Acute Routine Leukocyte and Neutrophil Counts Are Predictive of Poststroke Recovery at 3 and 12 Months Poststroke: An Exploratory Study

Abstract

Background and Aims. White blood cell (WBC) and neutrophil counts (NC) are common markers of inflammation and neurological stroke damage and could be expected to predict poststroke outcomes. Objective. The aim of this study was to explore the prognostic value of early poststroke WBC and NC to predict cognition, mood, and disability outcomes at 3 and 12 months poststroke. Methods. Routine clinical analyses WBC and NC were collected at 3 time points in the first 4 days of hospitalization from 156 acute stroke patients. Correlations using hierarchical or ordinal regressions were explored between acute WBC and NC and functional recovery, depression, and cognition at 3 and 12 months poststroke, after covarying for age and baseline stroke severity. Results. We found significant increases in NC between <12 hours and 24 to 48 hours time points (P = .05). Hierarchical regressions, covaried for age and baseline stroke severity, found that 24 to 48 hours WBC (P = .05) and NC (P = .04) significantly predicted 3-month cognition scores. Similarly, 24 to 48 hours WBC (P = .05) and NC (P = .02) predicted cognition scores at 12 months. Increases in WBC and NC were predictive of increased cognition scores at both 3 and 12 months (positive recovery) though there were no significant associations between WBC and NC and disability or depression scores. Conclusions. Routine acute stroke clinical laboratory tests such as WBC and NC taken between 24 and 48 hours poststroke are predictive of cognition poststroke. It is interpreted that higher rapid immunological activation in the acute phase is an indicator for the trajectory of positive stroke recovery.

Keywords

blood biomarkers poststroke recovery cognition white blood cell neutrophil leukocyte

Acute focal brain ischemia is characterized by rapid decrease in oxygen and adenosine triphosphate (ATP) availability leading to impairment of ionic pumps compromise of basement membranes and cellular swelling.¹ Such neuronal excitotoxicity immediately induces a sterile inflammatory response mediated by damage associated molecular patterns, increased permeability of the blood-brain barrier,² and local infiltration by peripheral blood leukocytes.³ Thus, the interaction of immune and inflammatory pathways has long been conceptualized as one of the mechanisms underlying long-term stroke recovery.^4,5 Indeed, it has also been proposed that recovery of neurological function after stroke may be mediated by the initial infiltration of leukocytes into the infarcted area and/or with the associated immune and inflammatory response within the first to second week poststroke.^5,6 However, general clinical blood tests such as the complete blood count or full blood examination⁷ and component leukocyte or white blood cell (WBC) count that are prioritized following suspected stroke admission⁸ are predominantly utilized to investigate the differential diagnoses presenting as stroke mimics such as the presence of infections or hypoglycemia⁹ rather than as a marker of tissue injury, recovery, and prognosis.

Previous studies in poststroke populations have shown that persistent leukocytosis (WBC > 11 × 10⁹ cells/L in adults) 48 hours after admission is associated with higher baseline scores on the National Institute of Health Stroke Scale (NIHSS) and worse functional recovery at discharge and longer lengths of stay (mean = 10.4 days vs mean = 6.6 days).¹⁰ Other studies have also shown that even after controlling for age and baseline stroke severity, relatively elevated leukocyte count (regardless of reference range defined leukocytosis¹¹), is a predictor of poststroke NIHSS severity, modified Rankin scale scores (mRS), and stroke volume.^12,13 Indeed, controlling for age and baseline stroke severity is extremely important to reveal changes in poststroke immune responses while addressing the natural progression of immunosenescence in elderly adults¹⁴ and the large effect of acute stroke on the peripheral immune system.¹⁵ However, knowing when the most clinically and scientifically valuable time points post stroke onset are to measure blood products such as WBC and neutrophil count (NC) for the prediction of poststroke recovery, has not yet been established. Studies examining the acute trajectory of poststroke WBC¹⁶ and NC¹⁷ have shown that cell counts increase rapidly within the first 0 to 2 days poststroke and then fall between 3 and 7 days. Indeed, both the timing and amplitude of acute poststroke immune responses have been found to correlate with stroke severity within 24 hours of symptom onset¹⁸ and also recently have been reported to correlate with improvement on NIHSS scores over the first week.¹⁹

A recent 2017 review and international consensus recommendation concludes that there is a relative lack of evidence for blood-based biomarkers to predict stroke outcomes and recovery,²⁰ when poststroke recovery is defined using the mRS disability and functional recovery scale.²¹ Yet this measure does not include mood or cognitive outcomes, despite the fact that they are commonly reported as unmet needs by stroke survivors.²² Recent studies examining the association of WBCs, neutrophils, and lymphocytes have determined that increased acute neutrophil-to-lymphocyte ratio is often predictive of poststroke depression.²³ Indeed, despite accumulating evidence for the use of acute routine investigations of blood products in the association and prediction of stroke outcomes, these measures have yet to be adopted by clinical guidelines.²⁴

The role of immune markers in depression and cognition regardless of stroke involves a complex pathophysiological system linking immune dysregulation, chronic inflammation, and neurobehavioral changes.²⁵ Previous meta-analyses have identified that increased levels of inflammatory mediators such as interleukin 1, interleukin 6, and C-reactive protein are associated with increased symptoms of depression.²⁶ Indeed, chronic inflammation has the potential to mediate neurotoxic response dysregulation of the glutamatergic system and upregulation of reactive oxygen and nitrogen species, with structural changes such as hippocampal and prefontal cortex matter loss.²⁵ Furthermore, there are sickness behavior similarities between subjects facing immune challenge (such as infection) and depression symptoms in many mammalian species.²⁷ The link between cognition and immunity has not been as well established as the physiological basis of immunity in depression.²⁷ Multiple large prospective studies examining elderly adults found both positive and nonsignificant associations between cognitive outcomes and leukocyte count^28,29 or inflammatory markers.^29,30 Studies in stroke immunity and cognition have identified that increases in myelin basic protein antibodies that respond to central nervous system antigens are associated with longitudinal cognitive decline.^4,31

