Elevated troponin levels as a predictor of mortality in patients with acute stroke: a systematic review and meta-analysis

Background and Aim The prognostic potential of cardiac troponin (cTn) in acute stroke patients has been a subject of ongoing debate. Our objective was to provide a comprehensive evidence for predicting mortality in acute stroke patients by using the elevated troponin levels. Methods We conducted an extensive literature search, including PubMed, EMbase, and Trip Databases, covering studies published up to September 30, 2023. We computed risk ratios (RR) with 95% confidence intervals (CIs), performed sensitivity analysis, and conducted trial sequential analysis (TSA). Results In total, 53 studies were analyzed, with 37 focusing on acute ischemic stroke (AIS), 11 on subarachnoid hemorrhage (SAH), and 7 on Intracerebral hemorrhage (ICH). Elevated cTn levels were significantly showed a higher predictive risk for In-hospital mortality in both AIS (RR=3.80, 95% CI; 2.82 to 5.12) as well as SAH (RR=2.23, 95% CI; 1.64 to 3.02). However, no significant predictive risk between elevated cTn levels and in-hospital mortality for ICH patients (RR=1.13, 95% CI: 0.46 to 2.79). A similar pattern was observed for elevated cTn levels, indicating an increased risk of last follow-up mortality for AIS (RR=2.41, 95% CI: 1.98 to 2.93) and SAH (RR=3.08, 95% CI: 2.25 to 4.21). Conclusion Elevated troponin levels can serve as a promising predictive marker for both in-hospital and last follow-up mortality in AIS and SAH patients but not in ICH patients. Further prospective studies are needed to validate our findings along with exploring the preventive management of mortality in acute stroke settings.


Introduction
Acute stroke represents a critical medical condition with substantial implications for patient outcomes and healthcare systems (1).In this context, the timely and accurate identification of prognostic markers capable of predicting stroke-associated mortality emerges as a paramount imperative (2).Such markers not only facilitate risk assessment but also guide precise and targeted clinical interventions, ultimately influencing patient outcomes and optimizing healthcare resource allocation (3).Cardiac Troponin (cTn), widely recognized as a cardinal biomarker in cardiology, has recently attracted attention for its potential role as an indicator of mortality risk in patients afflicted by acute stroke (4).
In the short term, the immediate post-stroke period, in-hospital mortality rates for acute stroke patients can exhibit alarming elevations, with figures occasionally reaching a staggering 50% (5).The major driving force behind such acute fatalities often lies in cardiac-related complications.These complications manifest through the release of cardiac troponin-T (cTnT), cardiac troponin-I (cTnI), High-sensitive cardiac Troponin I (hs-cTnI) and Highsensitive cardiac Troponin T (hs-cTnT) proteins into the circulatory system.Importantly, during the acute phase of ischemic stroke (IS), a marked elevation in serum levels of cTnT or cTnI is frequently observed, establishing a robust link between cardiac injury and the stroke itself (6).The precise and timely prognosis of this cardiac involvement not only impacts therapeutic strategies but also plays a pivotal role in the monitoring and management of patient outcomes, fundamentally shaping the quality of healthcare delivery (7).
However, despite the potential of cardiac troponin as a prognostic marker, the scientific literature presents contrasting findings regarding its association with the risk of all-cause mortality in patients afflicted with acute stroke (8).Moreover, the precision of risk estimates demonstrates considerable variability across individual research studies.Troponin elevations have been reported in a substantial proportion of stroke patients, with prevalence rates ranging from 27% to 34% (9).These elevations have been consistently linked to heightened mortality in various stroke subtypes, encompassing IS, Intracerebral hemorrhage (ICH), and subarachnoid hemorrhage (SAH).Nevertheless, previous meta-analyses (6,(10)(11)(12)(13)(14)(15)(16)(17) have failed to definitively establish troponin elevation as an independent and unequivocal prognostic factor in the context of acute stroke patients.As a result, the debate concerning the predictive utility of elevated cardiac troponin levels in the stroke population continues to be a subject of scientific scrutiny.Therefore, we aimed to conduct a systematic review and meta-analysis, aptly named "Elevated Troponin Levels as a Mortality Predictor in Acute Stroke, " in order to explore the utility of cTn in predicting mortality following acute stroke.

