Cardiovascular safety planning in NCCN-cited breast cancer trials: Audit and surveillance framework.

Abstract

e24022 Background: Cardiovascular (CV) and venous thromboembolism (VTE) risks differ across systemic breast cancer therapies. NCCN-cited trials underpin treatment recommendations but may embed inconsistent CV and VTE safety surveillance. Methods: Audit of phase II–III systemic-therapy trials cited in NCCN Breast Cancer Guidelines. We created a novel scoring system CVERQ-BC (Cardiovascular evaluation and reporting quality in Breast Cancer) (scored 0–12), capturing reported baseline CV risk; baseline imaging; prespecified serial imaging; baseline ECG; troponin/NP strategy; quantitative CV endpoint definitions; prespecified CV endpoints; central adjudication; CV AE grading; inclusion of stable CVD; BP monitoring with HTN criteria; CV dose-mod/hold rules. VTE score (0–3) captured validated VTE risk models, protocolized prophylaxis, and prespecified objective VTE endpoints. CVERQ-BC scoring was based on review of trial publications, full texts, study protocols, and trial registrations. Two independent reviewers scored each trial, with discrepancies resolved by consensus. Analyses were descriptive with exploratory comparisons by class, era, and sponsor. Results: We included 129 unique trials (1990–2026); 49.6% were industry-sponsored. Median CVERQ-BC was 2 (IQR 2–4); 56.6% scored ≤2 and 6.2% ≥8. Key CV elements were infrequently reported: baseline CV risk (28.7%), baseline imaging (32.6%), prespecified follow-up imaging (24.8%), ECG surveillance (10.9%), biomarker strategy (3.1%), prespecified CV endpoints (26.4%), and central adjudication (10.9%). CVERQ-BC did not differ by sponsor type (p = 0.39) or era among industry trials (p = 0.31). By class (median [IQR]): HER2-directed 6.0 [4.5–7.5], anthracycline 3.0 [2.0–6.0], taxane 3.0 [2.0–6.0], fluoropyrimidine 2.0 [2.0–5.0], alkylator 2.0 [1.25–5.0], platinum 2.0 [2.0–3.5], endocrine 2.0 [2.0–4.0], CDK4/6 inhibitor 2.0 [2.0–3.5]. In HER2-directed trials, baseline and serial cardiac imaging was reported in 77.8% (21/27), explicit quantitative CV endpoint definitions in 77.8% (21/27), and prespecified CV endpoints in 70.4% (19/27). Outside HER2-directed evidence, mechanism-aligned safety domains were uncommon, with QTc or ECG surveillance in QT-liability tyrosine kinase inhibitor and CDK4/6 inhibitor trials, blood pressure criteria in VEGF or VEGFR-directed regimens, and ischemia monitoring in fluoropyrimidine trials each reported in 20% or fewer studies. Conclusions: Reported CV/VTE safety planning in NCCN-cited trials is heterogeneous and concentrated in HER2-directed evidence. Without guideline-specified class-adapted surveillance standards (and possible under-reporting), variability may underestimate cardiotoxicity and limit generalizability to patients with baseline CVD. CVERQ-BC provides a reproducible, class-adapted checklist for trial protocols, reporting, and guideline citation.