We aimed to explore the associations between measures of acute poststroke WBC and NC acquired within the first week of hospitalization and longer-term poststroke recovery at 3 and 12 months defined by clinical scales for functional recovery (mRS), depression (Montgomery-Åsberg Depression Rating Scale [MADRS]),³² and cognition (Montreal Cognitive Assessment [MoCA]).³³ Investigation of predictive associations between routine blood tests conducted in the acute phase to clinical outcomes at 3 and 12 months poststroke was possible using the longitudinal START stroke cohort.³⁴ It was hypothesized that increases in hematological markers of immune function within 0 to 12 hours, 24 to 48 hours, and 3 to 7 days poststroke would be associated with poorer poststroke functional, cognitive, and depressive outcomes at 3 and 12 months poststroke. Furthermore, we investigated from the acute samples available, the optimal timing for blood draw and analysis that would be the most useful predictor of outcomes. Knowledge of an association between routine bloods tests, such as WBC and NC, combined with measures of recovery in the first year poststroke, may assist with longer term prediction and highlight the importance of treating early immunological dysfunction to reestablish immune and inflammatory homeostasis for better long-term outcome.

Method

Participants

Patients were recruited as part of START-EXTEND (Neuroscience Trials Australia: NTA 0901, NTA 0901, NCT01580839)³⁵ and START-PrePARE (NTA 0902)³⁴ cohort studies. These studies recruited 219 patients from a group of participating hospitals in Australia between 2011 and 2015. Briefly, the START-EXTEND study investigated the clinical characteristics of extension of the thrombolysis treatment window and included patients within 9 hours of stroke onset, and the START-PrePARE cohort allowed for recruitment of patients with stroke onset up to 3 days with the purpose of examining predictors of recovery, including the functional brain networks associated with longitudinal stroke recovery.¹⁶ A subset of 156 participants (104 male and 41 female) aged 71 (SD = 12.78) from whom clinical scores and blood test results (Table 3) were available were selected for the current study.

Blood Tests

Routine blood tests for a range of hematological markers were conducted at participating hospital sites by onsite commercial pathology services. Venipuncture was conducted at baseline (<12 hours), 24 to 48 hours, and 3 to 7 days poststroke, as outlined in the START study protocols.³⁴ Data presented are drawn from electronic patient records and standardized to units recommended by the Royal College of Pathologists Australia.¹¹ Only total WBC and NC were consistently available from data obtained from the START cohort.

Baseline Demographics and Clinical Characteristics

Patient demographics were obtained by semistructured interview at baseline admission as part of inclusion in the START cohort studies.³⁴ Routine demographic and clinical assessments included age, gender, stroke risk factors, and regular blood pressure monitoring. Stroke severity was assessed at baseline using the NIHSS according to clinical trial protocols.³⁶ Baseline patient characteristics including risk factors and physiological measures were obtained.

Clinical Measures

A series of standardized assessments of neurological function, disability status, cognition, and depression were conducted at 3 months and 12 months poststroke by double-blinded trained health professionals and encompassed several domains of poststroke recovery. Descriptive details of clinical measures and scoring criteria are presented in Table 1.

Table 1.

Clinical Assessments Administered at 3 Months and 12 Months Poststroke.

Clinical scale	Purpose	No. of items	Test scoring and interpretation^a
NIHSS—National Institute of Health Stroke Scale³⁶	Quantification of the degree of stroke severity by assessment of multiple neurological domains including motor, vision, speech, and cognition	11	0: No stroke symptoms 1-4: Mild stroke severity 5-15: Moderate stroke severity 16-20: Moderate to severe stroke 21-42: Severe stroke
mRS—Modified Rankin Scale²¹	Assessment of degree of disability or dependence in daily activities	Structured interview	0: No symptoms 1: No significant disability 2: Slight disability 3: Moderate disability 4: Moderately severe disability 5: Severe disability 6: Dead
MADRS—Montgomery-Åsberg Depression Rating Scale–Structured Interview³⁷	Assessment of presence and severity of depressive symptoms in the past week	10	0-6: Normal, no/minimal depressive symptoms 7-19: Mild depressive symptoms 20-34: Moderate depressive symptoms >34: Severe depressive symptoms
MoCA—Montreal Cognitive Assessment³³	Rapid screening assessment of generalized cognitive performance	30	⩾26: Normal <17-25: Mild cognitive impairment ⩽16: Cognitive impairment

Scoring and interpretation relates to total score obtained for each of the assessments. Total score is typically obtained following addition of item scores, as recommended for each of the tools. Details are provided to facilitate interpretation of test scores reported in this manuscript.

The MADRS³⁷ is a measure of depressive symptoms and has been widely used in major depressive disorders³⁸ and poststroke research.³⁹ The MADRS was administered using the structured interview guide, which is reported to have higher reliability than the self-report method.³⁷ A score above 8 on the MADRS for depressive symptom screening has been proposed as the normative cutoff in stroke cases³² for increased sensitivity (>0.80) and acceptable specificity (0.74) compared to the original recommended score of 12.³⁷ The mRS was used to describe functional disability of patients.²¹

The MoCA³³ is a screening measure of multiple cognitive domains and has been shown to be more sensitive to poststroke cognitive impairment than the Mini-Mental State Examination.⁴⁰ Normative cutoffs to determine cognitive impairment in poststroke research have been suggested at 19 to 22 for acute administration and 20 to 27 in the chronic phase.⁴¹

Data Analyses

A Pearson’s correlation analysis was first conducted to independently assess the associations between acute immune cell counts and longitudinal functional recovery, mood, and cognition status. A post hoc power analysis showed that our sample size (n = 156) was adequate to detect a medium correlation (r = .30) with a power of 0.8 and α level of .05 (n_pwr = 84). Normality assessment using the Kolmogorov-Smirnov (K-S) test was conducted after outlier removal at ±>2.5 interquartile range (IQR). The K-S demonstrated that normality was within acceptable bounds for parametric tests. Repeated-measures ANCOVAs were used to test for significant differences between immune cell counts at 0 to 12 hours, 24 to 48 hours, and 3 to 7 days time points poststroke, and t tests were used to test changes between clinical scores at 3 and 12 months. Post hoc ANCOVA analyses were used to determine changes between blood test time points.

