Postponement of Death by Pharmacological Heart Failure Treatment: A Meta-Analysis of Randomized Clinical Trials

Abstract

BACKGROUND: Outcome postponement has been proposed as an effect measure for preventive drug treatment. It describes the average delay of the investigated unwanted clinical event, achieved by taking medication. The objective was to estimate postponement of death for the following heart failure medications compared to placebo: beta-blockers, angiotensin-converting enzyme (ACE) inhibitors, angiotensin II receptor blockers (ARBs), ARB added to ACE inhibitors, aldosterone antagonists, ivabradine, and renin antagonists. METHODS: We searched Medline and Embase from inception of databases until October 2017. Eligibility criteria were randomized placebo-controlled heart failure trials, including at least 1000 participants, with survival as a prespecified outcome and a minimum trial duration of 1 year. We calculated the outcome postponement by modeling the area between survival curves. This area was modeled on the basis of the hazard ratio or relative risk, the rate of mortality in the placebo group, and the trial duration. All results were standardized to a 3-year trial duration to ensure comparability between treatments. RESULTS: We identified 14 eligible trials, with a total of 52,014 patients. The results in terms of postponement of all-cause mortality was: beta-blockers 43.7 days (95% confidence interval [95% CI], 20.8-66.5), ACE inhibitors 41.0 days (95% CI, 14.3,68.4). CONCLUSION: The modeled outcome postponement estimates reiterate beta-blockers, ACE inhibitors, and aldosterone antagonists as the mainstay of heart failure treatment. Furthermore, ivabradine or ARBs added to ACE inhibitors results in no statistically significant gain in survival.

Improving clinical trial efficiency using a machine learning‐based risk score to enrich study populations

“…As the cumulative impact of these diverse factors cannot be precisely quantified, the differences in relative mortality risk between community‐ and registry‐based cohorts and clinical trial‐based populations should be considered with these limitations in mind. There are, however, numerous reports in the medical literature that support our finding of a substantial difference in mortality rate between community‐based and registry populations and patients included in clinical trials 1–6,21–25,26 …”

“…There are, however, numerous reports in the medical literature that support our finding of a substantial difference in mortality rate between community-based and registry populations and patients included in clinical trials. [1][2][3][4][5][6][21][22][23][24][25]26 . Although mortality rates differed between patients with HFrEF and HFpEF in the clinical trials included in the present analysis, similar differences are not seen in community-based populations, 31,32 including the UCSD cohort.…”

“…Virtually all large‐scale phase 3 clinical trials in HF include mortality as an efficacy endpoint and many event‐driven trials compute sample size based on the anticipated reduction in death with the investigational therapy. Yet, partly by design which excludes individuals at highest risk, patients enrolled in clinical trials experience substantially lower rates of mortality when compared to community‐based HF cohorts 1–6 . Despite efforts to enrich clinical trial populations with high‐risk features, mortality remains relatively low and may thus limit the ability of a trial to detect a meaningful treatment effect on this important endpoint.…”

Consideration Regarding the Analysis of Randomized Controlled Trials in the Era of Evidence-based Medicine

“…When added to ACE-I, ivabradine did not improve survival. 35 Moreover, the marked decrease in heart rate in patients assigned to ivabradine may have broken the blind in the SHIFT trial. Regarding safety, the 2015-French Commission for Transparency of the Haute Autorité de Santé advised against ivabradine in the CAD treatment.…”

“…HF therapies, such as BARB or CRT, targeted at LV remodeling and dysfunction produce the most benefit in terms of life expectancy. 35 A longer duration of RCTs for interventions that do not improve LV function may resolve this important issue.…”