Abstract
Acute ischemic stroke (AIS) has a high rate of death and causes long-term disability, leading to a global economic burden annually. Therefore, discovering biomarkers to improve AIS patient prognosis is critical. Previous studies reported an association between serum cystatin C (CysC) levels and outcomes in AIS patients, but the results remain controversial. This systematic review and meta-analysis aimed to explore the relationship between serum CysC and AIS patient outcomes using currently available studies. The literature search included PubMed, Embase, Web of Science, Cochrane Library, China National Knowledge Infrastructure (CNKI), VIP, and Wan Fang databases. Outcomes included poor functional recovery, cognitive dysfunction and death. Weighted mean difference (WMD) with 95% confidence interval (95% CI) was used as an effect index for measurement data. Results demonstrated that serum CysC was significantly higher in AIS patients with poor functional recovery (WMD = 0.18, 95% CI: 0.08–0.28), cognitive dysfunction (WMD = 0.16, 95% CI: 0.09–0.23) and death (WMD = 0.32, 95% CI: 0.02–0.62) than in the control groups when follow-up time was <1 month. These findings show that high serum CysC levels were associated with poor AIS patient outcomes. Further studies are needed to examine whether reducing serum CysC can prevent poor outcomes in AIS patients.
Key words: risk factors, biomarkers, cognitive dysfunction, cystatin C, ischemic stroke
Introduction
Stroke has one of the highest global morbidity rates, with ischemic stroke accounting for approx. 87% of such events.1 Acute ischemic stroke (AIS) refers to disorders of the cerebral circulation which cause irreversible damage to local brain tissues, leading to brain tissue ischemia and hypoxic necrosis in the acute stage.2 Acute ischemic stroke has a high rate of death and causes long-term disability, resulting in a considerable annual economic burden worldwide.3
Acute ischemic stroke is caused by focal cerebral hypoperfusion, mainly from atherosclerotic diseases.2 During a transient ischemic attack, arterial flow to brain tissue is temporarily interrupted, leading to focal neurologic symptoms (such as hemiparesis), but spontaneous flow recovery can alleviate symptoms without permanent tissue damage.2 However, if arterial flow does not recover promptly, ischemia may develop into irreversible infarction, and patients may experience an AIS.2
Several neural substrates are thought to be involved in AIS, such as kynurenine, methylxanthine and N-methyl-D-aspartate,4, 5, 6 and have been shown to be associated with cognitive functions, motor skills and emotions.7, 8, 9, 10, 11, 12, 13 Post-stroke cognitive dysfunction is a common symptom affecting 20–80% of patients and impacting neurological function recovery.14 Depression is a common emotional syndrome after a stroke that substantially reduces patients’ quality of life and occurs in approx. 30% of cases.15 Quantitative analyses of electroencephalogram data have identified the psychophysiological correlates of cognition. However, current international guidelines do not endorse the use of electroencephalogram biomarkers in neuropsychiatric disorders and depression.16, 17 To improve the prognosis of AIS patients, it is necessary to identify other methods, such as simple and economical biochemical detection indicators, that may have additional clinical application value.
Serum cystatin C (CysC) is a member of the endogenous cysteine protease inhibitor family that can inhibit the activity of endogenous protein cysteine18; it is released from all nucleated cells at a constant rate, and can freely pass through the glomerular filtration membrane due to its low molecular weight.18, 19 Previous studies demonstrated that CysC was an effective biomarker for assessing kidney function.19 Kidney disease is a crucial risk factor for cerebrovascular diseases, and the proposed underlying mechanisms for this association include cerebral hypoperfusion.20 Recent studies found an association between CysC and cardiovascular diseases and showed its involvement in the pathophysiology of atherosclerosis in AIS patients.21, 22
Serum CysC levels were significantly elevated in patients with AIS.23 Zeng et al. found it to be an independent prediction biomarker for ischemic stroke,24 and a meta-analysis by Wang et al. reported an association between high serum CysC levels and increased ischemic stroke risk.25 However, the correlation between CysC and AIS patient outcomes also deserves attention, though the specific mechanisms by which CysC affects AIS patient outcomes remain unclear.