Pearson’s partial correlations were used to explore the associations between acute laboratory blood tests (ie, <12 hours, 24-48 hours, and 3-7 days) and mRS, MADRS, and MoCA at 3 and 12 months. To control for confounding factors such as age and baseline stroke severity (previously identified as being associated with WBC count and/or recovery^42,43), these factors were entered as covariates in the partial correlational matrix to allow for examination of the unique contribution of WBC and NC to poststroke clinical recovery. Hierarchical regressions were conducted following significant partial correlations to determine the changes in explained variance from the addition of immune cell counts to stroke and severity predicting MADRS and MoCA scores at 3 and 12 months. As this was an exploratory study, an α level of .05 was used to detect statistical significance after adjusting for covariates, for each of the 3 outcomes.

Results

Baseline patient characteristics including risk factors are reported in Table 2. Based on individual patient⁴⁴ NIHSS scores (mean of 8.47, range = 0-32, SD = 6.49, median = 6, IQR = 3-6) the neurological severity of ischemic stroke in this cohort at the baseline assessment (<12 hours) was primarily mild to moderate. At 3 and 12 months poststroke, mean stroke severity was still in the mild range (Table 3). Quantification of recovery outcomes showed relatively mild to moderate clinical scores on the mRS²¹ and MADRS⁴⁵ at 3 and 12 months (Table 3). However, the trajectory of recovery for cognitive outcomes poststroke was poor, with a significant decrease of MoCA scores from 3 months to 12 months post stroke (M_diff = 0.96, t = 2.72, degrees of freedom [df] = 137, P = .001), with a small effect size (d = 0.23). Clinically, this represents a trend from cognitive functioning in the unimpaired range toward mild cognitive impairment.⁴⁶ Stroke severity symptoms assessed using the NIHSS were significantly increased between 3 months and 12 months poststroke (M_diff = 0.64, t = 3.34, df = 180, P = .001, d = 0.21), indicating deterioration in neurological symptoms.

Table 2.

Patient Baseline Characteristics.

	Frequency (%)	Mean (SD)
Diabetes	27 (17.2%)
Lipid disorder	76 (51.3%)
Ischemic heart disease	41 (26.1%)
Hypertension	91 (57.6%)
Atrial fibrillation	41 (25.9%)
Height		162.56 (25.24)
Weight		78.76 (16.84)
Temperature		36.40 (0.65)
Heart rate		76.10 (13.90)
Systolic blood pressure		144.71 (23.07)
Diastolic pressure		79.44 (12.98)
Baseline NIHSS		8.47 (6.49)

Abbreviation: NIHSS, National Institute of Health Stroke Scale.

Table 3.

Clinical Scores at 3 Months and 12 Months Poststroke (n = 157).

	3 Months				12 Months				M_diff	P
	M (SD)	Range	Median (IQR)	25th, 75th percentiles	M (SD)	Range	Median (IQR)	25th, 75th percentiles	M_diff	P
NIHSS	1.33 (2.41)—None—very mild stroke symptoms^a	0-17	1 (2)	0-2	1.29 (2.78)	0-18	0 (1.75)	0-1	0.64	>.01**
mRS	1.34 (1.26)—None—slight disability	0-5	1 (2)	1-3	1.25 (1.21)	0-4	1 (2)	0-2	0.14	.08
MADRS	7.86 (7.94)—Mild depressive symptoms	0-33	5 (10)	1-11	6.50 (6.91)	0-29	4 (10)	0-10.25	0.83	.08
MoCA	24.61 (5.39)—Normal to MCI	6-31	26 (5.25)	23-28.25	21.13 (8.98)	0-31	26 (7)	21-28	0.96	>.01**

Abbreviations: M, mean; IQR, interquartile range; NIHSS, National Institute of Health Stroke Scale; mRS, Modified Rankin Scale; MADRS, Montgomery-Åsberg Depression Rating Scale; MoCA, Montreal Cognitive Assessment.

Scoring comments are based on descriptors given for scoring ranges based on original assessment. Paired samples t tests were conducted on all clinical scores between 3 and 12 months.

P < .01.

Details of acquisition, including the average time from stroke onset to blood draw for each time point, and counts of WBC and neutrophils at <12 hours, 24 to 48 hours, and 3 to 7 days, are provided in Table 4. WBC and NC showed an increasing pattern from <12 hours to 24 to 48 hours poststroke, followed by a trend back toward <12 hour levels at 3 to 7 days (Figure 1). A repeated-measures ANCOVA for NC at <12 hours, 24 to 48 hours, and 3 to 7 days, with age and stroke severity as covariates, was significant F(2, 76) = 4.52, P = .01, with post hocs showing that changes between <12 hours and 24 to 48 hours were significantly different (M_diff = 0.88, 95% confidence interval [CI] = [−0.38, 1.78], P< .05).

Table 4.

Routine Blood Tests: Details of Acquisition and Counts at 0 to 12 Hours, 12 to 48 Hours, and 3 to 7 Days.

	Reference ranges^a	0-12 Hours			24-48 Hours			3-7 Days
	Reference ranges^a	M (SD)	Range	Median (IQR)	M (SD)	Range	Median (IQR)	M (SD)	Range	Median (IQR)
Time from stroke to blood draw		5.12 (1.63) hours	1.42-9.18	5.13 (1.97)	29.62 (3.15)	21.25-35.75	30.23 (3.50)	3.29 (0.69) Days	1.92-5.33	3.28 (0.55)
Immune function
White blood cell count	4.3-10.3 × 10⁹/L	8.50 (2.85)	3.60-19.40	8.05 (3.28)	9.06 (3.37)	4.10-23.00	8.96 (4.08)	8.64 (2.65)	4.10-23	8.35 (3.33)
Neutrophil count	1.5-8.0 × 10⁹/L	5.96 (2.68)	1.87-15.30	5.3 (2.97)	6.48 (3.30)	1.60-21.10	6 (3.45)	6.18 (2.56)	2.20-16.40	5.80 (2.62)

Abbreviations: M, mean; IQR, interquartile range.

