Systemic inflammation and Oxidative stress with stroke mortality among patients admitted in tertiary Hospital in Uganda: a prospective cohort study in southwestern Uganda

Background: Stroke is an inflammatory state that causes death and chronic disability. Inflammation and oxidative stress are a predictor of poor clinical outcome, its effects are controversial and has not been evaluated in Sub-Saharan Africa (SSA). Methods: We conducted a prospective cohort study of CT head confirmed ischemic and hemorrhagic stroke admitted within 7 days of onset of motor weakness. Baseline CRP, NLR and baseline glucose was measured with subsequent modified Rankin Scale (mRS) score on day 14 post-stroke. Cox proportional hazard model was fitted to determine hazard ratios of mortality with CRP, NLR and blood glucose. Results: Out of 120 patients, 51.7% were female, 52.5% had ischemic stroke and the overall median age was 65 (IQR 54–80) years. Nineteen (15.8%) patients died within a median survival time of 7 days, while 32 (25.8%) died by day 14 after stroke. Conclusion: High C-reactive protein and stroke related hyperglycemia conferred statistically significant hazards of mortality among patients with acute and subacute stroke.


Introduction
Stroke affects millions of people annually with a high mortality and morbidity (1,2).There is a rise in stroke occurrence in low-income and middle-income countries (LMICs) (3).Hospital-based studies from Africa have shown a case fatality ranging from 16.2-46% among patients with stroke at 30 days (3).In Uganda, studies have shown a high 30 day mortality between 26.8% − 38.1% (4-6).Stroke has been described as a state of in ammation and oxidative stress (7).
Primary brain injury occurring in ischemia or hemorrhage is followed by secondary brain injury and Oxidative stress, which begins within minutes and persists for days to weeks or even longer (8).In primary brain injury; ischemic stroke refers to an abrupt cessation of blood supply in a vascular territory resulting in an ischemic core, surrounded by a penumbra and intracerebral hemorrhage refers to rupture of a blood vessel leading to extravasation of blood components directly into the brain parenchyma, forming a hematoma that provokes structural damage (8).Neural cellular damage and release of damage-associated molecular patterns (DAMPs) de nes a common pathway that provokes an innate immune response characterized by neutrophils, microglia cell activity and adaptive immune response characterized by glial activation, recruitment of peripheral immune cells, and release of cytokines and chemokines (8, 9).The peripheral leucocytes in ltrate the injured brain aggravating further disruption of the blood brain barrier by releasing proin ammatory cytokines and reactive oxygen species (10,11).An increased peripheral neutrophil count is independently predictive of severe stroke whereas lymphocytosis increases the regulation of anti-in ammatory cytokines like interleukin (IL) 10 and suppresses proin ammatory cytokines like IL 6 and tumor necrosis factor (TNF) alpha, thereby protecting the nerves (7,12).Systemic pressure after stroke leads to low lymphocyte count through activation of the reninangiotensin system, resulting in the release of cortisol and induction of lymphocyte apoptosis (13).A high NLR depicts neutrophilic elevation and lymphocytic depletion indicating a disproportionate interaction between central and peripheral in ammation.A high NLR is a negative prognostic indicator in acute ischemic stroke (AIS) and spontaneous intra-cerebral hemorrhage (ICH) (14).This differential cellular count is affected by the in ammatory process which can be assessed by in ammatory markers like C-reactive protein (CRP) (7).C-reactive protein (CRP) is a known biomarker of systemic in ammation among patients with stroke (15).A high CRP level is a predictor of mortality independent of stroke severity and infections (16).The level of brain in ammation is affected by the oxidative stress.
Oxidative stress de ned as an imbalance between anti-and pro-oxidants, which has been implicated in the stroke pathogenesis (17).The brain is sensitive to oxidative damage because of its high and speci c metabolic activity.High oxygen consumption, no energy reserves, almost exclusive oxidative phosphorylation, high lipid concentrations prone to peroxidation, and high levels of iron, all acting as a promotors of oxidative stress (18).Low blood ow reduces the amount of oxygen and glucose, following a cascade of events that leads to production of reactive oxygen species (ROSs) and free oxygen radicals that further worsen brain injury (17).Stroke related hyperglycemia mostly driven by stress hormones like cortisol, glucagon and adrenaline but some patients have underlying diabetes (19).Hyperglycemia subsequently induces intracellular ROS production resulting in an increased production of superoxide (20).Hyperglycemia exerts direct membrane lipid peroxidation and cell lysis in metabolically challenged tissue leading to global brain in ammation (19).
High biomarkers of in ammation and hyperglycemia are associated with poor short-term clinical outcomes (21).Studies on the association of in ammation and oxidative stress with all-cause mortality risk in patients with stroke have yielded inconsistent results.We set out to determine the effect in ammation and oxidative stroke on 14 day mortality among patients with stroke in Uganda

