Abstract Details

To determine whether blood based biomarkers within 72 hours of onset of stroke are significant predictor of 90-day mortality in patients with spontaneous intracerebral hemorrhage (sICH).

In a prospective multi-centric cohort study, patients with CT- proven ‘ICH’ were recruited within 72 hours of onset of symptoms. Venous blood samples (5 ml) were collected, serum levels of Troponin, Copeptin, C-reactive protein, GFAP (glial fibrillary acidic protein) and S100B were determined using method of Enzyme Linked Immunosorbent Assay (ELISA) by laboratory personnel masked for other clinical data. All the patients were telephonically followed using the modified Rankin Scale (mRS) at 3 months by an observer masked to the baseline and other clinical data. Study protocol has been published in BMC Neurology. Univariable and multivariable analyses were done to determine ‘discrimination’ of the predictive model using area under receiver operating curve (AUROC). All the statistical analyses were performed in STATA software (Version 13.1).

Data of 946 patients within 72 hours of onset of sICH were analysed. The mean age of patients was 56.72±13. AUROC for 90 day mortality were 0.54 (troponin), 0.52 (GFAP), 0.58 (Copeptin), 0.54 (S100B), 0.57(CRP) and 0.58(Total Leukocyte Count). In multi-variable model with age, volume of sICH, intraventricular hemorrhage and Glasgow Coma Scale the Area Under the curve was 0.77 ( 0.74 to 0.80)  and after addition of  biomarkers (CRP, s100b, TLC and copeptin)  the area improved significantly AUROC  0.7910; 95% CI 0.76 to 0.81,  chi2 = 7.15, P = 0.0075).

Our findings suggest that S100B, TLC, CRP and Copeptin significantly contribute to discrimination of prediction model.