Open Access Open Badges Research

Predictive value of circulating interleukin-6 and heart-type fatty acid binding protein for three months clinical outcome in acute cerebral infarction: multiple blood markers profiling study

So-Young Park1, Jinkwon Kim1, Ok-Joon Kim1, Jin-Kyeoung Kim2, Jihwan Song3, Dong-Ah Shin4 and Seung-Hun Oh1*

Author Affiliations

1 Department of Neurology, CHA Bundang Medical Center, CHA University, 351 Yatap-dong, Seongnam, 463-712, Republic of Korea

2 Department of Pharmacy, CHA University, 222 Yatap-dong, Seongnam, 463-070, Republic of Korea

3 Department of Biomedical Science, CHA Stem Cell Institute, CHA University, 606-16 Yeoksam-dong, Seoul, 135-081, Republic of Korea

4 Department of Neurosurgery, Yonsei University College of Medicine, 134 Sinchon-dong, Seoul, 120-752, Republic of Korea

For all author emails, please log on.

Critical Care 2013, 17:R45  doi:10.1186/cc12564

Published: 16 March 2013



There is no single blood marker for predicting the prognosis in ischemic stroke. A combination of multiple blood markers may enhance the ability to predict long-term outcome following ischemic stroke.


Blood concentrations of neuronal markers (neuron-specific enolase, visinin-like protein 1, heart type fatty acid binding protein (hFABP) and neuroglobin), astroglial markers (S100B and glial fibrillary acidic protein), inflammatory markers (IL-6, TNF-α, and C-reactive protein), blood-brain barrier marker (matrix metalloproteinase 9), and haemostatic markers (D-dimer and PAI-1) were measured within 24 hours after stroke onset. The discrimination and reclassification for favorable and poor outcome were compared after adding individual or a combination of blood markers to the clinical model of stroke outcome.


In multivariate analysis, natural log-transformed (log) IL-6 (odds ratio (OR): 1.75, 95% CI: 1.25 to 2.25, P = 0.001) and loghFABP (OR: 3.23, 95% CI: 1.44 to 7.27, P = 0.005) were independently associated with poor outcome. The addition of a single blood marker to the clinical model did not improve the discriminating ability of the clinical model of stroke outcome. However, the addition of the combination of logIL-6 and loghFABP to the clinical model showed improved discrimination (area under receiver operating characteristic (AUROC) curve: 0.939 versus 0.910, P = 0.03) and reclassification performance (net reclassification improvement index: 0.18, P = 0.005).


A combination of circulating IL-6 and hFABP level has an additive clinical value for the prediction of stroke outcome.