Abstract
Background
Immune checkpoint inhibitors (ICIs) may lead to immune-related adverse events, including potentially life-threatening cardiovascular (CV) complications. Despite guideline-recommended troponin monitoring, limited data exist on evaluating its predictive significance.
Objective
We aimed to assess the predictive value of serial high-sensitivity troponin I (hs-TnI) monitoring in patients treated with ICIs.
Methods
A retrospective, single-center, observational study including patients treated with ICIs and performing serial hs-TnI measurements. The primary endpoint was all-cause mortality, and the secondary endpoint was CV events, defined as the composite of myocarditis, pericarditis, acute coronary syndrome, heart failure, or arrhythmias. A tree classifier model identified the most predictive hs-TnI concentration for the main study endpoints.
Results
Overall, 455 patients performed baseline (T1) and follow-up hs-TnI (T2) assessments and were included in the cohort. During a mean follow-up of 25 months (IQR [12-36]), 253 (56%) patients died, and 70 (15%) developed CV events. T2 > 8 ng/L was significantly associated with increased all-cause mortality (64% vs 48%, P < .001) and CV events (22% vs 9%, P < .001), notably HF (12% vs 4%, P = .003) and myocarditis (3% vs 0%, P = .038). A multivariable analysis confirmed that T2 > 8 ng/L was an independent predictor for all-cause mortality (HR 1.67, 95% CI: 1.29-2.17, P < .001) and CV events (HR 2.59, 95% CI: 1.50-4.46, P = .001).
Conclusions
Our study emphasizes the significant role of serial troponin monitoring during ICIs therapy as an independent predictor of all-cause mortality and CV events, and suggests an optimal lower hs-TnI cutoff of >8 ng/L for risk stratification. Large prospective trials are needed to confirm these findings.
Keywords: high sensitivity troponin, immune checkpoint inhibitor, cardio-oncology, cardio-toxicity
Implications for practice: This study presents the largest cohort to date evaluating serial high-sensitivity troponin I measurement in patients treated with Immune checkpoint inhibitors. By identifying a novel troponin cutoff (>8ng/L) independently associated with all-cause mortality and cardiovascular events, the findings provide clinicians with a practical biomarker for early risk stratification. These results may support more personalized surveillance strategies, potentially improving treatment management and outcomes.
Introduction
Immune checkpoint inhibitors (ICIs) have revolutionized cancer treatment and are now extensively utilized across various cancer types and disease stages.1-3
ICIs are monoclonal antibodies designed to target inhibitory molecular pathways, including cytotoxic T-lymphocyte–associated protein 4 (CTLA-4; eg, ipilimumab), programmed death-1 inhibitor (PD-1; eg, nivolumab or pembrolizumab), and programmed cell death-ligand 1 inhibitor (PD-L1; eg, atezolizumab, avelumab, or durvalumab). By blocking these pathways, ICIs enhance immune system activation, enabling the recognition and destruction of cancer cells.4,5 However, while ICIs potentiate the immune response against cancer, they can also disrupt immune homeostasis, resulting in immune-related adverse events due to the accumulation of proinflammatory T cells in the organs.5 Cardiovascular (CV) events, although relatively uncommon, are potentially life-threatening.6 Among these, myocarditis has been the most extensively studied,7 but other manifestations include acute coronary syndrome (ACS), heart failure (HF), and arrhythmias.8 The underlying CV events are thought to involve immune-mediated myocardial inflammatory cell infiltration and increased myocardial fibrosis.5
Troponin, a highly sensitive and specific biomarker for myocardial injury, plays a predictive role as an early indicator of clinical outcomes in the general population.9 Among cancer patients,10,11 elevated troponin levels have been linked to left ventricular dysfunction in the context of chemotherapy.11 However, its role in patients treated with ICIs remains understudied and warrants further exploration. Recognizing its significance, the European Society of Cardiology (ESC) 2022 guidelines on cardio-oncology recommend routine serum troponin measurements during baseline assessment and in the initial weeks of ICIs therapy to facilitate early detection and prediction of CV events.7,12,13 Despite these recommendations, evidence supporting the routine use of troponin remains limited. Most data come from subgroup analyses of a randomized controlled trial (RCT)14 or retrospective observational studies,15 which demonstrate associations between elevated baseline troponin levels and increased risk of CV events and all-cause mortality.15 Furthermore, there is a paucity of research investigating the prognostic significance of elevated troponin concentrations following ICIs therapy.
While the upper reference limits and medical decision points for high-sensitivity cardiac troponin (hs-Tn) are primarily established for serial measurements in the diagnosis of acute myocardial infraction,16 evidence from non-acute populations suggests that the predictive value is associated with the absolute concentration of hs-cTn.17,18 This underscores the importance of identifying specific sub-populations to enable a more precise interpretation of results. Recognizing the potential clinical utility of hs-cTn, we implemented serial hs-cTn surveillance for patients undergoing ICIs therapy at our facility.