Cystatin C is an endogenous cathepsin inhibitor that plays a central role in regulating vascular wall proteases and antiproteases,26 and unbalanced expression of CysC and cysteine cathepsin may lead to atherosclerosis.26 In addition, increased serum CysC levels may indicate renal dysfunction in AIS patients,27 and those with chronic renal failure generally experience oxidative stress and systemic inflammation, which accelerates the progression of atherosclerosis.27
Cystatin C participates in inflammatory reactions and injures nervous system cells, creating a vicious cycle.28 Existing studies show that CysC plays a role in neuronal damage and dysfunction, and may be related to clinical manifestations of cognitive impairment in neurodegenerative disorders.29, 30 Meanwhile, other studies found that CysC can inhibit cerebral amyloid protein aggregation and potentially prevent cognitive dysfunction in neurodegenerative disorders.31, 32 Previous studies reported an association between CysC and AIS patient outcomes, but the results are inconsistent.29, 33, 34, 35 Zeng et al. found that higher serum CysC levels were associated with a higher risk of cognitive dysfunction,36 whereas Guo et al. reported a protective effect of increased CysC levels in cognitive dysfunction.33 The reason for this discrepancy may be the difference in serum CysC levels among the populations included in their studies (1.07 compared to 0.77).33, 36
Liu et al. found that a higher CysC level was related to an increased risk of recurrent stroke in AIS patients.34 Dong et al. also found that the risk of recurrence was elevated with increasing serum CysC in AIS patients.37 However, Zhu et al. found no association between CysC level and recurrence in AIS patients.35 Such a discrepancy may be due to the difference in follow-up time of their studies (1 year compared to 2 years).34, 35, 37
A meta-analysis is a powerful tool that combines the results of 2 or more separate studies to demonstrate good evidence strength and contribute to healthcare decision-making.38, 39 Considering the controversial results in previously reported studies, we aimed to conduct a systematic review and meta-analysis to comprehensively explore the association between CysC level and AIS patient outcomes to guide unfavorable outcome management in AIS patients.
Objectives
The study analyzed the association between CysC levels and outcomes in AIS patients.
Methods
The meta-analysis followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses Statement (PRISMA) guidelines.40
Literature search
The literature search included 4 English-language (PubMed, Web of Science, Cochrane Library, and Embase) and 3 Chinese databases (VIP, China National Knowledge Infrastructure (CNKI) and Wan Fang) from inception to August 21, 2023. Supplementary File 1 shows the search terms. The literature search was conducted by 2 independent researchers (CGH and SBC), and the 3rd person (Guosen Bu) provided consultation if conflicts arose.
Study selection
The inclusion criteria were 1) patients: patients with AIS, 2) reporting: serum CysC levels (at baseline); 3) outcomes: functional recovery, cognitive dysfunction, death, hemorrhagic transformation, vascular events, depression, and recurrence; 4) cohort studies or case-control studies; and 5) language: published in Chinese or English. Studies meeting 1 of the following criteria were excluded: 1) duplicated publication; 2) animal studies; 3) incomplete or inaccessible data, 4) not matching the topic; and 5) reviews or meta-analyses, conference abstracts, case reports, and letters.
Functional recovery was evaluated using the modified Rankin Scale (mRS)41 and the National Institute of Health Stroke Scale (NIHSS).42 According to the mRS, patients were divided into good recovery (mRS ≤2 points) and poor recovery (mRS >2 points) groups.41 Based on the NIHSS, patients were divided into basic recovery (91–100% decrease in NIHSS score), significant improvement (46–90% decrease in NIHSS score), improvement (18–45% decrease in NIHSS score), and ineffective/deterioration (≤17% decrease in NIHSS score) groups.42
The Montreal Cognitive Assessment (MoCA) evaluated cognitive function from a total score of 30 points, with <26 points defined as cognitive dysfunction.29
Data extraction and quality assessment
Two researchers (CGH and SBC) independently conducted data extraction, including the study authors, publication year, country, study design, patients, treatments, groups, sample size, gender, age, body mass index (BMI), comorbidities (hypertension, diabetes mellitus and hyperlipidemia), smoking, CysC levels, and follow-up time.
