterm infants.

OBJECTIVE

To estimate whether neonates with cerebral white matter injury have significant elevations in nucleated red blood cell counts and to estimate their predictive ability in identifying injury.

METHODS

This case-control study identified 176 infants born at 23-34 weeks of gestation between November 1994 and October 2004 at a single university hospital and with cerebral white matter injury characterized by periventricular leukomalacia (PVL) or ventriculomegaly due to white matter atrophy. A control was matched to each case using the subsequent delivery within 7 days of that gestational age without brain injury.

RESULTS

The gestational age at birth was 27 weeks for both groups, but the cases had a significantly lower birth weight (mean +/- standard deviation: 958 +/- 306 g compared with 1,038 +/- 381 g, P = .001). There was no difference in cesarean delivery (48% cases compared with 44% controls, P = .59). The cases had a significant increase in nucleated red blood cells per 100 white blood cells (WBC) (median, 5th percentile and 95th percentile: 22, 3 and 374 cases compared with 14, 1 and 312 controls; P = .02). Markers of chronic hypoxia, such as intrauterine growth restriction and oligohydramnios, and markers of acute hypoxia, such as an umbilical arterial pH less than 7.0 or base excess less than -12 mM, were both associated with significantly elevated neonatal nucleated red blood cell counts. A neonatal nucleated red blood cell count of 18 per 100 WBCs had a sensitivity of 56.9%, specificity of 57.9%, positive predictive value of 57.9%, and negative predictive value of 56.9% in predicting the development of cerebral white matter injury in this matched case-control sample.

CONCLUSION

Preterm neonates with cerebral white matter injury have significant increases in nucleated red blood cell counts. Both acute and chronic hypoxia-ischemia can increase these counts, which limits their usefulness in timing injury. The predictive value of nucleated red blood cell counts at birth in identifying injury is poor.

LEVEL OF EVIDENCE: II-2.