Hemoglobin, Albumin, Lymphocyte, and Platelet Score is Associated With Adverse Clinical Outcomes of Acute Ischemic Stroke: A Prospective Cohort Study

The combined index of hemoglobin, albumin, lymphocyte and platelet (HALP) is considered as a novel score to reect systemic inammation and nutritional status. This study aimed to investigate the association between HALP score and adverse clinical events in patients with acute ischemic stroke (AIS). Methods This study prospectively included patients with AIS within 24 hours of admission to the First Aliated Hospital of Zhengzhou University. The primary outcomes were all-cause death within 90 days and 1 year. The secondary outcomes included stroke recurrence and combined vascular events. The association between HALP score and adverse clinical outcomes was analyzed using Cox proportional hazards.


Background
Stroke is the leading cause of death and long-term disability worldwide; recurrent stroke worsens the outcome and even increases mortality [1,2]. Exploring new predictors of stroke prognosis can help improve patients' clinical outcomes and contribute to more effective secondary prevention. In ammation, abnormal blood coagulation, and poor nutritional status are associated with poor prognosis of acute ischemic stroke (AIS) [3][4][5]. Lymphocytes have key regulatory functions in post-stroke in ammation; lower lymphocyte counts were associated with increased infarct volume, neurological deterioration, and poor prognosis after ischemic stroke [6,7]. Platelet hyperactivity increases the risk of thromboembolism and atherosclerotic lesions and may lead to abnormal thrombosis which exacerbates in ammation response [8]. Anemia and hypoalbuminemia are manifestations of poor nutritional status and have been identi ed as risk factors for cerebrovascular events that may lead to poor prognosis in patients with AIS [9,10]. Anemia and lower albumin levels at admission may adversely affect the prognosis of stroke and can be used as indicators for stroke recurrence [11,12].
The hemoglobin, albumin, lymphocyte and platelet (HALP) score is considered to be an easily calculated marker of systemic in ammation and nutritional status [13,14] and has been con rmed as a signi cant prognostic factor in patients with pancreatic cancer, esophageal squamous cell carcinoma, and bladder cancer [15,16]. However, it is unknown whether HALP score is associated with the adverse clinical events of AIS. Therefore, our study aimed to investigate the predictive value of HALP score in a large cohort of patients with AIS.

Study Design and Population
The study was a prospective consecutive hospital-based cohort study. We enrolled patients with AIS within 24 hours of onset from the Ischemic Cerebrovascular Disease Database of the First A liated Hospital of Zhengzhou University from January 2015 to June 2018, which has been published previously [17,18]. The diagnosis of AIS was based on the criteria of the World Health Organization (WHO) [19]. The database was approved by the Ethics Committee and informed consent forms were obtained from all patients or their relatives.
The exclusion criteria were as follows: ( ) patients with active or chronic in ammatory disease; ( ) autoimmune diseases or using immunosuppressant drugs; ( ) patients with neoplastic hematologic disorder; ( ) patients without blood cell count data; and ( ) severe liver and kidney dysfunction.

Clinical Assessment
Baseline clinical information was obtained from case report forms at admission, including demographic characteristics, medical histories, imaging features, and medication use. The severity of AIS was assessed blindly according to the National Institutes of Health Stroke Scale (NIHSS) at admission [20].

Laboratory Assay
Laboratory tests were obtained within 24 hours of admission, including the serum albumin, hemoglobin, lymphocyte, and platelet levels. The blood counts were analyzed using an autoanalyzer (Beckman Coulter Hematology Analyzer LH750, USA). All the serum biochemical parameters were assayed using an automatic biochemical analyzer (Roche COBAS 8000 automatic biochemical analyzer). The HALP score was calculated according to the following formula: hemoglobin (g/L) × albumin (g/L) × lymphocytes (/L) / platelets (/L).

Outcome Assessments
The primary outcomes were all-cause death within 90 days and 1 year. The secondary outcomes included stroke recurrence and combined vascular events (stroke recurrence and all-cause death) within 90 days and 1 year. Patients enrolled were followed up by face-to-face or telephone interviews at 90 days and 1 year after stroke onset.