Reference ranges are obtained for adults but are slightly lower on average for elderly patients.⁷⁵

Figure 1.

Trend of mean white blood cell count and neutrophil count at <12 hours, 24 to 48 hours, and 3 to 7 days. Error bars represent SD.

A partial correlation matrix was used to explore the relationship between the early blood tests at <12 hours, 24 to 48 hours, and 3 to 7 days and clinical measures at 3 months and 12 months, adjusted for age and baseline NIHSS severity score (Table 5). Depression symptoms as measured by the MADRS were not significantly associated with any acute blood WBC or NC counts.

Table 5.

Correlation Between Immune Cell Counts (Multiple Acute Times) and Clinical Outcome Scores at 3 Months and 12 Months Poststroke (Covaried for Age and Baseline Stroke Severity)^a.

Immune cell counts	3 Months			12 Months
Immune cell counts	mRS	MADRS	MoCA	mRS	MADRS	MoCA
White blood cell count—0-12 hours	0.09	0.18	−0.03	0.12	0.16	0.06
White blood cell count—24-48 hours	0.13	0.04	0.26*	0.09	−0.10	0.26*
White blood cell count—3-7 days	0.09	0.08	−0.15	0.20*	−0.11	0.01
Neutrophil count—0-12 hours	0.09	0.15	−0.02	0.17*	0.13	−0.01
Neutrophil count—24-48 hours	0.08	0.10	0.29*	0.07	−0.09	0.31*
Neutrophil count—3-7 days	−0.03	0.05	−0.18	0.09	−0.07	−0.01

Abbreviations: mRS, Modified Rankin Scale; MADRS, Montgomery-Åsberg Depression Rating Scale; MoCA, Montreal Cognitive Assessment.

Pearson correlations covaried for age and NIHSS score.

p < .05.

As follow-up analyses to the significant correlations, hierarchical regressions were conducted between WBC and NC counts at <12 hours, 24 to 48 hours, and 3 to 7 days with 3-month (Table 6) and 12-month (Table 7) poststroke MoCA scores, with age and baseline NIHSS entered as covariates for all models in Step 1. In prediction of 3-month MoCA scores, the regression model with age and baseline NIHSS entered as Step 1, showed medium effect sizes (R² = .20, F[2, 54] = 6.84, P = .002). The addition of 24- to 48-hour WBC counts (B = 0.49, P = .05) to Step 1 improved explained variance with a small effect size (R² change = .06, F[1, 53] = 3.88, P = .05). Similarly, the addition of 24- to 48-hour NC counts (B = 0.45, P = .04) to Step 1 improved explained variance with a small effect size (R² change = .06, F[1, 53] = 4.71, P = .04). In prediction of 12-month MoCA scores, the regression model with baseline NIHSS entered as Step 1 showed large effect sizes (R² = .34, F[2, 56] = 14.70, P = >.001). The addition of 24- to 48-hour WBC counts (B = 0.74, P = .05) to Step 1 improved explained variance for 12-month MoCA scores with a small effect size (R² change = .05, F[1, 55] = 4.08, P = .048). The addition of 24- to 48-hour NC counts (B = 0.45, P = .02) to Step 1 improved explained variance with a small effect size (R² change = .06, F[1, 55] = 5.87, P = .02). As expected, age and baseline stroke severity predict a significant portion of 3- and 12-month MoCA scores, generally in the expected direction of cognitive decline. In contrast, given the unstandardized B of regressions models with the addition of immune cell counts, an increase of 1 × 10⁹ WBC count and NC at 24 to 48 hours poststroke is predictive of an approximate 0.53 increase in MoCA scores at 3 and 12 months.

Table 6.

Hierarchical Regressions for White Blood Cell Count and Neutrophil Count at 24 to 48 Hours With 3-Month MoCA Scores, Controlled for by Age and Baseline National Institute of Health Stroke Scale.

Explanatory variable	R ²	R² Δ	P	B	SE B	β	P
Step 1	.20	—	>.01*
Age				−0.10	0.05	−0.26	.04*
Baseline NIHSS				−0.23	0.09	−0.31	.02*
Step 2	.26	.06	>.01*
Age				−0.10	0.05	−0.25	.04*
Baseline NIHSS				−0.31	0.10	−0.42	>.01**
White blood cell count 24-48 hours				0.49	0.25	0.26	>.05*
Step 3	.27	.06	>.01*
Age				−0.11	0.05	−0.27	.03*
Baseline NIHSS				−0.33	0.10	−0.45	>.00*
Neutrophil count 24-48 hours				0.45	0.21	0.29	.04*

Abbreviations: NIHSS, National Institute of Health Stroke Scale; MoCA, Montreal Cognitive Assessment.

P < .05, **P < .01.

Table 7.

Hierarchical Regressions for White Blood Cell Count and Neutrophil Count at 24 to 48 Hours With 12-Month MoCA Scores, Controlled for by Age and Baseline National Institute of Health Stroke Scale.

Explanatory variable	R ²	R² Δ	P	B	SE B	β	P
Step 1	0.34	—	>.01**
Age				−0.14	0.07	−0.21	.06
Baseline NIHSS				−0.61	0.14	−0.50	>.00**
Step 2	0.39	.05	>.01**
Age				−0.13	0.07	−0.20	.07
Baseline NIHSS				−0.73	0.15	−0.60	>.00**
White blood cell count 24-48 hours				0.74	0.37	0.23	>.05*
Step 3	0.41	.06	>.01**
Age				−0.14	0.07	−0.22	.04*
Baseline NIHSS				−0.78	0.15	−0.64	>.01**
Neutrophil count 24-48 hours				0.73	0.30	0.29	.02*

Abbreviations: NIHSS, National Institute of Health Stroke Scale; MoCA, Montreal Cognitive Assessment.

