1259PSwitch to checkpoint inhibition (CPI) after targeted therapy (TT) at time of progression or during ongoing response: A retrospective analysis of patients with advanced BRAF mutated melanoma

Background: The outcome of patients (pts) with BRAFV600 mutated advanced melanoma has dramatically improved by the introduction of TT and CPI. Pts responding on CPI have a high chance of long-term benefit. Pts with fast progressive disease (PD), high LDH and/or symptomatic (brain) metastasis often start with TT to induce a fast response, although most pts will relapse. Short-term TT increases intratumoral T-cell infiltration, which is diminished at PD. In addition, LDH level normalization and improvement of performance status (PS) during response to TT, might also argue for an earlier switch to CPI before PD. Methods: We retrospectively identified pts that started treatment with TT and switched to subsequent CPI (either anti-PD-1 or anti-PD-1 plus anti-CTLA-4) at our institute. Progression was defined as radiological PD according to RECIST 1.1 or clinical PD defined by treating physician. Progression free survival (PFS) from start CPI (PFS-CPI), start TT (PFS-total) and overall survival (OS) were analyzed by Kaplan Meier methods and hazard ratio’s (HR) were calculated by Cox regression analysis. Results: In total 74 pts were included; 37 pts switched to CPI prior to PD (noPD-switch) and 37 pts switched at PD (PD-switch); 32 vs 51% had brain metastases and 41 vs 49% had an elevated LDH. Median time on TT in the noPD-switch and the PD-switch group was 3.2 and 4.5 months, respectively. PFS-CPI was 2.5 months in the noPD-switch group vs 1.2 months in the PD-switch group (p = 0.11). OS was superior in the noPD-switch group with a median of 30.6 vs 14.1 months (p = 0.01). After correcting for previous treatment, LDH, PS, brain metastasis and number of metastatic sites, HR for OS was 0.48 (95% CI 0.22–1.01). Subgroup analysis revealed that pts with brain metastasis seem to benefit most from switch prior to PD. In pts that had reinduction of TT after CPI failure, the noPD-switch group benefitted longer from this subsequent TT line, although total time of benefit on TT (before and after CPI) was the same in both groups. Conclusions: Pts that switch to CPI prior to PD upon TT have an increased OS, which can be attributed to the benefit from CPI and not the total time on TT. Legal entity responsible for the study: Netherlands Cancer Institute. Funding: Has not received any funding. Disclosure: J.V. van Thienen: Travel support: Roche. J.B.A.G. Haanen: Advisory boards: BMS, Merck, Roche, Neon Therapeutics, Pfizer, Ipsen; Grants to NKI from BMS, Merck, Novartis, Neon Therapeutics, outside the submitted work. C.U. Blank: Advisory boards: MSD, BMS, Roche, GSK, Novartis, Pfizer, Lilly; Grants: BMS, Novartis; Stock ownership: Verastem outside the submitted work. All other authors have declared no conflicts of interest.