Abstract 446P
Background
This study assessed overall survival (OS) of 1L mCRC pts with BRAF V600 and non-V600, KRAS G12 and G13, and NRAS mutations from RWD.
Methods
The Flatiron Health-Foundation Medicine Clinicogenomic Database (CGDB) was used, comprising de-identified EHR-derived clinical data linked to genomic alteration data for pts treated in the Flatiron Health network who received Foundation Medicine, Inc’s next-generation sequencing testing.1L treatments (txs) in 3462 adult mCRC pts newly diagnosed between 01/01/11 and 09/30/19 were categorized as: Anti-EGFR, anti-VEGF, other targeted, chemo, immuno, and other. OS was calculated from the start of 1L tx. Pts with BRAF V600, BRAF non-V600, KRAS G12, KRAS G13 or NRAS mutations were compared with wild type (WT) pts* using proportional-hazards regression models, either independent of, or stratified by, tx type. Analyses were adjusted for age, gender, race, stage, ECOG performance status, and microsatellite instability (MSI) status. Models were adjusted for left censoring to account for delayed entry into the study. *excluding the above mutations and other KRAS alterations of known/likely function.
Results
Relative to WT pts, BRAF V600, KRAS G12 and G13 pts had more females, more colon and less rectal tumors; BRAF V600 and KRAS G12 pts were older; BRAF V600 pts had higher MSI, and KRAS G12 pts had lower MSI. Unadjusted median OS and adjusted between-group differences across txs and stratified by tx with sufficient pts are listed (table). Table: 446P
| Pts | All | Anti-VEGF | Chemo | |||||||
| n | Median (CI) | HR (CI) | n | Median (CI) | HR (CI) | n | Median (CI) | HR (CI) | ||
| WT | 1002 | 21.3 (19.1, 22.9) | 1 | 565 | 21.8 (19.0, 24.1) | 1 | 257 | 20.0 (16.4, 23.8) | 1 | |
| BRAF | V600 | 214 | 10.0 (8.8, 12.1) | 2.3 (1.9, 2.9) | 112 | 10.1 (8.4, 14.4) | 2.4 (1.8, 3.3) | 61 | 9.5 (6.3, 13.0) | 2.9 (1.9 ,4.4) |
| non-V600 | 91 | 18.5 (14.3, 21.4) | 1.5 (1.1, 2.1) | 51 | 18.5 (13.9, 22.9) | 1.7 (1.2, 2.5) | 29 | 19.3 (11.6, NA) | 1.2 (0.7, 2.2) | |
| KRAS | G12 | 951 | 16.0 (14.8, 17.4) | 1.5 (1.3, 1.7) | 554 | 17.4 (15.6, 18.9) | 1.5 (1.2, 1.8) | 333 | 15.0 (12.9, 17.4) | 1.4 (1.1, 1.8) |
| G13 | 247 | 15.3 (12.9, 20.5) | 1.5 (1.3, 1.9) | 149 | 19.2 (16.0, 22.1) | 1.6 (1.2, 2.0) | 79 | 11.8 (6.5, 20.5) | 1.4 (1.0, 2.0) | |
| NRAS | 131 | 16.2 (12.0, 20.9) | 1.4 (1.1, 1.8) | 70 | 20.9 (17.8, 27.6) | 1.1 (0.8, 1.6) | 40 | 8.0 (2.4, 19.4) | 1.7 (1.1, 2.8) |
Conclusions
The results highlight the potential value of examining genomic alterations using RWD in CRC pts from community settings. OS differences between WT and BRAF, KRAS, or NRAS were observed by tx type. Whether differences in outcomes are determined by the mutations or txs remains to be investigated. Further understanding of the interplay between genomic alteration, tx type and outcome will enable personalized care to optimize each pt’s outcome.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
Roche.
Funding
Roche/Genentech.
Disclosure
A. Gershon, R. Machado, H. Trinh, Z. Shi, Y. Yan: Shareholder/Stockholder/Stock options: Roche. Q. Zhang: Shareholder/Stockholder/Stock options: Roche; Shareholder/Stockholder/Stock options: Regeneron; Shareholder/Stockholder/Stock options: BMS; Shareholder/Stockholder/Stock options: AbbVie. All other authors have declared no conflicts of interest.