Screening Experiments for Developing Dynamic Treatment Regimes

Individualized treatment rule characterization via a value function surrogate

Precision medicine is a promising framework for generating evidence to improve health and health care. Yet, a gap persists between the ever-growing number of statistical precision medicine strategies for evidence generation and implementation in real-world clinical settings, and the strategies for closing this gap will likely be context-dependent. In this paper, we consider the specific context of partial compliance to wound management among patients with peripheral artery disease. Using a Gaussian process surrogate for the value function, we show the feasibility of using Bayesian optimization to learn optimal individualized treatment rules. Further, we expand beyond the common precision medicine task of learning an optimal individualized treatment rule to the characterization of classes of individualized treatment rules and show how those findings can be translated into clinical contexts.

Synthesizing studies for comparing different treatment sequences in clinical trials

With advances in cancer treatments and improved patient survival, more patients may go through multiple lines of treatment. It is of clinical importance to choose a sequence of effective treatments (eg, lines of treatment) for individual patients with the goal of optimizing their long-term clinical outcome (eg, survival). Several important issues arise in cancer studies. First, cancer clinical trials are usually conducted by each line of treatment. For a treatment sequence, we may have first line and second line treatment data from two different studies. Second, there is typically a treatment initiation period varying from patient to patient between progression of disease and the start of the second line treatment due to administrative reasons. Additionally, the choice of the second line treatment for patients with progression of disease may depend on their characteristics. We address all these issues and develop semiparametric methods under the potential outcome framework for the estimation of the overall survival probability for a treatment sequence and for comparing different treatment sequences. We establish the large sample properties of the proposed inferential procedures. Simulation studies and an application to a colorectal clinical trial are provided.

Design and analysis considerations for a sequentially randomized HIV prevention trial

TechStep is a randomized trial of a mobile health interventions targeted towards transgender adolescents. The interventions include a short message system, a mobile-optimized web application, and electronic counseling. The primary outcomes are self-reported sexual risk behaviors and uptake of HIV preventing medication. In order that we may evaluate the efficacy of several different combinations of interventions, the trial has a sequentially randomized design. We use a causal framework to formalize the estimands of the primary and key secondary analyses of the TechStep trial data. Targeted minimum loss-based estimators of these quantities are described and studied in simulation.

Bayesian Adaptive Design for Finding the Maximum Tolerated Sequence of Doses in Multicycle Dose-Finding Clinical Trials

Purpose

Statistical designs for traditional phase I dose-finding trials consider dose-limiting toxicity in the first cycle of treatment. In reality, patients often go through multiple cycles of treatment and may experience toxicity events in more than one cycle. Therefore, it is desirable to identify the maximum tolerated sequence of three doses across three cycles of treatment.

Methods

Motivated by a three-cycle dose-finding clinical trial for a rare cancer with a JAK inhibitor, we proposed and implemented a simple Bayesian adaptive dose-cycle finding (BaSyc) design that allows intercycle and intrapatient dose modification. Because of the patient-specific dosing strategy over cycles, the BaSyc design is suited as a method in precision oncology.

Results

BaSyc is simple and transparent because its algorithm can be summarized as two tabulated decision rules before the trial starts, allowing physicians to visually examine these rules. In addition, BaSyc employs a time-saving enrollment scheme that speeds up the trial. Extensive simulation studies show that BaSyc has desirable operating characteristics in identifying the maximum tolerated sequence.

Conclusion

The BaSyc design provides a first-of-kind multicycle approach for dose finding and will likely lead to better and safer patient care and drug development.

LIBERTI: A SMART study in plastic surgery

BACKGROUND/AIMS

Laser treatment of burns scars is considered by some providers to be standard of care. However, there is little evidence-based research as to the true benefit. A number of factors hinder evaluation of the benefit of laser treatment. These include significant heterogeneity in patient response and possible delayed effects from the laser treatment. Moreover, laser treatments are often provided sequentially using different types of equipment and settings, so there are effectively a large number of overall treatment options that need to be compared. We propose a trial capable of coping with these issues and that also attempts to take advantage of the heterogeneous response in order to estimate optimal treatment plans personalized to each individual patient. It will be the first large-scale randomized trial to compare the effectiveness of laser treatments for burns scars and, to our knowledge, the very first example of the utility of a Sequential Multiple Assignment Randomized Trial in plastic surgery.

METHODS

We propose using a Sequential Multiple Assignment Randomized Trial design to investigate the effect of various permutations of laser treatment on hypertrophic burn scars. We will compare and test hypotheses regarding laser treatment effects at a general population level. Simultaneously, we hope to use the data generated to discover possible beneficial personalized treatment plans, tailored to individual patient characteristics.

