Biomarkers of emergent resistance to sotorasib plus panitumumab in <i>KRAS</i> G12C-mutated metastatic colorectal cancer (mCRC) from the randomized, phase 3 CodeBreaK 300 study.

Abstract

3540 Background: The use of sotorasib (soto; KRAS G12C inhibitor) alone in patients with KRAS G12C-mutated mCRC gives rise to several receptor tyrosine kinase (RTK) alterations, resulting in treatment resistance. In CodeBreaK 300 trial, the addition of panitumumab (pani; monoclonal anti-EGFR antibody) counters this resistance and significantly improves clinical outcomes compared with standard of care (SoC; trifluridine-tipiracil or regorafenib). Over time this combination (soto+pani) may lead to the emergence of new resistance patterns. This study reports the distribution, patterns, and prevalence of genetic alterations that arise after treatment with soto+pani. Methods: Patients from the phase 3 CodeBreaK 300 trial with chemorefractory KRAS G12C-mutated mCRC who had paired plasma samples at baseline and at progression, were included in the analysis. The samples were evaluated using the 753-gene Guardant Infinity ctDNA test. Gene alterations that were absent at baseline but present at progression were considered. Results: By December 2024, 99 patients (median age: 62 years, female: 48%) were evaluated (soto 960 mg-pani: 32 [60% of ITT population], soto 240 mg-pani: 33 [62% of ITT population], SoC: 34 [63% of ITT population]). Overall, 90% of patients had at least 1 emergent, likely pathogenic, genomic alteration at progression. Median time to progression in biomarker-evaluable patients was 3.8 months for soto 960 mg-pani arm, 3.6 months for soto 240 mg-pani arm, and 2.0 months for SoC arm. Overall, the distribution, patterns, and prevalence of pathogenic emergent alterations were similar across all treatment arms. The most common pathogenic emergent alterations included TP53 (34%), DNMT3A (17%), ERBB2 (12%), and LRP1B (11%), generally associated with the RTK, cell cycle control, DNA methylation, and DNA damage response pathways. The median copy number of emergent KRAS copy number variations (CNVs) was higher (p = 0.007) in the soto 960 mg-pani (4.18) and soto 240 mg-pani (4.24) arms compared with the SoC arm (2.05). Emergent KRAS CNVs were primarily present in the soto+pani arms (soto 960 mg-pani: 40.6% [n = 13], soto 240 mg-pani: 36.4% [n = 12], and SoC: 14.7% [n = 5]). The presence of emergent pathogenic variants in ALK (n = 7) and KMT2D (n = 4) was observed exclusively among patients treated with soto+pani. DNMT3A mutations, along with other diverse emergent alterations, were observed in all three evaluable patients with partial response in the soto 960 mg-pani arm. Conclusions: Dysregulation of the DNA methylation and RTK pathways and KRAS amplifications may contribute to the development of resistance to soto+pani combination. Further characterization of these acquired alterations, can help inform future therapeutic strategies. Clinical trial information: NCT05198934 .