This study aimed to investigate the utility of serial hs-cTn surveillance in patients treated with ICIs as a marker for high-risk stratification for increased all-cause mortality and CV events.
Methods
Study population and protocol
We conducted a retrospective, single-center, observational study at Tel-Aviv Sourasky Medical Center, a tertiary cancer center in Israel. We reviewed the consecutive medical records of all patients diagnosed with cancer and treated with ICIs therapy between October 2015 and October 2022. The cohort included patients who underwent serial high-sensitivity troponin I (hs-TnI) measurements during ICIs therapy. Importantly, this surveillance strategy is the standard institutional protocol implemented in our center and was not developed for the purposes of this study.
The study was approved by the Tel Aviv Sourasky Medical Center Helsinki regulatory ethics committee (Identifier: TLV-0228-16). Informed consent was waived for this study as it included retrospective analysis of anonymized patient data.
Data collection
Electronic medical charts were used to review baseline medical history and treatment. Cardiac risk factors included diabetes mellitus (DM), hypertension (HTN), hyperlipidemia, and chronic kidney disease (CKD), as defined by the 2022 ESC guidelines.12 CV diseases included ischemic heart disease (IHD), HF, and atrial fibrillation (AF).
Hs-TnI was measured by quantitative high-sensitivity troponin I (TNIH) assay in human plasma (lithium heparin) using the ADVIA Centaur XP system, Siemens Healthcare Diagnostics Inc. Our facility adopted Siemens recommendations for the cutoff 47 nanograms per liter (ng/L) [confidence interval (CI) 36-64], and set the cutoff to abnormal as >50ng/L. During the implementation of the method, we reevaluated the cut-off on our population. The exact frequency that the test yields, according to internal control, is 4%-5%. The actual accuracy in 2023 measured in the summary of external control stands at a deviation of 6% on average. In light of this, 6% + 4% is the total uncertainty of the test.
Hs-TnI was evaluated at baseline (T1) (before the initiation of ICIs therapy) and following the first cycle of ICIs (T2). Given that CV events typically manifest after 1-2 doses of ICIs, our study focused on the early stage of troponin concentrations.
In cases of abnormal hs-TnI concentrations (>50ng/L), we implemented a clinical protocol approach for the treating oncologist to facilitate primary diagnosis, regardless of our study, as shown in Supplementary Figure S1. The clinical decision regarding further investigation (echocardiography, cardiac computed tomography angiography (CTA), or cardiac magnetic resonance imaging (CMR)) was made at the discretion of the treating cardio-oncologist, based on the patient’s clinical presentation, CV risk profile, and concurrent laboratory findings. The imaging modality selection was not standardized, reflecting real-world clinical practice rather than a uniform diagnostic algorithm. According to this protocol, we identified 60 patients with elevated T1 and 64 patients with elevated T2. Of these, 9 patients underwent cardiac CTA, and 8 underwent CMR. Overall, 7 patients were diagnosed with ACS, 9 with HF exacerbation, and 5 with myocarditis. Steroid therapy was initiated on a case-by-case basis at the discretion of the treating physician.
Study endpoints
The primary endpoint was all-cause mortality, extracted from computerized patient charts and the population registry bureau.
The secondary endpoint was CV events, defined as the composite of myocarditis, pericarditis, ACS, HF exacerbation, or arrhythmias (including AF, atrial flutter, ventricular tachycardia, and ventricular fibrillation), and was determined by the treating physician and recorded in the electronic medical charts. These endpoints were selected based on the most commonly reported CV events of ICIs therapy.12
We assessed the prognostic value of hs-TnI levels at T2 in two manners:
Hs-TnI as a linear continuous variable on the primary and secondary endpoints.
Drawing from a literature review indicating that elevated hs-Tn concentrations are linked to CV risk, even within the normal range,19 we employed a tree classifier model to determine the most predictive hs-TnI (T2) concentrations associated with the occurrence of primary and secondary endpoints.
Statistical analysis
All continuous variables were displayed as mean (± standard deviation [SD]) for normally distributed variables or median (interquartile range [IQR]) for variables with non-normal distributions. Categorical variables were displayed as the number (%) of individuals within each group.
Continuous variables were compared by a two-tailed unpaired t-test for normally distributed variables and the Mann‐Whitney U test for non‐normally distributed ones. For associations among categorical variables, we used a Chi-square test. Bonferroni adjusted P-values were calculated for multiple comparisons. The median follow-up time for all-cause mortality was calculated using the reverse Kaplan-Meir method. The normality of distributions was assessed using the Kolmogorov-Smirnov test and Q-Q plots.
To determine which hs-TnI concentrations and corresponding cutoffs should be used to predict the primary and secondary endpoints, we used Chi-square automatic interaction detection (CHAID).20 The CHAID analysis builds a predictive model to determine the best cutoffs for the input variables for predicting an outcome. In CHAID, the continuous predictors are split into categories with an approximately equal number of observations. CHAID creates all possible cross-tabulations for each categorical predictor until the best outcome has been achieved and no further splitting can be performed. The patients were divided into groups according to the output cutoffs.
