The study is a randomised phase III trial that compares combination immunotherapy to combination BRAF MEK targeted therapy in the front line and if patients exhibit progression and MEK eligibility criteria, they could cross over to the alternative trial. The DREAMseq trial was taking treatment-naïve patients with metastatic BRAF mutant melanoma, stratifying them by performance status and LDH and randomising them to those two sequences. Because we thought the curves might actually cross, we chose two-year overall survival as the primary endpoint.

The study started in 2015 and in 2021, with 59% of patients having crossed the two-year timepoint, the Data Safety Monitoring Committee recommended closing the trial because of a clinically meaningful 20% difference in two-year overall survival, favouring the immunotherapy first sequence with 72% of patients being alive at two years with immunotherapy first and 52% with targeted therapy.

This data was presented at the inaugural ASCO virtual plenary session in 2021. We followed the patients to five years in this study, and at ASCO I am presenting the five-year follow-up clinical data. What we saw was a sustained benefit in terms of overall survival out to five years, with now a 30% absolute difference in overall survival for the immunotherapy first sequence: 60% of patients alive versus 30%.

We also saw a three-fold improvement in five-year PFS and a continued increase in duration of response favouring immunotherapy over targeted therapy. We then looked at the various subsets in the trial – age, sex, performance status, LDH levels, and a marker of good performance status which are the patients who do best with targeted therapy. In every group, the four-year PFS was significantly better for the immunotherapy arm and in almost all those groups the four-year overall survival was significantly better for the immunotherapy first sequence.

We then looked at incidence of CNS metastases, and we saw that there were double the number of CNS metastases as first site of relapse on the targeted therapy arm compared to the immunotherapy arm, indicating that immunotherapy was working better in the CNS. The patients who had CNS relapse by and large didn’t cross over and did very poorly.

We then looked at the response rates and saw a similar 51% response rate between the first line immunotherapy and first line targeted therapy. We were surprised a little bit that the response rate wasn’t higher for targeted therapy because it was higher in the registrational trials. We wanted to see what was the cause of the lower response rate, and we saw that if we looked at 12 weeks there were many more responses on the targeted therapy than the immunotherapy. By time the we got out to 24 weeks, 41% of those responses had progressed and so could not be confirmed, while for immunotherapy only about 16% of the responses couldn’t be confirmed, illustrating the better durability of benefit if you respond to immune therapy.

In summary, this long-term follow-up further supports the value of giving immunotherapy first for patients with treatment-naïve metastatic BRAF mutant melanoma. The reason why it was beneficial was better duration of response and fewer CNS relapses. This applied to practically every subgroup and so I think it’s now clear that for most patients who have BRAF mutant metastatic melanoma, they should be receiving combination immunotherapy first.

Is there anything else you would like to add?

I would add that we have the tissue and the blood from these patients and we’re looking at biomarkers in the tumour and the blood for two things: to see if we can identify a population that should get targeted therapy first, and a population who gets immunotherapy first who is progressing early and should be switched to targeted therapy earlier. That work is ongoing and I hope to present that in the next year.