Study design and participants
This was a prospective cohort study of patients with acute and subacute stroke admitted to Mbarara Regional Referral Hospital, a tertiary hospital in South Western Uganda.We included patients; 18 years and above with a sudden onset of one sided neurologic de cits within 7 days and non-contrasted CT head con rmation of ischemic stroke evidenced by a hypodense lesion or hemorrhagic stroke evidenced by a hyperdense lesion in the brain.We excluded patients with traumatic intracerebral hemorrhage such as hematomas, and traumatic brain injury.At admission, all patients were positioned with head of the bed elevated at 30 degrees to prevent aspiration and oxygen saturation was kept above 93% as the standard of care.Blood pressure at admission was measured using EDAN M3® (Edan USA 2014).Three blood pressure values were taken and an average of the last two was considered as the blood pressure at hospital admission (22).Socio-demographics (such as age, sex, marital status, address); behavioral factors (such as smoking and alcohol history) were captured.Past medical records were evaluated to capture history and duration of hypertension, diabetes mellitus, types of medications given, presence of co-morbid kidney disease and heart disease.A complete clinical examination was conducted which included; the Mayo Clinic Full Outline of Unresponsiveness score (FOUR score) to assess the level of consciousness and the National Institutes of Health Stroke Scale (NIHSS) score to assess stroke severity.

Laboratory procedures at admission
Capillary blood glucose was measured using Accuchek glucometer (Roche Diagnostics Inc.).Full blood count was measured using Mindray hematology analyzer, and total Cholesterol (TC) was measured by Enzymatic linked immunosorbent assay method, using Human 200 analyzer (German Design, Human Diagnostics), renal function tests, serum Sodium and potassium were measured using Sysmex XNL-550®.

Outcome Measures
The primary outcome of this study was de ned as mortality at day 14 of stroke onset.Modi ed Rankin Scales (mRS), which is a measure of the degree of neurological disability or dependence on daily activities was in addition assessed as an outcome at day 14 of stroke among those who were still alive (24).

Statistical Analysis
Clinical characteristics were computed as mean, and standard deviation for normally distributed variables.Categorical variables were summarized in frequencies and percentages.To determine the differences in baseline clinical characteristics between hemorrhagic stroke and ischemic stroke, were evaluated using a student's t-test for continuous variables and chi-square test for categorical variables.
Cox-proportional hazard regression analysis was tted to determine the hazard ratios of mortality at 14 days with 5% level of signi cance, 95% con dence interval and p -values.These were adjusted for baseline sociodemographic characteristics, Hypertension, type of stroke and NIHSS (stroke severity) against the clinical outcome (mortality at day 14)