P < .05, **P < .01.

An ordinal regression, with age and baseline NIHSS as covariates, between WBC count at 3 to 7 days and mRS scores at 12 months found an odds ratio of 1.02, 95% CI = [0.89, 1.18], Wald χ²(1) = .09, which was not significant (P = .76). An ordinal regression was conducted with age and baseline NIHSS entered as covariates with NC at <12 hours predicting mRS scores at 12 months found a significant odds ratio of 1.13, 95% CI = [1.00, 1.27], Wald χ²(1) = 3.91, P = .048.

Discussion

In this longitudinal cohort study, we have investigated the associations between hospital-based hematological markers of immune function taken within 1 week poststroke and outcomes of clinical functioning and recovery at 3 months and 12 months post vascular event. A significant increase in NC was found between <12 hours and 24 to 48 hours timepoints. The partial correlations and hierarchical regressions revealed that elevations in poststroke WBC and NC at 24 to 48 hours are associated with better cognitive outcome at 3 and 12 months poststroke, independent of age and baseline stroke severity. Although the convergence of regression models showed that both WBC and NC are predictive of increased MoCA scores at 2 separate time points, these data should be interpreted with caution given the relatively weak but consistent effect sizes and small sample sizes. To date, there has only been a single study showing that WBC and/or NC early (within 24-48 hours) are correlated or predictive of better cognition at longitudinal time points such as 3 or 12 months poststroke.⁴⁷ Rather, as recently indicated, the majority of research has examined novel biomarkers such as interleukins and apolipoproteins in relation to poststroke cognitive impairment and recovery (see review in Wahul et al⁴⁸).

The significant regression of immune-related white blood cells to cognitive outcomes can be interpreted as evidence for the involvement of immune system activation following cerebral ischemia in acute poststroke having long-term effects on functional recovery.² As shown in the hierarchical regression models, baseline stroke severity and age account for a large amount of variance in relation to longitudinal MoCA scores and are in the expected direction of predicting cognitive decline. Indeed, demographics, functional factors,⁴⁹ and lesion characteristics⁵⁰ often predict worsened cognitive outcomes poststroke. Interestingly, the addition of 24- to 48-hour WBC and NC to these models significantly explained additional variance and while the regression models overall were predictive of cognitive decline, WBC and NC as individual predictors were found to be in the unexpected direction of predicting positive cognitive outcomes. Previous studies examining the link between blood-based biomarkers and poststroke cognition have largely focused on the association of inflammatory markers to impairment.⁴⁸ Increases in serum C-reactive protein have been reported to predict lowered global cognitive function within the first month of stroke.⁵¹ Furthermore, prolonged erythrocyte sedimentation rates have been associated with cognitive decline.⁵² However, few studies have described a blood-based biomarker for positive cognitive outcomes, especially where blood-based measures are acquired early in admission and investigated in association with long term cognition.

The biophysiological correlates of worsened poststroke outcome are usually understood in the context of ongoing peripheral and central nervous system inflammation⁵³ and improved outcomes, including cognition, are understood in the context of neural repair and plasticity.⁵⁴ Although it is difficult to speculate on the functional neuro-immunomodulation of inflammation and neural repair in the context of gross peripheral leukocyte counts, acute innate immune responses in stroke have been suggested to be both inflammatory and neuroprotective.⁵⁵ Indeed, more rigorous studies with better defined timing for blood draw and increased sample size representative of more severe stroke cases are required to replicate the positive associations between immune cell counts and cognitive function found here.

Our study did not find any significant correlations between acute immune cell count and depressive symptoms, when adjusted for age and stroke severity, despite previous research suggesting an inflammation- and immune-based pathogenesis for poststroke depression.^56,57 The significant prediction of NC at <12 hours to mRS at 12 months is in agreement with previous studies that show an association between increased acute leukocyte counts and poor mRS functional recovery outcomes at both 3 months⁵⁸ and 12 months.¹⁸ These findings may appear inconsistent with our results that found an association with increased WBC and NC with positive cognitive outcomes in the same cohort. Although functional recovery might be expected to be associated with cognition, the mRS is perhaps not sensitive enough to adequately capture cognitive function or neuropsychological impairments in stroke patients in comparison to the MoCA.⁵⁹

Our results also found that an increase in NC at <12 hours poststroke is predictive of poorer functional recovery at 12 months. An increasing NC from 0 to 48 hours poststroke has been previously reported¹⁷ and may reflect changes in blood-brain barrier (BBB) permeability due to disassembly of the tight junctions by hypoxia-reoxygenation² and disruptions to both peripheral and central nervous system immune/inflammatory pathways (see review in Brouns and De Deyn⁶⁰). Although blood tests at 12 hours poststroke could not be compared to prestroke reference levels, this systemic change is confirmation of several known physiological responses to stroke such as penumbral excitotoxicity and perivascular edema⁶¹ and resulting increased glymphatic perfusion.⁶² Indeed, these results may be related to the molecular signaling timeline following neuronal oncosis driven by mitrochondrial and associated bioenergetic impairment (ie, reduced availability of ATP^63,64). Indeed, acute edema and inflammation following stroke-induced excitotoxity would be expected to lead to previously activated autophagy and necrosis pathways,⁶⁵ and be associated with the abrupt upregulation of immune activity that has been predicted to occur around 24 to 48 hours poststroke in conjunction with BBB permeability and be related to peripheral innate immunity as pro-inflammatory cytokines and other endocrine responses influence leukocyte recruitment across a vulnerable BBB.⁶⁶ The timeline of leukocyte levels presented here closely resembles dynamic contrast-enhanced magnetic resonance image defined BBB permeability pattern within the first 3 days of stroke.² Previous studies within this acute stage have shown a phasic increase in permeability-surface area product from 6 to 48 hours while subsequently returning to 6 hour levels beyond 48 hours poststroke.² Similarly, the return of WBC and NC to <12 hour levels at 3 to 7 days may suggest a shift in the pathophysiological cell death cascade, including functions such as cell necrosis, vascular leakage, cerebral edema, and maintenance processes such as the clearance of cellular debris, ionic osmoregulation, and restoration of energy processes—processes that have been well described in animal models.^5,63 The apparent reduction in WBC and NC may also be neuroprotective, as persistent antigen-specific immune cell recruitment and activity in the brain is predictive of poorer outcome and mortality.⁴