The Newcastle–Ottawa Scale (NOS) was used to assess the quality of cohort studies and case-control studies43 on a 9-point scale, with 0–3 points indicating poor quality, 4–6 points fair quality and 7–9 points good quality.43
Statistical analyses
Weighted mean difference (WMD) acted as an effect index for measurement data, with effect size expressed as 95% confidence intervals (95% CIs). The choice between using a fixed-effect or random-effect meta-analysis was performed a priori, with a random-effect model selected to analyze the association between CysC levels and functional recovery assessed using mRS, functional recovery assessed using NIHSS, cognitive dysfunction, and death. Heterogeneity was assessed for each outcome using Cochrane I2 statistics, with I2 ranging 0–100%, and divided into <50% (low heterogeneity) and ≥50% (high heterogeneity). Subgroup analysis was performed based on follow-up time to explore the source of heterogeneity for outcomes (functional recovery assessed using mRS) with high heterogeneity. The robustness of the pooled results for each outcome was assessed using sensitivity analysis by excluding each study independently. Publication bias was assessed using Begg’s test when the outcome included 10 or more studies.44 Funnel plots were provided for each outcome (functional recovery assessed using mRS, functional recovery assessed using NIHSS, cognitive dysfunction, and death). All statistical analysis employed Stata v. 15.1 software (StataCorp, College Station, USA), with p < 0.05 regarded as statistically significant.
Results
Selection and characteristics of studies
There were 535 studies identified from English-language databases and 410 from Chinese databases, with 232 duplicates excluded and 653 excluded due to being reviews or meta-analyses (n = 38), conference abstracts (n = 47), case reports (n = 4), letters (n = 4), animal experiments (n = 34), articles not in English or Chinese (n = 2), or the topic not meeting the requirements (n = 524). After screening full texts, 36 studies were excluded for not having access to the full text (n = 2) and the topic not meeting the requirements (n = 34). Finally, 24 eligible studies were included (Figure 1). Table 1 describes the characteristics of these studies, of which 23 were cohort studies15, 26, 29, 33, 34, 42, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61 and 1 was a case-control study.36 A total of 7,567 AIS patients were assessed.
Risk of bias assessment
The NOS evaluated the risk of bias by evaluating 3 methodological domains, including selection, comparability and outcomes (for cohort studies)/exposure (for case-control studies) assessment. In this meta-analysis, 1 study was assessed as having poor quality, 17 fair quality and 6 good quality (Supplementary Table 1).
Meta-analysis of the association between CysC levels and AIS patient outcomes
Functional recovery assessed by mRS showed no significant difference in CysC levels between the poor and good functional recovery groups (WMD = 0.36, 95% CI: –0.18–0.91, I2 = 99.9 %) (Figure 2A). Follow-up time was identified as the source of heterogeneity (follow-up time <1 month: WMD = 0.18, 95% CI: 0.08–0.28; follow-up time ≥ 1 month: WMD = 0.43, 95% CI: –0.29–1.14). For functional recovery assessed with the NIHSS, compared to the basic recovery group, the CysC levels were higher in the significant improvement group (WMD = 0.15, 95% CI: 0.06–0.24), improvement group (WMD = 0.27, 95% CI: 0.16–0.37) and ineffective/deterioration group (WMD = 0.37, 95% CI: 0.27–0.47) (Figure 2B). Furthermore, CysC was higher in AIS patients with cognitive dysfunction (WMD = 0.16, 95% CI: 0.09–0.23) (Figure 2C). In addition, high CysC levels were found in the death group (WMD = 0.32, 95% CI: 0.02–0.62) (Figure 2D). The results of the meta-analysis are summarized in Table 2.
