287 Predictive genetic biomarkers for the development of peritoneal metastases in colorectal cancer The clinical significance of NTRK3 fusion identified in our study, in the setting of CRC, as well as the possibility for targeted treatments should be explored in the future. Prior to our explorative study, we performed a systematic review to summarize the current knowledge on genetics and genomics in CRC-PM 19. An NGS analysis with 409 cancer genes showed several additional genetic mutations, i.e., ARID1A, PKHD1, UBR5, PAX5, TP53, ASXL1, and AR, presumably associated with PM 58. In our TSO500 NGS panel, ARID1A, PAX5, TP53, ASXL1, and AR were included as well. AR and ARID1A mutations occurred in one PM (11%) and one M0 sample (5%). Only one PAX5 mutation was found in one M0 patient, and ASXL1 mutations were not detected. Thereby, the suggested genes related to PM by Lee et al. are not confirmed in our paper. The latter may be explained by the difference in study population; Lee et al. included patients with small obstructing adenocarcinomas (≤3 cm) with synchronous or metachronous PM and compared them with large non-obstructing tumors without PM. Another explanation could be our small sample size. Other authors describe NEK2, MACC1, REG1A, KIF18A, RET, and TIP60 as possible PM-related cancer genes 59-64. In our TSO500 panel, only RET was investigated. In contrast to the suggestion of Yang et al. concerning the association of RET mutations and PM, we did not identify any mutation in this gene in our cohort 63. Another factor that can contribute to the difficulty of finding biomarkers is the genetic differences between the primary tumor and metastatic lesions. Studies investigating the differences between peritoneal lesions and their primary tumors reported some small unique differences 65, whilst other studies report high concordance 66-68. A very recent study by Lenos et al. showed that peritoneal lesions seemed to have much more similarity to their primary tumor compared to other metastases, and these lesions seemed to retain both clonal heterogeneity and transcriptional profile 67. A new way to look at CRC tumors is through dividing them into subtypes, for example, the previously described four consensus molecular subtypes (CMS 1-4). These subtypes aid in prognostication as well as in determining treatment strategies for individual patients based on the mutations, activated pathways, and phenotypic characteristics and responses to treatment of other tumors with similar signatures 69. The majority of PMs in their study are of the CMS4 subtype, known as the mesenchymal subgroup 70, 71. CMS4 is presented in 23% of CRC cases, which are most often distal tumors with poor relapse-free and overall survival and harbor prominent transforming growth factor β activation, stromal infiltration, and angiogenesis 72, 73. CMS4 tumors have extremely low levels of hypermutation, MSS, and very high somatic copy number alteration counts 69. The latter was also seen in our cohort. Unfortunately, we were not able to examine all of these characteristics in our study due to the limited content of our RNA NGS gene panel. Therefore, the translation to CMS subgroups was not possible in our study. 12
RkJQdWJsaXNoZXIy MTk4NDMw