Real-world evidence on genomic profiles and survival outcomes in patients with HER2-positive metastatic colorectal cancer.

Abstract

58 Background: Patients (pts) with HER2-positive (HER2+) metastatic colorectal cancer (mCRC) represent a small, but increasingly important subgroup due to emerging HER2-targeted therapies. Although HER2 positivity is associated with resistance to anti-EGFR therapy (EGFRi), it is unclear if EGFRi or bevacizumab (BEV) is preferable as first-line (1L) therapy. Here, we offer novel evidence on the genomic profiles of pts with HER2+ mCRC and examine differences in survival between key treatment groups. Methods: This study used the US-based deidentified Flatiron Health-Foundation Medicine mCRC clinico-genomic database. The deidentified data originated from approximately 280 US cancer clinics (~800 sites of care). Included pts were 18 years or older, had evidence of stage IV or recurrent mCRC diagnosed between 01 Jan 2012 and 31 Mar 2022, and underwent tissue-based Comprehensive Genomic Profiling testing. HER2+ was defined as an ERBB2 copy number of ≥5 for diploid tumors. A doublet regimen was defined as fluoropyrimidines plus oxaliplatin or irinotecan. Data cut-off was 31 Mar 2024. Kaplan-Meier methods were used to examine real-world overall survival (rwOS) and progression-free survival (rwPFS) overall, by 1L treatment class and across key subgroups defined by tumor sidedness and genomic profile, indexed to the start of 1L with pts censored at their last EHR-recorded activity. Results: Among a cohort of 5545 pts, HER2+ was detected in 171 (3.1%) pts. Median age (years) was 57 for HER2+ vs 61 for HER2-; 69.3% of HER2+ pts were White, compared to 72.1% in HER2-. Among pts with HER2+, 31 (18.1%) had RAS mutation (mt), BRAF V600E mt, or microsatellite instability-high. ERBB2 copy number was significantly lower in pts with RAS mt or BRAF V600E mt compared to those without these mutations (p < 0.01, p = 0.045, respectively). Among pts who received 1L treatment (n=4748), rwPFS significantly differed between HER2+ (n=144) and HER2-negative (HER2-; n=4604) groups (7.6 vs 8.7 months, unadjusted HR (uHR) 1.20, p=0.04). This difference was consistent in pts with left-sided RAS/BRAF V600E wild-type microsatellite stable (MSS) mCRC and treated with 1L doublet plus EGFRi (8.7 vs 12.5 months, uHR 2.18, p=0.02; n=10 HER2+,n=117 HER2-) or 1L doublet plus BEV (8.9 vs 10.5 months, uHR 1.65, p=0.04; n=20 HER2+, n=387 HER2-). No significant rwPFS difference was observed between EGFRi and BEV in the HER2+ subgroup. rwOS showed no significant difference between HER2+ and HER2- groups (overall: 26.1 vs 24.5 months, uHR 0.94, p=0.52; n=144 HER2+, n=4604 HER2-). Conclusions: Although about 80% of pts with HER2-positive mCRC are RAS/BRAF V600E wild-type and eligible for EGFRi, no survival benefit was observed in EGFRi relative to BEV as first-line combination. While the impact of potential confounders needs to be examined in further analyses, these findings highlight the need for specific treatments for pts with HER2+ mCRC.