Marie Davidian

I-SPY2 is a long-running phase 2 platform trial that evaluates neoadjuvant treatments for locally advanced breast cancer to identify those with high efficacy that are likely to be successful in phase 3 trials, assigning patients to novel agents using response-adaptive randomization (RAR). Recently, I-SPY2 was reconfigured as a sequential multiple assignment randomized trial (SMART), with up to three stages of therapy. At the first stage, a patient is assigned to a tumor-subtype-specific therapy. If the patient fails to show a satisfactory response, the patient is assigned to a second subtype-specific therapy, and receives a third, rescue therapy if response is still not achieved. The I-SPY2 SMART thus supports identification of highly efficacious entire treatment regimes. The transition of I-SPY2 to a SMART required development of a RAR scheme that updates randomization probabilities at each stage, aligned with the goal of maximizing the number of patients who achieve a pathological complete response (pCR). We present our Bayesian RAR approach, which updates randomization probabilities based on the posterior probability that treatments are part of the optimal regime. Empirical studies demonstrate that it results in more patients having treatment experience consistent with highly efficacious regimes, improves overall within-trial pCR rates, and identifies optimal regimes post trial at rates similar to or exceeding those under simple, uniform, nonadaptive randomization.

Widely used methods and software for group sequential tests of a null hypothesis of no treatment difference that allow for early stopping of a clinical trial depend primarily on the fact that sequentially‐computed test statistics have the independent increments property. However, there are many practical situations where the sequentially‐computed test statistics do not possess this property. Key examples are in trials where the primary outcome is a time to an event but where the assumption of proportional hazards is likely violated, motivating consideration of treatment effects such as the difference in restricted mean survival time or the use of approaches that are alternatives to the familiar logrank test, in which case the associated test statistics may not possess independent increments. We show that, regardless of the covariance structure of sequentially‐computed test statistics, one can always derive linear combinations of these test statistics sequentially that do have the independent increments property. We also describe how to best choose these linear combinations to target specific alternative hypotheses, such as proportional or non‐proportional hazards or log odds alternatives. We thus derive new, sequentially‐computed test statistics that not only have the independent increments property, supporting straightforward use of existing methods and software, but that also have greater power against target alternative hypotheses than do procedures based on the original test statistics, regardless of whether or not the original statistics have the independent increments property. We illustrate with two examples.

Response-adaptive randomization (RAR) has been studied extensively in conventional, single-stage clinical trials, where it has been shown to yield ethical and statistical benefits, especially in trials with many treatment arms. However, RAR and its potential benefits are understudied in sequential multiple assignment randomized trials (SMARTs), which are the gold-standard trial design for evaluation of multi-stage treatment regimes. We propose a suite of RAR algorithms for SMARTs based on Thompson Sampling (TS), a widely used RAR method in single-stage trials in which treatment randomization probabilities are aligned with the estimated probability that the treatment is optimal. We focus on two common objectives in SMARTs: (1) comparison of the regimes embedded in the trial and (2) estimation of an optimal embedded regime. We develop valid post-study inferential procedures for treatment regimes under the proposed algorithms. This is nontrivial, as even in single-stage settings standard estimators of an average treatment effect can have nonnormal asymptotic behavior under RAR. Our algorithms are the first for RAR in multi-stage trials that account for non-standard limiting behavior due to RAR. Empirical studies based on real-world SMARTs show that TS can improve in-trial subject outcomes without sacrificing efficiency for post-trial comparisons.

The sequential multiple assignment randomized trial (SMART) is the ideal study design for the evaluation of multistage treatment regimes, which comprise sequential decision rules that recommend treatments for a patient at each of a series of decision points based on their evolving characteristics. A common goal is to compare the set of so-called embedded regimes represented in the design on the basis of a primary outcome of interest. In the study of chronic diseases and disorders, this outcome is often a time to an event, and a goal is to compare the distributions of the time-to-event outcome associated with each regime in the set. We present a general statistical framework in which we develop a logrank-type test for comparison of the survival distributions associated with regimes within a specified set based on the data from a SMART with an arbitrary number of stages that allows incorporation of covariate information to enhance efficiency and can also be used with data from an observational study. The framework provides clarification of the assumptions required to yield a principled test procedure, and the proposed test subsumes or offers an improved alternative to existing methods. We demonstrate performance of the methods in a suite of simulation studies. The methods are applied to a SMART in patients with acute promyelocytic leukemia.

