LAUSR

e13163 Background: Tumor mutation burden (TMB) is recognized as a promising biomarker for PD-1/PD-L1 blockade therapy. However, a tissue biopsy is often not available, and hence a liquid biopsy using blood can be used to evaluate TMB (which is known as blood TMB). Unfortunately, the concordance of blood TMB (bTMB) and tissue TMB (tTMB) is not stable, and may be affected by the sample preparation methodology and sequencing depth. Therefore, e designed a study to explore the factors, which may significantly impact the concordance between bTMB and tTMB. Methods: From September 2018 to January 2019, 98 patients with pan-cancer were prospectively enrolled. For each patient, a tissue sample and paired plasma sample were collected. These samples were sequenced using a custom 605 cancer specificgene panel. Results: The data indicated that bTMB and tTMB were correlated, but the DNA input, sequencing depth and maximum somatic allele frequency (MSAF) might affect their concordance. The Spearman’s rank correlation (SRC) between bTMB and tTMB of 38 patients with > = 25ng cfDNA vs. 60 patients with < 25ng cfDNA was 0.77 vs. 0.15, respectively. Additionally, we explored the impact of sequencing depth and determined that the SRC (between bTMB and tTMB) of 25 patients with a sequencing depth > = 2,500x vs. 73 patients with a sequencing depth < 2,500x was 0.74 vs. 0.28, respectively. Finally, we showed that the allele frequency was critically important as well. We determined that the SRC (between bTMB and tTMB) of 50 patients with cfDNA MSAF > = 1% vs. 48 patients with cfDNA < 1% was 0.71 vs. -0.03. Conclusions: To improve concordance between bTMB and tTMB for better predicting efficacy of immunotherapy, we recommend that cfDNA input should be more than 25ng, and average effective sequencing depth should be higher than 2,500x, and the correlation hold for an allele frequency treating than 1%.