I will start with the data that was presented as poster in this ESMO 2020 Congress. There were several studies that were actually presented as posters, more than 50 posters were presented, and they showed mostly studies that were based in real-world data. Despite the fact that we have the best clinical evidence coming from clinical trials, being able to confirm the data in the real-world dataset remains important in my opinion.

So I’m just going to mention a couple of topics that were addressed in posters, one that refers to therapy in older patients. So although no information was provided regarding the comorbidities, there was data from a significant number of patients that showed that systemic therapy, namely targeted and immunotherapy, in this setting seems to be safe and also achieved the same outcomes as patients that are younger. So the message is older patients should not be excluded from receiving targeted or immunotherapy only due to age because they seem to derive the same benefit.
Then there were studies, or posters, that addressed sequential therapy. The caveat of this analysis is the fact that only patients that actually received at least two therapy lines were included in this analysis. Patients that have a very good prognosis on the first line of therapy or those that have progressed very fast under the first line of therapy were not included in this analysis. However, the data that was shown here and in the absence of clinical trials that address this question, or for which we have data from clinical trials, the data that we see here can be used to make some therapeutic decisions. So one of the works that was presented addressed the therapy sequencing of immunotherapy followed by targeted therapy or targeted therapy followed by immunotherapy in BRAF mutated melanoma. They analysed data from 300 patients and they showed that patients progressed relatively fast regardless of the first or second line therapy with a similar risk of progression and mortality.

Also another work that showed that patients that need a second line therapy seemed to do worse in terms of progression free survival and overall survival. Finally, another work with an Asian cohort that evaluated the efficacy of salvage therapy after failure of anti-PD-1 monotherapy that showed that actually the salvage therapy with ipilimumab-nivolumab might have more limited clinical efficacy in this Asian population after progression under PD-1 therapy, although they need a longer follow-up. Also they showed that the patients that progressed under PD-1 therapy and that received BRAF/MEK still appeared to have a good outcome compared to other salvage therapies in patients that have BRAF mutated melanoma.

The third aspect is the systemic therapy in another melanoma subtype. Two posters that evaluated therapy in uveal melanoma, the first one that showed that in uveal melanoma using the genomic profiling for treatment selection led to 65% of patients that received targeted therapy based on this genomic profiling although the outcomes were rather unsatisfactory. For other work that showed that the predictors of best response to immunotherapy in this subset are the low LDH at the time of diagnosis and also the time from the first diagnosis to stage 4 that needs to be more than two years. So these patients would have better outcomes to immunotherapy. Also for patients with NRAS mutated melanoma work from Italy showed that the patients that are treated with immune checkpoint inhibitors as first line have equal benefit, regardless of their NRAS mutation status.
Finally, data that came from real-world analysis that evaluated systemic therapy in patients with melanoma brain metastasis, so combined immunotherapy, targeted therapy and comparison between the population of patients included in clinical trials and treated in a real world setting. The number of patients included in these three reports is quite high and they deserve a second look in order to see the outcomes of these patients with melanoma brain metastases treated in real-world settings. There are more data presented as posters but these are my picks for this highlight.

Mini-oral presentations

Moving into the mini-oral presentations, there were two trials that evaluated vaccination in melanoma patients, one in the adjuvant setting and another one in patients with advanced melanoma. The first one evaluated the therapy of dendritic cells versus placebo in patients with stage 3b and stage 3c melanoma. The trial didn’t achieve its primary endpoint because there was no difference in terms of relapse free survival for patients treated with placebo or dendritic cells. The trial was also stopped earlier due to the approval and reimbursement of pembrolizumab in this setting. However, there seems to be a biological effect with a tumour-specific immune response that is clearly seen after three doses of therapy.

The second trial evaluated the combination of nivolumab plus IDO PD-L1 peptic vaccine in patients with metastatic melanoma. The results reported include 30 patients with metastatic melanoma that were therapy naïve and with no melanoma brain metastasis. The combi therapy seems to be feasible and safe with a high response rate, with a response rate that is approximately 80% and higher than the subgroup that was used as a comparison and treated with PD-1 monotherapy. The complete responses were also very high, approximately 45%. The authors also demonstrated that there were vaccine specific T-cells in the PBMCs and also in the tumour site. What is still to be determined is if these immune responses translate into clinical benefit. This is obviously something that we need to wait for more follow-up and for more data.

