The Global Kidney Patient Trials Network and the CAPTIVATE Platform Clinical Trial Design: A Trial Protocol

Abstract

Importance: Chronic kidney disease (CKD) is a global health priority affecting almost 1 billion people. New therapeutic options and clinical trial innovations such as adaptive platform trials provide an opportunity to efficiently test combination therapies.

Objective: To describe the design and baseline results of the Global Kidney Patient Trials Network (GKPTN) and the design and structure of the global adaptive platform clinical trial Chronic Kidney Disease Adaptive Platform Trial Investigating Various Agents for Therapeutic Effect (CAPTIVATE) to find new therapeutic options and treatments for people with kidney disease.

Design, setting, and participants: The GKPTN is a multicenter registry that started in May 2020 and is ongoing, while CAPTIVATE is a multicenter, multifactorial, phase 3, placebo-controlled adaptive platform randomized clinical trial that includes patients with CKD. The first participant was randomized in September 2024. The GKPTN recruits patients from kidney and endocrinology practices, and CAPTIVATE aims to recruit patients from GKPTN sites where possible. Both the GKPTN and CAPTIVATE recruit patients with nondialysis CKD.

Intervention: CAPTIVATE will test several investigational agents or combinations of agents, beginning with a mineralocorticoid receptor antagonist.

Main outcomes and measures: The GKPTN monitors clinical characteristics, treatment, and outcomes to identify eligible clinical trial participants and provide a contemporary global picture of patients with CKD. The primary outcome of CAPTIVATE is to identify investigational agents or combinations of agents to reduce the rate of chronic estimated glomerular filtration rate (eGFR) decline. The default maximum sample size per treatment arm in each domain, based on bayesian simulations, is 500 participants, providing approximately 90% power to detect a clinically meaningful improvement of 2.6 mL/min/1.73 m2 in eGFR at the end of the 104-week study period.

Results: The GKPTN has enrolled 4334 patients across 119 sites in 8 countries (US, Australia, Argentina, China, Italy, Canada, Spain, and Japan). The mean (SD) participant age at enrollment was 64.5 (16.2) years, 2542 participants (58.7%) were female, and diabetic kidney disease was most frequently reported among patients for CKD etiology (1875 [43.3%]). Among the participants, the mean (SD) eGFR was 52.9 (29.3) mL/min/1.73 m2, and the median urinary albumin-to-creatinine ratio was 89 mg/g (coefficient of variation, 20-420 mg/g). In the GKPTN cohort, the mean eGFR decline was steeper among participants with a baseline eGFR of 60 mL/min/1.73 m2 or more (-2.29 [95% CI, -3.14 to -1.44]) compared with those with an eGFR of less than 60 mL/min/1.73 m2 (-1.16 [95% CI, -1.77 to -1.44]) and was progressively steeper in more severe albuminuria subgroups.

Conclusions and relevance: The GKPTN registry and the CAPTIVATE trial have the potential to expand and optimize therapeutic options for people with CKD using an adaptive platform clinical trial design.

Trial registration: ClinicalTrials.gov Identifiers: NCT04389827 and NCT06058585.

Figure 1.. The Chronic Kidney Disease Adaptive Platform Trial Investigating Various Agents for Therapeutic Effect (CAPTIVATE) Study Design

DSA indicates domain-specific appendix; MRA, mineralocorticoid receptor antagonist.

Figure 2.. Interim Analyses for 1 Domain Within the CAPTIVATE Platform After 18 and 24 Months

Interim analyses for a 2-arm domain with the default-stopping rules for a single simulated example trial are shown. A-C, Interim analysis 1 is performed 18 months after the domain is open. A total of 745 participants have been randomized within the domain, and at least 100 participants per treatment arm have completed the 6-month visit. The intervention is assessed for futility based on the urinary albumin-to-creatinine ratio (UACR) end point. The treatment effect is estimated to reduce UACR by 31.2% (95% CI, 21.9%-39.3%), and the probability of providing a clinically important 25% reduction is 0.909; therefore, futility is not declared for the intervention, and the trial continues to recruit. While there is some preliminary data suggesting that the treatment also reduces the eGFR slope (difference in eGFR slope, 1.11 [95% CI, −1.73 to 3.95] mL/min/1.73 m²), there is insufficient follow-up to assess either success or futility on this end point. The probability of a slope effect greater than 0 mL/min/1.73 m² per year is 0.779; the probability of a slope effect greater than 0.8 mL/min/1.73 m² per year is 0.586. D-F, Interim analysis 2 is performed at 24 months. Enrollment is complete, and the domain has reached the maximum sample size (N = 1000). The treatment continues to show a strong benefit on the UACR and does not meet the futility criteria. The treatment effect is estimated to reduce UACR by 28.4% (95% CI, 21.2%-35.0%), and the probability of providing a clinically important 25% reduction is 0.830. The intervention is eligible for futility stopping on the eGFR end point, but the probability of providing a clinically important 0.8-mL/min/1.73 m² per year difference compared with placebo in the eGFR slope is 0.778; therefore, the intervention is not stopped for futility, and the trial continues to recruit. The difference in the eGFR slope is 1.38 (95% CI, −0.11 to 2.88) mL/min/1.73 m². The probability of a slope effect greater than 0 mL/min/1.73 m² per year is 0.965. Error bars indicate 95% CIs. In panels B and E, horizontal lines indicate the median change in UACR from baseline, and boxes indicate the IQR; the dots indicate outliers. CFB indicates change from baseline.

Figure 3.. Interim Analyses for 1 Domain Within the CAPTIVATE Platform After 30 and 36 Months

Two interim analyses after 30 and 36 months’ follow-up for domain-specific appendices, in which follow-up continues after the 24-month interim analysis described in Figure 2, are shown. A-C, For interim analysis 3, which occurs after 30 months’ follow-up, a total of 1000 participants have been randomized, and the treatment continues to show a strong benefit on the urinary albumin-to-creatinine ratio (UACR) (UACR reduction, 29.1% [95% CI, 22.7%-34.9%]; probability of providing a clinically important 25% reduction, 0.899) and the eGFR slope (1.32 [95% CI, 0.24-2.39] mL/min/1.73 m² per year). While the posterior probability of superiority on the eGFR is above the 0.985 threshold (probability of slope effect >0 mL/min/1.73 m² per year, 0.992), the number of participants assigned to the intervention with a complete 108-week follow-up is less than 100; thus, the intervention is not eligible to stop for success, and the trial continues. The probability of a slope effect greater than 0.8 mL/min/1.73 m² per year is 0.826. D-F, For interim analysis 4, which occurs after 36 months’ follow-up, a total of 1000 participants have been randomized, and there is now sufficient follow-up. The intervention is eligible for both futility and success stopping since more than 100 patients have completed 108 weeks’ follow-up. The treatment effect is estimated to reduce UACR by 29.1% (95% CI, 22.7%-34.9%), and the probability of providing a clinically important 25% reduction is 0.899. The treatment effect on the chronic eGFR slope is 1.08 (95% CI, 0.23-1.93) mL/min/1.73 m² per year. The posterior probability of a slope effect greater than 0 mL/min/1.73 m² per year is 0.994 and meets the requirement to stop the intervention for success. The probability of a slope effect greater than 0.8 mL/min/1.73 m² per year is 0.740. At this time, the domain is closed, and all patients discontinue the intervention and attend the end-of-treatment visit. Error bars indicate 95% CIs. In panels B and E, horizontal lines indicate the median change in UACR from baseline, and boxes indicate the IQR; the dots indicate outliers. CFB indicates change from baseline.