Considering the ionic, biochemical, and molecular cascades accompanying stroke damage,⁶⁷ the leukocyte pattern observed may be reflective of the mild nature of stroke in this cohort and an ongoing trajectory to positive poststroke recovery that has not been observed in other studies.⁴² Our regression analyses of acute immune cell counts predicting increased MoCA scores also suggest that trajectories of later poststroke recovery could be based on the extent of the early leukocyte response and the subsequent attempt to restore immune cell balance within the first week. The timing of the leukocyte infiltration response may also be indicative of the duration of BBB permeability, vessel occlusion, and therefore increased tissue damage, with longer duration strokes showing an earlier leukocyte infiltration.⁶⁸ On the contrary, poor poststroke recovery is commonly followed by continuing immunosuppression that may continue beyond 3 months.^5,69 This trend toward immunosuppression supports the prevalence and timeline of poststroke infections such as upper respiratory tract infection, pneumonia, and urinary tract infections that commonly present toward the end of the first week.^70,71 Indeed, the timeline for immunoreactivity seems to precede optimal trajectories for patients undergoing early physiotherapy rehabilitation at 15.6 days poststroke, followed by a period of 4 weeks of rapid functional recovery.⁷²

There are several limitations associated with the methodologies in this study. This study has examined a cross section of stroke patients obtained from a longitudinal stroke cohort study examining thrombolysis outcomes and stroke recovery outcomes: it is not inclusive of all patients as may be the case in national level data registries.^34,35 The nature of the original inclusion criteria and ability to participate in a year-long study may bias the characteristics of the sample, in particular the mild stroke severity, depressive symptoms, and cognitive outcomes. Age and baseline NIHSS are factors that have been shown in previous studies to correlate highly with clinical measures at 3 months^42,73 and thus were selected to adjust for in these analyses. It is acknowledged, however, that there are other baseline measures such as infection status or atherosclerosis in relationship to WBC and NC that could be addressed as covariates. Furthermore, there may be large variability in the amplitude of immune responses¹⁷ given the wide acute blood sampling times achieved in this study. It is also noted that this study did not collect lymphocyte cell counts that could be used to explore the neutrophil to lymphocyte ratio which has been suggested to be a better measure of acute inflammation than immune cell counts alone.⁷⁴ In order to provide more robust evidence for predictive value of immune cell counts in stroke recovery, future studies should examine national- or hospital-level data registries with similar blood test protocols. Furthermore, sample sizes should be greatly increased, in contrast to the relatively low sample size used in this exploratory study.

Conclusion

Blood tests are routinely prescribed as an investigative tool in acute stroke management but are seldom used as clinical tools to predict poststroke recovery in the longer term. We have found that increased WBC and neutrophils count within the first week of stroke are consistently independent predictors of better cognitive outcomes at 3 and 12 months. In conjunction with the observed trends in immune cell counts, these data suggest that the degree and timing of the peripheral immune response in the acute phase is likely to play a role in setting the trajectory for long-term recovery. Further studies should be conducted on acute blood tests and their role in similar long-term clinical outcomes, with the goal of eventual integration into stroke guidelines and best practice.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors acknowledge financial support for the conduct of the research from the Australian Commonwealth Scientific and Industrial Research Organisation (CSIRO) Flagship Collaboration Fund through the Preventative Health Flagship scheme, and support for the writing from the James S. McDonnell Foundation 21st Century Science Initiative in Cognitive Rehabilitation Collaborative Award (#220020413), NHMRC Program Grant: Saving Brain and Changing Practice in Stroke (#1113352), and La Trobe University Post-Graduate Scholarship from the Understanding Diseases research focus area awarded to VAN.

ORCID iD

Vinh A. Nguyen

References

Weerasinghe

Buja

. Oncosis: an important non-apoptotic mode of cell death. Exp Mol Pathol. 2012;93:302-308.

Merali

Huang

Mikulis

Silver

Kassner

. Evolution of blood-brain-barrier permeability after acute ischemic stroke. PLoS One. 2017;12:e0171558.

McCombe

Read

. Immune and inflammatory responses to stroke: good or bad? Int J Stroke. 2008;3:254-265.

Becker

Kalil

Tanzi

, et al. Autoimmune responses to brain following stroke are associated with worse outcome. Stroke. 2011;42:2763-2769.

Kamel

Iadecola

. Brain-immune interactions and ischemic stroke: clinical implications. Arch Neurol. 2012;69:576-581.

Planas

. Role of immune cells migrating to the ischemic brain. Stroke. 2018;49:2261-2267.

Lappé

Horne

Shah

, et al. Red cell distribution width, C-reactive protein, the complete blood count, and mortality in patients with coronary disease and a normal comparison population. Clin Chim Acta. 2011;412:2094-2099.

Kristensen

Iversen

AKS

Gerds

, et al. Routine blood tests are associated with short term mortality and can improve emergency department triage: a cohort study of >12,000 patients. Scand J Trauma Resusc Emerg Med. 2017;25:115.

Hosseininezhad

Sohrabnejad

. Stroke mimics in patients with clinical signs of stroke. Caspian J Intern Med. 2017;8:213-216.

10.

Boehme

Kumar

Lyerly

, et al. Persistent leukocytosis—is this a persistent problem for patients with acute ischemic stroke? J Stroke Cerebrovasc Dis. 2014;23:1939-1943.

11.

