Evaluation of cardiovascular status in clinical trials in breast cancer.
e14038 Background: With the expansion of treatment options in oncology, patients are now increasingly exposed to potentially cardio-injurious compounds. Nevertheless, many clinical trials do not assess the cardiac status of participants, from time of enrollment to study completion. Methods: We reviewed the landmark clinical trials of each of the single-agent chemotherapies specified by the NCCN for use in locally advanced/metastatic breast cancer. We analysed the cardiovascular exclusion criteria, baseline cardiac evaluation as well as cardiac surveillance of patients in each study. We then investigated the literature in the years following each index publication, scrutinising for evidence of cardiac toxicity. Results: 14 phase 2/3 studies were analysed. The total number of patients was 2,522. 42.8% (6/14) of the studies excluded patients with clinically-determined cardiovascular disease, 66.6% (4) of which specified normal EF as a pre-requisite for enrollment. 42.8% (6/14) of the studies evaluated the baseline cardiovascular function of participants using EKG or TTE. 35.7% (4/14) of the studies included a more rigorous monitoring that included serial EKGs and/or TTEs. New Heart Failure (HF) occurred in 1.63% (20/1226) and 0.87% (9/1024) of patients in the more- and less- rigorous trials, respectively (P value = 0.11). Subsequent follow-up studies (published a mean of 8.5 years later) disclosed significantly higher incidence of symptomatic HF (range 2.3-28.0%) with 3 (of 4) chemotherapies in the more-rigorous study group (doxorubicin, cyclophosphamide, epirubicin) and 1 (of 10) chemotherapies in the less-rigorous group (docetaxel). Conclusions: The majority of clinical studies in breast cancer do not prospectively assess cardiac status beyond a pure clinical assessment. The discrepancy between the toxicity profile shown by original studies and that detected in subsequent investigations implies that the tools currently used to evaluate the cardiovascular status of enrollees may be insufficiently sensitive in predicting the development of HF. Novel tools (circulating markers, imaging) and longer follow-up periods are needed to ascertain the cardiotoxic profile of new chemotherapies.