The relationship between neurological function trajectory,assessed by repeated NIHSS measurement,and long-term cardiovascular events,recurrent stroke,and mortality after ischemic stroke

Abstract

Background:

Clinically significant changes in neurological deficits frequently occur after stroke onset, reflecting further neurological injury or neurological improvement. However, the National Institutes of Health Stroke Scale (NIHSS) score is only evaluated once in most studies, usually at stroke onset. Utilizing repeated measures of NIHSS scores to identify different trajectories of neurological function may be more informative and provide more useful predictive information. We determined the association of neurological function trajectories with long-term clinical outcomes after ischemic stroke.

Methods:

A total of 4025 participants with ischemic stroke from the China Antihypertensive Trial in Acute Ischemic Stroke were included. Patients were recruited from 26 hospitals across China between August 2009 and May 2013. A group-based trajectory model was used to identify distinct neurological function trajectories, as measured by NIHSS at admission, 14 days or hospital discharge, and 3 months. Study outcomes were cardiovascular events, recurrent stroke, and all-cause mortality during 3–24 months after ischemic stroke onset. Cox proportional hazards models were used to examine the associations of neurological function trajectories with outcomes.

Results:

We identified three distinct subgroups of NIHSS trajectories: persistent severe (persistent high NIHSS scores during the 3-month follow-up), moderate (NIHSS scores started at around 5 and gradually reduced), and mild (NIHSS scores always below 2). The three trajectory groups had different clinical profiles and different risk of stroke outcomes at 24-month follow-up. Compared to the mild trajectory group, patients in the persistent severe trajectory group had a higher risk of cardiovascular events (multivariable-adjusted hazard ratios (95% confidence intervals) = 1.77 (1.10–2.86)), recurrent stroke (1.82 (1.10–3.00)), and all-cause mortality (5.64 (3.37–9.43)). Those with moderate trajectory had an intermediate risk: 1.45 (1.03–2.04) for cardiovascular events and 1.52 (1.06–2.19) for recurrent stroke.

Conclusion:

Longitudinal neurological function trajectories derived from repeated NIHSS measurements during the first 3 months after stroke provide additional predictive information and are associated with long-term clinical outcomes. The trajectories characterized by persistent severe and moderate neurological impairment were associated with increased risk of subsequent cardiovascular events.

Keywords

Neurological function changes NIHSS trajectories ischemic stroke long-term outcomes

Introduction

Stroke is the leading cause of adult mortality and disability in China, which also has profound and wide-ranging effects on physical and psychological wellbeing.^1,2 Stroke severity is a primary determinant of clinical outcomes in patients with acute ischemic stroke.³ The National Institutes of Health Stroke Scale (NIHSS) is one of the standard methods to measure the severity of stroke.^4–6 In practice, the NIHSS is useful for early prognostication and serial assessment of neurological deficits, with excellent reliability and validity.^7,8 Clinically significant changes in NIHSS scores frequently occurred after the initial onset of stroke, reflecting the secondary neurological injury or neurological improvement.^9–12 However, NIHSS score was evaluated only once in most studies, failing to account for the dynamic changes of neurological function.^7,13 Early neurological deterioration, frequently defined as two NIHSS score change of ⩾4 points in the acute period after ischemic stroke, occurs in approximately one-third of stroke patients and is consistently associated with adverse outcomes.^14–16 It was reported that the dynamic changes in early neurological function, as well as the impacts on clinical outcomes, are time-dependent.^17,18 Therefore, a single or two NIHSS score measurements may not adequately characterize the prediction of long-term stroke prognosis.^18,19

Utilizing repeated measures of NIHSS scores to identify different trajectories of neurological function may be more informative, revealing the complex predictive effects of neurological instability. A previous study reported that early neurological improvement trajectories defined by 48-h longitudinal NIHSS scores could predict 90-day stroke outcomes, and patients with large improvement were more likely to exhibit favorable functional outcomes than those with minimal improvement or no improvement.²⁰ However, little is known about the longitudinal changes of neurological function after hospital discharge, and the effects of distinct neurological fluctuations on clinical outcomes of ischemic stroke patients, especially with regard to long-term recurrent cardiovascular events and death, remain largely unknown. We hypothesized that different trajectories of neurological changes in the subacute phase of stroke may provide more prognostic information for the risk of subsequent poor outcomes.

In this study, we aimed to assess the fluctuations of neurological function within the first 3 months after ischemic stroke by group-based trajectory modeling and to identify patients with similar trajectories. In addition, we further explored the relationships between longitudinal trajectories of neurological function and long-term cardiovascular events, stroke recurrence, and all-cause mortality in patients with ischemic stroke.

