Abstract 5365: Comparison of whole genome sequencing and target capture sequencing in lung cancer

Abstract Introduction: Comprehensive genomic profiling (CGP) tests using target capture sequencing are commonly employed clinically to detect gene alterations and to recommend treatment. Whole-genome sequencing (WGS) is capable of detecting all mutations and structural variants of the genome and may offer advantages over CGP. On the other hand, WGS covers the entire genome with a low depth of coverage, which may limit its sensitivity. This study evaluates the concordance of variants detected by WGS and Todai OncoPanel (TOP), a CGP test targeting 737 DNA and 455 RNA cancer-related genes. Methods: We analyzed 32 consecutive lung cancer specimens collected through surgical resection, including two samples from the same patient, which underwent both WGS and TOP. Fresh frozen samples were used for WGS, while formalin-fixed paraffin-embedded samples were used for TOP. Variants annotated as "likely pathogenic/likely oncogenic" or higher in significance in databases such as ClinVar and COSMIC were compared between the two methods. The t-test and the Wilcoxon rank-sum test were used for comparison. Results: The clinical characteristics of the cohort were as follows: the median age of 74 years (range: 37-85); 21 males (66%); 10 never-smokers (31%); 20 adenocarcinomas (63%); 10 squamous cell carcinomas (31%). The most frequent variants identified were TP53 (22 cases, 69%), and EGFR (12 cases, 34%). Out of 134 single nucleotide variants (SNVs), short deletions, and insertions detected in at least one test, 65 variants (49%) were detected by both methods, 64 (48%) were exclusive to TOP, and 5 (4%) were exclusive to WGS. The median number of variants detected per sample between these two tests was significantly different (TOP: 3.5 (IQR: 2-6) vs WGS: 2 (IQR:1-3), p < 0.001). Twenty-three samples had variants that were only detected by TOP, and 9 samples did not have any. Comparing the variant allele frequency (VAF) of variants detected by TOP, those detected only by TOP had a significantly lower VAF compared with variants detected by both methods (median, 14% (IQR: 10-22%) vs. 22% (IQR: 17-40%)) (p < 0.001). There were no significant differences between the two groups in terms of DNA yield (mean 7.0 µg vs. 6.1 µg, p = 0.37), DNA Integrity Number (DIN, median 9.2 vs. 9.3, p = 0.35), TOP sequencing depth (mean 1095× vs. 1058×, p = 0.44), or WGS depth (median 120× vs. 120×, p = 0.93). Conclusion: Significant differences were observed in the variants detected by target capture sequencing and WGS. However, the specimens analyzed using these two methods were obtained from different tumor sections. Therefore, the discrepancies in test results between the two assays may be due to intratumoral heterogeneity, different sample preparation methods, or low VAF. We will further investigate the causes underlying these findings and make a presentation. Citation Format: Koki Fujii, Masahiro Sekiguchi, Hiroaki Ikushima, Kousuke Watanabe, Aya Shinozaki-Ushiku, Shuichi Tsutsumi, Kenji Tatsuno, Kuniko Sunami, Masaaki Sato, Tetsuo Ushiku, Katsutoshi Oda, Hiroyuki Aburatani, Motohiro Kato, Hidenori Kage. Comparison of whole genome sequencing and target capture sequencing in lung cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 5365.