All-cause mortality was evaluated using univariate and multivariable Cox proportional hazard regressions. Kaplan-Meir cumulative incidence curves divided by hs-TnI cutoff groups were presented. The composite of CV events was analyzed, considering death as a competing event and using competing risk survival analysis. We reported sub-distribution hazard ratios (sHRs) using the Fine‐Gray model. Our models were adjusted for age, sex, prior CKD, HTN, hyperlipidemia, IHD, and use of medications (angiotensin-converting enzyme inhibitor (ACEi), beta-blockers (BB), calcium-channel blockers (CCB), furosemide and statins).
A two-tailed P < .05 was considered statistically significant. All analyses were performed with the SPSS (IBM SPSS Statistics, version 28, IBM Corp., 2016), The R statistical package (version 3.3.1) (R Foundation for Statistical Computing, Vienna, Austria), and GraphPad Prism version 9.00 (GraphPad Software).
Results
Baseline patient characteristics
We reviewed the medical records of 2933 consecutive patients diagnosed with cancer and treated with ICIs therapy between October 2015 and October 2022. Our study cohort comprised 455 patients who underwent serial hs-TnI measurements during ICIs therapy. The median time between T1 and T2 was 41 days. The mean age was 71 ± 10 years, with 67% female representatives. The most common cancer diagnosis was non-small cell lung cancer (NSCLC) (20%), and the most common type of ICIs therapy was pembrolizumab (62%). Only 4% of the cohort were treated with combined ICIs therapy.
Cardiac risk factors were relatively common, as expected among this elderly population, including HTN (53%), hyperlipidemia (39%), and DM (27%). Overall, 28%, 6% and 4% of the patients had a diagnosis of IHD, AF and HF at baseline (Table 1).
Table 1.
Baseline characteristics according to follow-up hs-TnI.
| hs-TnI ≤ 8ng/L (n = 230) |
hs-TnI > 8ng/L (n = 225) |
P value | |
|---|---|---|---|
| Demographics | |||
| Female Gender n (%) | 155 (67.4) | 149 (66.2) | .869 |
| Age, mean (SD) | 70 (10.1) | 72 (10.7) | .011 |
| Cancer type, n (%) | NaN | ||
| Hematologic | 11 (4.8) | 10 (4.4) | |
| Lung | 49 (21.3) | 40 (17.8) | |
| GI | 15 (6.5) | 17 (7.6) | |
| Renal | 18 (7.8) | 23 (10.2) | |
| Bone | 4 (1.7) | 4 (1.8) | |
| Breast | 7 (3.0) | 11 (4.9) | |
| CNS and Spine | 10 (4.3) | 7 (3.1) | |
| Prostate | 8 (3.5) | 5 (2.2) | |
| Gynecology | 12 (5.2) | 7 (3.1) | |
| Urologic | 18 (7.8) | 20 (8.9) | |
| Pancreas and Cholangiocarcinoma | 1 (0.4) | 4 (1.8) | |
| Sarcoma | 0 (0.0) | 0 (0.0) | |
| Melanoma | 14 (6.1) | 18 (8.0) | |
| BCC/SCC/other skin | 12 (5.2) | 11 (4.9) | |
| Liver | 11 (4.8) | 5 (2.2) | |
| Head and Neck | 12 (5.2) | 20 (8.9) | |
| Endocrinologic | 2 (0.9) | 3 (1.3) | |
| Other | 26 (11.3) | 20 (8.9) | |
| ICIs type, n (%) | .463 | ||
| Pembrolizumab | 132 (57.4) | 148 (65.8) | |
| Nivolumab | 43 (18.7) | 31 (13.8) | |
| Avelumab | 5 (2.2) | 4 (1.8) | |
| Atezolizumab | 19 (8.3) | 20 (8.9) | |
| Durvalumab | 20 (8.7) | 13 (5.8) | |
| Ipilimumab | 11 (4.8) | 9 (4.0) | |
| Cardiovascular risk factors | |||
| CKD, n (%) | 10 (4.3) | 27 (12.0) | .005 |
| Hypertension, n (%) | 98 (42.6) | 143 (63.6) | <.001 |
| Hyperlipidemia, n (%). | 78 (33.9) | 101 (44.9) | .021 |
| Diabetes, n (%) | 62 (27.0) | 60 (26.7) | 1.000 |
| BMI, mean (SD) | 29.1 (26.0) | 27.3 (9.2) | .351 |
| Ischemic heart disease, n (%) | 52 (22.6) | 77 (34.2) | .008 |
| Atrial fibrillation, n (%) | 5 (2.2) | 21 (9.3) | .001 |
| Heart Failure, n (%) | 5 (2.2) | 12 (5.3) | .756 |
| Cardiovascular Medications, n (%) | |||
| Statins | 105 (45.7) | 141 (62.7) | <.001 |
| ACEi | 46 (20.0) | 83 (36.9) | <.001 |
| ARB | 38 (16.5) | 42 (18.7) | .633 |
| Beta-blockers | 75 (32.6) | 103 (45.8) | .005 |
| Fusid | 16 (7.0) | 33 (14.7) | .012 |
| Baseline Laboratory Parameters | |||
| Hematocrit %, mean (SD) | 36.3 (5.9) | 34.1 (6.1) | <.001 |
| WBC K/uL, mean (SD) | 8.7 (4.8) | 8.8 (4.6) | .902 |
| Platelet K/uL, mean (SD) | 252.2 (103.6) | 256.6 (116.2) | .668 |
| Creatinine mg/dL, median [IQR] | 0.8 [0.7, 1.1] | 0.9 [0.8, 1.3] | .001 |
ACEi, angiotensin converting enzyme inhibitor; ARB, angiotensin receptor blocker; BCC, basal cell carcinoma; BMI, body mass index; CKD, chronic kidney disease; CNS, central nervous system; GI, Gastrointestinal; hs-TnI = high-sensitivity Troponin I; ICIs, immune checkpoint inhibitor; IHD, ischemic heart disease; n, number; n = number; SCC, squamous cell carcinoma; SD, standard deviation; WBC, white blood cells.
