Targeted therapies in melanoma
Dr Dirk Schadendorf - University Hospital Essen, Essen, Germany
My topics I have to cover are resistance but also kinase inhibitors for rare mutations and I think one of the most interesting topics is also on patients who have got combination checkpoint blockade and patients who discontinue from treatment. How do they do in terms of safety and efficacy?
Could we hear more about resistance?
Understanding resistance obviously will open some ways to get around it, possibly prevent it. On the other hand if there is a molecular basis for resistance and there are certain indications, for example, the checkpoint blockade might be dependent on the interferon signalling pathway and if there are mutations or epigenetic modifications it might be possible, it’s not so easy to overcome or to prevent. The same with the targeted therapies we know that certain amplification and mutations of certain key signalling molecules do occur. I think the only way to prevent that is to hit hard and early in order to overcome and possibly delay resistance. There are multiple ways to overcome resistance; the key is to understand resistance. We know for example for checkpoint blockade that obviously the interferon signalling pathways are of critical importance and mutations or epigenetic modifications cannot easily be prevented. The same if we analyse the MAP kinase pathway in the context of BRAF MEK inhibition, also here we see mutations, amplifications, which can be in part prevented by combined approaches. I think in summary that we need to combine approaches, hit hard and early and possibly also to combine targeted and immuno oncology approaches.
Could we hear more about rare mutational subgroups?
Rare mutation subgroups are obviously more challenging. I think we have learned by the sequencing approaches that there are rare mutations. The problem is that there will never be firm clinical evidence how to treat these rare mutations because we will have on one hand preclinical science supporting that there is resistance or sensitivity to certain drugs. There will be possibly also some patients case reports or a small series of patients treated with certain drugs but there will be never a clinical trial supporting really clearly the clinical evidence and the magnitude of the evidence so that’s challenging. It’s more based on preclinical data, drugs which are available, testing in the lab and then testing efficacy in the clinical setting.
One very nice example is, for example, the experience we have with c-Kit. This is one of these rare mutations which we know nowadays occurs in acral melanoma, in mucosal melanoma, in chronically sun-damaged melanomas. Roughly 10% of these are already rare subgroups and this is not one mutation like the BRAF mutation. We know that c-Kit mutations occur in various forms and also amplification of c-Kit has been seen and other trials which have been done with c-Kit have included most of the mutations or also amplification. There is some early indication now that amplification obviously as a sum do not respond to inhibitors like imatinib or nilotinib. On the other hand, if you analyse the different mutations which are there and their sensitivity to drug exposure there is a clear tendency that only two or three of these mutations really seem to be sensitive, in the range between 20-50%. That’s already an indication but it’s a solution only for a very small fraction of our patients.
What are your thoughts on people discontinuing from checkpoint therapy?
One very interesting piece which was presented here is the experience of patients who have been treated with nivolumab and ipilimumab, the PD-1 antibody in combination with the CTLA-4 antibody. This regimen is already approved in Europe and the US and has been tested in two larger trials, the CheckMate 069 and the CheckMate 067 study.
One of the major challenges of this treatment approach combining checkpoint inhibitors is toxicity. More than 50% of these patients will experience grade 3 or 4 toxicity and almost 40% of the patients need to discontinue treatment early. The question which was asked is how is the efficacy of this combination in the patient group who has discontinued? How does that compare to the patient group who is continuing and staying on treatment?
If one looks on this meta-analysis now which was presented, more than 400 patients were included. The group who has discontinued, a group of more than 170 patients stopped treatment quite early in the range of 1.5 months. The group who stayed on treatment had more than 9 months of treatment so that’s already a substantial difference. So if we look now on the efficacy then the surprise is actually quite large because efficacy in terms of tumour control measured by progression free survival time is better in the group where it’s discontinued early. It’s more in median ten months for the patient group who continues versus more that sixteen months for the patient who has discontinued early. This is statistically highly significant. If you look at response rate, the response rate is a respectable 50% for the patient group who has discontinued but is 68% for the patient group who has discontinued early.
There is clear efficacy so if we now look at what is also of great importance: durability of response. The durability of response is not reached, the follow-up time is more than twenty months and so usually the clinical response is achieved within eight weeks. For the first staging you see that the patient is benefiting but we know after one year if we look at the patient groups, 90% of the patients who had discontinued early still do not need any treatment. This is in comparison to patients who have continued treatment, close to 80%. The only difference in favour of the group who has continued treatment is the duration of response, ongoing responses. Possibly for longer duration of response you need some more clinical treatment.
It’s currently open what to do with these results. I think the caveat is it’s a retrospective analysis of a large group of patients. There is clear evidence of the patients who discontinue seem to benefit. Toxicity in these patients is much, much higher, the burden of toxicity. Almost 50% of the patients who had to discontinue had two or more organ sites involved by their own immunity in contrast to 80% of the patients who continued treatment, had either no toxicity or only one organ system involved. This is also visualised by the fact that the patients who discontinued needed much more immuno-suppressive agents like infliximab -10% in that group, 1% in the other group. So this is quite an interesting result but it poses obviously a lot of questions now what to do. Should we treat patients with a combination only for eight to twelve weeks or is this toxicity, this severe toxicity, and the good clinical response just an indicator, some sort of biomarker, for a particularly sensitive patient group? I think that needs to be addressed in a prospective study.