Systematic review of the association between CysC levels and AIS patient outcomes
Dong et al. divided the CysC level by quartile and found no statistical difference between the CysC and functional recovery in the 4 groups.47 Regarding cognitive dysfunction, Guo et al. reported no association between CysC and cognitive dysfunction,33 while Li et al. found that high CysC was a risk factor for cognitive dysfunction.50 We also observed hemorrhagic transformation, vascular events, depression, and recurrence, with 1 study showing that AIS patients with hemorrhagic transformation had higher CysC levels than those without hemorrhagic transformation (1.28 ±0.27 mg/L compared to 1.13 ±0.27 mg/L, p < 0.01).56 Xing et al. reported high CysC levels as an independent risk factor for depression in AIS patients.15 For stroke recurrence, Ding et al. found that CysC was higher in the recurrence group than in the non-recurrence group.46 Meanwhile, Liu et al. reported no association between CysC and recurrence but found a higher risk of vascular events in patients with increased CysC.34
Sensitivity analysis and publication bias
The influence of a single study on the effect estimate was assessed using sensitivity analysis by removing each study in turn. The results showed no significant influence of any individual study on the results of the meta-analysis. Begg’s test was used to evaluate the publication bias for functional recovery assessed with the mRS, showing no publication bias (Z = 0.18, p = 0.858) (Table 2). Also, funnel plots indicated no evidence of publication bias for functional recovery assessed using the mRS (Supplementary Fig. 1A), functional recovery assessed with the NIHSS (Supplementary Fig. 1B), cognitive dysfunction (Supplementary Fig. 1C), and death (Supplementary Fig. 1D).
Discussion
Acute ischemic stroke is a cause of long-term disability and has a high mortality rate, which leads to a significant annual economic burden globally.3 Previous studies reported controversial results on the association between CysC and AIS patient outcomes. Therefore, this systematic review and meta-analysis comprehensively explored the association between CysC and AIS outcomes based on currently available studies. Results showed that CysC level was higher in AIS patients with cognitive dysfunction, significant improvement of function recovery, improvement of function recovery and ineffective/deterioration of function recovery, and in patients who died. Although there was no significant difference between CysC levels and functional recovery assessed using mRS in the total population, subgroup analysis showed a significant difference during follow-up time <1 month. For the systematic review, we found that high CysC levels were associated with hemorrhagic transformation,56 vascular events,34 depression,15 and recurrence.46 The overall results suggest that high CysC levels may be one of the risk factors for poor outcomes in AIS patients.
Cystatin C has been reported to influence endogenous neuroprotection and may be a candidate drug for treating stroke through lysosomal membrane integrity maintenance.62 However, some studies reported CysC as a risk factor for stroke.25, 27 Compared to patients without stroke, CysC was higher in AIS patients, indicating that CysC may be an AIS risk prediction factor.25 Cognitive dysfunction is a common symptom after stroke, and the incidence reaches up to 56.6% in China within 3 months after stroke,63, 64 though current studies on the association between CysC and cognitive dysfunction remain inconsistent.29, 33, 36, 65 Nonetheless, our meta-analysis found higher CysC in AIS patients with cognitive dysfunction. Some pathways may explain the mechanisms by which CysC affects cognitive function after AIS. The stroke event may elevate serum CysC level, and a high CysC level is associated with dementia and cognitive dysfunction.29, 36 Furthermore, there is an association between CysC and symptomatic common carotid artery stenosis, which is related to cognitive impairment in patients with or without stroke.21, 66, 67, 68 Moreover, CysC is involved in the pathogenesis of cerebral amyloidosis, which may result in early cerebral hemorrhage.69 Previous research indicated higher serum CysC in AIS patients complicated with cerebral microbleeds than in those without, highlighting CysC as a risk factor for cerebral microbleeds.70 Cerebral microbleeds are closely associated with cognitive dysfunction, resulting in damage to memory, abstract thinking and visual, spatial and executive functions.71
Functional recovery is of substantial concern to patients. The 7-level mRS covers a range of functional outcomes, from no symptoms to death, and the categories are intuitive and easy for clinicians and patients to understand.41 Results of our meta-analysis showed that, although there was no association between CysC levels and functional recovery assessed using mRS in the total AIS population, a significant association was found with a follow-up time <1 month. There are no studies on serum CysC levels at different time points, and the mechanism behind this finding remains unclear. Further studies are needed to explore the relationship between serum CysC and functional recovery at different time points to help understand the mechanisms underlying these associations.
