BAYESIAN STATISTICAL METHODS FOR ADAPTIVE BIOSIMILARITY CLINICAL TRIALS AND JOINT MODELS

Many methods exist to jointly model either recurrent and related terminal survival events or longitudinal outcome measures and related terminal survival event. However, few methods exist which can account for the dependency between all three outcomes of interest, and none allow for the modeling of all three outcomes without strong correlation assumptions. We propose a joint model which uses subject-specific random effects to connect the survival model (terminal and recurrent events) with a longitudinal outcome model. In the proposed method, proportional hazards models with shared frailties are used to model dependence between the recurrent and terminal events, while a separate (but correlated) set of random effects are utilized in a generalized linear mixed model to model dependence with longitudinal outcome measures. All random effects are related based on an assumed multivariate normal distribution. The proposed joint modeling approach allows for flexible models, particularly for unique longitudinal trajectories, that can be utilized in a wide range of health applications. We evaluate the model through simulation studies as well as through an application to data from the Atherosclerosis Risk in Communities (ARIC) study. Separately, we consider approaches for clinical trials for biosimilars. Biosimilars are biological products with no clinically meaningful difference in safety, purity, and potency when compared to an approved biologic. Biosimilars are interchangeable when the biosimilar has the same expected risk, in terms of safety and efficacy, when compared to the reference biologic. The FDA regards biosimilarity and interchangeability approval based on totality of evidence approaches. The nature of biosimilars, and their comparison in clinical trials to reference products (RP), leads to a natural utilization of historical information on the RP in the elicitation of prior distributions as well as the efficient utilization of study participants in an adaptive clinical trial for both biosimilarity and interchangeability designations. We thus propose a two-stage clinical trial. Stage 1 consists of a 2-arm randomized clinical trial with clinical efficacy endpoint, utilizing an informative robust Meta-Analytic-Predictive (MAP) prior on the reference product arm estimated with historical information on the RP, allowing for a reduction in the RP arm. Stage 2 consists of a 2-arm randomized switching study, where participants with clinical success from the RP arm of Stage 1 are carried forward to determine interchangeability. We similarly utilize an informative robust MAP prior on the RP arm with available PK data. We demonstrate the methodology for the design and analysis of a biosimilar clinical program through simulation. We consider the Rheumatoid Arthritis clinical space, as might be feasible for a biosimilar to adalimumab. Additionally, we will consider a two-stage adaptive clinical trial for biosimilarity and interchangeability involving multiple therapeutic indications. We will utilize a correlated mixture prior to induce information sharing across both indications and trial stages, while also incorporating informative robust MAP priors based on historical data of the reference product. We demonstrate the methodology for the design and analysis of a biosimilar clinical program through simulation and consider a trial emulation as might be feasible for a biosimilar to adalimumab.