Our cohort observed a higher prevalence of baseline CV disease and risk factors, compared to patients treated with ICIs and not performing serial hs-TnI measurements (Supplementary Table S1). We believe this reflects the initial phase of implementation of our hs-TnI surveillance protocol, during which the practice was more rigorously applied to patients with pre-existing CV disease and risk factors, who were also referred more frequently for assessment by a cardio-oncologist.
Median (IQR) baseline (T1) and follow-up (T2) hs-TnI levels were 6 ng/L (4.0, 14.0) and 8 ng/L (5.0, 22.5). Due to the limited data and high cost, our routine follow-up did not include N-terminal-pro-brain natriuretic peptide (NT-proBNP).
Study outcomes
During a median follow-up of 25 months [IQR: 12-36], 253 (56%) patients died, and 70 (15%) patients developed CV events. These included 30 patients (7%) with arrhythmias, 36 (8%) with HF, 15 (3%) with ACS, and 6 (1%) with myocarditis. Among the arrhythmia cases, only 4 patients had a baseline diagnosis of AF. Similarly, 9 patients who experienced HF exacerbations had a known history of HF. No events of pericarditis were noted. Among the patients diagnosed with myocarditis, 5 patients died during follow-up, with 4 (67%) patients attributable to immune-related adverse events, supporting past reports on high mortality.6
When assessing the predictive value of hs-TnI as a linear continuous variable, we found that hs-TnI at T2 emerged as a significant predictor for all-cause mortality (HR 1.02 (95% CI 1.00-1.03) per 100-unit increase in hs-TnI, P = .024), while there was only a trend for higher risk for CV events (HR 1.01 (95% CI 1.00-1.03) per 100-unit increase in hs-TnI, P = .157). The only significant differences in baseline clinical characteristics observed between patients who did or did not develop CV events were baseline HF and elevated hs-TnI levels (Supplementary Table S2).
Using the CHAID tree classifier model, we identified that the optimal cutoff value of hs-TnI at T2 for predicting the development of CV events was 8 ng/L. This encouraged us to further investigate the prognostic significance of this cutoff value for the primary and secondary outcomes.
Overall, 225 (50%) patients presented with hs-TnI at T2 > 8 ng/L. These patients were characterized by older age (72 ± 11 vs 70 ± 10, P = .011) and higher prevalence of baseline cardiac risk factors, including HTN (64% vs 43%, P < .001), hyperlipidemia (45% vs 34%, P = .021) and CKD (12% vs 4%, P = .005). Baseline IHD (34% vs 23%, P = .008) and AF (21% vs 5%, P = .001) were also higher in this group. Accordingly, higher use of cardioprotective treatments was observed in this group, including ACEi (37% vs 20%, P < .001), BB (46% vs 33%, P = .005), and statins (63% vs 46%, P < .001). No significant differences were observed between the two groups regarding the type of cancer or cancer therapy (Table 1).
