In the world of clinical trials, the pace of innovation is accelerating, and approaches such as Bayesian methods are gaining traction. These methods bring flexibility and speed to clinical trial design and analysis, and with increased access to the necessary computational power, are transforming today’s clinical research [1]. However, the number of simulation and modeling tools necessary to perform Bayesian computations requires statisticians to be well-resourced technologically. Many biostatisticians may not readily have access to the cloud computing power to make these design approaches practical within the time constraints afforded for statistical design.
There are many dose-finding designs that have been developed over the past 30 years and several more are anticipated [2]. Sponsors often face the dilemma of choosing from the various design options available today. Finding the right dose in Phase 2 gives a potential new therapy its best chance to demonstrate efficacy during Phase 3, and Bayesian techniques prove to be useful for optimal dose-finding.
Benefits of using i3+3 in Phase 1
In 2019, a new rule-based design was proposed, namely the i3+3 design which abandoned statistical modeling altogether. It is based on simplified but more advanced set of rules that account for the variabilities in the observed data. The i3+3 is a superior design in terms of trial safety and the ability to identify the true Maximum Tolerated Dose (MTD). It is very flexible and powerful in terms of simulation performance. This design uses Simple Bayesian models and can produce decision tables which can help with real-world clinical trials.
Multiple Cohort Expansion (MuCE) designs
MUCE is a Bayesian solution for cohort expansion trials where multiple dose(s) and multiple indication(s) are tested in parallel. Such methods are particularly important for areas like oncology where several doses and several indications must be tested for successful completion of early phase trials, and optimal choice of dose and population to move on from early phase to a reasonable dosage for Phase 3.
Built on Bayesian hierarchical models with multiplicity control, MUCE adaptively borrows information across patient groups from different indications treated with different doses. A hierarchical model accounts for the fact that when aggregating data across patient groups, some treatment arms might have more significant differences than others, and this might require statisticians to make adjustments by weighting to achieve unbiased measurement. This enables MUCE to Control Type 1 Error while increasing power and reducing sample size. These efficient designs can be applied in any clinical trials with two or more arms. For an expansion cohort trial in the US, the MUCE design showed saving in sample size of up to 16.67% compared to Simon’s 2-stage design.
The Bayesian Logistic Regression Model (BLRM)
BLRM is a Bayesian method that uses a logistic model for dose-toxicity relationship. Its use in early phase dose escalation trials was pioneered in a paper by Neuenschwander et al [3]. This model can incorporate all relevant pre-trial information into the priors and improve decision-making.
It is generally agreed that the BLRM offers clinicians more flexibility compared to the more traditional designs like the 3+3. BLRM was designed to make better use of evidence collected during a clinical trial. It allows clinicians to make decisions about dosages as a trial proceeds, thereby aiming to ensure that each patient receives the best dose given the data available.
Learn more about BLRM in a Cytel webinar presented by Dr. Satrajit Roychoudhury, Senior Director and a member of Statistical Research and Innovation group in Pfizer Inc.
East Bayes*
Cytel is continuing to invest in and improve their flagship East software. Developed by accomplished study design experts, East creates clinical trials that best address key questions confronted by clinical trial sponsors. East Bayes, an enhanced web-based extension of East, combines Bayesian design capabilities with a growing suite of advanced and innovative clinical trial designs. The complex algorithms necessary to compute Bayesian calculations are finally accessible to statisticians, including novel Cytel-invented and Cytel-curated designs.
*Available March 8th, 2021
Watch a webinar by Professor Yuan Ji, University of Chicago, to learn more.
References:
1. Employing the Power of Bayesian Methods to Expedite Learning
2. Bayesian Dose-Finding Designs – An Overview
3. Neuenschwander, B., Branson, M. and Gsponer, T. (2008) ‘Critical aspects of the Bayesian approach to phase I cancer trials’,Statistics in Medicine, 27(13), pp. 2420–2439. doi: 10.1002/sim.3230.
