Leveraging a hybrid tumor-informed and tumor-agnostic ctDNA assay for optimal ctDNA-MRD detection: A real-world study in Southeast Asia.

3058 Background: While tumor-informed ctDNA assays are the gold standard for minimal residual disease (MRD) detection, their utility is frequently limited by suboptimal tissue quality and quantity. This study evaluated a hybrid strategy combining personalized tumor-informed mutations and a tumor-agnostic hotspot panel to improve ctDNA-MRD detection in a real-world patient cohort. Methods: We retrospectively analyzed 1015 blood samples of 577 patients diagnosed with stage I-IV lung (n=222), colorectal (n=172), and breast (n=143) cancers. The hybrid assay (K-4CARE, Gene Solutions) integrated patient-specific variants with a cancer-specific tumor-agnostic hotspot panel (K-4CARE, Gene Solutions). ctDNA status was correlated with clinical outcomes in 200 early-stage patients (median follow-up >6 months) and 59 metastatic-stage patients. Results: The hybrid approach significantly outperformed the tumor-informed method, increasing pre-treatment ctDNA detection rates by 33.3–48.8% in cases with suboptimal tissue quality. In the early stage, post-operative ctDNA was detected in 93.5% (29/31) of the patients having recurrence, with the median lead time of 8.4 months; while 97.6% (165/169) of those with no recurrence had negative results. ctDNA positivity was an independent prognostic factor (p<0.001) that increased recurrence risk in lung (HR=145.4; 95% CI: 37.4-564.5), colorectal (HR=173.6; 95% CI: 19.7->300.0), and breast (HR=147.8; 95% CI: 26.3->500.0) cancers. In the metastatic stage, at 12 weeks after therapy initiation, 85.3% (29/34) of the molecular responders, who had ctDNA cleared or reduced by >50% of baseline VAF level, achieved clinical partial response or stable disease; while 88.0% (22/25) of the molecular non-responders had confirmed progressive disease. Notably, the tumor-agnostic hotspot panel helped identify new and resistance mutations, including EGFR T790M in lung (20.0%), ESR1 Y537N (7.1%), ESR1 Y538G (3.6%), ESR1 P535H (1.8%) in breast (12.5%), and KRAS G12D in colorectal (25.0%) cancers. Conclusions: This hybrid approach improved sensitivity to detect ctDNA-MRD, and the capability to identify emergent resistance mutations provides expanded clinical utility for longitudinal monitoring.