Incorporating biomarkers into clinical trial design

Bookmark and Share
Published: 11 May 2018
Views: 947
Dr Amar Gajjar - St. Jude Children's Research Hospital, Memphis, USA

Dr Gajjar speaks with ecancer at the Cancer Research UK Brain Tumour Conference 2018 in London about trial designs to match targeted therapies to molecular subgroups of brain cancers.

He highlights the need for international collaboration to widen access to trial data for clinicians, and that this will also improve patient access and involvement in experimental therapies.

Dr Gajjar cites experience from an international group working with medulloblastoma patients, who are now hoping to reach new countries and more patients.

What I have been requested to do is to have some provocative ideas about design of clinical trials and how should we be positioning ourselves for the future so that we can make rapid progress. So I’m going to be speaking on novel designs of clinical trials that we have otherwise not considered before and with the avalanche of clinical molecular data how we could design these trials because it’s possible now.

The other area that I’ll be focussing on is what is the kind of clinical trials research infrastructure that needs to be built so that we can carry out these clinical trials. Because now we understand that these brain tumours are molecularly very distinct subgroups so to have rapid enrolment we’ve got to work across the Atlantic, Europe and North America together. And to be able to do this and meet the regulatory requirements of the European agencies and the FDA on our side we need to really hit the refresh button and think how we’re going to achieve this.

Traditionally we have designed trials to improve the cure rate and although that’s a very important point we are not ready for those large randomised trials as of yet. What we need to do is shorter, more efficient trials which are called umbrella trials, basket trials, adaptive trial designs which can quickly match up molecular targeted therapies or precision therapies with the subgroup of tumours and get an early signal whether these approaches are working or not and if they’re not working we move on. For the next 5-7 years if we conduct such trials then we will be able to frame the questions in bigger trials down the road. We are not ready yet to do the large randomised trials that we were used to doing a decade ago.

How could infrastructure be improved to benefit clinical trials?

The first thing is the tissue processing and handling. We need labs which are centralised and have standardised tests. In the US we call it a CLIA environment, labs which are certified so that the results are consistent throughout. Then I’m sure in Europe, in the UK I know for medulloblastoma there are three Centres of Excellence; in Germany now they’ve got certain centres, in France they’re getting certain centres. So all these are speaking the same language so when tissue comes either from the UK or from Germany or from France the output is the same so they can enrol on the study. I think that’s one very important step – it’s no longer just looking at the slide under the microscope, it’s generating the molecular genomics analysis of these tumours.

The second part of doing this is having trial databases set up, again across North America and across Europe, so that data is being remotely entered by web-based systems. So it doesn’t matter if you enrol a patient in Newcastle or we enrol a patient in Memphis, the data is coming in simultaneously. And we need to start thinking globally and using technology and infrastructure, web-based, that allows us to do this.

It’s going to happen because meetings like this will give us an impetus to start working together. It’s just a mental block that investigators have not thought about it in that way. Part of that pressure is going to come from the market. By market I mean parents. I get a lot of requests from family, in fact I just got one last night from a mother in the UK. I feel that if we work together and have these systems in place parents don’t have to come running halfway around the world to get the therapy we should be able to establish in three or four centres in each country which can do these high-end sophisticated studies so that we can merge the data and move forward. It’s going to take a little bit of effort but once we break that barrier it will become the standard

One trial that I’ve been working on for the last two decades is we’ve now got a medulloblastoma which is the most common malignant paediatric brain tumour. We’ve been gradually expanding our footprint so this trial is the first trial in North America to really assign treatment based on the clinical and molecular risk features. We have now expanded this trial into Canada, New Zealand and Australia. So that’s one of the models we’ve created that we can actually do this. The tissue comes to St Jude and then we analyse the tissue and then assign the risk stratum. Now we are working with colleagues in Europe, we’ve done this part of the world, we want to start working with this part of the world. We’ve initiated negotiations, our best friends, the lawyers, are working out the contracts so that we can start making that happen on this side too.