Perspectives
Proposal for a new clinical end point to evaluate the efficacy of drugs and devices in the treatment of chronic heart failure*

https://doi.org/10.1054/jcaf.2001.25652Get rights and content

Abstract

Clinical trials designed to evaluate the effect of drugs and devices on the symptoms and clinical status in chronic heart failure have frequently produced conflicting, inconclusive, or misleading results. These difficulties can be explained by the fact that previous studies have relied on efficacy measures that have inherent limitations and have been analyzed using statistical approaches that ignored episodes of clinical deterioration. Recognition of these pitfalls has led to the development of a new clinical composite score, which combines changes in the New York Heart Association class and the global assessment together with the information provided from the occurrence of major clinical events. Use of this score would have correctly distinguished active therapy from placebo in earlier trials and thus would have avoided some of their misleading conclusions. The new clinical composite score has been prospectively incorporated into the design of studies evaluating the efficacy of endothelin antagonists, cytokine antagonists, vasopressin antagonists, and cardiac resynchronization in the treatment of chronic heart failure. In the trials that have been completed to date, the clinical composite score has been more sensitive than conventional approaches in discerning the presence or absence of a true treatment effect.

Section snippets

Evaluation of clinical status

Some investigators have proposed that the inconsistent results of intermediate-term trials can be explained by inherent limitations of the measures used to evaluate changes in clinical status. Several different measures have been used to evaluate the functional capacity of patients with heart failure including 1) the direct assessment of symptoms, 2) exercise tolerance, 3) New York Heart Association (NYHA) functional class, 4) global assessment of progress, and 5) quality-of-life assessments.

Analyzing the effects of drugs for heart failure

Some investigators have proposed that the inconsistent results of intermediate-term trials are caused by inherent limitations in the statistical approaches that have been used to analyze data. Two areas of specific concern have been identified: analysis of missing data and analysis of efficacy data collected after a major nonfatal clinical event.

Experience in clinical trials using a conventional approach

Until recently, intermediate-term clinical trials identified a single measure of efficacy (designated as the primary end point) and the results were evaluated using a completers analysis or the last-observation-carried-forward method, which failed to appropriately account for the occurrence of major clinical events or missing data. Consequently, these trials frequently produced results that proved to be inconsistent with the biological and clinical effects of the drug established in other

Development of a clinical composite score

These examples highlight the critical flaws in our current approach to conducting intermediate-term studies in the evaluation of new drugs for the treatment of chronic heart failure. Experience to date indicates that a strategy that minimizes the importance of major clinical events in analyzing the effects of treatment on a single clinical end point is inherently biased and carries an unacceptable risk of yielding results that do not reflect the true potential of the new treatment. Recognition

Conclusions

A new clinical composite score has been developed to assess the clinical efficacy of new drugs for heart failure. Retrospective analysis of earlier trials suggests that the use of the composite score would have avoided many of the misleading conclusions of these studies. The new clinical composite score has been and is being used prospectively in many recently completed and ongoing trials, including those evaluating the efficacy of endothelin antagonists, cytokine antagonists, vasopressin

Cited by (0)

*

Reprint requests: Milton Packer, MD, Division of Circulatory Physiology, College of Physicians and Surgeons, 630 W 168th St, Columbia University, New York, NY 10032.

View full text