Divergent cardiovascular risk profiles of immune checkpoint inhibitors versus platinum-based chemotherapy in NSCLC: a real-world study of 84,097 patients

Background Both immune checkpoint inhibitors (ICIs) and platinum-based doublet chemotherapy carry cardiovascular risk in non-small cell lung cancer (NSCLC), yet existing comparative data are conflicting,¹,² derived largely from heterogeneous trial meta-analyses. Platinum agents cause direct endothelial injury and vascular toxicity; taxanes are associated with myocardial ischemia and arrhythmia.³ Whether ICIs confer net greater or lesser cardiovascular harm in direct clinical practice comparison is unknown. Methods Retrospective active-comparator cohort study using Epic Cosmos, a de-identified real-world EHR network of 300,262,705 patients (diagram). Among 1,405,504 patients receiving systemic cancer therapy (2018–2023), 148,743 carried an NSCLC diagnosis; 84,097 initiated either a PD-1/PD-L1 inhibitor (pembrolizumab or nivolumab; n=53,833) or platinum-based doublet chemotherapy (carboplatin/cisplatin + taxane or pemetrexed within ±28 days; n=30,264). Index date was the first qualifying dose. Patients were censored at last encounter or cross-over to the alternate treatment class to prevent exposure misclassification. Propensity scores were estimated by logistic regression; stabilized IPTW truncated at the 1st/99th percentiles was applied, with balance verified by standardized mean differences (all post-IPTW SMDs <0.10). The primary analysis used an IPTW-weighted Cox proportional hazards model with robust variance. IPTW-weighted Kaplan-Meier curves and competing-risk cumulative incidence functions (death) were generated as complementary analyses. Results 11,825 incident heart failure (HF) events occurred (12.4%). ICI was associated with a 24% lower hazard of HF compared with platinum-based chemotherapy (HR 0.76, 95% CI 0.72–0.80; p<0.001, see Fig.), with consistent reductions in both HFpEF (HR 0.68, 0.63–0.74) and HFrEF (HR 0.76, 0.70–0.83). Among pre-specified secondary outcomes, ICI was associated with lower rates of atrial fibrillation (HR 0.65, 0.62–0.68), coronary artery disease (HR 0.70, 0.66–0.73), and unstable angina (HR 0.70, 0.56–0.87). In contrast, ICI carried significantly higher risk of myocarditis (HR 2.03, 1.01–4.09; p=0.047) and stroke (HR 1.27, 1.15–1.39; p<0.001). Conclusions In this large real-world active comparator analysis, ICI therapy was associated with substantially lower risk of heart failure, arrhythmia, and ischemic cardiovascular events compared with platinum-based chemotherapy, consistent with the intrinsic vascular and cardiodepressant burden of platinum and taxane agents. The trade-off of elevated immune-mediated myocarditis and stroke risk with ICIs reinforces the need for vigilant cardio-oncology monitoring, particularly early in the treatment course4. These findings challenge the framing of ICIs as categorically more cardiotoxic and provide actionable real-world evidence for cardiovascular risk stratification in NSCLC treatment planning.Consort DiagramFig.