OBJECTIVES

The combination dabrafenib (BRAFi) and trametinib (MEKi) was approved in Canada for treatment of BRAF V600-mutated metastatic NSCLC based on a single arm phase II trial (NCT01336634). Given the rarity of this mutation and lack of available comparative trial evidence, the current analysis compares overall survival (OS) of dabrafenib+trametinib within the 3-year trial follow up with a similar cohort of real world patients receiving standard of care.

METHODS

The nationwide Flatiron Health electronic health record (EHR)-derived de-identified database was queried for patients with BRAFV600 mNSCLC receiving first-line (1L) platinum-based-chemotherapy, 1L pembrolizumab+carboplatin+pemetrexed, or second-line (2L) PD(L)1 monotherapy. Key inclusion and exclusion criteria were consistent with NCT01336634. Weighting by odds was used to adjust for differences between the cohorts at baseline. OS from initiation of therapy was compared using Kaplan-Meier and Cox proportional-hazards analysis.

RESULTS

Median 1L pre-weighted OS for dabrafenib+trametinib (n=36) vs platinum-based chemotherapy (n=64) was 17.3 vs. 14.5 months (p=0.11, HR: 0.65 [0.39, 1.1]) Weighting by odds resulted in an adjusted OS of 17.3 vs. 9.7 months (weighted n= 36.9, p=0.01, HR: 0.51 [0.29, 0.92]).

Median 2L+ unweighted OS for dabrafenib+trametinib (n=57) vs 2L PD(L)1 (n=42) was 18.2 vs 11.1 months (p=0.09, HR: 0.65 [0.39, 1.07]). Weighting by odds resulted in an adjusted OS of 18.2 vs 11.2 months (weighted n=54.3, p=0.55, proportionality assumptions not met).

Data for the 1L pembrolizumab+carboplatin+pemetrexed analysis were challenging to interpret due to high censoring

CONCLUSIONS

Patients with BRAFV600 mNSCLC receiving dabrafenib+trametinib in the NCT01336634 study had better 1L survival than observed in similar real-world patients receiving platinum-based chemotherapy, and numerically higher but not statistically different 2L+ survival to real-world patients receiving 2L PD(L)1 monotherapy. Limitations of this analysis include limited patient numbers; and incomplete BRAF V600E status and missing information on presence of baseline brain metastases in the real world cohort.