Statistical Analyses
Continuous variables with normal distribution were expressed as mean ± standard deviation (SD) and skewed distribution were expressed as medians with interquartile ranges (median, IQR). Categorical variables were expressed as frequencies and percentages (n, %). Patients were classi ed into three groups based on the HALP score tertiles. The nonparametric Wilcoxon test or Kruskal-Wallis test was used to compare group differences for continuous variables, and the χ 2 test was used for categorical variables. The cumulative incidence risks of adverse clinical events across baseline HALP score were calculated with Kaplan-Meier curves and compared by log-rank test. Multivariate Cox proportional hazards models were used to assess the risk of adverse clinical outcomes with HALP score tertiles. C statistics, net reclassi cation index (NRI), and integrated discrimination improvement (IDI) were used to evaluate the incremental predictive ability of HALP score beyond the conventional model. Optimal HALP score cut-off points were obtained using receiver operating characteristic curve (ROC) analysis. Finally, subgroup analyses were conducted to assess the robustness of association between HALP score and adverse clinical outcomes of AIS. Interactions between HALP score and subgroup variables on the adverse clinical outcomes were tested in the models with interaction terms by the likelihood ratio test, adjusting for the aforementioned covariates unless the variable was used as a subgroup variable. Statistical analysis was performed using IBM SPSS software version 24.0 (SPSS, Inc, Chicago, IL, USA) and R (version 3.5.0). A two-tailed value of P < 0.05 was considered statistically signi cant.

Baseline Characteristics
A total of 1337 patients were included in our analysis (see Additional le 1). The baseline characteristics were balanced between the patients included and excluded (see Additional le 2). The mean age of the patients was 61 years and 30.5% of them were female. Compared with the higher HALP score patients, those with a lower HALP score were more likely to be older, had lower lymphocyte, hemoglobin and albumin levels, and had higher platelet counts and baseline NIHSS scores (Table 1). The cumulative incidence of death, stroke recurrence, and combined vascular events within 90 days of follow-up were 2.8%, 1.8%, and 4.5%, respectively, while the cumulative incidence within 1 year were 4.6%, 4.7%, and 8.8%. All Kaplan-Meier curves showed that patients in the lowest tertile of HALP score had the highest incidence of death, stroke recurrence and combined vascular events within 90 days and at 1 year (log-rank P < 0.05 for all, Fig. 1 and see Additional le 3).
After adjustment for age, sex, smoking, alcohol consumption, history of hypertension, diabetes, ischemic stroke, coronary heart disease, atrial brillation, and baseline NIHSS score, higher levels of HALP score were associated with a decreased risk of death within 90 days and 1 year (

Incremental Predictive Ability Of Halp Score
The incremental predictive ability of HALP score to predict adverse clinical outcomes after AIS is presented in Table 3. Adding HALP score to the conventional model, which included age, sex, smoking, alcohol consumption, history of hypertension, diabetes, ischemic stroke, coronary heart disease, atrial brillation, and baseline NIHSS scores, signi cantly improved the predictive ability for death within 90 days and 1 year (NRI: 38.63% and 38.68%; IDI: 2.43% and 2.57%, P < 0.02 for all). Similar results were found in combined vascular events but not for stroke recurrence within 90 days and 1 year after AIS onset.

Subgroup Analyses
An optimal HALP score cut-off point level was obtained from the ROC analysis and low HALP score levels were associated with the death after adjustment for age, sex, smoking, alcohol consumption, history of hypertension, diabetes, ischemic stroke, coronary heart disease, atrial brillation, and baseline NIHSS score (Fig. 2). The adjusted Hazards ratio were 0.30 (95% CI: 0.15-0.59) within 90 days and 0.35 (95% CI: 0.21-0.60) at 1 year. In the subgroup analyses, negative associations between HALP score and death were observed in most subgroups. No statistical signi cance between HALP score and these factors on primary outcome was observed (all P for interaction > 0.05). Similar results were observed for stroke recurrence and combined vascular events within 90 days and 1 year (see Additional le 5).

