Prognostic Performance of a Modified HFA-ICOS Tool for Prediction of Cardiovascular Complications in Cancer Patients

Background: The 2022 European Society of Cardiology Cardio-oncology Guidelines recommend the Heart Failure Association and International Cardio-Oncology Society risk stratification tool to predict cardiovascular disease (CVD) risk in cancer patients. Developed from expert consensus rather than mathematical derivation, its quantitative prognostic value in unselected cancer populations remains uncertain. Objectives: The objective of the study was to assess the prognostic utility of a modified Heart Failure Association and International Cardio-Oncology Society tool, excluding biomarkers and imaging, in predicting CVD complications in cancer patients, thereby providing a quantitative correlate to an expert consensus-based method. Methods: We included 2,290 cancer patients from the third follow-up of the Trøndelag Health study, diagnosed within 4 years of participation, and followed through 2023. Cancer therapy details were obtained from the Norwegian Cancer Registry. The primary outcome was a composite of myocardial infarction, heart failure hospitalization, stroke, or CVD mortality, identified via national registries. Results: The mean age was 65.8 years; 48% were female. The most common cancers were gastrointestinal (23%), prostate (19%), and breast (12%). The median follow-up was 6.9 years. The modified tool classified 28% as low-, 35% moderate-, 34% high-, and 3% very high-risk. CVD incidence rose across categories: 7%, 21%, 29%, and 37% (P < 0.001). The C-statistic was 0.696 (95% CI: 0.674-0.718), and 0.746 (95% CI: 0.676-0.812) among anthracycline-treated patients. Within the low-to-moderate group, event rates ranged from 6% to 8% (0-1 points) to 24% to 27% (3-4 points). Conclusions: The tool stratifies CVD risk in cancer patients and may guide tailored follow-up and cardioprotection. Heterogeneity within moderate-risk patients highlights the need for refinement to improve individualized assessment.