Comparing Analytical Methods for Composite Endpoints in Clinical Trials: Insights from the Vericiguat Global Study in Subjects with Heart Failure with Reduced Ejection Fraction (VICTORIA) Trial.
J Card Fail
; 2024 Aug 23.
Article
en En
| MEDLINE
| ID: mdl-39182825
ABSTRACT
BACKGROUND AND AIMS:
In the VICTORIA trial of participants with heart failure (HF) and reduced ejection fraction, vericiguat (V) reduced the primary composite outcome [time to first HF hospitalization (HFH) or cardiovascular death (CVD)] (897 events) compared to placebo (P) (972 events) (hazard ratio, 0.90; 95% confidence interval [CI], 0.82-0.98; p=0.02). In this prespecified secondary analysis, we applied the weighted composite endpoint (WCE) and the win ratio (WR) methods to provide complementary assessments of treatment effect. METHODS ANDRESULTS:
The WCE method estimated the mean HFH-adjusted survival based on prespecified weights from a Delphi panel of the VICTORIA executive committee and national leaders mild (weight per event 0.39), moderate (0.5) or severe (0.67) HFH, and CVD (1.0). The unmatched WR was estimated for the descending hierarchy of CVD, then recurrent HFH. The WCE used all 3412 primary clinical events 875(V416/P459) severe HFH, 1614(767/847) moderate HFH and 68(38/30) mild HFH, 855(414/441) CVD. Improved HFH-adjusted survival occurred with vericiguat [mean 78.2% vs. 75.6%; difference (95% CI) 2.4% (1.7%-3.2%); p<0.0001]. Based on the comparison of 6,375,624 pairs, the WR of 1.13 (95% CI, 1.03-1.24, p=0.01) also indicated improved clinical outcomes with vericiguat.CONCLUSIONS:
The results of the WCE and WR methods were consistent with the primary analysis of the time to first HFH or CVD. Whereas both WCE and WR assessed recurrent events, the WCE allowed inclusion of all recurrent events, insights on the severity of HFH events, and an absolute measure of the participant-treatment experience. This approach complements conventional assessment, better informing consumers of new therapeutics and future trial designs.
Texto completo:
1
Colección:
01-internacional
Base de datos:
MEDLINE
Idioma:
En
Revista:
J Card Fail
Asunto de la revista:
CARDIOLOGIA
Año:
2024
Tipo del documento:
Article
País de afiliación:
Canadá
Pais de publicación:
Estados Unidos