Atezolizumab combinations with targeted therapy for anaplastic thyroid carcinoma (ATC).

Abstract

6514 Background: ATC is a rare/aggressive cancer with dismal outcome. Dabrafenib/trametinib is approved for BRAF-mutated ATC but pts eventually develop resistance. There are no approved drugs for pts with BRAF-wild type ATC. Better treatments (tx) are needed. Methods: ATC pts with PS < 3 enrolled on a prospective trial at a single center, and tx was assigned by driver mutation: BRAF (cohort 1), RAS, NF1, or NF2 (cohort 2), or none of these (cohort 3). Cohort 4 with paclitaxel was exploratory for pts who did not qualify for 1-3. All pts received atezolizumab (A) IV + targeted therapy. Cohort 1 had run-in with vemurafenib (V) 960mg BID/cobimetinib (C) 60mg QD po for 28 days, followed by A 840mg Q2 weeks, at which time V dose was decreased to 720mg BID. Cohort 2: A + C (same doses as cohort 1); cohort 3: A 1200 mg Q3 weeks + bevacizumab 15 mg/kg q3 weeks. Pts unable to swallow used alternative drug preparation (ADP; crushed vemurafenib, suspension cobimetinib). Primary objective is to determine whether the tx in cohorts 1-3 leads to improved overall survival (OS). The trial was designed to enroll 36 pts but we are reporting early due to positive findings. Response rate (RR) was measured by RECISTv1.1. Median OS was estimated by Kaplan-Meier method. cfDNA and biopsy were obtained at baseline, course 2 and progression. Pts were allowed to undergo surgery and radiation while on trial. Results: From August 2017-January 2020, 34 ATC pts were enrolled in cohorts 1-3 and 9 in cohort 4. Cohort 3 closed early for futility. 3 pts had ADP. Median follow-up time was 7.51 mos (range: 0.43 – 27.37). Median OS in cohorts 1-3 was 18.23 mos (CI 10.45-NE) and 1-year OS was 67% (95%CI: 45%, 82%). See table. Response rate (RR) in cohort 1 was 71%: CR 1/17 (6%), PR 11/17 (65%), SD 4/17 (23%), 1 never restaged; in cohort 2 RR was 7%: PR 1/14 (7%), SD 7/14 (50%), PD 4/14 (29%), 2 died early. 8 (24% of cohort 1-3) pts had complete tumor resection after tx with VCA (n = 7) or CA (n = 1); all but 1 of these pts are alive. AEs as expected. cfDNA data will be reported at meeting. Conclusions: Atezolizumab + vemurafenib/cobimetinib for BRAF-mutated or + cobimetinib for NF1/2 or RAS-mutated ATC is effective, as evidenced by the long OS in these pts (13 mos > historical control). A significant number of patients, particularly in cohort 1, were able to undergo complete tumor resection due to a favorable response to tx. Clinical trial information: NCT03181100 . [Table: see text]