Chiara Zecchin 1, Anna Largajolli 2, Klervi Golhen 3, Nicholas Medjeral-Thomas 1, Sinziana Cristea 4, Matt Zierhut 5, Monica Simeoni 6
1 GSK (Stevenage, ), 2 Certara (Treviso, Italy), 3 GSK (Basel, Switzerland), 4 Certara (Amsterdam, Netherland), 5 Certara (San Diego, USA), 6 GSK (London, United Kingdom)
Introduction/ Objectives: Chronic kidney disease (CKD) represents a significant unmet medical need, with diverse aetiologies and disease severities requiring tailored treatments. In developing new therapies for CKD, a central challenge is balancing the use of rapid but variable short-term endpoints like urinary albumin-to-creatinine ratio (UACR) with registrational slow-to-develop endpoints e.g. end-stage kidney disease, significant decline in estimated Glomerular Filtration Rate (eGFR), or across the diverse composite clinical endpoints reported in the trials, e.g. including kidney failure, sustained 40% or 50% reduction in eGFR, or death from kidney or cardiovascular causes [1].
Model-Based Meta-Analysis (MBMA) is a quantitative framework that integrates aggregated data from multiple randomized controlled trials (RCTs) using pharmacologically and statistically structured models [2].
This study leverages the CODEx [3] CKD Database and MBMA to explore the correlation between early surrogate endpoints, such as UACR, measured at 6 months, and long-term composite clinical outcomes measured at 1 and 2 years. The goal is to provide insights to inform clinical trial design, optimize drug development, and refine patient stratification for CKD and its subtypes.
Methods: The database [3] was developed using systematic methods outlined in the Cochrane Handbook and PRISMA guidelines. Safety and efficacy average level data from RCTs investigating interventions such as angiotensin-converting-enzyme inhibitors, angiotension II receptor inhibitors, SGLT2 inhibitors, and endothelin receptor antagonists were extracted.
The methodology to bridge across different clinical endpoints definitions was investigated.
A MBMA assessed the relationship between UACR changes at early timepoints (6 months) and composite clinical endpoints at 1 and 2 years. The model was developed to account for inter-study heterogeneity and variability in CKD etiology and severity.
Results: All renal function related endpoints available in the database (i.e. UPCR, urine protein excretion over 24 hours, urine albumin excretion over 24 hours) were converted to UACR after adjusting for different endpoints units. The conversion used the correction factor of 0.6 as described in [4]. Geometric mean ratio of the change from baseline of active treatment vs control was calculated for all studies available. With respect to the clinical endpoint, the hazard ratio of active vs control was calculated for each study and the different composite endpoints were assumed comparable after this transformation. Exploratory analyses suggest that reductions in UACR at 6 months correlates with improved composite clinical outcomes at 1 and 2 years. These findings support the use of UACR as intermediate markers to support futility, and further work detailing and quantifying the translational relationship can enable more robust adaptive trial designs and enable Go/No-Go decisions in interim analyses with smaller sample sizes. This analysis includes data from all phases of clinical development, thus addressing limitations of prior studies that focussed uniquely on Ph3 [4].
Conclusions: The integration of CODEx CKD Database and MBMA offers a promising framework to explore the correlation between surrogate short-term endpoints and clinical long-term endpoints in CKD drug development. Insights into the relationship between UACR and clinical outcome, including heterogeneity in CKD progression, variability across patient populations and between Ph2 and Ph3 trials could enhance clinical trial design, improve efficiency, and accelerate development of targeted therapies for CKD and its subtypes. Future work will focus on characterizing the time course of UACR and clinical endpoints and their variability across CKD subtypes, aetiology and standard of care.
References:
[1] Levey AS, Gansevoort RT, Coresh J, Inker LA, Heerspink HL, Grams ME, Greene T, Tighiouart H, Matsushita K, Ballew SH, Sang Y, Vonesh E, Ying J, Manley T, de Zeeuw D, Eckardt KU, Levin A, Perkovic V, Zhang L, Willis K. Change in Albuminuria and GFR as End Points for Clinical Trials in Early Stages of CKD: A Scientific Workshop Sponsored by the National Kidney Foundation in Collaboration With the US Food and Drug Administration and European Medicines Agency. Am J Kidney Dis. 2020 Jan;75(1):84-104.
[2] Mandema JW, Gibbs M, Boyd R, Wada D, Pfister M. Model-based meta-analysis for comparative efficacy and safety: application in drug development and beyond. Clin Pharmacol Ther. 2011;90(6):766-769.
[3] Certara. CODEX: AI-powered analytics. Published 2025. Available at: https://www.certara.com/codex-clinical-trial-outcomes-databases/
[4] Heerspink HJL, Greene T, Tighiouart H, Gansevoort RT, Coresh J, Simon AL, Chan TM, Hou FF, Lewis JB, Locatelli F, Praga M, Schena FP, Levey AS, Inker LA; Chronic Kidney Disease Epidemiology Collaboration. Change in albuminuria as a surrogate endpoint for progression of kidney disease: a meta-analysis of treatment effects in randomised clinical trials. Lancet Diabetes Endocrinol. 2019 Feb;7(2):128-139. doi: 10.1016/S2213-8587(18)30314-0. Epub 2019 Jan 8. PMID: 30635226.
Reference: PAGE 34 (2026) Abstr 11981 [www.page-meeting.org/?abstract=11981]
Poster: Drug/Disease Modelling - Other Topics