Then there were two presentations on therapy in melanoma brain metastasis, the first one that looked into patients that progressed under BRAF/MEK and that received further ipilimumab-nivolumab. The authors looked not only into the response but also into the biology and the characterisation of the metastatic lesions that progress under this targeted therapy. The patients actually that received this second line of ipilimumab-nivolumab didn’t derive a large benefit from this second systemic therapy. They looked into the median intracranial progression free survival and it was only 5.5 weeks for these patients that received this second line. When looking into the lesions that actually progressed and having an NGS characterisation, the authors showed that there was a signature that is similar to the innate PD-1 resistant signature in this metastasis that progressed under targeted therapy. When they looked into the immunohistochemistry they also showed that this non-responsive metastasis also had an enrichment in terms of myeloid cell activation that is also known to induce less good responses to immunotherapy.

Another trial that also evaluated systemic therapy in melanoma brain metastasis, so in the same line of the first one that I just mentioned, is the NIBIT-M2 trial that evaluated ipilimumab plus nivolumab versus ipilimumab plus fotemustine or fotemustine in patients that were treatment naïve and asymptomatic and had melanoma brain metastasis. This follow-up, this longer follow-up, showed that there is a clear benefit in terms of median overall survival for the patients that were treated with the combination of ipilimumab nivolumab and also in terms of median progression free survival. The four-year overall survival rate was also higher for the patients that received combination immunotherapy, 41% versus 15% for the ipilimumab plus fotemustine versus 8% for the patients that received only fotemustine.  With this longer follow-up from this trial one can say that probably the first line therapy for patients with melanoma brain metastasis that did not receive previous therapy and that have no symptoms related to melanoma brain metastasis is the combination of ipilimumab plus nivolumab.

Then there were also three works presented on the therapy that can be provided to patients that progress during or after immune checkpoint inhibitor. The first one was the evaluation of local therapy given to patients that progressed under or after immune checkpoint inhibitor. The colleagues looked into these two subgroups, patients that progressed under immunotherapy and patients that progressed after stopping immunotherapy, with a similar number of patients in each group. What they’ve shown is that the patients that progressed during therapy are more often treated with local therapy and continuous systemic therapy. This seems to be also the approach that has more advantage in terms of overall survival. Patients that progress after stopping immunotherapy are more frequently treated with local therapy but those who benefit more in terms of time to second progression are those that receive local therapy but also restart immune checkpoint inhibitors.

However, the results that were presented so far do not show that there is an improvement in terms of overall survival for patients that receive local therapy plus restarting checkpoint inhibitors compared to the ones that received a single modality. What we can conclude is that probably the local therapy might benefit patients with stage 4 who progress with solitary lesions; what is missing is the localisation of these solitary lesions. So we don’t know if these localisations might be the reason for these patients to receive local therapy and if the local therapy is feasible based on the localisation of these progressive lesions.

A second trial that evaluated ? therapy in patients that progressed under PD-1 inhibitor was the ILLUMINATE-2 04. This trial evaluated the combination of ipilimumab plus intratumoural tilsotolimod. They included data from 67 patients and the results that were presented here on a heavily pre-treated population, so the patients included could have received up to two therapies if they were BRAF wildtype or up to three therapies if they were BRAF mutant patients. This might be a potential second line therapy for patients progressing under PD-1 therapy. The results that were presented here led also to the design of a phase III trial that will evaluate this combination versus ipilimumab alone.

What is missing, obviously, is how was this progressive disease determined, how was it confirmed? Because we say that we evaluate second line therapies in patients that progressed under PD-1 inhibitor but the definition of this progression is not homogeneous among the trials so it’s difficult to compare the results coming from all these trials that evaluate therapy after progression under PD-1.

