Atezolizumab, vemurafenib, and cobimetinib in patients with melanoma with CNS metastases (TRICOTEL): a multicentre, open-label, single-arm, phase 2 study.

BACKGROUND: Targeted therapy and immunotherapy have shown intracranial activity in melanoma with CNS metastases, but there remains an unmet need, particularly for patients with symptomatic CNS metastases. We aimed to evaluate atezolizumab in combination with cobimetinib or vemurafenib plus cobimetinib in patients with melanoma with CNS metastases. METHODS: TRICOTEL was a multicentre, open-label, single-arm, phase 2 study done in two cohorts: a BRAFV600 wild-type cohort and a BRAFV600 mutation-positive cohort, recruited at 21 hospitals and oncology centres in Brazil, France, Germany, Hungary, Italy, Spain, and Switzerland. Eligible patients were aged 18 years or older with previously untreated metastatic melanoma, brain metastases of 5 mm or larger in at least one dimension, and an Eastern Cooperative Oncology Group performance status of 2 or less. Patients in the BRAFV600 wild-type cohort received intravenous atezolizumab (840 mg, days 1 and 15 of each 28-day cycle) plus oral cobimetinib (60 mg once daily, days 1-21). Patients in the BRAFV600 mutation-positive cohort received intravenous atezolizumab (840 mg, days 1 and 15 of each 28-day cycle) plus oral vemurafenib (720 mg twice daily) plus oral cobimetinib (60 mg once daily, days 1-21); atezolizumab was withheld in cycle 1. Treatment was continued until progression, toxicity, or death. The primary outcome was intracranial objective response rate confirmed by assessments at least 4 weeks apart, as assessed by independent review committee (IRC) using modified Response Evaluation Criteria in Solid Tumours version 1.1. Because of early closure of the BRAFV600 wild-type cohort, the primary endpoint of intracranial objective response rate by IRC assessment was not done in this cohort; intracranial objective response rate by investigator assessment was reported instead. Efficacy and safety were analysed in all patients who received at least one dose of study medication. This trial is closed to enrolment and is registered with ClinicalTrials.gov, NCT03625141. FINDINGS: Between Dec 13, 2018, and Dec 7, 2020, 65 patients were enrolled in the BRAFV600 mutation-positive cohort; the BRAFV600 wild-type cohort was closed early after enrolment of 15 patients. Median follow-up was 9·7 months (IQR 6·3-15·0) for the BRAFV600 mutation-positive cohort and 6·2 months (3·5-23·0) for the BRAFV600 wild-type cohort. Intracranial objective response rate was 42% (95% CI 29-54) by IRC assessment in the BRAFV600 mutation-positive cohort and 27% (95% CI 8-55) by investigator assessment in the BRAFV600 wild-type cohort. Treatment-related grade 3 or worse adverse events occurred in 41 (68%) of 60 patients who received atezolizumab plus vemurafenib plus cobimetinib in the BRAFV600 mutation-positive cohort, the most common of which were lipase increased (15 [25%] of 60 patients) and blood creatine phosphokinase increased (11 [18%]). Eight (53%) of 15 patients treated with atezolizumab plus cobimetinib in the BRAFV600 wild-type cohort had treatment-related grade 3 or worse adverse events, most commonly anaemia (two [13%]) and dermatitis acneiform (two [13%]). Treatment-related serious adverse events occurred in 14 (23%) of 60 patients who received triplet therapy in the BRAFV600 mutation-positive cohort and two (13%) of 15 in the BRAFV600 wild-type cohort. No treatment-related deaths occurred. INTERPRETATION: Atezolizumab plus vemurafenib and cobimetinib provided intracranial activity in patients with BRAFV600-mutated melanoma with CNS metastases. FUNDING: F Hoffmann-La Roche.