Methods

Study population

This study was based on data from the CATIS (China Antihypertensive Trial in Acute Ischemic Stroke), a multicenter, single-blind, blinded end-point randomized clinical trial conducted in 26 hospitals across China. The design and methods have been described in detail previously.²¹ In brief, from August 2009 to May 2013, 4071 patients aged 22 years or older who had ischemic stroke confirmed by computed tomography or magnetic resonance imaging of the brain within 48 h of symptom onset and who had systolic blood pressure (BP) between 140 and 220 mmHg were recruited. Patients with a BP ⩾220/120 mmHg, severe heart failure, acute myocardial infarction or unstable angina, atrial fibrillation, aortic dissection, cerebrovascular stenosis, resistant hypertension, or unconsciousness or treated with intravenous (IV) thrombolytic therapy were excluded from the CATIS trial. In this study, we further excluded participants without available NIHSS measurements at 14 days/discharge and 3 months, or experienced death, stroke recurrence, or cardiovascular events within the first 3 months; a total of 3781 subjects were included for neurological function trajectory analysis and 4025 for neurological deterioration analysis (Supplemental Figure S1).

This study was approved by the Institutional Review Boards or Ethical Committees at Soochow University in China and Tulane University in the United States (approval number 2007IRB1), as well as participating hospitals. All participants provided written informed consent.

NIHSS measurement and data collection

Baseline data, including demographic characteristics, clinical features, medical and medication history, and ischemic stroke subtype, were collected at the time of enrollment. Three BP measurements were collected at admission by trained nurses using a standard mercury sphygmomanometer according to a standard protocol adapted from procedures recommended by the American Heart Association. Plasma glucose and other clinical laboratory measurements were performed on fresh blood samples at each participating hospital at admission. Data collection was performed by trained interviewers through face-to-face interviews.

Stroke severity was assessed using the NIHSS score by trained neurologists.⁴ The total NIHSS score ranges from 0 to 42, with the higher score indicating a more severe neurological deficit. In this study, the NIHSS scores were evaluated at baseline, 14 days or hospital discharge, and at a 3-month post-treatment follow-up visit. Therefore, three NIHSS scores were used to fit neurological function trajectories.

Study outcomes

Participants were followed up in person at 24 months after stroke by trained neurologists (Figure 1). The primary endpoint was a composite of cardiovascular events during 3–24 months after ischemic stroke onset, including vascular deaths, nonfatal stroke, nonfatal myocardial infarction, hospitalization and treatment for angina, congestive heart failure, and peripheral arterial disease. The secondary outcomes included death from any cause (all-cause mortality) and recurrent fatal and nonfatal stroke, and the composite outcome consisted of all-cause mortality and cardiovascular events from 3 to 24 months after stroke onset. Death certificates were obtained for deceased participants, and hospital data were abstracted for participants who experienced study outcomes. The causes and date of death were verified by examining hospital medical records. Recurrent stroke was defined as a new neurological deficit or a deterioration of the previous deficit that lasted longer than 24 h and fitted the definitions for ischemic or hemorrhagic stroke. The study outcome assessment committee, blinded to treatment assignment, reviewed and adjudicated subsequent outcomes based on the established criteria.

Figure 1.

Timeline of NIHSS measurements and follow-up assessment.

Statistical analysis

We used a group-based trajectory modeling approach to identify subgroups with similar longitudinal stroke severity trajectories, as measured by the NIHSS at admission, 14 days or hospital discharge, and 3 months (Figure 1). This method fits longitudinal data as a discrete mixture of two or more latent trajectories via maximum likelihood using SAS Proc Traj procedure in SAS software (version 9.4, SAS Institute, Cary, NC, USA).²² The optimal number of groups and the shapes of trajectories were selected for best fit to the data using a two-stage approach, as assessed by change in the Bayesian Information Criterion (BIC).²³ The first stage was to determine the number of trajectories using the cubic form for all trajectory groups. We initiated a model with one group and compared the BIC to that with two, three, four, and five groups, respectively. The minimum sample size of each trajectory was specified to be more than 5% of the total population. Finally, we had identified that the model with three groups fit best. Once the optimal number of groups was identified, the shape of each trajectory was determined by fitting each group with a cubic degree polynomial and reduced the polynomial orders until the highest order polynomial for each group remained significant. We named these three trajectories based on the visual patterns of dynamic change of NIHSS scores: persistent severe trajectory, moderate trajectory, and mild trajectory (Figure 2). We calculated the posterior predicted probability for each individual of being a member of each of the three trajectories. Participants were assigned into the trajectory group to which they hold the highest posterior membership probability. The average posterior probability for each trajectory group was tested to assess the adequacy of our trajectory model. Moreover, we characterized a dynamic metric of ischemic stroke instability, defined by change in NIHSS scores from baseline to 14 days or discharge (NIHSS_baseline–NIHSS_discharge = ΔNIHSS), to examine its relevance to long-term outcomes. Neurological improvement was defined as ΔNIHSS ⩾4.^9,14,24

Figure 2.