Search strategy
The Cochrane Handbook (version 5.1.0)was used to conduct a systematic review and meta-analysis (18).The preferred reporting items for systematic reviews and meta-analyses (PRISMA) were followed throughout the systematic review's creation (19).Electronic searches were conducted in databases including PubMed, EMbase, and Trip Databases up to 30th September 2023.Additionally, the reference list of retrieved studies and previous meta-analyses, was manually search for collecting more relevant studies often missed while performing the electronic search.The search aimed to identify studies the prediction of cTn level with the mortality rate in patient with acute stroke.The search utilized a set of specific keywords, including ("Troponin" OR "Cardiac Troponin" OR "cTnT" OR "cTnI" OR "Troponin I" OR "Troponin T" OR "High Sensitive Troponin" OR "High Sensitive Troponin I" OR "hTnI" OR "High Sensitive Troponin T" OR "hTnT") AND ("Stroke" OR "Brain Stroke" OR "Acute Stroke" OR "Ischemic Stroke" OR "Cerebral Infarction" OR "Intracerebral Hemorrhage" OR "Hemorrhagic Stroke" OR "Subarachnoid Hemorrhage" OR "Aneurysmal Subarachnoid Hemorrhage") AND ("Mortality" OR "Mortality Rate" OR "Death" OR "Death Rate" OR "Prognosis" OR "Outcome").

Data extraction
Two independent authors ("A" and "MS") conducted a meticulous evaluation of the identified articles to assess their eligibility for inclusion in our study.Upon the initial screening, the authors proceeded to a comprehensive examination of the full-text articles to validate their eligibility and to systematically extract relevant data.To facilitate this process, a standardized data extraction form was utilized, ensuring uniformity and thoroughness in data collection process.The following data was extracted from studies: First Author's Name, Published year, Country, Study design, Study period, Sample size of elevated cTn and normal cTn groups, source, cTn cut-off value, cTn assessment time point, cTn type, cTn estimation method, In-hospital mortality, Last follow-up mortality outcome measured.The values of cTn levels were reported with different units in the included studies and were converted to similar units for analysis purpose using online unit conversion tools. 1 , 2 Any disagreement were resolved through discussion with the corresponding author.

Quality assessment
The quality of the included studies was assessed using the Newcastle-Ottawa Scale (NOS) for observational cohort studies (20).The NOS evaluates study quality across three domains: selection of participants, comparability of groups, and ascertainment of outcomes.Studies were assigned scores ranging from 0 to 9, with higher scores indicating higher quality.Any discrepancies in quality scores were resolved through consensus with the corresponding author.

Publication bias
To assess publication bias, we employed a funnel plot analysis (21).Egger's regression test was used to ascertain the asymmetry of funnel plots (22).

Statistical analysis
A fixed/random-effects model was used to calculate the pooled risk ratio (RR) with 95% confidence interval (CI).Heterogeneity was calculated with the I 2 statistic.The heterogeneity was considered as significant in case of I 2 more than 50% for which random effects model was applied, on the other hand, if I 2 was less than 50%, then fixed-effect model was applied.Sensitivity analysis was performed by sequentially omitting a single study in each turn, to validate the pooled observed effect.In order to ascertain the sufficiency of our sample size and the statistical power of our meta-analysis, we conducted Trial Sequential Analysis (TSA).Tests were considered statistically significant at a p-value less than 0.05.All statistical analyses were carried out using STATA, version 12.0 (Stata Statistical Software, Release 12; StataCorp LP, College Station, TX).

