There has been increasing interest in multi-arm multi-stage trials with treatment selection and sample size re-estimation at interim analysis. The East 6.4 release incorporates new Multi-Arm Multi-Stage (MAMS) module to support statisticians designing these studies. In this blog, we take a high level look at some of the features and advantages of this approach.
Challenging the Paradigm
Clinical trial failure rates make for sobering reading with only around 50% of confirmatory trials estimated to show a beneficial effect. Against this backdrop, challenging the paradigm and creating innovative approaches to clinical development is key. In areas of high unmet need, it is particularly important that promising treatments are prioritized to improve patient outcomes. It is also important that existing therapies can be compared against one another and provide evidence to clinicians for their treatment decisions.
The multi-arm approach evaluates a number of experimental arms against a common control within a single trial. This is combined with a ‘multi-stage’ adaptive approach which can increase efficiency by reviewing outcomes at key activity stages through interim analyses. As with other adaptive approaches, the decision-making process may be accelerated, allowing for progression of promising treatments and attrition of less promising ones. As Jaki points out in the paper Multi-arm clinical trials with treatment selection: what can be gained and at what price? ‘Many trials fail simply because ineffective treatments are identified too late’.
The ‘multi-stage’ piece can be handled in a number of different ways, and as such there are several classes of MAMS design. These include group-sequential, multi-stage drop-the-losers designs, and multi-arm adaptive randomization.
Advantages of the MAMS Approach
There are a number of potential advantages associated with the MAMS approach. These are summarized by James Wason at al in the paper Some recommendations for multi-arm multi-stage trials as:
- The use of a shared control group can be used so fewer patients required overall.
- Head-to-head comparison of treatments is conducted, minimizing biases which could be introduced from making comparisons between treatments tested in separate trials
- Interim analyses allow ineﬀective treatments to be dropped early, or early stopping of the trial if one treatment is clearly superior (although Wason points out that this advantage applies also in the case of separate trials of each treatment through use of group-sequential designs)
There are also logistical and practical gains since there is only one protocol, ethics committee application, grant application etc required.
To watch Cyrus Mehta's presentation on Multi-arm Multi- stage designs in East, click the link below.
Multi-arm clinical trials with treatment selection: what can be gained and at what price? Thomas Jaki, 2015, Clinical Investigation
Issues in applying Multi-Arm Multi Stage approaches to a Clinical Trial in Prostate Cancer: The STAMPEDE trial. Matthew Sydes et al, June 2009, Biomed Central
Some recommendations for multi-arm multi-stage trials Wason et al, December 2012 , Statistical Methods in Medical Research