Trajectories of neurological function within the first 3 months of ischemic stroke.

Baseline clinical characteristics of study participants were summarized and compared according to the trajectory groups using χ² test for categorical variables and analysis of variance (ANOVA) or Kruskal–Wallis test for continuous variables, as appropriate. The cumulative risks of three trajectory groups for study outcomes during 3–24 months were estimated with Kaplan–Meier curves and compared using the log-rank test. Cox proportional hazard models were used to investigate the associations of NIHSS trajectories or neurological improvement with risk of cardiovascular events, recurrent stroke, all-cause mortality, and the composite outcome. The proportional hazards assumption of the Cox models was tested using Schoenfeld residuals, which showed no significant departure from proportionality (p > 0.05). Hazard ratios (HRs) and 95% confidence intervals (CIs) for persistent severe trajectory and moderate trajectory compared to mild trajectory were calculated after being adjusted for age, sex, systolic BP at admission, current cigarette smoking and alcohol drinking, time from onset to hospitalization, history of hypertension, hyperlipidemia, coronary heart disease and diabetes, fasting plasma glucose, total cholesterol, family history of stroke, ischemic stroke subtype, and use of antihypertensive medication.

In the sensitivity analyses, to test the robustness of our findings, we further adjusted for baseline NIHSS score, or systolic and diastolic BP at 3 months, or use of antihypertensive medication at 3 months after stroke onset in the multivariable model. All statistical analyses were performed with SAS. Hypothesis testing was two-sided and conducted at the 5% level of significance.

Results

NIHSS trajectories and baseline characteristics

During the study period, a total of 4025 patients (2579 men and 1446 women) with the mean age (SD) at the time of enrollment of 61.9 (10.9) years were included in the analysis. The median baseline NIHSS score was 4 (interquartile range, 2–7). Through group-based trajectory modeling, the study participants were classified into three groups based on their NIHSS trajectories (Figure 2): group 1 (mild trajectory) included 866 (22.9%) participants with NIHSS scores below 2 during the first 3 months after stroke onset; group 2 (moderate trajectory) included 2541 (67.2%) participants with NIHSS scores started at around 5 and gradually reduced to 3 during hospitalization and then maintained this level afterward; group 3 (persistent severe trajectory) included 374 (9.9%) participants with persistent higher NIHSS scores during the 3-month follow-up. The average posterior probability for each trajectory group was 0.93, 0.97, and 0.83 for persistent severe, moderate, and mild group, respectively. Using 0.70 as the recommended criteria, our model demonstrated good discrimination in classifying individuals into distinctive trajectory groups. In all, 753 patients (18.7%) experienced neurological improvement during hospitalization. The median ΔNIHSS was 1 (interquartile range, 0–3).

Participants in three NIHSS trajectory groups had distinct clinical profiles (Table 1). Patients with mild trajectory were more likely to be young and drink and had lower admission NIHSS score, but had higher total cholesterol and a higher proportion of lacunar stroke, whereas patients with persistent severe trajectory were more likely to be older and had higher systolic BP, fasting plasma glucose, and higher proportions of coronary heart disease history and embolic stroke.

Table 1.

Characteristics of the study participants at baseline by neurological function trajectories.

