Neutrophil-to-Lymphocyte Ratio Is an Independent Predictor for In-Hospital Mortality in Spontaneous Intracerebral Hemorrhage

Abstract

Background and Purpose: Stroke-associated immunosuppression and inflammation are increasingly recognized as factors that trigger infections and thus, potentially influence the outcome after stroke. Several studies demonstrated that elevated neutrophil-to-lymphocyte ratio (NLR) is a significant predictor of adverse outcomes in patients with ischemic stroke. However, little is known about the impact of NLR on short-term mortality in intracerebral hemorrhage (ICH). Methods: This observational study included 855 consecutive ICH-patients. Patient demographics, clinical, laboratory, and in-hospital measures as well as neuroradiological data were retrieved from institutional databases. Functional 3-months-outcome was assessed and categorized as favorable (modified Rankin Scale [mRS] 0-3) and unfavorable (mRS 4-6). We (i) studied the natural course of NLR in ICH, (ii) analyzed parameters associated with NLR on admission (NLROA), and (iii) evaluated the clinical impact of NLR on mortality and functional outcome. Results: The median NLROA of the entire cohort was 4.66 and it remained stable during the entire hospital stay. Patients with NLR ≥4.66 showed significant associations with poorer neurological status (National Institute of Health Stroke Scale [NIHSS] 18 [9-32] vs. 10 [4-21]; p < 0.001), larger hematoma volume on admission (17.6 [6.9-47.7] vs. 10.6 [3.8-31.7] mL; p = 0.001), and more frequently unfavorable outcome (mRS 4-6 at 3 months: 317/427 [74.2%] vs. 275/428 [64.3%]; p = 0.002). Patients with an NLR under the 25th percentile (NLR <2.606) - compared to patients with NLR >2.606 - presented with a better clinical status (NIHSS 12 [5-21] vs. 15 [6-28]; p = 0.005), lower hematoma volumes on admission (10.6 [3.6-30.1] vs. 15.1 [5.7-42.3] mL; p = 0.004) and showed a better functional outcome (3 months mRS 0-3: 82/214 [38.3%] vs. 185/641 [28.9%]; p = 0.009). Patients associated with high NLR (≥8.508 = above 75th-percentile) showed the worst neurological status on admission (NIHSS 21 [12-32] vs. 12 [5-23]; p < 0.001), larger hematoma volumes (21.0 [8.6-48.8] vs. 12.2 [4.1-34.9] mL; p < 0.001), and higher proportions of unfavorable functional outcome at 3 months (mRS 4-6: 173/214 vs. 418/641; p < 0.001). Further, NLR was linked to more frequently occurring infectious complications (pneumonia 107/214 vs. 240/641; p = 0.001, sepsis: 78/214 vs. 116/641; p < 0.001), and increased c-reactive-protein levels on admission (p < 0.001; R2 = 0.064). Adjusting for the above-mentioned baseline confounders, multivariable logistic analyses revealed independent associations of NLROA with in-hospital mortality (OR 0.967, 95% CI 0.939-0.997; p = 0.029). Conclusions: NLR represents an independent parameter associated with increased mortality in ICH patients. Stroke physicians should focus intensely on patients with increased NLR, as these patients appear to represent a population at risk for infectious complications and increased short-mortality. Whether these patients with elevated NLR may benefit from a close monitoring and specially designed therapies should be investigated in future studies.