Then in this topic of therapy after progression under PD-1, a trial that was presented as a proofed paper but that I will mention here because it’s on the same context as the two ones that I just mentioned, which is the LEAP-004 that evaluated lenvatinib plus pembrolizumab in patients that had progressed on the PD-1 and PD-L1 inhibitor. This was not a randomised trial, all the patients received this combination, and the overall response rate and the median overall survival for the patients that were included in this study, and particularly those that had progressed under PD-1 or PD-L1, it’s quite impressive, particularly when compared to other studies that also evaluated second line therapies in patients that had already progressed under PD-1 based therapies. The median overall survival in this case is actually almost 14 months.

To finalise the oral presentations there was a presentation evaluating the outcomes of patients with complete response under immune checkpoint inhibitor. This was a five-year characterisation of this subgroup of patients that were treated in the CheckMate 066, 067 and 069. Almost 400 patients, or more than 400 patients, were treated with ipilimumab plus nivolumab and more than 500 patients were treated with nivolumab monotherapy. 96 patients had a complete response when they were treated with ipilimumab plus nivolumab and from these almost 50% discontinued therapy due to toxicity and three due to progressive disease. On the other side, patients that received nivolumab monotherapy, 102 patients had a complete response and only 20% discontinued therapy due to toxicity and 6% due to progressive disease. The twelve-month landmark analysis was presented here both for patients with and without complete response. In terms of overall survival we don’t see a difference in the two subgroups of treatment regarding overall survival for those that had a complete response but for those that didn’t have a complete response there was a higher overall survival rate for patients that received ipilimumab plus nivolumab. For the PFS the same goes for the patients that had a complete response but for the patients that didn’t have a complete response a higher progression free survival rate was seen in the patients that were treated with ipilimumab plus nivolumab.
So what we see is that actually the patients that have a complete response have a higher five year overall survival in both therapy subgroups. For the patients that achieve a complete response there is a higher percentage of patients in the nivolumab plus ipilimumab group that are free of treatment at five years compared to those that receive nivolumab monotherapy. The patients that are off treatment in these two subgroups are more than 80% in the nivolumab plus ipilimumab arm and 68% in the nivolumab arm. There seems to be an association in terms of baseline characteristics of the patients and the complete response to these two immunotherapies, so normal LDH and lower tumour burden are associated with complete response in both subgroups, lower M stage is associated with complete response in the nivolumab plus ipilimumab arm. Interestingly, in the patients that have a PD-L1 expression of more than 5% it seems to be associated with a higher percentage of patients that achieve a complete response but only in the nivolumab arm. The maximum tumour reduction is important and probably incorporating on-therapy assessments may provide information on clinical outcomes. The benefits seem to exist regardless of therapy discontinuation. However, there was no information on how this complete response, or this cut-off of complete response, at twelve months was established and also there was no information or no further characterisation into the patients that had a complete response after these twelve months that was a group of more than 40% in both subgroups.

Proffered papers

Moving into the proofed papers section or in the proofed papers session there were three trials presented in the metastatic setting. The LEAP-004 I already mentioned, the two others were the COMBI-i trial and the SECOMBIT trial. The COMBI-i trial was a trial evaluating the combination of spartalizumab plus dabrafenib and trametinib in patients that were previously untreated and had a BRAF V600 mutation. This study didn’t meet its primary endpoint although it was very similar to two other trials, the KEYNOTE-024 that was also a negative study, or didn’t meet its primary endpoint, and a study that met its primary endpoint that is IMspire150.
During the discussion there was a very elegant explanation on why these so similar trials lead to a positive or negative, if they meet their primary endpoint or not. When we look into the curves they are actually overlapping, the curves from these three trials, so probably very small differences in terms of trial design and very small differences in terms of the population that is included might be the reason why we see some trials being positive and achieving the primary endpoint and others being negative and not achieving the primary endpoint.

It was also pointed during the discussion, and I find a very important question, whether we should be addressing if immunotherapy improves the outcomes of targeted therapy or the other way around – if we should look into immunotherapy and see if targeted therapy can improve or not the outcomes of immunotherapy since we know that targeted therapy is also immunogenic. So probably we need to consider the question in a different way and there are trials that will evaluate this question, as I said.

