Development of novel methods to enable robust comparison of real-world Progression Free Survival (rwPFS) and Clinical Trial PFS in Multiple Myeloma

点击翻译按钮获取中文摘要

Project Summary In multiple myeloma (MM), real-world evidence (RWE) has been successfully used in the post-approval setting to characterize real-world efficacy and provide important safety information. Recent legislation aimed at advancing RWE to support drug development has created opportunities for considering RWE in pre-approval regulatory decision-making.1 Despite these efforts, RWE has been unimpactful in the pre- approval MM setting. In the ABECMA memo, the FDA stated the accompanying RWE study was inconclusive because of measurement bias in real-world Progression Free Survival (rwPFS) limited comparisons to trial data.2 We propose to develop novel methods for comparing rwPFS and trial PFS in MM. The measurement bias that limits comparison of rwPFS and trial PFS can be disaggregated into misclassification bias & surveillance bias. We will address each of these biases as follows: AIM 1: Develop novel methods to correct for misclassification bias of IMWG response criteria among (a) all real-world MM patients and in (b) historically under-represented MM patients • Characterize real-world patterns of missingness in assays collected for MM disease assessment per the International Myeloma Working Group (IMWG) response criteria3 • Replicate real-world patterns of missingness in IMWG response criteria in the control arms (CA) of 3 Janssen trials & quantify the effect of this missingness on progression (miscPROG) • Test likelihood-based and weighting methods to correct misclassification bias through simulation. AIM 2: Develop novel methods to correct for surveillance bias in disease assessment among (a) all real-world MM patients and in (b) historically under-represented MM patients • Characterize the real-world cadence(s) of MM disease assessment • Replicate real-world cadence(s) in CA of 3 Janssen trials & quantify the effect on PFS (survPFS) • Develop surveillance bias correction methods through simulations AIM 3: Validate all bias correction methods developed in AIM 1 & 2 in holdout RWD and compare to the control arms from at least two other Janssen MM trials • Assess comparison of trial CA PFS to bias-corrected and trial-emulated RWD CA PFS Janssen will work closely with renowned academic methods experts, MM researchers and Flatiron Health, a private sector leader in MM RWD, to complete this research (please see ‘Consortium Agreement and Key Personnel’). We intend to make these novel methods available through peer-reviewed publications. This research is fully aligned with Janssen’s Credo and our first responsibility to serve the needs of patients.