Hawkins

Badrick

. Reporting unit size and measurement uncertainty: current Australian practice in clinical chemistry and haematology. Pathology. 2015;47:462-465.

12.

Nardi

Milia

Eusebi

Paciaroni

Caso

Agnelli

. Admission leukocytosis in acute cerebral ischemia: influence on early outcome. J Stroke Cerebrovasc Dis. 2012;21:819-824.

13.

Zhou

Zhang

, et al. Association of stroke clinical outcomes with coexistence of hyperglycemia and biomarkers of inflammation. J Stroke Cerebrovasc Dis. 2015;24:1250-1255.

14.

Fulop

Larbi

Dupuis

, et al. Immunosenescence and inflamm-aging as two sides of the same coin: friends or foes? Front Immunol. 2018;8:1960.

15.

Kammersgaard

Jørgensen

Nakayama

Reith

Raaschou

Olsen

. Leukocytosis in acute stroke: relation to initial stroke severity, infarct size, and outcome: the Copenhagen stroke study. J Stroke Cerebrovasc Dis. 1999;8:259-263.

16.

Haeusler

Schmidt

WUH

Föhring

, et al. Cellular immunodepression preceding infectious complications after acute ischemic stroke in humans. Cerebrovasc Dis. 2008;25:50-58.

17.

Petrone

Eisenman

Steele

Mosmiller

Urhie

Zdilla

. Temporal dynamics of peripheral neutrophil and lymphocytes following acute ischemic stroke. Neurol Sci. 2019;40:1877-1885.

18.

Quan

Wang

Zhang

Wang

. Leukocyte count and adverse clinical outcomes in acute ischemic stroke patients. Front Neurol. 2019;10:1240.

19.

Vogelgesang

Witt

Heuer

, et al. Clinical improvement following stroke promptly reverses post-stroke cellular immune alterations. Front Neurol. 2019;10:414.

20.

Boyd

Hayward

Ward

, et al. Biomarkers of stroke recovery: consensus-based core recommendations from the Stroke Recovery and Rehabilitation roundtable. Neurorehabil Neural Repair. 2017;31:864-876.

21.

Banks

Marotta

. outcomes validity and reliability of the modified Rankin scale: implications for stroke clinical trials. Stroke. 2007;38:1091-1096.

22.

Rohde

Gaynor

Large

, et al. Stroke survivor cognitive decline and psychological wellbeing of family caregivers five years post-stroke: a cross-sectional analysis. Top Stroke Rehabil. 2019;26:180-186.

23.

Chen

Luan

Zhao

, et al. The association between neutrophil-to-lymphocyte ratio and post-stroke depression. Clin Chim Acta. 2018;486:298-302.

24.

Powers

Rabinstein

Ackerson

, et al. 2018 guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2018;49:e46-e110.

25.

Leonard

. The concept of depression as a dysfunction of the immune system. Curr Immunol Rev. 2010;6:205-212.

26.

Howren

Lamkin

Suls

. Associations of depression with C-reactive protein, IL-1, and IL-6: a meta-analysis. Psychosom Med. 2009;71:171-186.

27.

Dantzer

. Cytokine, sickness behavior, and depression. Immunol Allergy Clin North Am. 2009;29:247-264.

28.

Taniguchi

Shinkai

Nishi

, et al. Nutritional biomarkers and subsequent cognitive decline among community-dwelling older Japanese: a prospective study. J Gerontol A Biol Sci Med Sci. 2014;69:1276-1283.

29.

Kao

Chang

Chou

Kuo

. White blood cell count and psychomotor cognitive performance in the elderly. Eur J Clin Invest. 2011;41:513-520.

30.

Todd

. Inflammation and cognition in older adults: evidence from Taiwan. Biodemography Soc Biol. 2017;63:309-323.

31.

Becker

Tanzi

Zierath

Buckwalter

. Antibodies to myelin basic protein are associated with cognitive decline after stroke. J Neuroimmunol. 2016;295-296:9-11.

32.

Sagen

Vik

Moum

Mørland

Finset

Dammen

. Screening for anxiety and depression after stroke: comparison of the Hospital Anxiety and Depression Scale and the Montgomery and Åsberg Depression Rating Scale. J Psychosom Res. 2009;67:325-332.

33.

Nasreddine

Phillips

Bédirian

, et al. The Montreal Cognitive Assessment, MoCA: a brief screening tool for mild cognitive impairment. J Am Geriatr Soc. 2005;53:695-699.

34.

Carey

Crewther

Salvado

, et al. STroke imAging pRevention and Treatment (START): a longitudinal stroke cohort study: clinical trials protocol. Int J Stroke. 2013;10:636-644.

35.

Parsons

Christensen

, et al. A multicentre, randomized, double-blinded, placebo-controlled phase III study to investigate Extending the Time for Thrombolysis in Emergency Neurological Deficits (EXTEND). Int J Stroke. 2011;7:74-80.

36.

Lyden

Raman

Liu

Emr

Warren

Marler

. NIHSS certification is reliable across multiple venues. Stroke. 2009;40:2507-2511.

37.

Williams

Kobak

. Development and reliability of a structured interview guide for the Montgomery-Åsberg Depression Rating Scale (SIGMA). Br J Psychiatry. 2008;192:52-58.

38.

Fantino

Moore

. The self-reported Montgomery-Åsberg Depression Rating Scale is a useful evaluative tool in major depressive disorder. BMC Psychiatry. 2009;9:26.

39.

Welin

Bjälkefur

Roland

. Open, randomized pilot study after first stroke. Stroke. 2010;41:1555-1557.

40.

Blackburn

Bafadhel

Randall

Harkness

. Cognitive screening in the acute stroke setting. Age Ageing. 2013;42:113-116.

41.

Chiti

Pantoni

. Use of Montreal Cognitive Assessment in patients with stroke. Stroke. 2014;45:3135-3140.

42.

Furlan

Vergouwen

MDI

Fang

Silver

. White blood cell count is an independent predictor of outcomes after acute ischaemic stroke. Eur J Neurol. 2014;21:215-222.

