Circulating Tumor Dna Clone Analysis To Predict Disease Progression/Indicates Trastuzumab-Resistant Mechanism In Advanced Gastric Cancer.

e24030 Background: Circulating tumor DNA (ctDNA) detection could improve the accuracy of imaging methods for evaluating therapy efficacy in advanced gastric cancer (AGC). However, the feasibility of ctDNA analysis based on single genes or a few mutation positions is challenging in heterogeneous tumors. In addition, the trastuzumab resistance mechanism of AGC is incompletely understood. Methods: This prospective study enrolled 49 AGC patients (pts), including 21 pts receiving chemotherapy plus trastuzumab and 18 pts receiving chemotherapy alone. Serial peripheral blood samples (10 mL) were collected from each patient at baseline and every 2–4 treatment cycles. Fourteen matched formalin-fixed, paraffin-embedded tumor tissues were obtained. Pan-cancer panel sequencing was used to analyze somatic mutations. PyClone was employed to analyze the clonal structure. Molecular tumor burden index (mTBI) were calculated with clonal mutations. Results: A total of 121 and 146 functional mutations were identified in 14 paired tumor tissue and pretreatment plasma samples, respectively. Clonal mutations in plasma samples were derived from the matched tumor, with a median predicted cancer cell fraction of 89%. Compared with single gene analysis (TP53, patient coverage = 71%), mTBI based pan-cancer panel sequencing improved patient coverage to 100%. Serial mTBI analysis identified progressive disease (PD) 3–48 weeks (mean = 18 weeks) before imaging results. In validation cohort of 16 pts, mTBI increased to exceed the cutoff before or at clinical PD, demonstrating 94% sensitivity. We identified 32 expanding mutations potentially related to trastuzumab resistance by clonal structure analysis of ctDNA during anti-HER2 treatment. Multiple pathways activating in the same patients revealed the heterogeneity of the trastuzumab resistance mechanism. In addition, a higher mTBI in pretreatment ctDNA was also shown to be a risk factor for progression-free survival (HR, 13.75; 95% CI, 2.69–70.24; P= 0.002). Conclusions: Analysis of ctDNA clones based on pan-cancer panel sequencing is a promising approach to enhance the clinical management and improve therapeutic strategy for AGC patients.