Characteristics*	Neurological function trajectories			p value
Characteristics*	Mild	Moderate	Persistent severe	p value
No. of patients	866 (22.9)	2541 (67.2)	374 (9.9)
Age, y	59.0 ± 10.8	62.1 ± 10.7	65.0 ± 10.6	<0.001
Male sex	541 (62.5)	1654 (65.1)	228 (61.0)	0.16
Current cigarette smoking	342 (39.5)	913 (35.9)	144 (38.5)	0.14
Current alcohol consumption	314 (36.3)	759 (29.9)	112 (29.9)	0.002
Clinical features
SBP, mmHg	165.7 ± 17.2	165.6 ± 16.6	168.6 ± 17.3	0.01
DBP, mmHg	97.1 ± 11.3	96.5 ± 11.0	97.1 ± 11.2	0.48
FPG, mmol/L	6.4 ± 2.6	6.7 ± 2.8	7.1 ± 2.7	<0.001
Triglyceride, mmol/L	5.1 ± 1.1	5.1 ± 1.2	5.2 ± 1.2	0.18
Total cholesterol, mmol/L	2.1 ± 5.0	1.8 ± 1.3	1.7 ± 1.3	<0.001
LDL cholesterol, mmol/L	2.9 ± 0.9	2.9 ± 0.9	3.1 ± 1.1	0.06
HDL cholesterol, mmol/L	1.3 ± 0.6	1.3 ± 0.4	1.3 ± 0.5	0.06
Body mass index, kg/m²	25.1 ± 3.0	24.9 ± 3.1	24.9 ± 3.2	0.25
Admission NIHSS score	2 (1–3)	5 (3–7)	13 (11–17)	<0.001
Time from onset to hospital, h	12 (5–24)	12 (5–24)	7 (3–20)	<0.001
Medical history
Hypertension	675 (77.9)	2010 (79.1)	289 (77.3)	0.61
Hyperlipidemia	62 (7.2)	176 (6.9)	23 (6.1)	0.81
Diabetes mellitus	142 (16.4)	447 (17.6)	74 (19.8)	0.35
Coronary heart disease	78 (9.0)	264 (10.4)	59 (15.8)	0.002
Family history of stroke	188 (21.7)	459 (18.1)	66 (17.6)	0.05
Prior use of medications
Antihypertensive	432 (49.9)	1246 (49.0)	157 (42.0)	0.03
Lipid-lowering	30 (3.5)	85 (3.3)	9 (2.4)	0.60
Ischemic stroke subtype
Thrombotic	659 (76.1)	1986 (78.2)	294 (78.6)	0.41
Embolic	17 (2.0)	112 (4.4)	43 (11.5)	<0.001
Lacunar	215 (24.8)	506 (19.9)	47 (12.6)	<0.001

SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucose; LDL, low-density lipoprotein; HDL, high-density lipoprotein; NIHSS, National Institute of Health Stroke Scale.

Continuous variables are expressed as mean ± SD or as median (interquartile range). Categorical variables are expressed as frequency (percentage).

NIHSS trajectories and long-term outcomes

During 12 months of follow-up, 119 (3.1%) cardiovascular events and 108 (2.9%) recurrent strokes were identified, and 94 (2.5%) participants died (Table 2). In all, 135 participants (3.4%) were lost to follow-up. Compared to the mild trajectory, the multivariable-adjusted HRs (95% CIs) for patients in moderate trajectory were 1.80 (1.02–3.19) for recurrent stroke and 1.62 (1.01–2.59) for the composite outcome of cardiovascular events and all-cause mortality; the multivariable-adjusted HRs (95% CIs) for those in the persistent severe group were 2.06 (1.05–4.06) for cardiovascular events, 2.35 (1.12–4.93) for recurrent stroke, 6.86 (3.24–14.53) for all-cause mortality, and 3.87 (2.26–6.63) for the composite outcome. In addition, neurological improvement (ΔNIHSS ⩾4) was significantly associated with decreased risks of cardiovascular events (HR = 0.23, 95% CI = 0.15–0.35), recurrent stroke (HR = 0.44, 95% CI = 0.23–0.84), all-cause mortality (HR = 0.16, 95% CI = 0.11–0.24), and the composite outcome (HR = 0.22, 95% CI = 0.16–0.31).

Table 2.

Hazard ratios of neurological function trajectories for clinical outcomes at 12 months after ischemic stroke.

	Mild trajectory(n = 866)	Moderate trajectory(n = 2541)	Persistent severe trajectory(n = 374)	p trend	Neurological improvement(ΔNIHSS⩾4)
Cardiovascular events
No. (%) of cases	18 (2.1)	84 (3.2)	17 (4.5)
Unadjusted model	1.00 (Ref.)	1.59 (0.96–2.65)	2.28 (1.18–4.42)	0.01	0.21 (0.14–0.32)
Adjusted model^a	1.00 (Ref.)	1.48 (0.89–2.48)	2.06 (1.05–4.06)	0.03	0.23 (0.15–0.35)
Recurrent stroke
No. (%) of cases	14 (1.6)	79 (3.1)	15 (4.0)
Unadjusted model	1.00 (Ref.)	1.93 (1.09–3.40)	2.59 (1.25–5.37)	0.01	0.42 (0.23–0.79)
Adjusted model^a	1.00 (Ref.)	1.80 (1.02–3.19)	2.35 (1.12–4.93)	0.02	0.44 (0.23–0.84)
All-cause mortality
No. (%) of cases	9 (1.0)	50 (2.0)	35 (9.4)
Unadjusted model	1.00 (Ref.)	1.89 (0.93–3.85)	9.35 (4.49–19.45)	<0.001	0.13 (0.09–0.19)
Adjusted model^a	1.00 (Ref.)	1.59 (0.78–3.25)	6.86 (3.24–14.53)	<0.001	0.16 (0.11–0.24)
Cardiovascular events or all-cause mortality
No. (%) of cases	21 (2.4)	112 (4.4)	41 (11.0)
Unadjusted model	1.00 (Ref.)	1.82 (1.14–2.90)	4.71 (2.78–7.96)	<0.001	0.19 (0.14–0.27)
Adjusted model^a	1.00 (Ref.)	1.62 (1.01–2.59)	3.87 (2.26–6.63)	<0.001	0.22 (0.16–0.31)

HR: hazard ratio; CI: confidence interval; NIHSS: National Institutes of Health Stroke Scale.