Hs-TnI at T2 > 8 ng/L was associated with a higher rate of all-cause mortality (64% vs 47%, P < .001) and CV events (22% vs 9%, P < .001), with a specifically higher rate of HF exacerbations (12% vs 4%, P = .003) and myocarditis (3% vs 0%, P = .038) (Table 2). Furthermore, increased hs-TnI at T2 showed a 73% increased risk of all-cause mortality (unadjusted HR 1.73, 95% CI: 1.35-2.22, P < .001). After adjustment for age, sex, existing co-morbidities, and use of medications, the multivariable Cox model showed that hs-TnI at T2 > 8 ng/L remained an independent significant predictor for higher all-cause mortality (HR 1.67, 95% CI: 1.29-2.17, P < .001) (Figure 1). Older age was also found to be a significant predictor for higher all-cause mortality (HR 1.017, 95% CI: 1.0-1.03, P = .011), while statin therapy was significantly associated with lower all-cause mortality (HR = 0.65, 95% CI: 0.48-0.90, P = .008) (Table 3). Similarly, hs-TnI at T2 > 8 ng/L was found to be associated with a 2.6-fold increased risk of CV events (unadjusted HR 2.62, 95% CI: 1.58- 4.35, P < .001) and remained the only significant predictor of CV events following multivariable Cox regression model (HR 2.59, 95% CI: 1.50-4.46, P = .001) (Figure 2). We observed no differences in the occurrence of CV events concerning the type of ICIs administered (P = .974). By stratifying our cohort into four groups based on troponin concentrations > 8 ng/L at T1 and T2, we noted that the most significant predictor for all-cause mortality emerged as troponin measurements at T2 (Supplementary Figure S2). Notably, patients who presented with troponin concentrations > 8 ng/L both at T1 and T2 exhibited a significantly higher incidence of CV events (25%) compared to those with troponin concentrations > 8 ng/L only at T2 (14%) or T1 (13%) (P < .001) (Supplementary Figure S3).
Table 2.
Outcomes.
| hs-TnI ≤ 8ng/L | hs-TnI > 8ng/L | P value | |
|---|---|---|---|
| Atrial fibrillation/flutter, n (%) | 10 (4.3) | 15 (6.7) | .379 |
| Arrhythmias composite events, n (%) | 12 (5.2) | 18 (8) | .314 |
| HF exacerbation, n (%) | 9 (3.9) | 27 (12) | .003 |
| ACS events, n (%) | 4 (1.7) | 11 (4.9) | .105 |
| Myocarditis, n (%) | 0 (0) | 6 (2.7) | .037 |
| CV events, n (%) | 21 (9.1) | 49 (21.8) | <.001 |
| All-cause mortality, n (%) | 109 (47.4) | 144 (64) | .001 |
n, number; hs-TnI, high-sensitivity Troponin I; HF, heart failure; ACS, acute coronary syndrome; CV, cardiovascular.
Figure 1.
Overall mortality Kaplan-Meier curves by follow-up troponin concentrations
Table 3.
Time to event analysis*
| HR | 95.0% CI | P-value | |
|---|---|---|---|
| All-Cause Mortality | |||
| Follow-up hs-TnI (T2) > 8 ng/L | 1.670 | 1.288 - 2.166 | <.001 |
| Age | 1.017 | 1.004 - 1.031 | .011 |
| Statins therapy | 0.652 | 0.475 - 0.895 | .008 |
| Cardiovascular events | |||
| Follow-up hs-TnI (T2) > 8 ng/L | 2.540 | 1.479 - 4.360 | .001 |
*Adjusted for age, sex, CKD, HTN, hyperlipidemia, IHD, Statin, BBs, fusid, ACEi, CCBs.
ACEi, angiotensin converting enzyme inhibitor; BB, beta-blockers; CCB, calcium channel blockers; CKD, chronic kidney disease; hs-TnI, high-sensitivity Troponin I; HR, hazard ratio; HTN, hypertension; IHD, ischemic heart disease; n, number.
Figure 2.
Composite of cardiovascular (CV) events Kaplan-Meier curves by follow-up troponin concentrations
Discussion
We performed a large retrospective observational study of patients treated with ICIs who underwent serial hs-TnI assessment. Our findings support the role of hs-TnI following ICIs therapy as a significant and independent risk factor for all-cause mortality and CV events.
ICIs have significantly altered the field of cancer therapy, with nearly 50% of cancer patients being found eligible for immunotherapy.21 However, ICIs therapy may come with a potential cost of CV events. While CV events are not uncommon in this typically elderly and high-risk population, regardless of ICIs therapy, emerging evidence suggests these events may be potentially directly related to the mechanism of action of ICIs.22 Troponin is a specific indicator of cardiac tissue injury, first described in the late 1970s,23 and has been shown to have a prognostic value even in asymptomatic patients within the general population.16 Cardinale et al. showed that high levels of troponin were associated with the development of cardiac dysfunction during cancer therapy.24
While the European Society of Cardiology (ESC) 2022 guidelines on cardio-oncology incorporated serum troponin measurements as a screening tool at baseline assessment and during the initial weeks of ICIs therapy,12 this recommendation is based on limited data. Rini BI et al14 showed that among patients diagnosed with RCC and treated with combined ICIs and vascular endothelial growth factor receptor (VEGFR) protocol therapy, elevated baseline troponin concentrations were correlated with a higher incidence of CV events. While this RCT is innovative and essential, it has several limitations, including non-standardized biomarker assays between the different sites, the inability to separate cardiotoxicity associated with ICIs without VEGFR exposure, and merely evaluating avelumab therapy, which is less frequently used than other ICIs therapies. Supporting the prognostic role of troponin among ICIs-treated patients, primary analysis of our group15 has shown that both baseline and short follow-up abnormal hs-TnI were prognostic predictors for the development of CV events. However, this cohort was relatively small and included only patients treated with pembrolizumab.
