Improving trial design
Clinical trials seek to evaluate the potential benefits and
harms of new treatments. They should therefore be designed and
conducted carefully to ensure patient safety and produce valid
results.
Many exciting new therapies are becoming available, and
efficient designs to identify effective treatment regimens in a
timely fashion are essential. For this purpose, the Unit has
pioneered the multi-arm,
multi-stage (MAMS) design. This allows several regimens to be
assessed simultaneously against a single control group in a
randomised fashion. In general, MAMS is an adaptive
design and can be regarded as one type of group sequential
design.
One advantage the MAMS design has over other types of adaptive
design is that it utilises an ‘intermediate’ outcome measure at the
interim stages. This makes the design more efficient, because the
expected number of patients and the length of time taken to
complete the trial will be dramatically reduced.
CTU has successfully applied MAMS to our
ICON5 and STAMPEDE cancer
trials. A user-friendly Stata
program is available to help others design such trials. We also
plan to explore the practical issues involved in running ‘optimal’
phase II designs to increase the use and impact of MAMS
designs.
Unlike other sample size programs, our
in-house ART (Analysis
of Resources for Trials) software covers survival outcomes
in conjunction with complex design issues for superiority trials.
However, the CTU has been implementing more non-inferiority and
equivalence trials than before. Accordingly, we are still
developing the ART software further to include sample size
calculations for non-inferiority and more complex trial
designs.
Key projects
- Speeding up evaluation of new therapies and improving success
rate for identification of effective therapies: further
developments of MAMS designs, and identifying the optimal size and
design of phase II trials
- Complex sample size calculations allowing
cross-over, non-proportional hazards, non-inferiority designs:
extension of theory and the ART sample size
software
Selected publications
- Royston P, Barthel FM-S, Parmar MKB, Choodari-Oskooei B, Isham
V, Designs for clinical trials with time-to-event outcomes based on
stopping guidelines for lack of benefit. Trials 2011;
12:81 (doi:10.1186/1745-6215-12-81)
- Royston P, Barthel FM-S Projection of power and events in
clinical trials with a time-to-event outcome. Stata Journal 2010;
10(3): 386-394.
Barthel F-MS, Royston P, Parmar MKB. A menu-driven facility for
sample-size calculation in novel multiarm, multistage randomized
controlled trials with a time-to-event outcome. Stata Journal
2009; 9 (4): 505-523
- Sydes MP, Parmar M.K.B, James ND, Clarke NW, Dearnaley DP,
Mason MD, Morgan RC, Sanders K, Royston P. Issues in applying
multi-arm multi-stage methodology to a clinical trial in prostate
cancer: the MRC STAMPEDE trial. Trials 2009; 10:39
- Barthel FMS, Babiker A, Royston P, Parmar MKB. Evaluation of
sample size and power for multi-arm survival trials allowing for
non-uniform accrual, non-proportional hazards, loss to follow-up
and cross-over. Statistics in Medicine 2006; 25:2521-2542