Data are HRs (95% CIs). ΔNIHSS = NIHSS_baseline–NIHSS_discharge.

Adjusted for age, sex, current smoking, alcohol consumption, time from onset to randomization, systolic blood pressure, medical history (hypertension, hyperlipidemia, diabetes mellitus, coronary heart disease), use of antihypertensive medication, fasting plasma glucose, total cholesterol, family history of stroke, and ischemic stroke subtype.

During a maximum 24-month follow-up period, 257 (6.8%) incident cardiovascular events, 232 (6.1%) recurrent strokes, and 186 (4.9%) deaths were identified (Table 3). In all, 212 participants (5.3%) were lost to follow-up. Kaplan–Meier curves showed that participants in the severe group had significantly higher cumulative incidence rates of all study outcomes (all log-rank p value <0.005) (Figure 3). Compared with mild group, the multivariable-adjusted HRs (95% CIs) of patients in the moderate group were 1.45 (1.03–2.04) for cardiovascular events, 1.52 (1.06–2.19) for recurrent stroke, and 1.46 (1.07–1.98) for the composite outcome of cardiovascular events or deaths, and the adjusted HRs (95% CIs) for the persistent severe group were 1.77 (1.10–2.86) for cardiovascular events, 1.82 (1.10–3.00) for recurrent stroke, 5.64 (3.37–9.43) for all-cause mortality, and 3.11 (2.15–4.49) for the composite outcome. Neurological improvement was also associated with decreased risks of cardiovascular events (HR = 0.34, 95% CI = 0.23–0.50), all-cause mortality (HR = 0.22, 95% CI = 0.15–0.31), and the composite outcome (HR = 0.31, 95% CI = 0.23–0.42) after adjusted for age, sex, admission systolic BP, medical history, and other potential confounders.

Table 3.

Hazard ratios of neurological function trajectories for clinical outcomes at 24 months after ischemic stroke.

	Mild trajectory(n = 866)	Moderate trajectory(n = 2541)	Persistent severe trajectory(n = 374)	p trend	Neurological improvement(ΔNIHSS ⩾4)
Cardiovascular events
No. (%) of cases	41 (4.7)	185 (7.3)	31 (8.3)
Unadjusted model	1.00 (Ref.)	1.54 (1.10–2.17)	1.91 (1.20–3.04)	0.003	0.32 (0.22–0.48)
Adjusted model^a	1.00 (Ref.)	1.45 (1.03–2.04)	1.77 (1.10–2.86)	0.01	0.34 (0.23–0.50)
Recurrent stroke
No. (%) of cases	36 (4.2)	168 (6.6)	28 (7.5)
Unadjusted model	1.00 (Ref.)	1.62 (1.13–2.32)	1.97 (1.20–3.22)	0.003	0.58 (0.33–1.04)
Adjusted model^a	1.00 (Ref.)	1.52 (1.06–2.19)	1.82 (1.10–3.00)	0.01	0.62 (0.34–1.10)
All-cause mortality
No. (%) of cases	20 (2.3)	103 (4.1)	63 (16.8)
Unadjusted model	1.00 (Ref.)	1.72 (1.06–2.77)	7.94 (4.80–13.13)	<0.001	0.18 (0.13–0.25)
Adjusted model^a	1.00 (Ref.)	1.43 (0.88–2.32)	5.64 (3.37–9.43)	<0.001	0.22 (0.15–0.31)
Cardiovascular events or all-cause mortality
No. (%) of cases	50 (5.8)	240 (9.5)	74 (19.8)
Unadjusted model	1.00 (Ref.)	1.63 (1.20–2.22)	3.74 (2.61–5.35)	<0.001	0.27 (0.20–0.38)
Adjusted model^a	1.00 (Ref.)	1.46 (1.07–1.98)	3.11 (2.15–4.49)	<0.001	0.31 (0.23–0.42)

HR: hazard ratio; CI: confidence interval; NIHSS: National Institutes of Health Stroke Scale.

Data are HRs (95% CIs). ΔNIHSS = NIHSS_baseline–NIHSS_discharge.

