Methods
Study Design
The SUPPORT trial was a prospective, randomized, open-label blinded endpoint study, which was conducted according to the ethical principles described in Declaration of Helsinki. The study protocol was approved by the institutional ethics committees of the 17 participating institutions in the Tohoku District of Japan (Appendix). The primary objective of the study was to examine whether an additive treatment with an ARB, olmesartan, reduces the mortality and morbidity of hypertensive patients with stable CHF (NCT00417222). The inclusion criteria were designed to enroll hypertensive patients aged 20–79 with symptomatic CHF but in stable condition and were treated with ACE inhibitors and/or β-blockers, while the exclusion criteria were designed to exclude patients with substantive confounding medical conditions or an inability to meaningfully participate in the SUPPORT trial ( Table 1 ). Finally, a total of 1147 symptomatic CHF patients with a history of hypertension who met the inclusion and exclusion criteria and gave written informed consent for the trial were assigned to either the olmesartan or the control group according to a 1 : 1 ratio of olmesartan to control, through stratification by participating institute, sex and age between October 2006 and March 2010. The patients were followed up until the study ended on 31 March 2013 (Figure 1). If contact could not be made at the end of the study, data of these patients were censored at the date when they were known to be alive last. Olmesartan was initiated at a dose of 5–10 mg/day, and then up titrated to 40 mg/day, if tolerable, in the olmesartan group, while no ARB use was allowed in the control group. The diagnosis of HF was made based on the criteria of the Framingham study by an attending physician at each hospital. All physicians were encouraged to control blood pressure of the patients in each group according to the recommendations in the JNC7. The primary endpoint was the composite of all-cause death, non-fatal acute myocardial infarction, non-fatal stroke, and worsening HF requiring hospitalization, while secondary endpoints consisted of the modes of death, hospitalization for cardiovascular reasons, surrogate markers for HF and development of cardiovascular disease, atrial fibrillation, diabetes, and renal dysfunction (see Supplementary material online, Table S1http://eurheartj.oxfordjournals.org/content/suppl/2015/01/26/ehu504.DC1).
(Enlarge Image)
Figure 1.
Trial profile.
Interim Analysis and Data Monitoring
Meetings of the Data Safety Monitoring Board, which is independent of the study committees, were held twice a year throughout the trial period to monitor the study safety until the end of the trial. Interim analyses were conducted at the end of and 2 years after the end of the registration period, namely March 2010 and September 2011, to evaluate the primary endpoint and the safety for the continuation of the trial.
Statistical Analysis
All analyses were performed according to a predefined statistical analysis plan. Y.S., S.M., and H.S. had a full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of data analysis. The primary and secondary endpoints were analysed based on the time to the first occurrence, according to the intention-to-treat principle, including all patients of lost to follow-up censored at the day of the last contact. Survival curves were estimated using the Kaplan–Meier procedure and compared with a two-sided log-rank test. Based on the results of the CHART-1 Study, we assumed that the annual incidence of the primary composite endpoint would be ~12% in the SUPPORT trial and thus 480 patients would be required on the condition of 3-year follow-up for each arm to provide 80% power to detect 30% risk reduction by olmesartan, using a two-sided significance level of 0.05 by the log-rank test. Considering that ~15% of cases would be lost during follow-up or unsuitable for analysis, we estimated that >565 patients in each group (control and olmesartan) would be needed to complete the trial.
The P-value of the interim analysis was 0.005 while the P-value of the final analysis was 0.048. The effects of olmesartan were examined using Cox proportional hazards models. Subgroup analyses were performed according to baseline medication and other baseline characteristics and clinical parameters. Continuous variables are presented as means ± standard deviations except brain natriuretic peptide (BNP). Brain natriuretic peptide levels are presented as medians and interquartile ranges. Categorical variables were presented as numerals and percentages. Group comparisons were made with the Mann–Whitney test for continuous variables, and the χ-test without continuity correction for categorical variables. All statistical analyses were performed using IBM SPSS Statistics 21.0 (IBM, Somers, NY, USA) and R 3.0.2 (R Foundation for Statistical Computing, Vienna. http://www.R-project.org/). A two-sided probability values of <0.05 and P-values for interaction <0.1 were considered to be statistically significant.