Tumor-agnostic therapies in a high-volume Brazilian center: Real-world challenges and opportunities.

e23304 Background: Advances in precision oncology led to the approval of tumor-agnostic therapies (TAT), challenging conventional drug development. Despite promising Phase I/II data leading to multiple drug approvals, real-world data remains scarce, hindering confirmation of reproducibility, especially in Latin America. Methods: In this retrospective analysis, we collected data from patients (pts) who underwent comprehensive genomic profiling (CGP) between July 2015 and September 2023 at Hospital Israelita Albert Einstein (São Paulo, Brazil). Clinical and genomic data were collected for patients whose biomarkers lacked tumor-specific approvals and who had at least one FDA-approved or emerging agnostic biomarker, including MSI-H, TMB-H, gene fusions (NTRK, ALK, RET, NRG1, FGFR), mutations (BRAF V600E, HER2, KRAS G12C), or HER2 amplification. Results: Of 1271 pts, 164 (12.9%) would be eligible for TATs; of these,78% had FDA-approved agnostic-biomarkers while 22% had emerging agnostic indications. Median age was 66 years (IQR: 55-75). TMB-H was the most frequent biomarker (53%), followed by KRAS G12C (14.6%), HER-2 amplification, MSI-H and BRAF (10.9% each). Most pts eligible for TATs had lung cancer (21%), followed by 14% CUP, 12% colorectal, 11% pancreatobiliary, 9% breast, 5% skin and 22% other histologies. Twenty-seven pts (16%) had more than one tumor-agnostic biomarker, mostly TMB-H and MSI-H. Germline testing was available for 10.3%, while treatment data was available for 68%. Forty pts (24%) received TAT: 37% checkpoint inhibitors, 12.5% Trastuzumab deruxtecan, 7.5% trastuzumab plus pertuzumab, crizotinib and sotorasib, 5% dabrafenib plus trametinib and 23% others. The objective response rate was 53% (34% partial and 17% complete responses). Most pts received TAT in the third-line setting (28%), followed by first-line (26%). Median PFS on TAT was 11 months. Pts who received TAT had a higher median OS compared to those who did not (92 versus 37 months, p = 0.0095). Outcomes by histology are shown in Table 1. Conclusions: In this real-world retrospective analysis of a high-volume Latin American center, CGP enabled the identification of agnostic biomarkers, offering targeted therapeutic options and improved outcomes for select individuals. The analysis was limited by unmatched distribution of tumor types, lacking treatment history and comorbidities among groups. Limited access to TAT may reflect the timing of tests, pre-dating some agnostic approvals, as well as the high cost and restricted availability of these therapies. Histology mOS TAT+ (mo) mOS TAT- (mo) P-value Lung (n 36) 35 (13-NA) 24 (18-NA) 0.75 CUP (n 23) 6 (3-NA) 8 (5-NA) 0.3 Colorectal (n 20) 16 (11-NA) 19 (8-NA) 0.57 Pancreatobiliary (n 18) 37 (37-NA) 8 (3-NA) 0.017 Sarcoma (n 5) 14 (14-NA) 12 (12-NA) 0.16