BTOG 2021 | NSCLC biomarker testing in the UK

I gave a presentation today at BTOG about where we are with biomarker testing in lung cancer, taking account of all the changes we’ve had, both in testing strategies and access to new treatments. So, we’re at a point now where, for immunotherapy, we have PD-L1 testing and there’s a group of patients, 20 to 30%, that have high PD-L1 expression. And we know a proportion of those will get really nice long-term outcomes and that’s been demonstrated in recent presentations.

What’s slightly more challenging is the testing for oncogenic abnormalities. So, in particular, for lung adenocarcinoma, we now have access within the NHS to treatments targeting EGFR, ALK, ROS1 and NTRK, with some others coming forward to the point where if you look at American incidence there may be about 50% of our patients that have targeted treatment.

Actually, what I showed in my presentation was in the UK environment, it’s not quite as good as that because our patients have often smoked and, you know, had exposure to industrial carcinogens, such as many of my patients were ex-miners or have worked in the shipyards. And so, in my patient group, it’s probably about a third of patients might have something that we can potentially target.

I think that was the very easy bit of the presentation. The question then was sort of, you know, who should we do this biomarker testing on? And I was arguing that actually if we’re going to biopsy a patient, at least with a lung adenocarcinoma, we should probably should be doing the molecular testing as we’ve shown benefits for targeted treatments, at least in patients even with poor performance status, although we already have issues about access to appropriate treatments because of the strictures laid down on us by NHS England.

But, also this sort of question about, which came up quite a lot in discussion afterwards, about whether we should be doing a molecular profiling on early stage cancers. Is this a wasted test? And the argument I made in my presentation was, well, many of these patients will relapse at some point relatively soon anyway, and it’d be useful to have the information.

The second point that I made was that with advances that we now have in, with the ongoing advances we have in early-stage disease, so we have targeted therapies coming through such as potentially adjuvant osimertinib, but that’s not yet available in UK. There’re all the studies of neoadjuvant and adjuvant immunotherapy coming through. We’re going to need the molecular profile of these patients anyway, so we may as well get used to start doing it.

And Professor Kerr from Aberdeen made a point in discussion of, actually from a lab point of view, sort of reflex testing, so doing this biomarker testing on all these patients actually is probably more efficient and a more effective way to use lab resources.

And then, I made the point that, you know, how can we have an MDT discussion nowadays about a lung cancer patient if we don’t have all their molecular profile information available because that’s got huge implications for prognosis for treatment. And so, to have a rational discussion, we need that information at time of MDT discussion rather than using the MDT to guide when testing should take place.

And sort of, I guess, the last highlight of my presentation, at least I thought, was discussions about the role of circulating free DNA and analysis in the NHS. There’s a number of companies out there that now offer this as a service. And I presented some data that I generated in Newcastle in conjunction with The Christie and we presented at ESMO last year, and that Mary O’Brien and Sanjay Popat had developed and had presented from their experience of using these technologies at The Marsden, showing that we can get our results back very quickly using these tests with a high degree, high chance that they’ll work, at least in the stage four setting, and often, we’ll get results back more quickly than the present pathways of doing tumor biopsies and panel testing.

What we haven’t yet worked out, at least at the NHS is which patients we should be doing these tests on and how we’re going to fund it if it’s commercial assay or whether we should be building up these tests within our own genomic laboratory hub setting.

The other bit I touched on in my presentation was the formation of a genomic laboratory hub, so these are seven hubs that in England NHS England has set up to provide testing, more complicated genomic testing, including large next generation sequencing panels and whole genome sequencing for some indications such as sarcomas, leukemias and pediatric tumors.

And so, this is the centralization of services and we’ll be adjusting how we test our lung cancer patients moving away from single gene tests to an analysis on a 500 gene plus panel using both DNA and RNA inputs. And we have to think about how we’re going to deal with the more complicated reports that we get.

But I was arguing that knowing more about your patient’s tumor has got to be useful information in terms of both giving access to additional treatments, but also giving potential additional information, such as prognostic information, that could be helpful in discussions. For example, having mutations in KEAP1 and STK11, along with other mutations, may predict for a poor response to immunotherapy. And it’s not that we, we might still use immunotherapy in the setting, but it’d be helpful to guide the discussions with the patients as to the likely long-term outcomes.