Discussion
We explored the prognostic value of the novel index HALP score in patients with AIS in this prospective cohort study and found that higher levels of HALP score at admission were highly associated with a decreased risk of death, stroke recurrence and combined vascular events within 90 days and 1 year.
Ischemic stroke initiates with gradual or sudden cerebral hypoperfusion, including oxidative stress, hemostatic activation, in ammation, and eventually leads to a corresponding loss of neurological function [21]. Ischemic brain tissue activates leukocytes and promotes their migration to the ischemic site by releasing pro-in ammatory chemokines [22]. The in ammation triggers the process of thrombosis, in which platelets participate in adhesion, release reaction, and aggregation [8]. Lymphocytes play an important role in the elimination and repair of in ammation [23]. Albumin has a neuroprotective effect because of its antagonism of oxidation, stagnant, thrombosis, and leukocyte adhesion [24,25]. Hemoglobin has oxygen-carrying capacity and can affect the energy balance in the penumbra [26]. Studies have shown that each of these indicators is a predictor of stroke prognosis [7,9,27,28]. Further, serum albumin and hemoglobin concentration have predictive value for stroke recurrence and combined events [11,12]. The new indicator HALP score is based on the combination of the four hematological parameters mentioned above. Our study showed that HALP score was associated with the risk of death, stroke recurrence and combined vascular events within 90 days and 1 year after AIS. These results suggested that decreased HALP score could be an independent risk factor of adverse clinical outcomes.
So far, no study has investigated the association between HALP score and AIS. Recent studies have shown that the HALP score can re ect the in ammation-nutritional status of patients [14,29], and has been proved to be an important prognostic indicator for patients with pancreatic cancer [13], esophageal squamous cell carcinoma [15], and bladder cancer [16]. Anemia and thrombosis could exacerbate in ammation while lymphocytes reduce in ammation [30]. Since Seltzer [31] proposed that admission serum albumin levels and total lymphocyte counts evaluate immediate nutritional status of patients, serum albumin has been considered as an indicator of nutritional status. Some studies also suggest that albumin re ects the severity of in ammation and illness in acute diseases [32]. It is widely accepted that the in ammatory response and nutritional status are correlated with the prognosis of patients with stroke [5,33,34]. HALP score is obtained by hemoglobin (g/L) × albumin (g/L) × lymphocytes (/L) / platelets (/L), which makes it a cost effective, simple parameter to easily assess the in ammationnutritional status. This nding may be signi cant because instant in ammation-nutritional status assessment can help clinicians assess prognosis and formulate appropriate treatment plans.
There were some limitations to this study. First, this was a single-center study; therefore, there may be a selection bias. Second, the potential in uence of previously received treatments, such as thrombolytic therapy, oral anti-platelet, and statins were not considered. Third, we only obtained the hemoglobin, albumin levels, lymphocyte, and platelet count at admission but did not present the dynamic change of HALP score at different stages. Therefore, multicenter cohort studies are still needed to validate the ndings.

Conclusion
Our study indicated association between HALP score and risk of death, stroke recurrence and combined vascular events within 90 days and 1 year, suggesting that HALP score at admission may act as a

Declarations
Ethics approval and consent to participate The study was approved by the ethics committee of the First A liated Hospital of Zhengzhou University. Informed consent forms were obtained from all patients or their relatives.All informed consent obtained from study participants is written.

Consent for publication
Not applicable Availability of data and materials The datasets used during the current study are available from the corresponding author on reasonable request, subject to permission from the relevant ethics committees at the hospital and university.

Competing interests
The authors declare that they have no competing interests.  Kaplan-Meier curves of cumulative incidence (%) of death by tertiles of HALP score at 90-days and 1-year follow-up.

Figure 2
Subgroup analyses of the association between HALP score and death within 90 days and 1 year. In the multivariate models, confounding factors, such as age, sex, smoking, alcohol consumption, history of hypertension, diabetes, ischemic stroke, coronary heart disease and atrial brillation and baseline NIHSS scores were included unless the variable was used as a subgroup variable. HR, Hazard Ratio; CI, Con dence Interval; HALP = hemoglobin, albumin, lymphocyte, and platelet