Response-adaptive randomization (RAR) has been studied extensively in conventional, single-stage clinical trials, where it has been shown to yield ethical and statistical benefits, especially in trials with many treatment arms. However, RAR and its potential benefits are understudied in sequential multiple assignment randomized trials (SMARTs), which are the gold-standard trial design for evaluation of multi-stage treatment regimes. We propose a suite of RAR algorithms for SMARTs based on Thompson Sampling (TS), a widely used RAR method in single-stage trials in which treatment randomization probabilities are aligned with the estimated probability that the treatment is optimal. We focus on two common objectives in SMARTs: (i) comparison of the regimes embedded in the trial, and (ii) estimation of an optimal embedded regime. We develop valid post-study inferential procedures for treatment regimes under the proposed algorithms. This is nontrivial, as (even in single-stage settings) RAR can lead to nonnormal limiting distributions of estimators. Our algorithms are the first for RAR in multi-stage trials that account for nonregularity in the estimand. Empirical studies based on real-world SMARTs show that TS can improve in-trial subject outcomes without sacrificing efficiency for post-trial comparisons.

Sepsis, a complex medical condition that involves severe infections with life-threatening organ dysfunction, is a leading cause of death worldwide. Treatment of sepsis is highly challenging. When making treatment decisions, clinicians and patients desire accurate predictions of mean residual life (MRL) that leverage all available patient information, including longitudinal biomarker data. Biomarkers are biological, clinical, and other variables reflecting disease progression that are often measured repeatedly on patients in the clinical setting. Dynamic prediction methods leverage accruing biomarker measurements to improve performance, providing updated predictions as new measurements become available. We introduce two methods for dynamic prediction of MRL using longitudinal biomarkers. In both methods, we begin by using long short-term memory networks (LSTMs) to construct encoded representations of the biomarker trajectories, referred to as “context vectors.” In our first method, the LSTM-GLM, we dynamically predict MRL via a transformed MRL model that includes the context vectors as covariates. In our second method, the LSTM-NN, we dynamically predict MRL from the context vectors using a feed-forward neural network. We demonstrate the improved performance of both proposed methods relative to competing methods in simulation studies. We apply the proposed methods to dynamically predict the restricted mean residual life (RMRL) of septic patients in the intensive care unit using electronic medical record data. We demonstrate that the LSTM-GLM and the LSTM-NN are useful tools for producing individualized, real-time predictions of RMRL that can help inform the treatment decisions of septic patients.

The sequential multiple assignment randomized trial (SMART) is the gold standard trial design to generate data for the evaluation of multistage treatment regimes. As with conventional (single-stage) randomized clinical trials, interim monitoring allows early stopping; however, there are few methods for principled interim analysis in SMARTs. Because SMARTs involve multiple stages of treatment, a key challenge is that not all enrolled participants will have progressed through all treatment stages at the time of an interim analysis. Wu et al. (2021) propose basing interim analyses on an estimator for the mean outcome under a given regime that uses data only from participants who have completed all treatment stages. We propose an estimator for the mean outcome under a given regime that gains efficiency by using partial information from enrolled participants regardless of their progression through treatment stages. Using the asymptotic distribution of this estimator, we derive associated Pocock and O'Brien-Fleming testing procedures for early stopping. In simulation experiments, the estimator controls type I error and achieves nominal power while reducing expected sample size relative to the method of Wu et al. (2021). We present an illustrative application of the proposed estimator based on a recent SMART evaluating behavioral pain interventions for breast cancer patients.