Genomic characterization of primary cervical cancer and lung metastasis.

e17527 Background: Cervical cancer ranks fourth in incidence and mortality among female tumors in the world. Lung metastases are the most common metastatic foci of cervical cancer, which were associated with poor prognosis. Exploring the different genomic landscape between primary tumor and lung metastasis would help to understand the mechanism of cervical cancer lung metastasis. Methods: A total of 52 Chinese cervical cancer patients with lung metastasis were enrolled between 2016 and 2020. Among whom 37 patients had paired samples (primary tumor and lung metastasis), 6 had only primary tumor samples, and 9 had only lung metastasis samples. Lung metastasis samples were confirmed as cervical cancer metastasis via immunohistochemistry. All the formalin-fixed paraffin-embedded (FFPE) tumor samples and matched peripheral blood were collected and underwent whole-exome sequencing (WES). The Illumina NovaSeq 6000 (Illumina, San Diego, CA, USA) sequencing platform was used to perform 150-nt paired-end multiplex sequencing of DNA samples. The average coverage depth was 340× in FFPE samples and 170× in white blood cell controls. Normally distributed data were compared using the t-test and data with abnormal distribution were compared using the Mann–Whitney U test. Results: There were 4,212 and 6,341 somatic mutations in primary tumor and lung metastasis, respectively. The consistencies for SNVs/indels and CNVs were 41.22% (2.86%–76.11%) and 13.65% (3.69%–76.11%). The most frequently altered genes in primary tumor and lung metastasis were PIK3CA (23.26% vs. 36.96%; p = 0.26), TTN (23.26% vs. 39.96%; p = 0.18), MUC4 (16.28% vs. 39.96%; p = 0.03), ARMCX4 (25.58% vs. 26.09%; p = 1), and MUC16 (18.60% vs. 26.09%; p = 0.45). MUC4 had a significantly higher mutation frequency in lung metastasis than in primary tumor. To explore the mutation characteristics, co-mutations in primary tumor and lung metastasis were analyzed. EP300/MUC16 co-mutation (p < 0.01) was identified in primary tumor. KMT2D/MUC16 co-mutation (p < 0.01) was identified in lung metastasis. Among the paired samples, the Cosmic Signature analysis had shown that Signature 2, which was attributed to activity of the AID/APOBEC family of cytidine deaminases, was higher in lung metastasis than primary tumor (p = 0.042). Meanwhile, the amplification of 17q12 was more frequently observed in lung metastasis in contrast to primary tumor (p = 0.035). Furthermore, the GO/KEGG analysis had revealed that the mutant genes which only observed in lung metastasis were significantly enriched in ATPase activity and inositol phosphate metabolism. Conclusions: In summary, comprehensive genomic profiling of primary cervical cancer and lung metastasis can be of help to identify metastasis-specific features and understand the mechanism of cervical cancer lung metastasis.