Supplementary Material for: Cardiac troponin I for early detection of cancer therapy–related cardiac dysfunction after chemotherapy: a diagnostic test accuracy meta-analysis.

Introduction
Early myocardial injury during chemotherapy may precede declines in left ventricular ejection fraction (LVEF). We assessed the diagnostic accuracy of cardiac troponin I (cTnI) for early detection of cancer therapy–related cardiac dysfunction (CTRCD).
Methods
We conducted a diagnostic test-accuracy meta-analysis of prospective cohorts measuring cTnI within 3 months after chemotherapy initiation following PRISMA-DTA. Searches were conducted across PubMed, Embase, Cochrane Library, CNKI, and Wanfang databases through August 2025. A bivariate random-effects hierarchical summary receiver operating characteristic (HSROC) model estimated pooled sensitivity and specificity with confidence/prediction regions; risk of bias was appraised using QUADAS-2. Clinical implications were illustrated with Fagan nomograms at prespecified pre-test probabilities (10% and 20%).
Results
A total of 5 prospective cohort studies were included. The HSROC summary operating point yielded pooled sensitivity 0.669 (95% CI 0.444–0.836) and specificity 0.839 (95% CI 0.634–0.940). The corresponding positive likelihood ratio was 4.15 and negative was 0.39. At a 10% pre-test probability, post-test probabilities were 32% after a positive result and 4% after a negative result; at a 20% pre-test probability, they were 51% and 9%, respectively. Leave-one-out analyses indicated stable estimates.
Conclusions
Within the first 3 months of chemotherapy, cTnI shows moderate sensitivity and high specificity for early CTRCD detection. These performance characteristics support cTnI as a confirmatory biomarker to inform risk-stratified surveillance alongside imaging under standardized assay procedures.