Characterizing Heart Failure in Diabetes Trials

Study Questions:

What are the gaps in heart failure (HF) data capture within cardiovascular (CV) outcome trials of glucose-lowering therapies and strategies for improving HF characterization?


The investigators did a systematic review of large (N >1,000) published phase III/ IV CV outcome trials evaluating glucose-lowering therapies through June 2017. Data were abstracted from publications, Food and Drug Administration (FDA) Advisory Committee records, and FDA labeling documents. Descriptive analyses were performed, ranges were presented, and proportions were assessed. In circumstances where rates of baseline or incident HF for the overall study population were not provided, these rates were manually calculated from the raw trial data, when available.


The initial query yielded a total of 8,447 potentially relevant abstracts, of which 4,478 remained after removing duplicates. Based on manual screen of each of the remaining articles, 4,457 articles did not meet the systematic review eligibility criteria and were excluded. The remaining 21 articles were included in the systematic review, which included a total of 152,737 patients. Rates and definitions of baseline HF and incident HF were inconsistently provided. Baseline ejection fraction data were provided in three studies, but not specific to patients with HF. No trial reported functional class, ejection fraction, or HF therapy at time of incident HF diagnosis. HF hospitalization data were available in 15 trials, but only two included a HF-related event within the primary composite endpoint.


The authors concluded that even recently completed large CV outcome trials of novel glucose-lowering agents lack sufficient details to fully appraise treatment effects on a HF endpoint or relative safety in patients with prevalent HF.


This systematic review of 21 large CV outcome trials testing a broad range of glucose-lowering therapies highlights the limited characterization of the baseline prevalence, severity, or treatment optimization of HF, with less than a third of trials describing new-onset HF among patients without baseline HF and only two trials including HF in a primary composite endpoint. Future trials should include detailed profiling of patients with baseline HF and a rigorous assessment of downstream incident and worsening HF events using prespecified and adjudicated endpoints. This would help optimize diabetes management by better defining the safety profile of glucose-lowering therapies among the general type 2 diabetes mellitus population with respect to HF.

Clinical Topics: Anticoagulation Management, Diabetes and Cardiometabolic Disease, Heart Failure and Cardiomyopathies, Prevention, Acute Heart Failure

Keywords: ACC18, ACC Annual Scientific Session, Diabetes Mellitus, Type 2, Glucose, Heart Failure, Metabolic Syndrome X, Outcome Assessment (Health Care), Polysaccharides, Primary Prevention, Stroke Volume

< Back to Listings