Last follow-up mortality
Elevated cTn levels were also significantly showed a higher predictive risk for last follow-up mortality in both AIS (RR = 2.41, 95% CI: 1.98 to 2.93; Figure 3A) and SAH (RR = 3.08, 95% CI: 2.25 to 4.21; Figure 3B).Subgroup analysis showed hs-cTnT stood out with a significantly higher prediction risk (RR: 6.02; 95% CI: 2.60 to 13.93) in AIS, compared to cTnT and cTnI (Supplementary Table S2A).Both prospective and retrospective study designs, indicated an elevated predictive role of elevated troponin for last follow-up mortality in AIS and SAH subgroup analysis (Supplementary Tables S2A,B).Predictive risk remained stable for studies examining troponin levels within 24 h after AIS and SAH.Troponin estimation methods varied for AIS studies, with most utilizing CIA method, while for SAH, the predominant method was ELISA.Analysis on last follow-up mortality

Publication bias
Funnel plot analysis did not reveal any apparent signs of asymmetry.The distribution of studies around the estimated effect appeared to be even, suggesting the absence of significant publication bias that could impact the results.Funnel plots for the in-hospital mortality in patients with AIS (p = 0.45), SAH (p = 0.10), and ICH (p = 0.19), and last follow-up mortality for AIS (p = 0.13), and SAH (p = 0.42), are represented in Supplementary Figures S1A-C, S2A,B.

Sensitivity analysis
Sensitivity analysis indicated that the pooled risk estimates for in-hospital mortality in patients with AIS, SAH, and ICH was not significantly affected by the removal of any individual study (Supplementary Figures S3A-C).Similarly, the pooled risk estimates for last follow-up mortality in patients with AIS and SAH was not significantly affected by the removal of any individual study (Supplementary Figures S4A,B).Our findings suggest that the results of the meta-analysis are robust and are not driven by any single study which provides additional confidence in the validity of the findings.

Trial sequential analysis
Our Trial sequential analysis (TSA) showed a very promising strength with 77% power with total sample size of 4,939 AIS patients and 55% power for the studies with 3,062 HS subjects.TSA plot are represented in Figures 4A,B.The high power of the TSA suggests that the meta-analysis is unlikely to be influenced by chance or random fluctuations in the data.Overall, the TSA results provide strong evidence that the meta-analysis is well-powered and that the findings are reliable and generalizable.This reinforces the importance of considering troponin elevation as a crucial factor in risk stratification and treatment decisions for patients with acute stroke.

Discussion
Our findings highlights the significance of routinely assessing admission cTn in stroke patients, emphasizing its association with higher mortality rates.Existing literature has consistently reported the association between elevated troponin levels and all-cause mortality across various stroke types (10, 12).Studies have shown that baseline troponin levels predict poor outcomes in both in-hospital and last follow-up scenarios, irrespective of acute coronary syndrome (69,71,72).Furthermore, higher blood troponin levels during the acute phase of acute spontaneous ICH are linked with unfavorable outcomes (69,71,72).Similar findings are observed in patients with spontaneous subarachnoid hemorrhage, where an increased risk of cerebral ischemia and death is demonstrated (13,14,60).The current meta-analysis supports and reinforces these findings, indicating that elevated troponin levels serve as predictors for both in-hospital and follow-up mortalities for AIS with additional finding of other subtypes of stroke including SAH and ICH.Elevated cTn levels were significantly showed a higher predictive risk for In-hospital mortality and last follow-up mortality in both AIS as well as SAH.However, no significant predictive risk between elevated cTn levels and in-hospital mortality for ICH patients.However, it's crucial to note that this observation may be influenced by the limited inclusion of only seven studies in the overall analysis.