Given the limited data on the role of follow-up troponin measurements during the early stages of ICIs therapy, our study specifically focused on evaluating early follow-up troponin levels. Our analysis revealed that troponin, assessed as a linear continuous variable in a single measurement, was a significant predictor of all-cause mortality and demonstrated a positive trend toward predicting the development of CV events.
The advent of highly sensitive assays has enabled the detection of minimal troponin concentrations with high precision,23 promoting the need to identify optimal troponin cutoff values for risk stratification and the prediction of adverse clinical outcomes. Finke et al25 reported that troponin T levels above a median value of 7 ng/L were an independent marker of all-cause mortality in cancer patients. Similarly, Petricciuolo et al.26 found that high-sensitivity troponin T (hs-TnT) levels exceeding 14 ng/L predicted a composite CV endpoint and progression of cardiac involvement at three months. In our study, we identified that a follow-up hs-TnI level (T2) > 8 ng/L was the optimal cutoff value associated with both increased all-cause mortality and CV events. We do acknowledge that troponin elevations may result from various non-CV mechanisms, including the underlying malignancy, infectious or inflammatory processes, and other comorbidities. Therefore, after adjustment for those key comorbidities, hs-TnI (T2) > 8 ng/L remained a significant independent predictor for both outcomes. Notably, no significant differences in cancer type or cancer therapy were observed between the groups. In this context, we view troponin as a marker of overall vulnerability and increased risk for poorer outcomes, rather than as a standalone diagnostic tool for CV events. Identifying a specific troponin cutoff with the potential to predict mortality and CV events could provide cardio-oncologists with a valuable tool for risk stratification. This could lead to the implementation of closer monitoring or even the initiation of cardioprotective therapies in high-risk patients. However, further studies are warranted to validate these findings and explore the efficacy of potential cardioprotective strategies based on this low troponin threshold.
By evaluating baseline and serial follow-up hs-TnI, we learned that follow-up troponin is a more significant predictor of all-cause mortality and CV events. Furthermore, hs-TnI (T2) > 8 ng/L both at baseline and following ICIs therapy has the highest significant impact on the development of CV events.
To our knowledge, this is the largest cohort to evaluate serial troponin measurement during ICIs therapy. The strengths of our study include the assessment of various types of ICIs therapies and cancer types, the use of a standardized hs-TnI assay, and a focus on several follow-up troponin measurements. Additionally, our study expands the role of troponin beyond its established use in diagnosing myocarditis by evaluating its prognostic value for the broader spectrum of CV events and all-cause mortality. This is particularly relevant given that CV events are underdiagnosed in this population. Moreover, the potential for overdiagnosis and its consequences, such as unnecessary interruption of life-saving cancer therapy, must be carefully considered.
Although the cut-off identified falls within the normal range, we propose that it serves as a risk stratification marker to identify patients at risk of developing CV complications due to immunotherapy rather than as a diagnostic threshold for cardiotoxicity.
Limitations
Our study has several limitations. First, it is a single-center study, and thus, the generalization of our results is limited. Second, this is a retrospective study, and therefore, our results are subject to the effects of potential confounders inherent to the nature of such studies and may be biased by their design. Third, our results are applicable specifically to hs-TnI performed by ADVIA Centaur XP system, Siemens Healthcare Diagnostics Inc.; therefore, validations will be needed for other assays and hs-TnT. Fourth, the limited number of patients performing more than two tests of hs-TnI measurements did not allow us to evaluate the role of elevated hs-TnI at a later stage. Last, we must note that using computerized patient charts from the Population Registry Bureau did not provide information on the cause of death. Consequently, we could not assess whether troponin correlates to CV death.
Conclusions
We report, for the first time to the best of our knowledge, the largest cohort of serial hs-TnI surveillance in patients diagnosed with cancer and treated with ICIs therapies. Our results revealing the significant and independent role of hs-TnI as a predictor for all-cause mortality and CV events and implying for a possible lower optimal cutoff of hs-TnI > 8 ng/L. Future prospective studies are essential to evaluate the clinical utility of standardized management protocols for elevated troponin, including the potential role of cardio-protective therapies and strategies for cancer therapy management.
Supplementary Material
Acknowledgments
None.
Contributor Information
Moran Gvili Perelman, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Rafael Y Brzezinski, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Barliz Waissengrin, Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel; Division of Oncology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel.
Yasmin Leshem, Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel; Division of Oncology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel.
Ari Raphael, Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel; Division of Oncology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel.
Maxim Perelman, Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel; Internal Medicine T, Chaim Sheba Medical Center, Ramat-Gan 6423906, Israel.
Noam Weiss, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Maor Tzuberi, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Moshe Stark, Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel; Division of Clinical Laboratories, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel.
Ilana Goldiner, Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel; Division of Clinical Laboratories, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel.