Adjusted for age, sex, current smoking, alcohol consumption, time from onset to randomization, systolic blood pressure, medical history, use of antihypertensive medication, fasting plasma glucose, total cholesterol, family history of stroke, and ischemic stroke subtype.

Figure 3.

Cumulative risks of clinical outcomes by neurological function trajectories. Panel (a): cardiovascular events; panel (b): recurrent stroke; panel (c): all-cause mortality; panel (d): cardiovascular events or all-cause mortality.

Sensitivity analyses

In the sensitivity analyses, the significant results remained after further adjustment for baseline NIHSS score, 3-month systolic and diastolic BP, or use of antihypertensive medication at 3 months after stroke onset based on multivariable models (Supplemental Table S1). The competing risk regression analyses showed NIHSS trajectories were also associated with increased risk of 24-month cardiovascular events and recurrent stroke (Supplemental Table S2).

Discussion

In this large-scale cohort study of patients with ischemic stroke, we first observed heterogeneous NIHSS trajectory patterns during the first 3 months after stroke onset and identified three NIHSS trajectories, in which participants shared similar patterns of neurological function changes. The three trajectories were characterized by stable low NIHSS scores, gradually decreasing and maintaining medium NIHSS scores, and persistent high NIHSS scores, respectively. Patients with persistent severe neurological deficits had the highest risk of cardiovascular events, recurrent stroke, and all-cause mortality at both 12- and 24-month follow-up. In addition, patients with a gradual decline but maintaining moderate neurological deficits still had a higher 24-month risk of cardiovascular events and recurrent stroke compared to patients with stable low neurological deficits. Furthermore, neurological improvement during hospitalization was associated with a reduced risk of cardiovascular events and all-cause mortality at 24 months. To our knowledge, this is the first study to explore the relationships of early longitudinal changes of neurological function with long-term outcomes in patients with acute ischemic stroke.

Neurological deficit severity is a main determinant of clinical outcomes after stroke. Due to the substantial changes in neurological function in the acute phase of ischemic stroke, early neurological deterioration is also widely known as a serious event for the prognosis of stroke. For example, a report of the Trial of Org 10172 in Acute Stroke Treatment (TOAST) showed that the baseline NIHSS scores strongly predicted the likelihood of stroke recovery, with one additional point on the NIHSS decreasing the likelihood of excellent outcomes at 7 days by 24% and at 3 months by 17%.¹³ In a study of 2555 patients with acute ischemic stroke, a dynamic change of neurological deficit in the first 24 hours after stroke onset was observed, and the changes in stroke severity, defined by change in ΔNIHSS_6–24h, were independently associated with 90-day favorable outcome.¹⁸ Our findings confirmed these previous studies and further extended the predictive significance of neurological changes to long-term stroke outcomes. On the other hand, the severity of neurological deficits often varies during the initial period after onset, the prognostic value of the NIHSS score on final functional outcome improved with increasing time to stroke onset, whereas the impact of very early NIHSS measurements on final outcome was attenuated. Furthermore, the dynamic changes in neurological function are also time-dependent, with a higher likelihood of change earlier in acute phase of stroke. Therefore, single or two assessments of neurological function may be limited for the prediction of unfavorable stroke outcomes.

In the secondary analysis of ESCAPE trial (Endovascular Treatment for Small Core and Anterior Circulation Proximal Occlusion With Emphasis on Minimizing CT to Recanalization Times), based on the repeated assessments of neurological deficit, the authors identified three distinct subgroups of potential trajectories and found that early trajectory of neurological improvement within the first 48 h predicted 90-day functional outcome.²⁰ However, the predictive effects of NIHSS trajectories in the subacute phase on the long-term clinical outcomes of ischemic stroke remain to be clarified. This study expanded this information; we found that there were three different patterns of neurological fluctuations within 90 days after the onset of stroke, and discrete NIHSS trajectories were significantly associated with the risk of long-term adverse consequences of ischemic stroke. Our results indicated that patients with persistent severe trajectory had the highest risk of poor outcomes, including cardiovascular events, recurrent stroke, and all-cause mortality at both 12 and 24 months after stroke; it is worth noting that patients who had a gradual decline in neurological deficit but remained at moderate levels are still at increased risk of cardiovascular events and recurrent stroke.

