The inadequacy of cardiovascular safety reporting in breast cancer clinical trials.

Abstract

e24100 Background: Although some cardiovascular (CV) toxicity is detected in early phase trials of therapeutic agents, most CV toxicities are not appreciated until later phase clinical trials or general clinical practice. The high prevalence of cardiovascular disease (CVD) in the general population warrants study of how later phase oncologic pharmacotherapy clinical trials address eligibility and assess for treatment-related cardiotoxicity. We focused our study on phase 2 and 3 breast cancer trials to limit disease-specific variations. Methods: A methodical Embase search was performed for phase 2 and 3 trials administering breast cancer pharmacotherapy and including > 50 participants. Firstly, we examined CVD as an exclusion criterion. Secondly, we identified what CV safety assessments were reported such as troponin, brain natriuretic peptide (BNP), lipid profile, electrocardiogram (ECG), or CV imaging. Thirdly, we examined whether trials reported adverse CV events. Fischer’s exact test was applied to compare reporting amongst trial characteristics (trial funding source and type of agent administered), and a 2 sided p-value of < 0.05 was considered significant. Results: 50 published articles from 46 separate clinical trials were included. Patients with prevalent CVD were excluded in 67.4% of trials, the most frequently excluded CVD was heart failure/NYHA class > 1 (65.2% of trials). At least one pre or post-treatment CV safety assessment was reported in only 58.7% of trials, most commonly an ECG (52.2% of trials). BNP or troponin assessments were not reported in any trial. Adverse CV events were reported in 82.6% of trials, the most frequently reported event was hypertension (43.5% of trials). CV Safety assessments were more frequently reported in pharmaceutically funded trials compared to non-pharmaceutically funded trials (86.4 vs 10.0%, p < 0.001), and in trials administering targeted/immunotherapy agents compared to those administering only traditional chemotherapy/hormonal therapy (85.2 vs 21.1%, p < 0.001). Conclusions: Later stage breast cancer clinical trials are conducted in populations under-representative of patients with prevalent CVD, a common comorbidity. Trial funding source and type of agent administered impact whether or not CV toxicity is monitored. Despite the frequent exclusion of patients with CVD and poor safety assessment reporting, adverse CV events are frequently encountered and reported. Thus, these trials may underestimate CV toxicity risks for the greater population receiving such agents in clinical practice.