RESULTS

We show that the proposed trial has good power to detect laser treatment effect at the overall population level, despite comparing a large number of treatment combinations. The trial will simultaneously provide high-quality data appropriate for estimating precision-medicine treatment rules. We detail population-level comparisons of interest and corresponding sample size calculations. We provide simulations to suggest the power of the trial to detect laser effect and also the possible benefits of personalization of laser treatment to individual characteristics.

CONCLUSION

We propose, to our knowledge, the first use of a Sequential Multiple Assignment Randomized Trial in surgery. The trial is rigorously designed so that it is reasonably straightforward to implement and powered to answer general overall questions of interest. The trial is also designed to provide data that are suitable for the estimation of beneficial precision-medicine treatment rules that depend both on individual patient characteristics and on-going real-time patient response to treatment.

Bayesian Dose-Finding in Two Treatment Cycles Based on the Joint Utility of Efficacy and Toxicity

A phase I/II clinical trial design is proposed for adaptively and dynamically optimizing each patient's dose in each of two cycles of therapy based on the joint binary efficacy and toxicity outcomes in each cycle. A dose-outcome model is assumed that includes a Bayesian hierarchical latent variable structure to induce association among the outcomes and also facilitate posterior computation. Doses are chosen in each cycle based on posteriors of a model-based objective function, similar to a reinforcement learning or Q-learning function, defined in terms of numerical utilities of the joint outcomes in each cycle. For each patient, the procedure outputs a sequence of two actions, one for each cycle, with each action being the decision to either treat the patient at a chosen dose or not to treat. The cycle 2 action depends on the individual patient's cycle 1 dose and outcomes. In addition, decisions are based on posterior inference using other patients' data, and therefore the proposed method is adaptive both within and between patients. A simulation study of the method is presented, including comparison to two-cycle extensions of the conventional 3+3 algorithm, continual reassessment method, and a Bayesian model-based design, and evaluation of robustness.

Introduction to SMART designs for the development of adaptive interventions: with application to weight loss research

The management of many health disorders often entails a sequential, individualized approach whereby treatment is adapted and readapted over time in response to the specific needs and evolving status of the individual. Adaptive interventions provide one way to operationalize the strategies (e.g., continue, augment, switch, step-down) leading to individualized sequences of treatment. Often, a wide variety of critical questions must be answered when developing a high-quality adaptive intervention. Yet, there is often insufficient empirical evidence or theoretical basis to address these questions. The Sequential Multiple Assignment Randomized Trial (SMART)-a type of research design-was developed explicitly for the purpose of building optimal adaptive interventions by providing answers to such questions. Despite increasing popularity, SMARTs remain relatively new to intervention scientists. This manuscript provides an introduction to adaptive interventions and SMARTs. We discuss SMART design considerations, including common primary and secondary aims. For illustration, we discuss the development of an adaptive intervention for optimizing weight loss among adult individuals who are overweight.

Optimization of individualized dynamic treatment regimes for recurrent diseases

Patients with cancer or other recurrent diseases may undergo a long process of initial treatment, disease recurrences, and salvage treatments. It is important to optimize the multi-stage treatment sequence in this process to maximally prolong patients' survival. Comparing disease-free survival for each treatment stage over penalizes disease recurrences but under penalizes treatment-related mortalities. Moreover, treatment regimes used in practice are dynamic; that is, the choice of next treatment depends on a patient's responses to previous therapies. In this article, using accelerated failure time models, we develop a method to optimize such dynamic treatment regimes. This method utilizes all the longitudinal data collected during the multi-stage process of disease recurrences and treatments, and identifies the optimal dynamic treatment regime for each individual patient by maximizing his or her expected overall survival. We illustrate the application of this method using data from a study of acute myeloid leukemia, for which the optimal treatment strategies for different patient subgroups are identified.

Pattern-mixture-type Estimation and Testing of Neuroblastoma Treatment Regimes

Sequentially randomized designs are commonly used in biomedical research, particularly in clinical trials, to assess and compare the effects of different treatment regimes. In such designs, eligible patients are first randomized to one of the initial therapies, then patients with some intermediate response (e.g. without progressive diseases) are randomized to one of the maintenance therapies. The goal is to evaluate dynamic treatment regimes consisting of an initial therapy, the intermediate response, and a maintenance therapy. In this article, we demonstrate the use of pattern-mixture model (commonly used for analyzing missing data) for estimating the effects of treatment regimes based on familiar survival analysis techniques such as Nelson-Aalen and parametric models. Moreover, we demonstrate how to use estimates from pattern-mixture models to test for the differences across treatment regimes in a weighted log-rank setting. We investigate the properties of the proposed estimators and test in a Monte Carlo simulation study. Finally we demonstrate the methods using the long-term survival data from the high risk neuroblastoma study.