For the SECOMBIT trial this is also a very important trial that is addressing the question of how should we sequence therapy mostly in patients with BRAF mutation. They looked into three possible sequences, the first one that started with targeted therapy, a second one that started with immunotherapy and the third one that had targeted therapy before immunotherapy shortly after and then targeted therapy after that is mentioned as the sandwich approach for this study. Although the data is very preliminary, what we can see is that the twelve months progression free survival seems to be better for those that started with immunotherapy but after twelve months the curves cross. The median duration of response is higher for the patients that start with immunotherapy but we see also more toxicity in this arm. The total progression free survival seems to be better for those that have this sandwich approach and these were very preliminary reports so we really don’t know if we will continue to see this trend once we have an update of the trial.
One of the aspects that was mentioned and I find important is that also the time-point when this trial was designed was a time-point where we didn’t know the final results, or the updated results, of the overall survival of the combination of nivolumab plus ipilimumab and encorafenib plus binimetinib so it might be that the design that was proposed for this trial could be an issue when having a positive trial in the end. Also based on the data that we have now we cannot say that there is a better sequence approach in this setting.

Finally in the adjuvant setting there was an update on the trials on the KEYNOTE-054 and on the CheckMate 238. The trials are not exactly similar but they included patients that have received a complete lymph node dissection which is no longer the standard of care. In the CheckMate 238 there was an active comparator arm that was ipilimumab 10mg/kg. The update of the KEYNOTE-054 showed that the distant metastasis free survival the hazard ratio favours the therapy with pembrolizumab. This benefit is higher or more significant in patients with BRAF mutation with a hazard ratio of 0.54. This hazard ratio is quite similar to what we have seen in the COMBI-AD trial for the BRAF mutated patients.

There is an interesting aspect which is the fact that the benefit of pembrolizumab was not so significant in terms of local regional recurrences and this might be an aspect that will even be more evident because the majority of the patients that will receive this therapy nowadays will not have received the complete lymph node dissection. There are still pending results in terms of overall survival and the data from the crossover to pembrolizumab after relapse.

The update of the CheckMate 238, the authors also provided an update with 48 months of follow-up. The benefit of nivolumab continues to be evident compared to ipilimumab at this time-point. For the first time was presented data on patients with in-transit metastases and also there seems to be a benefit for nivolumab treatment compared to ipilimumab.  They reported for the first time data on overall survival and what we see is that at 48 months there seems to be no difference in terms of overall survival for the patients that receive nivolumab or ipilimumab and with an overall survival of approximately 78% in both groups, 77% for the ipilimumab. The difference, or the absence of difference, in these two subgroups is probably explained by the fact that when the patients progressed they were unblinded and more patients in the ipilimumab arm received PD-1 therapy subsequently. This subsequent therapy probably dilutes the benefit in terms of overall survival in the nivolumab group.

Take home messages

Take home messages – there is a lot of information coming from multicentre real-world analyses that are needed to understand how good drugs perform in real life and if the benefit is conserved across subgroups. This can be achieved by well-designed post-approval monitoring, excellent collaborative studies that have their basis on data from national registries.

Adjuvant studies are now assessing toxicity of regimens since efficacy has been proven, however, still more than 50% of the patients relapse and 35% of the patients relapse with distant metastasis. So we definitely need to improve either systemic therapy by using triple combinations or potentially the neoadjuvant approach or by better selecting the patients with these biomarkers that we are still awaiting.

In the metastatic setting the best sequential therapy is still yet to be defined, particularly for patients with BRAF mutations. Patients who progress under first line therapy, particularly if they progress with a solitary metastasis, seem to benefit from local therapy. Other therapeutic approaches with other therapeutic combinations, namely local therapy or intralesional therapy, might be an option but definitely the definition of progressive disease under first line therapy needs to be clear and needs to be homogeneous among the different trials that evaluate the second line of therapy. Achieving a complete response in the first twelve months of immunotherapy seems to be predictive of long-term survival and seems to be associated with a high possibility of not needing further systemic therapies at five years.

For patients that are treated with nivolumab, the expression of PD-L1 of more than 5% seems to be associated with a higher response rate with a higher number of patients that achieve a complete response but this doesn’t seem to be true for the patients that receive the combination therapy.

Thank you for listening, this is our highlights from the ESMO 2020 Congress, the first weekend and the data on the melanoma trials that have been presented. Thank you.