Mutational Patterns in Colorectal Cancer: Do PDX Models Retain the Heterogeneity of the Original Tumor?

Colorectal cancer (CRC) remains a leading cause of cancer-related mortality worldwide, highlighting the need for a deeper understanding of the genetic mechanisms driving its development and progression. Identifying genetic mutations that affect key molecular pathways is crucial for advancing CRC diagnosis, prognosis, and treatment. Patient-derived xenograft (PDX) models are essential tools in precision medicine and preclinical research, aiding in the development of personalized therapeutic strategies. In this study, a comparative analysis was conducted on the most frequently mutated genes-<i>APC</i>, <i>TP53</i>, <i>KRAS</i>, <i>BRAF</i>, <i>NRAS</i>, and <i>ERBB2</i>-using data from publicly available databases (<i>n</i> = 7894) and models from University Medicine Rostock (<i>n</i> = 139). The aim of this study was to evaluate the accuracy of these models in reflecting the mutational landscape observed in patient-derived samples, with a focus on both individual mutations and co-occurring mutational patterns. Our comparative analysis demonstrated that while the ranking of individual mutations remained consistent, their overall frequencies were slightly lower in the PDX models. Interestingly, we observed a notably higher prevalence of <i>BRAF</i> mutations in the PDX cohort. When examining co-occurring mutations, <i>TP53</i> and <i>APC</i> mutations-both individually and in combination with other alterations-were the most frequent in both datasets. While the PDX models showed a greater prevalence of single mutations and a slightly higher proportion of tumors without detectable mutations compared to the public dataset, these findings present valuable insights into CRC's mutational landscape. The discrepancies highlight important considerations, such as selective engraftment bias favoring more aggressive tumors, differences in sample size between the two cohorts, and potential bottleneck effects during PDX engraftment. Understanding these factors can help refine the use of PDX models in CRC research, enhancing their potential for more accurate and relevant applications in precision oncology.