Optimization of behavioral dynamic treatment regimens based on the sequential, multiple assignment, randomized trial (SMART)

BACKGROUND AND PURPOSE

A behavioral intervention is a program aimed at modifying behavior for the purpose of treating or preventing disease, promoting health, and/or enhancing well-being. Many behavioral interventions are dynamic treatment regimens, that is, sequential, individualized multicomponent interventions in which the intensity and/or type of treatment is varied in response to the needs and progress of the individual participant. The multiphase optimization strategy (MOST) is a comprehensive framework for development, optimization, and evaluation of behavioral interventions, including dynamic treatment regimens. The objective of optimization is to make dynamic treatment regimens more effective, efficient, scalable, and sustainable. An important tool for optimization of dynamic treatment regimens is the sequential, multiple assignment, randomized trial (SMART). The purpose of this article is to discuss how to develop optimized dynamic treatment regimens within the MOST framework.

METHODS AND RESULTS

The article discusses the preparation, optimization, and evaluation phases of MOST. It is shown how MOST can be used to develop a dynamic treatment regimen to meet a prespecified optimization criterion. The SMART is an efficient experimental design for gathering the information needed to optimize a dynamic treatment regimen within MOST. One signature feature of the SMART is that randomization takes place at more than one point in time.

CONCLUSION

MOST and SMART can be used to develop optimized dynamic treatment regimens that will have a greater public health impact.

Dynamic Treatment Regimes

A dynamic treatment regime consists of a sequence of decision rules, one per stage of intervention, that dictate how to individualize treatments to patients based on evolving treatment and covariate history. These regimes are particularly useful for managing chronic disorders, and fit well into the larger paradigm of personalized medicine. They provide one way to operationalize a clinical decision support system. Statistics plays a key role in the construction of evidence-based dynamic treatment regimes - informing best study design as well as efficient estimation and valid inference. Due to the many novel methodological challenges it offers, this area has been growing in popularity among statisticians in recent years. In this article, we review the key developments in this exciting field of research. In particular, we discuss the sequential multiple assignment randomized trial designs, estimation techniques like Q-learning and marginal structural models, and several inference techniques designed to address the associated non-standard asymptotics. We reference software, whenever available. We also outline some important future directions.

Designing a pilot sequential multiple assignment randomized trial for developing an adaptive treatment strategy

There is growing interest in how best to adapt and readapt treatments to individuals to maximize clinical benefit. In response, adaptive treatment strategies (ATS), which operationalize adaptive, sequential clinical decision making, have been developed. From a patient's perspective an ATS is a sequence of treatments, each individualized to the patient's evolving health status. From a clinician's perspective, an ATS is a sequence of decision rules that input the patient's current health status and output the next recommended treatment. Sequential multiple assignment randomized trials (SMART) have been developed to address the sequencing questions that arise in the development of ATSs, but SMARTs are relatively new in clinical research. This article provides an introduction to ATSs and SMART designs. This article also discusses the design of SMART pilot studies to address feasibility concerns, and to prepare investigators for a full-scale SMART. We consider an example SMART for the development of an ATS in the treatment of pediatric generalized anxiety disorders. Using the example SMART, we identify and discuss design issues unique to SMARTs that are best addressed in an external pilot study prior to the full-scale SMART. We also address the question of how many participants are needed in a SMART pilot study. A properly executed pilot study can be used to effectively address concerns about acceptability and feasibility in preparation for (that is, prior to) executing a full-scale SMART.

Designing a sequentially randomized study with adherence enhancing interventions for diabetes patients

Adaptive treatment strategies can change treatment prescription over time in response to intermediate outcomes. They are the natural choice for treating chronic diseases or for prevention, since the condition of subjects tends to change over the long term. Similarly, flexible intervention strategies are vital for generating or sustaining better adherence in long term treatment settings. When a cost-efficient first-line treatment is available, for instance, good adherence is expected to help delay or avoid second-line treatment.Sequentially randomized trials enable unbiased evaluation of how to best adapt adherence supporting interventions to a history of outcomes and adherence with the goal to optimize future treatment response. In this paper we propose and study different sequential designs targeting cost-efficient control of type II diabetes under first-line treatment through two different classes of adherence support: by (bio)technical and by behavioural means. We study their respective and joint impact first through double factorial adaptive designs, where interventions are triggered by an elevated risk of current treatment failure predicted by poor surrogate response.We develop the double factorial design and several derived designs that are more cost-efficient in the context of managed care of diabetes patients. We evaluate the marginal responses over time to different adaptive treatment strategies by means of doubly robust estimators. We consider sample sizes needed to thus detect realistic and worthwhile effects and discuss the relative practical and theoretical merits of the separate designs.