Detecting <i>MTAP</i> loss in liquid biopsies from patients with clinically advanced colorectal cancer (CRC).

3553 Background: Homozygous loss of the MTAP gene ( MTAP loss) is an emerging biomarker guiding investigational use of PRMT5 and MAT2A inhibitors in a wide variety of tumor types, including CRC. Detecting copy number changes, including homozygous losses, in solid tumors is often challenging and requires robust assay and analytic pipeline especially when the amount of extracted DNA is small. In clinically advanced CRC, small endoscopic and needle-based biopsies are often used for diagnosis, imposing significant limitations on immunohistochemistry (IHC) and comprehensive genomic profiling (CGP) required for treatment selection and clinical trial enrollment. We assessed whether liquid biopsy (LBx) CGP could provide treatment-relevant information on MTAP loss compared to tissue biopsy (TBx) CGP for patients with CRC. Methods: Hybrid capture-based CGP was performed on 46,173 clinically advanced CRC (stage III, IV) TBx cases using the FoundationOne CDx assay and on 9,049 LBx cases using the FoundationOneLiquid CDx assay. Tumor fraction (TF) for each LBx sample was determined using assessments of aneuploidy and variant allele frequencies, as previously described. Results: MTAP loss was detected in 498 (1.08%) TBx cases. For LBx, detection of MTAP loss increased with higher TF, with detections observed only at TF ≥5% (Table 1). With TF ≥5%, the frequency of MTAP loss detection was comparable to or exceeds that observed with TBx. Regarding complete vs partial MTAP exon loss, TBx and LBx showed similar patterns, with complete loss of all 8 MTAP exons observed in 85.7% of TBx cases and 84.5%-89.8% in LBx cases, increasing in higher TF levels. Conclusions: LBx emerges as a promising tool for detecting MTAP loss in clinically advanced CRC, with MTAP loss detection rates approaching those of TBx when LBx TF is ≥5%, while detection is limited at lower TF levels. Incorporation of LBx for MTAP loss detection may expand patient access to biomarker-driven clinical trials involving PRMT5 and MAT2A inhibitors for patients with clinically advanced CRC when sufficient tissue for tissue-based CGP is not feasible. MTAP loss detected by TBx and LBx in clinically advanced CRC. Cohort (cumulative)* MTAP no loss MTAP loss MTAP loss frequency TBx (n=46173) 45675 498 1.08% LBx TF ≥0% (100% of LBx, n=9049) 9002 47 0.52% LBx TF ≥1.0% (62.7% of LBx) 5624 47 0.83% LBx TF ≥5.0% (48.7% of LBx) 4359 47 1.07% LBx TF ≥10.0% (41.5% of LBx) 3712 46 1.22% LBx TF ≥20.0% (31.5% of LBx) 2812 38 1.33% LBx TF ≥30% (24.8% of LBx) 2214 29 1.29% *TF thresholds represent cumulative subsets of the LBx cohort.