Shafik Khoury, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Ofer Havakuk, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Yan Topilsky, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Shmuel Banai, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Rabea Asleh, Heart Institute, Hadassah University Medical Center, Faculty of Medicine, Hebrew University of Jerusalem, Jerusalem 6423906, Israel.
Ido Wolf, Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel; Division of Oncology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel.
Michal Laufer-Perl, Division of Cardiology, Tel-Aviv Sourasky Medical Center, Tel Aviv 6423906, Israel; Tel Aviv School of Medicine, Tel Aviv University, Tel Aviv 6423906, Israel.
Author contributions
Moran Gvili Perelman: Conception and Drafting of the manuscript; Rafael Y. Brzezinski: Analysis and interpretation of data; Barliz Waissengrin: Conception and Design; Yasmin Leshem: Conception and Design; Ari Raphael: Conception and Design; Maxim Perelman: Conception and Design; Noam Weiss: Conception and Design; Maor Tzuberi:; Moshe Stark: Conception and Design; Ilana Goldiner: Conception and Design; Shafik Khoury: Conception and Design; Ofer Havakuk: Conception and Design; Yan Topilsky: Conception and Design; Shmuel Banai: Conception and Design; Rabea Asleh: Conception and Drafting of the manuscript; Ido Wolf: revising for critically important intellectual content; Michal Laufer-Perl: conception and design, revising for critically important intellectual content and final approval of the manuscript
Funding
None.
Conflicts of interest
None.
Data availability
The data underlying this article are not publicly available in accordance with the institutional policies regarding patient confidentiality and data sharing.
References
- 1. Provencio M, Nadal E, Insa A, et al. Neoadjuvant chemotherapy and nivolumab in resectable non-small-cell lung cancer(NADIM): an open-label, multicentre, single-arm, phase 2 trial. Lancet Oncol. 2020;21:1413-1422. https://doi.org/ 10.1016/S1470-2045(20)30453-8 [DOI] [PubMed] [Google Scholar]
- 2. Cascone T, William WN Jr, Weissferdt A, et al. Neoadjuvant nivolumab or nivolumab plus ipilimumab in operable non-smallcell lung cancer: the phase 2 randomized NEOSTAR trial. Nat Med. 2021;27:504-514. https://doi.org/ 10.1038/s41591-020-01224-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Felip E, Altorki N, Zhou C, et al. IMpower010 Investigators. Adjuvant atezolizumab after adjuvant chemotherapy in resected stage IB-IIIA non-smallcell lung cancer (IMpower010): a randomised, multicentre, openlabel,phase 3 trial [published correction appears in Lancet. Lancet. 0308;398:1344-1357. https://doi.org/ 10.1016/S0140-6736(21)02098-5 [DOI] [PubMed] [Google Scholar]
- 4. Guo CW, Alexander M, Dib Y, et al. A closer look at immune-mediated myocarditis in the era of combined checkpoint blockade and targeted therapies. Eur J Cancer 2020;124(none):15-24. https://doi.org/ 10.1016/j.ejca.2019.09.009 [DOI] [PubMed] [Google Scholar]
- 5. Lyon AR, Yousaf N, Battisti NML, Moslehi J, Larkin J.. Immune checkpoint inhibitors and cardiovascular toxicity;. Lancet Oncol. 2018;19:e447-e458. https://doi.org/ 10.1016/S1470-2045(18)30457-1 [DOI] [PubMed] [Google Scholar]
- 6. Moslehi JJ, Salem JE, Sosman JA, Lebrun-Vignes B, Johnson DB.. Increased reporting of fatal immune checkpoint inhibitor-associated myocarditis. Lancet. 2018;391:933. https://doi.org/ 10.1016/S0140-6736(18)30533-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Mahmood SS, Fradley MG, Cohen JV, et al. Myocarditis in patients treated with immune checkpoint inhibitors. J Am Coll Cardiol. 2018;71:1755-1764. https://doi.org/ 10.1016/j.jacc.2018.02.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Dolladille C, Akroun J, Morice PM, et al. Cardiovascular immunotoxicities associated with immune checkpoint inhibitors: a safety meta-analysis. Eur Heart J. 2021;42:4964-4977. https://doi.org/ 10.1093/eurheartj/ehab618 [DOI] [PubMed] [Google Scholar]
- 9. van der Linden N, Klinkenberg LJJ, Bekers O, et al. Prognostic value of basal high-sensitive cardiac troponin peats on mortality in the general population: a meta-analysis. Medicine (Baltim). 2016;95:e5703. https://doi.org/ 10.1097/MD.0000000000005703 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Cardinale D, Sandri MT.. Role of biomarkers in chemotherapy-induced cardiotoxicity. Prog Cardiovasc Dis. 2010;53:121-129. https://doi.org/ 10.1016/j.pcad.2010.04.002 [DOI] [PubMed] [Google Scholar]
