Interrogating mechanisms of fluoropyrimidine cardiotoxicity: a prospective stress CMR study

Background Fluoropyrimidine-associated cardiotoxicity (FAC) often presents with myocardial ischaemia despite unobstructed epicardial coronaries, implicating functional mechanisms such as vasospasm, endothelial dysfunction, or microvascular impairment rather than fixed coronary disease. In vivo mechanistic data are limited. Stress cardiovascular magnetic resonance (CMR) enables quantitative assessment of myocardial perfusion and myocardial mechanics to interrogate these pathways. Purpose To determine whether fluoropyrimidine therapy induces coronary microvascular dysfunction detectable by stress CMR and whether early myocardial mechanical alterations occur in the absence of overt systolic dysfunction. Methods This prospective single-centre cohort study enrolled consecutive patients with gastrointestinal cancer receiving fluoropyrimidine chemotherapy. Baseline evaluation included stress perfusion CMR with quantification of global myocardial blood flow (MBF) and myocardial perfusion reserve (MPR), CT coronary angiography with calcium scoring, and cardiac biomarkers. CMR was repeated after cycle 1 and at treatment completion. LV volumes, LVEF, and global longitudinal (GLS), circumferential (GCS), and radial strain (GRS) were derived. Paired statistical testing was used as appropriate. Results Fifty-one patients were recruited (age 60.2±12.4 years; 66.7% male). Coronary atherosclerotic burden was low (median calcium score 3 [IQR 0–198.5]). Forty-six patients completed baseline and post–cycle 1 imaging; 43 completed all timepoints. Baseline stress MBF was normal (2.9±0.7ml/g/min) with no meaningful reduction after cycle 1 (2.7±0.7ml/g/min). MPR was similarly unchanged (2.93±0.8 vs 2.82±0.7ml/g/min). LV end-diastolic and stroke volumes decreased, while LVEF and GLS remained normal and unchanged. Small but significant reductions occurred in GCS (–19.2±2.4% to –18.6±2.2%) and GRS (32.4% [IQR 28.5–36.7] to 31.5% [27.1–35.2]). GCS and GRS reductions persisted at treatment completion. One patient developed incident myocardial ischaemia with normal stress MBF (3.28ml/g/min) and normal MPR (4.15). Conclusions Fluoropyrimidine exposure was not associated with impaired stress CMR–derived myocardial blood flow or myocardial perfusion reserve, arguing against fixed coronary microvascular dysfunction as the dominant mechanism of FAC. Early reductions in circumferential and radial strain occurred despite preserved LVEF and normal perfusion; however, concurrent decreases in LV volumes suggest these changes may partly reflect altered loading conditions during treatment rather than isolated myocardial injury. Overall, the pattern supports dynamic vascular mechanisms (e.g., vasospasm or endothelial dysfunction) rather than structural microvascular disease. Myocardial strain alterations may represent an early imaging phenotype of fluoropyrimidine cardiac effects, but their mechanistic specificity requires further investigation.Central illustration Baseline and post cycle 1 CMR findings.