The neurological function changes, reflecting the secondary neurological injury or neurological improvement, were common in the acute and subacute stages of ischemic stroke.^25,26 Many potential mechanisms may contribute to neurological instability. Perfusion abnormalities were suggested to be a main explanation for early neurological injury after ischemic stroke.²⁷ Decreased cerebral perfusion pressure, proximal occlusions, and inadequate collaterals may affect the hemodynamics, inability to achieve sustained vessel patency, and causing infarct growth and cerebral edema.²⁸ Excitotoxicity, hyperthermia, hyperglycemia, and other metabolic factors were associated with secondary neurological deterioration.²⁹ In addition, neurological and systemic complications, such as hemorrhagic transformation and infections, were more likely to be associated with late deterioration.²⁷

The advantage of our trajectory approach can assign participants sharing similar NIHSS trajectories, which reflects the longitudinal changes of neurological function. This approach simultaneously estimates the average, variability, and the direction of variability to investigate the heterogeneity in NIHSS trajectories and may provide additional prognostic information. Furthermore, the collection of stroke severity, using the well-validated NIHSS score, is routine during hospitalization, as well as for post-discharge follow-up assessment. Understanding the longitudinal changes of neurological function could advance prognostication and identify high-risk patients of unfavorable long-term outcomes. Based on our findings of distinct prognostic significance of neurological function trajectories, this study highlights the importance of neurological monitoring and management during hospitalization and after discharge, and has great clinical implications.

However, some limitations of our study are worth mentioning. First, this study was a post hoc analysis of CATIS trial, and patients with BP ⩾ 220/120 mmHg or treated with intravenous thrombolytic therapy were excluded. Therefore, a selection bias may unavoidably be present. Second, data were collected before the introduction of endovascular clot retrieval; this may limit the external validity of results. Third, there were only three timepoints in which NIHSS data were collected, and participants without available NIHSS measurements at the three timepoints were excluded, whereas more NIHSS data can more accurately assess NIHSS trajectories. Fourth, due to insufficient statistical power, it may be underpowered to detect a statistically significant difference in outcomes between mild and moderate trajectories. In addition, patients included in this trial had a lower median NIHSS score of 4 (interquartile range, 2–8), compared with 7 (interquartile range, 2–10) in Chinese national registry data; this may also limit the generalizability of our findings to all acute ischemic stroke patients.³⁰ Last, our patients were from China, and the findings should be extrapolated cautiously to other populations with different social and culture backgrounds.

In conclusion, we found that neurological function trajectories within the first 3 months of ischemic stroke onset were associated with risk of long-term adverse outcomes. The trajectories characterized by persistent severe and moderate neurological impairment were associated with increased risk of adverse clinical outcomes. Our findings indicated that monitoring trajectories of neurological function may provide an important approach to identify stroke patients at higher risks of long-term clinical outcomes.

Supplemental Material

sj-pdf-1-wso-10.1177_17474930231180446 – Supplemental material for The relationship between neurological function trajectory, assessed by repeated NIHSS measurement, and long-term cardiovascular events, recurrent stroke, and mortality after ischemic stroke

Supplemental material, sj-pdf-1-wso-10.1177_17474930231180446 for The relationship between neurological function trajectory, assessed by repeated NIHSS measurement, and long-term cardiovascular events, recurrent stroke, and mortality after ischemic stroke by Jigang Du, Yan Wang, Bizhong Che, Mengyuan Miao, Anran Bao, Yanbo Peng, Zhong Ju, Tan Xu, Jiang He, Yonghong Zhang and Chongke Zhong in International Journal of Stroke

Footnotes

Availability of data and materials

Data are available from the corresponding author on reasonable request.

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: This study was supported by the National Natural Science Foundation of China (grant number 82273706).

ORCID iDs

Jigang Du

Bizhong Che

Chongke Zhong

Supplemental material

Supplemental material for this article is available online.

References

Feigin

Nguyen

Cercy

, et al. Global, regional, and country-specific lifetime risks of stroke, 1990 and 2016. N Engl J Med 2018; 379: 2429–2437.

Golomb

Vickrey

Hays

. A review of health-related quality-of-life measures in stroke. Pharmacoeconomics 2001; 19: 155–185.

Rost

Bottle

Lee

, et al. Stroke severity is a crucial predictor of outcome: an international prospective validation study. J Am Heart Assoc 2016; 5: e002433.

Brott

Adams

Jr Olinger

, et al. Measurements of acute cerebral infarction: a clinical examination scale. Stroke 1989; 20: 864–870.

Meyer

Hemmen

Jackson

Lyden

. Modified National Institutes of Health Stroke Scale for use in stroke clinical trials: prospective reliability and validity. Stroke 2002; 33: 1261–1266.

Kasner

. Clinical interpretation and use of stroke scales. Lancet Neurol 2006; 5: 603–612.

Hilz

Moeller

Akhundova

Marthol

Pauli

De Fina

Schwab

. High NIHSS values predict impairment of cardiovascular autonomic control. Stroke 2011; 42: 1528–1533.