- 11. Cardinale D, Sandri MT, Martinoni A, et al. Left ventricular dysfunction predicted by early troponin I release after high-dose chemotherapy. J Am Coll Cardiol. 2000;36:517-522. https://doi.org/ 10.1016/s0735-1097(00)00748-8 [DOI] [PubMed] [Google Scholar]
- 12. Lyon A, Lopez-Fernandez T, Couch L, et al. ESC Guidelines on cardio-oncology developed in collaboration with the European Hematology Association (EHA), the European Society for Therapeutic Radiology and Oncology (ESTRO) and the International Cardio-Oncology Society (IC-OS). Eur Heart J. 20222000;00:1-133. https://doi.org/ 10.1093/eurheartj/ehac244 [DOI] [PubMed] [Google Scholar]
- 13. Lyon A, Dent S, Stanway S, et al. Baseline cardiovascular risk assessment in cancer patients scheduled to receive cardiotoxic cancer therapies: a position statement and new risk assessment tools from the Cardio-Oncology Study Group of the Heart Failure Association of the European Society of Cardiology in collaboration with the International Cardio-Oncology Society. Eur J Heart Fail. 2020;22:1945-1960. https://doi.org/ 10.1002/ejhf.1920 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14. Rini BI, Moslehi JJ, Bonaca M, et al. Prospective cardiovascular surveillance of immune checkpoint inhibitor-based combination therapy in patients with advanced renal cell cancer: data from the phase III JAVELIN Renal 101 Trial. J Clin Oncol. 2022;40:1929-1938. https://doi.org/ 10.1200/jco.21.01806 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Waissengein B, Abu AB, Merimsky O, et al. The predictive value of high sensitivity troponin measurements in patients treated with immune checkpoint inhibitors. Clin Res Cardiol. 2023;112:409-418. https://doi.org/ 10.1007/s00392-022-02118-8 [DOI] [PubMed] [Google Scholar]
- 16. van der Linden N, Klinkenberg LJJ, Bekers O, et al. Prognostic value of basal high-sensitive cardiac troponin levels on mortality in the general population: a meta-analysis. Medicine (Baltim). 2016;95:e5703. https://doi.org/ 10.1097/MD.0000000000005703 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Neumann JT, Twerenbold R, Ojeda F, et al. ; COMPASS-MI Study Group. Application of high-sensitivity troponin in suspected myocardial infarction. N Engl J Med. 2019;380:2529-2540. https://doi.org/ 10.1056/NEJMoa1803377 [DOI] [PubMed] [Google Scholar]
- 18. Reichlin T, Twerenbold R, Maushart C, et al. Risk stratification in patients with unstable angina using absolute serial changes of 3 high-sensitive troponin assays. Am Heart J. 2013;165:371-378. https://doi.org/ 10.1016/j.ahj.2012.11.010 [DOI] [PubMed] [Google Scholar]
- 19. Muscente F, Caterina RD.. New insights from the MESA study: increased high sensitivity troponins as a cardiovascular risk factor. Eur Heart J Suppl. 2021;2021(none):23. https://doi.org/ 10.1093/eurheartj/suab092 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Magidson J,. Inc. S. SPSS for Windows, CHAID, release 6.0. Chicago, IL: SPSS, 1993 [Google Scholar]
- 21. Lehmann LH, Cautela J, Palaskas N, et al. Clinical strategy for the diagnosis and treatment of immune checkpoint inhibitor-associated myocarditis: a narrative review. JAMA Cardiol. 2021;6:1329-1337. https://doi.org/ 10.1001/jamacardio.2021.2241 [DOI] [PubMed] [Google Scholar]
- 22. Vasbinder A, Chen Y, Procureur A, et al. Biomarker trends, incidence, and outcomes of immune checkpoint inhibitor–induced myocarditis. JACC: CardioOncol. 2022;4:689-700. https://doi.org/ 10.1016/j.jaccao.2022.11.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Delombaerde D, Vervloet D, Franssen C, et al. Clinical implications of isolated troponinemia following immune checkpoint inhibitor therapy. ESMO Open. 2021;6:100216. https://doi.org/ 10.1016/j.esmoop.2021.100216 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Cardinale D, Sandri MT.. Role of biomarkers in chemotherapy-induced cardiotoxicity. Prog Cardiovasc Dis. 2010;53:121-129. https://doi.org/ 10.1016/j.pcad.2010.04.002 [DOI] [PubMed] [Google Scholar]
- 25. Finke D, Romann SW, Heckmann MB, et al. High sensitivity cardiac troponin T determines all-cause mortality in cancer patients: a single center cohort study. ESC Heart Failure 2021;8:3709-3719. https://doi.org/ 10.1002/ehf2.13515 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Petricciuolo S, Delle Donne MG, Aimo A, Chella A, De Caterina R.. Pre-treatment high-sensitivity troponin T for the short-term prediction of cardiac outcomes in patients on immune checkpoint inhibitors. Eur J Clin Invest. 2021;51:e13400. https://doi.org/ 10.1111/eci.13400 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data underlying this article are not publicly available in accordance with the institutional policies regarding patient confidentiality and data sharing.