Dr. Maria Pia Sormani
Department of Health Sciences, University of Genova, Italy
Victoria Murphy:
My guest today is Dr. Maria Pia Sormani from the department of health sciences, university of Genova, Italy. Today we are here to discuss a paper you authored in The Multiple Sclerosis Journal in August 2012. The title of your paper was ‘Time to first relapse as an endpoint in Multiple Sclerosis clinical trials’. Dr. Sormani, what is the context of this study?
Dr. Sormani:
This study inserts into the large discussion around the new strategies that are needed for future clinical trials design in MS. In fact up to now the standard design for testing new drugs in phase 3 trials was the placebo controlled study lasting 1 or 2 years.In these last few years many new drugs for the treatment of MS have been tested and there’s now an increasing number of active drugs on the market. So, the use of placebo in clinical trials has become unethical, at least for trials lasting more than 6 months. This is the reason why there is the need of new strategies for future clinical trials for drug approval in MS.
Victoria Murphy:
What were the objectives of your study?
Dr. Sormani:
The main objective of our study was to investigate the statistical properties of a very simple outcome, that is time to first relapse. The time to first relapse is the time elapsing from treatment start to the first relapse experienced by the patients. This outcome has the great advantage to allow patients who were randomised to placebo to switch to the active treatment once they relapsed preserving the validity of the study. They can switch because they reached the primary outcome.
Victoria Murphy:
How was it designed?
Dr. Sormani:
We conducted simulations to estimate the sample size needed when using the time to first relapse as the primary outcome and comparing these numbers with those needed when using the ARR. These simulations are based on a new statistical model that we derived and that fits very well the distribution of time to first relapse in MS patients. The details of this statistical model have been published in a statistical journal.
Victoria Murphy:
What were the results?
Dr. Sormani:
We estimated the sample sizes of trials based on time to first relapse and on annualized relapse rate and compared them. There is the general opinion that trial based on the time to first relapse are less powerful than trials based on the full number of relapses counted over a fixed period of time. We showed that this is not always true: in trials with a low rate of events (as modern trials) and lasting one year, time to first relapse is more powerful than the number of relapses: in fact, in this situation we have a low probability to see more than 1 relapse per patient and the time of occurrence of relapses gives more information than counting the relapses. In summary, we showed that there are situations where the time to first relapse is even a more powerful outcome than the classic annualized relapse rate.
Victoria Murphy:
Are there any limitations to the study?
Dr. Sormani:
Yes, there are 2 major limitations to the use of time to first relapse as an endpoint. The first one is when we have to evaluate treatments with a delayed effect. If the treatment effect is not immediate we can observe a number of early first relapses that are similar in the 2 arms, because the treatment requires some time to show its effect. The second limitation is for the analyses of other endpoints assessed at fixed time points, like MRI variables. The analysis of these endpoints is contaminated by the switch from the placebo to the active arm of relapsing patients.
Victoria Murphy:
What conclusions can be drawn from these results?
Dr. Sormani:
This study shows that clinical trials in specific conditions can be based on the time to first relapse as the primary outcome. The advantage is that in placebo controlled trials, when we use such an endpoint, pts are randomized to an active treatment or to a delayed active treatment (after the first relapse). So, also recruitment can be easier. This is not the solution to the problems of future clinical trials in MS, but just a small step to try to find feasible alternative to the standard placebo controlled design in MS.
Victoria Murphy:
Dr. Sormani thank you for joining us today, it has been a pleasure.
Dr. Sormani:
You’re welcome, thank you.