DISCUSSION
In this study we set out to determine the effect of in ammation and oxidative stress assessed using C -Reactive Protein, random blood sugar and Neutrophil-lymphocyte ratio at admission among patients with acute and subacute stroke on mortality at day 14 of stroke onset.We found that stroke related hyperglycemia and high C -reactive Protein did signi cantly relate to mortality.Patients with hemorrhagic stroke had higher Neutrophil Lymphocyte Ratio (NLR) and C -Reactive Protein (CRP) in comparison with ischemic stroke but a lower random blood sugar (RBS) at admission.These ndings support earlier studies that demonstrated that systemic in ammation and oxidative stress increase the risk of death and/or dependency after acute stroke.This study is among the few studies in sub-Saharan Africa that have evaluated in ammation and oxidative stress in stroke correlating it with mortality at day 14 of stroke onset.
Systemic in ammation described by a high Neutrophil Lymphocyte Ratio (NLR) which was not statistically signi cant but a high C-reactive protein (CRP) with signi cant adjusted hazards for mortality of 9.2 in our study.This implies that stroke is an in ammatory state and stroke related infections may also exacerbate the CRP value.In ammation starts early and plays a central role not only in ischemic damage but also in endothelial progenitor cells in angiogenesis (2).When local in ammation occurs, it can worsen a secondary injury and evoke global brain in ammation (2).This means that higher baseline in ammation level is a good predictor of short-term poor outcome in strokes (25).C-reactive protein is a peripheral biomarker of in ammation and it is an acute phase protein which has been assessed as a biomarker for mortality and poor clinical outcome among patients with stroke (26).Yu et al found that a high CRP had a 2.07 fold risk of all-cause mortality in acute ischemic stroke (16).This study only assessed ischemic stroke patients yet in our study, patients with hemorrhagic stroke contributed a higher CRP than patients with ischemia.The odds for neurological improvement measured by mRS decrease as the level of plasma CRP increase after adjustment for age, sex, baseline neurological severity, and stroke subtypes (26).These ndings may be partly explained by in increased risk of recurrent stroke and cardiac ischemia during the early period post stroke (27)(28)(29).Elevated CRP has been related to poor functional outcome, mortality and also to the occurrence of post-stroke infections (30).However even when patients with early infection are excluded during hospitalization, this does not signi cantly eliminate the association of CRP with mortality (31).
Stroke related hyperglycemia resulted in 3.2 high hazards of mortality among patients with stroke in our cohort yet only 11% had diabetes mellitus.The cause of the hyperglycemia in patients with stroke is mostly driven by stress hormones like cortisol, and catecholamines which play a big role in glucose regulation (32).Stroke induced hyperglycemia regardless of diabetes mellitus points to a physiological stress and also relative insulin resistance, which is linked to increased lipolysis (33).A study done in Egypt that was evaluating 24 hour hyperglycemia in stroke de ned by blood sugar of 8.3 mmol/dl and found 1.2 adjusted risk for mortality among patients with stroke (34).This study had a lower cutoff compared to our study.These varying cutoffs for hyperglycemia in different studies among patients with stroke still show that stroke related hyperglycemia is strong predictor mortality (34).Experts recommend a target blood glucose between 7.8 mmol/L and 10 mmol/dl patients that have suffered a stroke (35).Tight glucose control (< 6.1mmol/dl) versus loose glucose control, there was no difference in the clinical outcome of patients with stroke (36).Persistent stroke induced hyperglycemia has been shown to decrease cerebral blood ow through vascular dilation leading to an increase intracranial pressure, causing cerebral edema, in ammation, and neuronal death hence resulting in blood brain barrier disruption, hemorrhagic transformation in acute ischemic stroke and growth of hematoma size in hemorrhagic stroke (34,(37)(38)(39).
In conclusion, we have provided evidence that high CRP and stroke related hyperglycemia in acute and subacute stroke were predictors of mortality.Setting up of stroke units in sub-Saharan Africa.
We recommend interventions for controlling systemic in ammation and blood sugars among patients with stroke to reduce mortality.
This study also has some limitations.Due to late presentation of patients to the hospital might have affected rate of in ammation and oxidative stress considering that the risk of infections from catheters and aspiration pneumonias are high during the late presentation hence producing a higher state of in ammation and oxidative stress.

Figures
Figures

Table 1
Baseline clinical characteristics of patients with stroke.
use of alcohol was elicited in 23.3% and 40.8% of all participants respectively (Table