43.

Chien

Lin

, et al. Total white blood cell count or neutrophil count predict ischemic stroke events among adult Taiwanese: report from a community-based cohort study. BMC Neurol. 2013;13:7.

44.

Fonarow

Saver

Smith

, et al. Relationship of national institutes of health stroke scale to 30-day mortality in Medicare beneficiaries with acute ischemic stroke. J Am Heart Assoc. 2012;1:42-50.

45.

Montgomery

Åsberg

. A new depression scale designed to be sensitive to change. Br J Psychiatry. 1979;134:382-389.

46.

McLennan

Mathias

Brennan

Stewart

. Validity of the Montreal Cognitive Assessment (MoCA) as a screening test for mild cognitive impairment (MCI) in a cardiovascular population. J Geriatr Psychiatry Neurol. 2011;24:33-38.

47.

Tsai

Berry

Beneyto

, et al. A year-long immune profile of the systemic response in acute stroke survivors. Brain. 2019;142:978-991.

48.

Wahul

Pranav Kumar

Chakravarty

. Association of diagnostic stroke biomarkers with post stroke cognitive impairment. J Neurol Disord Stroke. 2018;6:1134-1140.

49.

Zulkifly

MFM

Ghazali

Din

Singh

DKA

Subramaniam

. A review of risk factors for cognitive impairment in stroke survivors. Scientific World J. 2016;2016:3456943.

50.

Yatawara

Chander

Kandiah

. Associations between lesions and domain-specific cognitive decline in poststroke dementia. Neurology. 2018;91:e45-e54.

51.

Rothenburg

Herrmann

Swardfager

, et al. The relationship between inflammatory markers and post stroke cognitive impairment. J Geriatr Psychiatry Neurol. 2010;23:199-205.

52.

Kliper

Bashat

Bornstein

, et al. Cognitive decline after stroke: relation to inflammatory biomarkers and hippocampal volume. Stroke. 2013;44:1433-1435.

53.

Barrington

Lemarchand

Allan

. A brain in flame; do inflammasomes and pyroptosis influence stroke pathology? Brain Pathol. 2017;27:205-212.

54.

Carey

Walsh

Adikari

, et al. Finding the intersection of neuroplasticity, stroke recovery, and learning: scope and contributions to stroke rehabilitation. Neural Plast. 2019;2019:5232374.

55.

Amantea

Greco

Micieli

Bagetta

. Paradigm shift to neuroimmunomodulation for translational neuroprotection in stroke. Front Neurosci. 2018;12:241.

56.

Feng

Fang

Liu

. The neurobiological pathogenesis of poststroke depression. Scientific World J. 2014;2014:521349.

57.

Pascoe

Crewther

Carey

Crewther

. Inflammation and depression: why poststroke depression may be the norm and not the exception. Int J Stroke. 2011;6:128-135.

58.

Pagram

Bivard

Lincz

Levi

. Peripheral immune cell counts and advanced imaging as biomarkers of stroke outcome. Cerebrovasc Dis Extra. 2016;6:120-128.

59.

Rautalin

Sebök

Germans

, et al. Screening tools for early neuropsychological impairment after aneurysmal subarachnoid hemorrhage. Neurol Sci. 2020;41:817-824.

60.

Brouns

De Deyn

. The complexity of neurobiological processes in acute ischemic stroke. Clin Neurol Neurosurg. 2009;111:483-495.

61.

Zador

Stiver

Wang

Manley

. Role of aquaporin-4 in cerebral edema and stroke. Handb Exp Pharmacol. 2009;(190):159-170.

62.

Gaberel

Gakuba

Goulay

, et al. Impaired glymphatic perfusion after strokes revealed by contrast-enhanced MRI: a new target for fibrinolysis? Stroke. 2014;45:3092-3096.

63.

Samary

Pelosi

Silva

Rocco

PRM

. Immunomodulation after ischemic stroke: potential mechanisms and implications for therapy. Crit Care. 2016;20:391.

64.

Liesz

Dalpke

Mracsko

, et al. DAMP signaling is a key pathway inducing immune modulation after brain injury. J Neurosci. 2015;35:583-598.

65.

Chen

Sun

Liu

Sun

. Autophagy: a double-edged sword for neuronal survival after cerebral ischemia. Neural Regen Res. 2014;9:1210-1216.

66.

Yilmaz

Granger

. Leukocyte recruitment and ischemic brain injury. Neuromolecular Med. 2010;12:193-204.

67.

Xing

Arai

Hommel

. Pathophysiologic cascades in ischemic stroke. Int J Stroke. 2012;7:378-385.

68.

Grønberg

Johansen

Kristiansen

Hasseldam

. Leukocyte infiltration in experimental stroke. J Neuroinflammation. 2013;10:115.

69.

Nguyen

Carey

Giummarra

, et al. A pathway proteomic profile of ischemic stroke survivors reveals innate immune dysfunction in association with mild symptoms of depression—a pilot study. Front Neurol. 2016;7:85.

70.

Emsley

HCA

Hopkins

. Acute ischaemic stroke and infection: recent and emerging concepts. Lancet Neurol. 2008;7:341-353.

71.

Grabska

Gromadzka

Członkowska

. Infections and ischemic stroke outcome. Neurol Res Int. 2011;2011:691348.

72.

Lee

Lim

Kim

, et al. Six-month functional recovery of stroke patients: a multi-time-point study. Int J Rehabil Res. 2015;38:173-180.

73.

Huber

Kleine

Kaesmacher

, et al. Blood leukocytes as prognostic parameter in stroke thrombectomy. Cerebrovasc Dis. 2016;42:32-40.

74.

Xue

Huang

Chen

, et al. Neutrophil-to-lymphocyte ratio is a prognostic marker in acute ischemic stroke. J Stroke Cerebrovasc Dis. 2017;26:650-657.

75.

Aminzadeh

Parsa

. Relationship between age and peripheral white blood cell count in patients with sepsis. Int J Prev Med. 2011;2:238-242.