Applications of Multistate Model on Complex Disease Processes

A multistate platform model was developed to describe time-to-event (TTE) endpoints in an oncology trial through the following states: initial, tumor response (TR), progressive disease (PD), overall survival (OS) event (death), censor to the last evaluable tumor assessment (progression-free survival [PFS] censor), and censor to study end (OS censor), using an ordinary differential equation framework. Two types of piecewise functions were used to describe the hazards for different events. Piecewise surge functions were used for events that require tumor assessments at the scheduled study visit times (TR, PD, and PFS censor). Piecewise constant functions were used to describe hazards for events that occur evenly throughout the study (OS event and OS censor). The multistate TTE model was applied to describe TTE endpoints from a published phase III study. The piecewise surge functions well-described the observed surges of hazards/events for TR, PD, PFS, and OS occurring near scheduled tumor assessments and showed good agreement with all Kaplan-Meier curves. With the flexibility of piecewise hazard functions, the model was able to evaluate covariate effects in a time-variant fashion to better understand the temporal patterns of disease prognosis through different disease states. This model can be applied to advance the field of oncology trial design and optimization by: (1) enabling robust estimations of baseline hazards and covariate effects for multiple TTE endpoints, (2) providing a platform model for understanding the composition and correlations between different TTE endpoints, and (3) facilitating oncology trial design optimization through clinical trial simulations. 

We introduce a computationally efficient and general approach for utilizing multiple, possibly interval-censored, data streams to study complex biomedical endpoints using multistate semi-Markov models. Our motivating application is the REGEN-2069 trial, which investigated the protective efficacy (PE) of the monoclonal antibody combination REGEN-COV against SARS-CoV-2 when administered prophylactically to individuals in households at high risk of secondary transmission. Using data on symptom onset, episodic RT-qPCR sampling, and serological testing, we estimate the PE of REGEN-COV for asymptomatic infection, its effect on seroconversion following infection, and the duration of viral shedding. We find that REGEN-COV reduced the risk of asymptomatic infection and the duration of viral shedding, and led to lower rates of seroconversion among asymptomatically infected participants. Our algorithm for fitting semi-Markov models to interval-censored data employs a Monte Carlo expectation maximization (MCEM) algorithm combined with importance sampling to efficiently address the intractability of the marginal likelihood when data are intermittently observed. Our algorithm provide substantial computational improvements over existing methods and allows us to fit semi-parametric models despite complex coarsening of the data. 

Upon completion of a phase II clinical trial, critical decision needs to be made regarding the viability of advancing to phase III development. In early phase oncology trials for NSCLC, the overall response rate (ORR) is commonly employed to assess the likelihood of success. However, the relationship between ORR and the long-term endpoint overall survival (OS) remains uncertain, rendering ORR an inadequate surrogate for OS. Moreover, OS data collected during phase II trials typically lack maturity due to limited follow-up time. In this project, we explore the use of multistate models for prediction of long-term OS based on the available phase II tumor shrinkage/response data and information borrowed from historical/external control arm. The multistate models account for different patient states. We illustrate the method via data from JDQ443 KontRASt-01 trial and historical control-docetaxel arm of CANOPY-2 trial for non-small cell lung cancer (NSCLC) and show how the median OS and the hazard ratio can be effectively predicted and ultimately prove how instrumental is the method in informing the go/no-go decision for future phase III trials. In the meantime, analysis based on simulation also shows favorable operating characteristics of the method and suggests that the method would lead to efficient early go/no-go decision in the presence of limited follow-up and small sample size.  

Clinical trials are costly and require significant commitment. Maximizing the use of collected data is crucial so we can learn as much as possible. A multistate model describes longitudinal events, allowing multiple clinical endpoints to be treated as outcomes and covariates to be estimated simultaneously. Proportional hazards models are fitted for each transition, enabling calculations of absolute risks, probability of being in a state, expected number of visits to a state, and time spent in a state. For this presentation, we use a publicly available clinical trial dataset, myeloid, from the survival package in R to demonstrate the application of multistate hazards models. In the myeloid dataset, treatment B results in a longer duration of complete response (CR) compared to treatment A. Mutation status does not affect the rate of transition to CR but significantly influences the duration of CR. We also found that more patients in treatment A received transplants without achieving complete response (CR); in contrast, more patients in treatment B received transplants after achieving CR. Additionally, the mutation status significantly influences the transition rate from CR to transplant, whereas the treatment has no impact on this particular transition. Our observations on these datasets were made possible by multistate models. Clinical trial data offers much more than a simple yes/no answer if we, the statisticians, are willing to explore it.