Neutrophil-to-Lymphocyte Ratio Predicts the Outcome of Acute Intracerebral Hemorrhage
Background and Purpose—Increasing evidence suggests that inflammatory mechanisms are involved in the intracerebral hemorrhage–induced brain injury. We evaluated the prognostic role of the peripheral leukocyte counts and neutrophil-to-lymphocyte ratio (NLR) in patients with intracerebral hemorrhage.
Methods—Patients with acute spontaneous intracerebral hemorrhage were retrospectively identified. Total white blood cells, absolute neutrophil count, and absolute lymphocyte count were obtained and the NLR computed from the admission blood work. The study end point was the occurrence of death or major disability at 3 months.
Results—One hundred seventy-seven subjects were enrolled. Ninety-four (53.1%) had unfavorable outcome. The absolute neutrophil count, absolute lymphocyte count, and NLR were independently associated with the 3-month status. The NLR resulted the best discriminating variable and the best predictive cut-off value was 4.58.
Conclusions—In patients with acute intracerebral hemorrhage, higher neutrophils, lower lymphocytes, and higher NLRs predicted worse 3-month outcome.
Spontaneous intracerebral hemorrhage (ICH) accounts for 10% to 30% of all strokes and is characterized by high rates of mortality and disability. The inflammatory response contributes to the ICH-induced secondary brain injury although the mechanisms are unknown.1 The aim of this study was to evaluate the relationships between the total and differential leukocyte counts and the neutrophil-to-lymphocyte ratio (NLR) at admission with the 3-month outcome in ICH patients.
Participants and Study Outcome
We retrospectively identified consecutive patients hospitalized at the Stroke Unit of the Marche Polytechnic University, Ancona, Italy from January 2008 to September 2015 for stroke syndrome caused by acute spontaneous ICH who underwent admission routine blood sampling and cranial computed tomographic neuroimaging within 24 hours from symptom onset. Demographics, medical history, admission/discharge National Institutes of Health Stroke Scale2 scores, baseline ICH topography and volume,3 admission blood pressure and 24-hour blood pressure variability by means the coefficient of variation4 were considered. Total white blood cells (WBC), absolute neutrophil count (ANC), absolute lymphocyte count (ALC), and erythrocyte sedimentation rate were collected from admission blood work. The outcome measure was the 3-month functional status: poor outcome was the occurrence of death or major disability (modified Rankin Scale score, ≥3).5
Values are presented as mean±SD, median (interquartile range) or number (%) of subjects. Comparisons were made through the Student t test, Mann–Whitney U test or χ2 test. Spearman correlation was used to correlate continuous variables. The associations between the WBC, ANC, ALC, NLR, and the study end point were determined using the logistic regression; the variables with P<0.05 from comparison of baseline characteristics and selected variables (age, sex, initial National Institutes of Health Stroke Scale score, baseline volume, location, and intraventricular extension of ICH)6 were forced into the multivariate analysis. The analysis was performed after categorization of the WBC, ANC, and ALC values into higher and lower groups with respect to the normal reference ranges.7 The receiver operating characteristic analysis evaluated the ability of the WBC, ANC, ALC, and NLR to predict the outcome. Results were significant for P<0.05 (2 sided). Analysis was performed using STATA/IC 13.1 (StataCorp LP, TX).
Standard Protocol Approvals
The local ethical committee approved this study. The board allowed the study to be conducted without patients’ consent because of the retrospective nature of the study.
