Comparison of gene alterations detected with tumor tissue and plasma in a Chinese population with advanced lung adenocarcinoma

e20511 Background: Currently, plasma cell-free DNA has been recommended as an alternative biomarker for guiding precision treatment in lung cancer. In this study, we retrospectively reviewed gene alterations detected by matched plasma and tumor tissue to investigate the concordance and sample-specific observations. Methods: The inclusion criterions included 1) samples collected between January and October 2022, 2) clinical stage IV, lung adenocarcinoma, 3) plasma taken within 4 weeks of tumor biopsy, 4) cell-free DNA and tumor DNA were extracted for next-generation sequencing (sequencing depth, 4000X for tissue and 10000X for plasma), with a pan-cancer 539-gene panel in a CAP-certificated laboratory. Alterations including mutations (SNV, indel), fusion and CNV of driver genes, including EGFR, ALK, ROS1, RET, MET, NTRK, BRAF, ERBB2, KRAS (9-gene), were emphatically reviewed and compared. Concordance was defined as tissue and plasma reported alterations within the same gene, or both negative. Tumor mutation burden (TMB) was calculated as ratio between total number of non-synonymous variants and total size of coding region. Results: After reviewing 630 patients with matched plasma and tumor samples, 121 patients diagnosed with stage IV lung adenocarcinoma were included. The top five altered genes in tissue were TP53 (71.07%), EGFR (57.86%), MYC (33.88%), MUC16 (23.97%), CDK6(23.97%), those in plasma were TP53 (52.89%), EGFR (40.5%), DNMT3A (28.93%%), MUC16 (18.18%), LRP1B (15.7%). The detection rates of driver gene (9-gene) alterations were 92.56% (112/121) for tissue and 69.42% (84/121) for plasma. The overall concordance of 9-gene alteration was 69.42%, it increased to 74.38% excluding fusions and CNVs. Among mutations, EGFR (61/121), KRAS (14/121), ERBB2 (9/121) and BRAF (3/121) were more often reported in tissue or plasma. The concordances were 71.13% ( EGFR), 66.67% ( BRAF), 55.56% ( ERBB2) and 42.86% ( KRAS) respectively. We also observed five mutations ( EGFR G719A/S768I, KRAS G12A/Q61L/A146P) that were only detected in plasma in four patients. One patient was detected with STRN-ALK only in plasma. These plasma-only mutations might provide extra efficacy predicting (sensitive or resistance) for EGFR/KRAS/ALK inhibitors. TMB of plasma and tissue showed strong correlation (Pearson’s Correlation Coefficient of 0.74). 14 and 13 patients were divided into TMB-H ( > 10 mutations/Mb) in tissue and plasma respectively, the concordance rate was 94.21% (114/121). Among qualified samples for MSI evaluations, non MSI-H was detected, the concordance rate was 100%. Conclusions: The sequencing results of matched plasma and tissue presented good concordance of driver gene alterations (especially for EGFR mutations), TMB-H and MSI-H. These suggested ctDNA testing might be a reliable commentary approach to tissue to guide target and immunotherapy in advanced lung adenocarcinoma.