Our Institute of Personalised Cancer Therapy programme is a very robust effort at trying to do a better job of connecting biomarkers to therapy decision making. So we are interested in both discovery as well as implementation of biomarkers for clinical utility. One of the early aspects that we wanted to embark on is trying to figure out how we can better use genomics that have now come to be available in the clinic to better use for clinical treatment selection, especially for ongoing clinical trials.

As you know, for only a handful of diseases there’s actually a drug already approved for a specific genetic mutation. However, there are a large number of clinical trials now available where specific genetic mutation is being used as eligibility criteria for entry. Further, even in early phase I trials we are interested in better matching patients, their molecular profile, to a drug that we think they will benefit more from. So there’s a greater interest even for genomically non-selected trials to offer genomically matched therapies.

So early in 2012 we initiated a protocol that we refer to as clearing house where we started offering genomic testing in the institution to really any patient who had advanced disease where clinical trial selection was being considered early in their treatment course. So we offered genomic testing using a hotspot mutation panel followed by then some exploratory research testing with targeted exome sequencing. At this meeting we were presenting the results of our first 2,000 patients enrolled. Everything is bigger in Texas so we’ve enrolled a little over 5,000 patients to date but we’ve been tracking to see the utility of the genomic testing, the utilisation overall for decision making and now we’re getting to a point where we’re going to have a robust genomic database that we can link to clinical data to start making genotype phenotype correlations, even for standard care therapies. So this is a really exciting time in genomics.

And you were quite successful?

Yes, we used a targeted hotspot panel so that alone, of course, focuses on genes that we understand a little bit more about and they’re genes that are more actionable and that there are more trials for them. Using the panel we have we had about a 39.5% rate of identifying a potentially actionable mutation, or rather a mutation that’s in a potentially actionable gene. That being said, of course, this number differs based on the tumour type being tested as well as how we define actionability

However, we quickly noticed the drop-off from identifying a patient that has a potentially actionable alteration to getting to a point where we can identify a trial that they would be eligible for. So this allowed us to do a critical review, really, of the challenges of personalised therapy and then devise solutions for it. So not only were there not enough trials for especially rarer histologies but also not enough recognition maybe, even when their genomic test results came back. A clear need for increased awareness for therapeutic implications of a genomic alteration when it is found.

So, in addition to critically analysing the data we put forward a series of decision support tools within our institution. So we started offering a decision support service where the treating physicians that get a genomic report can do an immediate consultation getting insight from a researcher scientist about the meaning of a specific variant. We all know, of course, that not all genes are the same but not every mutation has the same functional impact. More and more reports come in stating a patient has a mutation but really not giving information about what is known about that mutation. Do we know that it affects the function of the mutation? Those are very critical things when we’re thinking about enrolling patients into a clinical trial. So we want to not only personalise that choice but really make sure that the level of evidence for each individual gene’s impact as well, which variance impact, is considered while deciding between other therapeutic options or a clinical trial or, if there are multiple alterations, allowing to use this information to prioritise the gene that’s going to be targeted or optimal combinations.

What are the next steps?

This is a very exciting time. I think we all realise that getting an increased portfolio of trials, especially with the approach of basket trials, was a very important first step but we’re all looking at rational combinations as our next step.