A total of 177 patients were recruited, whose 94 (53.1%) had a modified Rankin Scale score of ≥3 at 3 months (Table 1). The poor outcome patients had higher WBC, ANC, and NLR and lower ALC (Table I in the online-only Data Supplement); no difference was found in the admission erythrocyte sedimentation rate values (28±2 versus 26±2 mm/h for good and poor outcome, respectively; P=0.518). The WBC strongly correlated to the ANC (ρ=0.917; P<0.001); the WBC, ANC, and ALC were moderately correlated to the baseline ICH volume (ρ=0.274, 0.389, and −0.301; P<0.001) and initial National Institutes of Health Stroke Scale score (ρ=0.256 [P=0.002], 0.375 [P<0.001], and −0.304 [P<0.001]). The WBC, ANC, ALC, and NLR were associated with the 3-month functional outcome; after the adjustment for potential confounders, the ANC, ALC, and NLR remained significantly associated, whereas the WBC did not (Table 2; Table II in the online-only Data Supplement). The ANC above the upper and ALC below the lower limits of normal ranges were associated to 1.7- and 2-fold increase in likelihood of poor outcome, respectively (Table III in the online-only Data Supplement). The receiver operating characteristic analysis and AUCs with respect to the outcome are shown in the Figure. The best discriminating variable was the NLR; the best predictive cut-off value was 4.58 (sensitivity 62.2%, specificity 84.9%, positive likelihood ratio, 4.12; negative likelihood ratio, 0.45). At 3 months, 86.4% and 40.8% (P<0.001) of the patients with NLR ≥4.58 and <4.58 had a modified Rankin Scale score, ≥3, respectively (odds ratio, 9.25 [95% confidence interval, 3.86–22.19]; P<0.001; adjusted odds ratio, 6.37 [95% confidence interval, 2.33–17.40]; P<0.001).
The main findings of our study are that higher neutrophils and lower lymphocytes at admission were independently associated with poor ICH outcome, and the NLR represented a readily available prognostic predictor. The inflammatory response occurs soon after the ICH: the neutrophils are actively recruited because the first hours around the hematoma and contribute to cellular injury and disruption of the viable tissue.1 The neutrophil increase could be the result of the acute phase reaction and the marker of the severity of disease and tissue inflammation. Higher leukocyte count have been associated to hematoma growth and early neurological deterioration. Our data suggested that early inflammation may be detrimental on the functional outcome. Because the regenerative capacities of the neural cells are limited, the secondary-induced brain damage may influence the prognosis. The selective neutrophil depletion before and the early inhibition of the neutrophil-derived matrix metalloproteinases after the ICH decreased the microglial/macrophage response and the neural apoptosis in the perihematoma region and improved the neurobehavioral recovery.8
The acute damage of vulnerable areas within the CNS induces the apoptosis and functional deactivation of peripheral lymphocytes,9 which are key players in the host defense against pathogens. In-hospital infections are common after stroke and may worsen the clinical course by inducing hyperthermia, increasing cerebral metabolic demands, and favoring acidosis and hypoxia.
The NLR was directly and independently related to the risk of adverse 3-month outcome and outperformed the predictive value of the WBC, ANC, and ALC. The NLR represents a reliable composite marker and dynamic index of systemic inflammation that reflects the immune response and combines information of innate and adaptive pathways. It could represent the hyperacute inflammatory response to cerebral hematoma and integrate the likelihood of the secondary brain injury and the susceptibility to the poststroke complications.
Interpretations of our findings should take into account limitations as the relatively small sample size and the retrospective analysis. The study sought to assess the associations between leukocytes and ICH outcome: prospective investigations providing serial cell counts measurements, sampling of sensitive biomarkers as the interleukin-6 or C-reactive protein, and information about the hematoma expansion, cerebral edema, development of fever, infections, and poststroke complications would be warranted to understand the underlying mechanisms. The main strengths of the study included the enrollment of patients despite the ICH location, the use of widely accessible laboratory variables, the cost effectiveness of the NLR.
Higher neutrophils, lower lymphocytes, and higher NLR values predicted worse outcome after ICH and could aid in the risk stratification of patients. Further understanding of the mediators of the immune orchestra may help to identify therapeutic strategies.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.013627/-/DC1.
- Received March 31, 2016.
- Revision received March 31, 2016.
- Accepted April 12, 2016.
- © 2016 American Heart Association, Inc.
- Aronowski J,
- Zhao X
- Wityk RJ,
- Pessin MS,
- Kaplan RF,
- Caplan LR
- Broderick JP,
- Brott TG,
- Duldner JE,
- Tomsick T,
- Huster G
- Banks JL,
- Marotta CA
- Hemphill JC 3rd,
- Bonovich DC,
- Besmertis L,
- Manley GT,
- Johnston SC
- Hoffbrand AV,
- Catovsky D,
- Tuddenham EGD
- Wang J,
- Tsirka SE