Intraventricular Hemorrhage Clearance in Human Neonatal Cerebrospinal Fluid: Associations With Hydrocephalus

Abstract

Background and Purpose- Preterm neonates with intraventricular hemorrhage (IVH) are at risk for posthemorrhagic hydrocephalus and poor neurological outcomes. Iron has been implicated in ventriculomegaly, hippocampal injury, and poor outcomes following IVH. We hypothesized that levels of cerebrospinal fluid blood breakdown products and endogenous iron clearance proteins in neonates with IVH differ from those of neonates with IVH who subsequently develop posthemorrhagic hydrocephalus. Methods- Premature neonates with an estimated gestational age at birth <30 weeks who underwent lumbar puncture for clinical evaluation an average of 2 weeks after birth were evaluated. Groups consisted of controls (n=16), low-grade IVH (grades I-II; n=4), high-grade IVH (grades III-IV; n=6), and posthemorrhagic hydrocephalus (n=9). Control subjects were preterm neonates born at <30 weeks' gestation without brain abnormality or hemorrhage on cranial ultrasound, who underwent lumbar puncture for clinical purposes. Cerebrospinal fluid hemoglobin, total bilirubin, total iron, ferritin, ceruloplasmin, transferrin, haptoglobin, and hemopexin were quantified. Results- Cerebrospinal fluid hemoglobin levels were increased in posthemorrhagic hydrocephalus compared with high-grade IVH (9.45 versus 6.06 µg/mL, P<0.05) and cerebrospinal fluid ferritin levels were increased in posthemorrhagic hydrocephalus compared with controls (511.33 versus 67.08, P<0.01). No significant group differences existed for the other cerebrospinal fluid blood breakdown and iron-handling proteins tested. We observed positive correlations between ventricular enlargement (frontal occipital horn ratio) and ferritin (Pearson r=0.67), hemoglobin (Pearson r=0.68), and total bilirubin (Pearson r=0.69). Conclusions- Neonates with posthemorrhagic hydrocephalus had significantly higher levels of hemoglobin than those with high-grade IVH. Levels of blood breakdown products, hemoglobin, ferritin, and bilirubin correlated with ventricular size. There was no elevation of several iron-scavenging proteins in cerebrospinal fluid in neonates with posthemorrhagic hydrocpehalus, indicative of posthemorrhagic hydrocephalus as a disease state occurring when endogenous iron clearance mechanisms are overwhelmed.

Keywords: biomarkers; cerebrospinal fluid; hemorrhage; hydrocephalus; iron.

Figure 1.. (A-D) Comparison of CSF levels of blood break-down products by group.

CSF levels of A. hemoglobin, B. total bilirubin, C. ferritin, and D. total iron in controls, low grade IVH, high grade IVH and post-hemorrhagic hydrocephalus (PHH) participants. (One-way ANOVA with Tukey’s Post-hoc multiple comparisons test)

Figure 2.. (A-D) Comparison of CSF levels of iron handling proteins by group.

CSF levels of A. ceruloplasmin, B. transferrin, C. haptoglobin, and D. hemopexin in controls, low grade IVH, high grade IVH and post-hemorrhagic hydrocephalus (PHH) participants. No significant group differences observed. (One-way ANOVA)

Figure 3.. (A-D) Ventricular size: differences between patient groups and correlation of ventricular size with CSF levels of blood-breakdown products.

A. Comparison of frontal-occipital horn ratio (FOHR, measurement of ventricular size) by group: control (n=11), low grade IVH (n=4), high grade IVH (n=6), and post-hemorrhagic hydrocephalus (PHH) (n=8), validating expected between groups differences in ventricular size. B-D. Correlation of FOHR with CSF ferritin, n=26 (B), hemoglobin, n=29 (C), and total bilirubin, n=14 (D) levels, demonstrating increased ventricular size correlated with higher levels of CSF ferritin, hemoglobin, and bilirubin.

Figure 4.. (A-B) ROC curves for CSF hemoglobin and ferritin.

Accuracy of CSF hemoglobin (A) and ferritin (B) in discriminating between post-hemorrhagic hydrocephalus (PHH) and non-PHH. True positive rate (sensitivity) is plotted as a function of false positive rate (1-specificity). The area under the ROC curve for hemoglobin is 0.90 (p<0.05) and for ferritin is 0.81 (p<0.05). Diagnostic cutoff values for association with high risk of post-hemorrhagic hydrocephalus were selected at 6.5(μg/ml) for hemoglobin (sensitivity 0.88, and specificity 0.81), and 555 (ng/ml) for ferritin (sensitivity 0.71, specificity 0.89). Cutoff values were selected based on optimal Youden index.