My presentations will focus on advances in the use of precision techniques for immunotherapy, focussing on both an immunohistochemistry approach that involves PD-L1, a ligand for anti-PD-1 therapies, as well as using assays that measure defects in mismatch repair. These are the two biomarkers that have had the greatest success in being able to identify patients most likely to benefit from checkpoint inhibitors. So my talk will focus on that; some of the challenges in their use, in particular the fact that there are multiple assays that are available for both of these areas. Then finally I’ll also touch upon personalised cancer vaccines which are also, of course, an area that involves individualised patient approaches.
How is PD-L1 changing the landscape of checkpoint therapies?
PD-L1, as measured by immunohistochemistry, is the biomarker that has been most successful in identifying patients most likely to benefit from immunotherapies and has actually been instrumental in certain approvals. In fact, I go over that with pembrolizumab where, among the many approvals that have occurred with that agent, several have actually required the use of a PD-L1 test to demonstrate the clinical benefit relative to standard of care types of treatments. So that type of work is continuing across multiple cancer types, both for monotherapy pembrolizumab as well as other checkpoint inhibitors, anti-PD-1 agents, but also, importantly, in combinations.
That type of work is just beginning but it will be important to sort out whether for combinations you need the use of PD-L1, for example, or not. One of the points I make is for a recent study, again with pembrolizumab in lung cancer where, when combined with chemotherapy it turns out that actually patients with low levels of PD-L1 benefit significantly. So in that case it’s an area where you wouldn’t want to exclude patients for treatment when you’re using that combination with a PD-L1 test.
The other area where biomarkers have been useful is looking at the other side, again, of elements that we think are important in predicting response to anti-PD-1 agents. So PD-L1 is, of course, the ligand, that’s easy to understand, but the other element would be alterations that lead to high tumour mutational burden which in turn would lead to production of new proteins and neoantigens, new antigens, that could be recognised by the immune system.
The most practical manifestation of that theory that has worked is the tissue-agnostic approval that pembrolizumab obtained with the US FDA recently where that was done by using assays that measure microsatellite instability, either by a protein assessment, where you’re actually looking for a loss of protein expression of genes involved in mismatch repair, or you can also use another assay that measures the phenotype there where you’re measuring microsatellites, which are short DNA repeats that we all have, that become unstable when you’re lacking these mismatch repair genes. So those two assays are actually widely available and they were used in the registrational studies that became the basis for this approval. So this would be a second class of tests that have now become available to be able to identify patients who are most likely to benefit from anti-PD-1 therapies.
Tell us about the use of vaccines.
Another area that is quite interesting under investigation by a number of companies, including Merck. We have a collaboration with Moderna and their premise is they use DNA sequencing, next gen sequencing techniques, to fully sequence a patient’s tumour, looking at DNA, looking for mutations. Then using an algorithm that would then predict which of those mutations would result in an immunogenic antigen they construct an RNA-based vaccine with multiple antigens present from that prediction algorithm and inject that into a patient. The idea there is, combined with checkpoint inhibitors such as pembrolizumab, you would then target the immune system towards those neoantigens, prevent the inhibition with the pembrolizumab part and that this would result in a very personalised approach to a particular patient’s cancer treatment.
What should we be focussing on within immunotherapy?
First of all I’ll just step back and say that I think the whole era of immunotherapy has really ushered in a very dramatic change in the way we treat cancer. Previously most of our focus was on chemotherapy or targeted agents that targeted the cancer, immunotherapy is actually, in some sense, not specifically directed against the cancer, it’s directed against the patient’s own immune system to unlock the immune system to realise its ability to attack the cancer.
However, despite those multiple successes there are many patients who don’t benefit from these treatments and so it’s back to where there needs to be focusses on continued advances in both identifying tests such as PD-L1 or the MSI test that identified patients with very high potential for benefit, as well as moving towards combinations that will increase the likelihood of an individual patient having a dramatic outcome to these agents. Those can include a variety of modalities, I mentioned chemotherapy as being one that it’s clear that improves outcomes compared to pembrolizumab alone in lung cancer patients. Many other combinations are under investigation including the personalised cancer vaccines that I mentioned previously.