Selection
Previous findings have reported that elevation of cardiac troponin has been reported to occur in ICH patients along with only 1.2% of them died of cardiac causes (69).Although previous studies have been tried to address hypothesis that elevated cardiac troponin might serve as prognostic markers for prediction of adverse clinical events, the results were largely inconsistent and inconclusive with regard to ICH.The previous study demonstrated that elevated troponin levels were associated with higher mortality following ICH (69).Subsequently, the utility of elevated troponin levels for prediction of mortality was confirmed in surgical ICH patients (74), but was not found to be consistently associated with in-hospital mortality in Chinese ICH patients (71).Small sample sizes (less than 240 stroke patients) and ethnic variability probably contribute to the negative results and discrepancies.Similarly in our meta-analysis, only seven ICH studies were included and we observed non-significant association with the elevated level of Troponin.Approximately 18%-20% (75-77) of ischemic stroke patients present with elevated high-sensitive Troponin T (hsTnT) levels on admission, which can be attributed to various factors including renal failure, chronic heart failure, myocardial infarction (MI), and stressrelated cardiomyopathies such as neurogenic stunned myocardium (NSM).NSM, characterized by contraction band necrosis at the cellular level, manifests with elevated hsTnT levels and electrocardiographic abnormalities (78)(79)(80)(81).While some patients can be differentiated using ECG and echocardiography, overlap in diagnostic criteria complicates diagnosis for others.Despite common assumptions in stroke centers, attributing elevated hsTnT levels solely to brain or heart origin may not always hold true, as MI patients have an increased risk of stroke (82), even with marginally elevated hsTnT levels, raising questions about predictive value in differentiating NSM from MI (83).
As per the 2023 Guideline for the Management of Patients with SAH by AHA/ASA (84), certain medical parameters, such as BMI, hypertension, hyperglycemia, troponin levels, hyperthermia, peak white blood cell count, C-reactive protein, and high neutrophil counts, have been linked to clinical outcomes in SAH.However, it is emphasized that additional investigation is necessary to determine their prognostic value and influence on treatment outcomes.Furthermore, the guideline underscores the active exploration of novel biomarkers, encompassing imaging, serum, and cerebrospinal fluid (CSF), in the field of SAH.Ongoing research incorporating advanced proteomic, genomic, and other biological marker methods, in conjunction with existing clinical, radiographic, and physiological monitoring data, is deemed essential.These efforts aim to shed light on the potential use of biomarkers for prognosis and interventions, ultimately contributing to improved outcomes in patients with SAH.
The COVID-19 pandemic has placed unprecedented strain on stroke services globally, disrupting healthcare delivery and raising concerns about its impact on stroke outcomes (85).Despite these challenges, our meta-analysis did not identify any explicit instances linking the pandemic with influential impacts on stroke management or patient outcomes across included studies.This highlights the need for further research to understand the nuanced interplay between COVID-19 and stroke care delivery, while also emphasizing the importance of developing adaptive strategies and fostering interdisciplinary collaborations to mitigate the pandemic's adverse effects on stroke services and ensure optimal patient care.
Clinical decision-making regarding troponin elevations in stroke patients is a multifaceted challenge that requires careful consideration of various factors.While elevated troponin levels in stroke patients can indicate myocardial injury, termed "troponitis, " it's essential to recognize that other contributing factors such as stress, sepsis, or renal dysfunction may also lead to troponin

Study limitations
While the study has strengths, it acknowledges limitations, including: (1) using a single baseline cTn assessment time point and non-standardized measurement procedures, potentially introducing misclassification; (2) lacking information on important confounders like infarction locations, stroke severity, time from symptom onset, and type of recanalization therapies, which could bias the relationship between elevated cTn and mortality; (3) absence of data on functional outcomes like the modified Rankin Scale; and (4) variations in cut-off values, assay methods, and follow-up periods across individual studies contributed to high heterogeneity from pooled analysis by troponin assays, complicating result interpretation.

Conclusion
Elevated troponin levels can serve as a promising predictive marker for both in-hospital and last follow-up mortality in AIS and SAH patients but not in ICH patients.Further prospective studies are needed to validate our findings along with exploring the preventive management of mortality in acute stroke settings.

FIGURE 1 PRISMA
FIGURE 1PRISMA flow diagram for the selection of studies and specific reasons for exclusion from the present meta-analysis.

FIGURE 2 (
FIGURE 2 (A-C) Forest plot for the prediction of In-hospital mortality in (A) AIS, (B) SAH, and (C) ICH with respect to elevated cTn Levels.

FIGURE 3 (
FIGURE 3 (A,B) Forest plot for the prediction of last follow-up mortality in (A) AIS and (B) SAH with respect to elevated cTn Levels.

TABLE 1
Baseline characteristics for the included studies investigating for elevated Troponin level in predicting mortality for acute stroke patients.

TABLE 2
Quality assessment of included studies based on Newcastle-Ottawa Scale (NOS).