Cystatin C was higher in significant improvement, improvement and ineffective/deterioration groups than the basic cure group, and there are several explanations for this. Cystatin C participated in the inflammatory reaction, destroyed nerve cells and aggravated nervous system damage, which affected the recovery of patients with cerebral infarction.72 In addition, CysC was involved in blood vessel wall remodeling and aggravated atherosclerosis, causing cerebral infarction.73 These mechanisms are supported by our finding that high CysC levels were associated with vascular events.34
Stroke is the 2nd leading cause of death and the primary reason for long-term disability worldwide.3 Our results showed that serum CysC was higher in the death group than in the survival group. Luan et al. and Li et al. reported that serum CysC was higher in the death group than in the AIS patient survival group.52, 53 Serum CysC level may increase in AIS patients with disease aggravation,53 as it may cause neuronal cell apoptosis and neuron loss, leading to delayed neuronal damage and resulting in death.53
We included available studies on the relationship between CysC levels and AIS patient outcomes, and found consistent and meaningful results through the meta-analysis. Publication bias was not found, and sensitivity analyses showed stable overall effect sizes. To minimize factors that may affect the results and explore potential sources of heterogeneity, we applied subgroup analyses and found that follow-up time may be a source of heterogeneity. Furthermore, high CysC levels in AIS patients with cognitive dysfunction, unfavorable functional recovery and death suggest that it may be related to AIS patient outcomes. Our meta-analysis highlighted the importance of early monitoring and management of CysC and provided evidence for improving AIS patient outcomes. Future studies should explore whether decreasing serum CysC is a therapeutic target for preventing poor outcomes in AIS patients.
Limitations
Limitations of this meta-analysis include the small number of studies on some outcomes, which may affect the stability of the results. All included studies were conducted in China, meaning the results cannot be generalized to global populations. As such, the findings of this meta-analysis should be verified by studies in other countries. Cystatin C is a highly accurate measure of kidney decline, and since data on kidney function were not recorded in the included studies, we were unable to perform analysis based on renal function. In addition, patient age, AIS course and severity, and comorbidities may affect the prognosis, but current data are insufficient to support further analysis. As such, further meta-analyses with more studies are needed to explore the impact of these factors on the association between serum CysC and AIS patient outcomes.
Conclusions
The current systematic review and meta-analysis found that higher levels of serum CysC are associated with poor outcomes of patient with AIS. The results demonstrate that higher serum CysC may be a risk factor for poor outcomes in patients with AIS, which provides a new direction for preventing such outcomes. However, more prospective studies are needed to confirm these findings and reduce study limitations.
Since AIS patient outcomes are affected by many factors, it is not clear whether CysC is an independent predictor. Additional research is needed to explore whether single or combined measurements of this biomarker, with or without other demographic, clinical and biochemical variables, can further increase early risk stratification and clinical decisions in this population. Close attention should also be paid to studies of CysC as a therapeutic target for preventing poor outcomes in patients with AIS. Further clinical studies are important for examining whether decreasing serum CysC can prevent or treat poor AIS patient outcomes.
Supplementary data
The Supplementary materials are available at https://doi.org/10.5281/zenodo.10674816. The package includes the following files:
Supplementary File 1. Search terms.
Supplementary Table 1 Risk of bias summary.
Supplementary Fig. 1. A. Funnel plot for publication bias regarding function recovery assessed with mRS; B. Funnel plot for publication bias regarding to function recovery assessed using NIHSS; C. Funnel plot for publication bias regarding cognitive dysfunction; D Funnel plot for publication bias regarding death.
