Pan-immune-inflammation value and neutrophil-to-HDL-C ratio for prognosis in checkpoint inhibitor myocarditis for gastrointestinal malignancies: a retrospective analysis

Objectives Immune checkpoint inhibitor-related myocarditis (ICI-M) is rare in patients with gastrointestinal (GI) cancers, yet it may be life-threatening. Rapid and cost-effective risk stratification biomarkers remain lacking. This study sought to assess the prognostic utility of six ratio-based biomarkers in this patient group. Methods This retrospective study involved 88 GI cancer patients with clinically diagnosed ICI-M at the Fourth Hospital of Hebei Medical University from June 2021 to June 2026, with a 1-year follow-up period. Patients were stratified by Common Terminology Criteria for Adverse Events (CTCAE) grade (≤2 low-grade, n=59; ≥3 high-grade, n=29). Ratio biomarkers were calculated from complete blood counts and high-density lipoprotein cholesterol (HDL-C). 7-day major adverse cardiovascular events (MACEs) and 1-year all-cause mortality were the primary clinical endpoints. Receiver operating characteristic (ROC) analyses and univariate logistic regression (LR) for 7-day MACEs, and time-dependent ROC analysis, univariate/multivariate Cox regression, and calibration curve analysis for 1-year overall survival (OS) were performed. Restricted cubic spline (RCS) curves and Kaplan–Meier analyses further demonstrated the prognostic performance of ratio biomarkers. Results Pan-Immune-Inflammation Value (PIV) demonstrated prognostic value for 7-day MACEs [AUC=0.786; OR=8.81, 95 % confidence interval (CI):2.24–34.64]. Neutrophil-to-HDL-C ratio (NHR) emerged as the prognostic risk factor for 1-year OS in the univariate (HR=1.26, 95 % CI: 1.14–1.40, p=0.001) and multivariate (HR=1.27, 95 % CI: 1.15–1.41, p<0.001) Cox regression, with an optimal threshold of 5.70 effectively stratifying low- and high-risk patients (log-rank p<0.05). Both PIV and NHR showed significant positive correlations with cardiac biomarkers (cTnI and NT-proBNP: both R>0, both p<0.05) and were higher in the high-grade ICI-M group (both p<0.01). The combination of PIV with cTnI significantly improved the prediction of 7-day MACEs compared with cTnI alone (p=0.0307), while the combination of NHR with cTnI improved the prediction of 1-year OS at 90 days compared with cTnI alone (p=0.037). Conclusions PIV discriminated 7-day MACEs, whereas NHR served as a stratifier for 1-year OS in GI cancer patients with ICI-M. These easily obtainable ratio biomarkers may support risk-adapted monitoring, though external validation in larger cohorts is still needed.