Straight-to-Phase 3: MIDD accelerates depemokimab clinical development in Type 2 inflammatory conditions
Stein Schalkwijk1, Chiara Zecchin2, Isabelle Pouliquen3,4, Alienor Berges5, Daren Austin6
1GSK, 2GSK, 3GSK, 4RD5 Consulting Limited, 5GSK, 6GSK
Objectives: We describe the accelerated development of depemokimab, the first ultra-long-acting anti-IL-5 biologic for Type 2 inflammatory conditions, by transitioning from Phase 1 directly to Phase 3 studies using Model-informed Drug Development (MIDD). Methods: IL-5 is a central cytokine in Type 2 inflammation, with multidirectional effects on eosinophils along with multiple immune and structural cells. Blood eosinophil count (BEC) is considered a predictive pharmacodynamic marker for therapeutics that target this cytokine (e.g., mepolizumab dosed 100mg Q4W in asthma). A Phase 1 study [1] in participants with mild-to-moderate asthma and BEC =200 cells µL^-1 was designed to support a depemokimab Straight-to-Phase 3 strategy allowing identification of a suitable Phase 3 dosing regimen with precision and high confidence. MIDD was integrated into the GSK Quantitative Decision-making (QDM) framework with go/no-go criteria, based on Phase 1 results and precedented aIL-5 target pharmacology. Using QDM, for any proposed dosing regimen, depemokimab would transition to Phase 3 asthma studies if the probability that BEC reductions exceeded a minimum value (MV) was >80%, and not transition if the probability of exceeding a target value (TV) was <10%. MV and TV were set by two positive mepolizumab Phase 3 studies in asthma: MENSA [2] and MUSCA [3]. The MIDD strategy included Bayesian non-linear mixed effects dose-time response analysis, with subsequent population PKPD analyses for other eosinophilic indications. Clinical trial simulations were also conducted to evaluate the performance of conventional Phase 2b efficacy approaches to identify Phase 3 dose regimens relative to the Pharmacology-based MIDD approach. Results: Using Phase 1 results, MIDD predicted =80% probability of achieving MV with a single dose =60 mg; 100 mg Q6M was selected as robust dosing regimen. For CRSwNP, and EGPA and HES indications, simulations recommended 100 mg and 200 mg Q6M, respectively, to satisfy pharmacodynamic criteria in each indication. Clinical trial simulations demonstrated that a traditional Phase 2b study with an efficacy endpoint was unlikely (<3% probability) to provide a more robust and precise estimate of the Phase 3 dosing regimen and highlighted sample size inadequacies. Altogether, the use of MIDD based on Phase 1 results, and the well-established target pharmacology, safety and efficacy of the aIL-5 treatments provided a strong rationale to transition depemokimab directly to confirmatory Phase 3 programs in asthma, CRSwNP, HES and EGPA. Consultations with Health Authorities (HA) were held prior to Phase 3 start. Depemokimab registrational Phase 3 studies in asthma and CRSwNP completed in 2024 [4,5]. Pharmacodynamic results were aligned with MIDD predictions. Importantly, all primary efficacy endpoints were met, validating the Pharmacology-based approach. Depemokimab marketing applications in asthma and CRSwNP are currently under review by HAs in all major markets. Conclusions: The QDM/MIDD approach accelerated depemokimab clinical development timelines by approximately 2-3 years. Key to the success of the approach was the demonstration of the inadequacies of conventional dose-finding studies for severe asthma, against a background of precedented mechanism of action, efficacy, and a large safety database. This combination offered ideal conditions for acceleration. Notwithstanding, late-stage development timelines remain long, and institutional memory was key to successful implementation and execution. Various HAs may have different views and requirements regarding QDM/MIDD strategy employed by a sponsor, hence dedicated Pharmacometric and statistical review is essential. In the absence of Phase 2b and/or repeat dose studies, it may be challenging to interpret unexpected results and delineate their cause. These risks need to be flagged and mitigated or accepted. Also, in the absence of Phase 2 data, Phase 3 recruitment can be more challenging, and the potential impact on timelines needs to be weighed against the time saved by skipping clinical studies. This case study demonstrates how MIDD can streamline drug development, bring new medicine to patients faster, and improve R&D productivity. If similar conditions present in the future, it serves as a framework for the development of new therapeutics. Funding: GSK (NCT04719832, NCT04718103, NCT05274750, NCT05281523, NCT03287310)