Combination of lysine-specific demethylase 6A (KDM6A) and mismatch repair (MMR) status is a potential prognostic factor in colorectal cancer.

PURPOSE To evaluate the relationship between the DNA mismatch repair (MMR) status and histone lysine-specific demethylase 6A (KDM6A) on the prognosis of colorectal cancer (CRC). METHODS About 594 patients with CRC from The Cancer Genome Atlas (TCGA) were enrolled in this retrospective study. Subsequently, a series of different classification tests for MMR status, cancer types, and target gene expression was conducted. RESULTS After filtering out the KDMs group of genes, we selected KDM6A as the target gene. A significant difference in the performance of KDM6A in tumor and normal tissues were confirmed. Our results showed a lower KDM6A expression, lower KDM6A exon expression, and higher KDM6A DNA methylation than their corresponding normal tissues in colon adenocarcinoma (COAD). Notably, the main MMR genes were highly expressed in tumor tissues than normal tissues both in COAD and rectum adenocarcinoma (READ). Moreover, proficient DNA mismatch repair (pMMR) was found to be an important poor prognostic factor in COAD (p = 0.0064) and the low KDM6A expression was an important factor for poor prognosis in READ (p = 0.0217). Based on these results, we consequently relate MMR status with KDM6A expression in predicting the prognosis of patients with CRC. Moreover, patients with pMMR exhibited a low KDM6A expression in COAD (p = 0.0250). Samples were divided into two groups based on the KDM6A expression. Interestingly, the group with low KDM6A expression showed no difference between pMMR and deficient DNA mismatch repair (dMMR) in prognosis, whereas the group with high KDM6A expression was closely related to MMR status in OS (p = 0.0082). Besides, COAD patients with high KDM6A expression and pMMR status had poor OS (p = 0.0082). CONCLUSIONS The KDM6A/MMR classification-based subtypes of low KDM6A expression/READ, high KDM6A expression/pMMR, and COAD/pMMR were associated with poor prognosis. This classification can be a novel prognostic approach in CRC.