Goldstein

Samsa

. Reliability of the National Institutes of Health Stroke Scale. Extension to non-neurologists in the context of a clinical trial. Stroke 1997; 28: 307–310.

Irvine

Battey

Ostwaldt

, et al. Early neurological stability predicts adverse outcome after acute ischemic stroke. Int J Stroke 2016; 11: 882–889.

10.

Saver

Altman

. Relationship between neurologic deficit severity and final functional outcome shifts and strengthens during first hours after onset. Stroke 2012; 43: 1537–1541.

11.

Wouters

Nysten

Thijs

Lemmens

. Prediction of outcome in patients with acute ischemic stroke based on initial severity and improvement in the first 24 h. Front Neurol 2018; 9: 308.

12.

Wityk

Pessin

Kaplan

Caplan

. Serial assessment of acute stroke using the NIH Stroke Scale. Stroke 1994; 25: 362–365.

13.

Adams

Jr Davis

Leira

, et al. Baseline NIH Stroke Scale score strongly predicts outcome after stroke: a report of the Trial of Org 10172 in acute stroke treatment (TOAST). Neurology 1999; 53: 126–131.

14.

Arenillas

Rovira

Molina

Grivé

Montaner

Alvarez-Sabín

. Prediction of early neurological deterioration using diffusion- and perfusion-weighted imaging in hyperacute middle cerebral artery ischemic stroke. Stroke 2002; 33: 2197–2203.

15.

Rödén-Jüllig

. Progressing stroke: epidemiology. Cerebrovasc Dis 1997; 7: 2–5.

16.

Toni

Fiorelli

Gentile

, et al. Progressing neurological deficit secondary to acute ischemic stroke. A study on predictability, pathogenesis, and prognosis. Arch Neurol 1995; 52: 670–675.

17.

National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. Tissue plasminogen activator for acute ischemic stroke. N Engl J Med 1995; 333: 1581–1587.

18.

Heitsch

Ibanez

Carrera

, et al. Early neurological change after ischemic stroke is associated with 90-day outcome. Stroke 2021; 52: 132–141.

19.

Mistry

Yeatts

de Havenon

, et al. Predicting 90-day outcome after thrombectomy: baseline-adjusted 24-hour NIHSS is more powerful than NIHSS score change. Stroke 2021; 52: 2547–2553.

20.

Sajobi

Menon

Wang

, et al. Early trajectory of stroke severity predicts long-term functional outcomes in ischemic stroke subjects: results from the ESCAPE trial (endovascular treatment for small core and anterior circulation proximal occlusion with emphasis on minimizing CT to recanalization times). Stroke 2017; 48: 105–110.

21.

Zhang

, et al. Effects of immediate blood pressure reduction on death and major disability in patients with acute ischemic stroke: the CATIS randomized clinical trial. JAMA 2014; 311: 479–489.

22.

Jones

Nagin

. Advances in group-based trajectory modeling and an SAS procedure for estimating them. Sociol Method Res 2007; 35: 542–571.

23.

Jones

Nagin

Roeder

. A SAS procedure based on mixture models for estimating developmental trajectories. Sociol Method Res 2001; 29: 374–393.

24.

Mori

Naganuma

Okada

, et al. Early neurological deterioration within 24 hours after intravenous rt-PA therapy for stroke patients: the stroke acute management with urgent risk factor assessment and improvement rt-PA registry. Cerebrovasc Dis 2012; 34: 140–146.

25.

Yeo

Paliwal

Teoh

, et al. Timing of recanalization after intravenous thrombolysis and functional outcomes after acute ischemic stroke. JAMA Neurol 2013; 70: 353–358.

26.

Seners

Turc

Oppenheim

Baron

. Incidence, causes and predictors of neurological deterioration occurring within 24 h following acute ischaemic stroke: a systematic review with pathophysiological implications. J Neurol Neurosurg Psychiatry 2015; 86: 87–94.

27.

Alawneh

Moustafa

Baron

. Hemodynamic factors and perfusion abnormalities in early neurological deterioration. Stroke 2009; 40: e443–e450.

28.

Zhang

Liu

. Cerebral hemodynamic evaluation after cerebral recanalization therapy for acute ischemic stroke. Front Neurol 2019; 10: 719.

29.

Parsons

Barber

Desmond

, et al. Acute hyperglycemia adversely affects stroke outcome: a magnetic resonance imaging and spectroscopy study. Ann Neurol 2002; 52: 20–28.

30.

Wang

Zhao

, et al. In-hospital medical complications associated with patient dependency after acute ischemic stroke: data from the China National Stroke Registry. Chin Med J 2013; 126: 1236–1241.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.17 MB