Immune checkpoint inhibitor (ICI) therapy has achieved impressive success in deficient mismatch repair (dMMR)/microsatellite instability-high (MSI-H) colorectal cancer (CRC). ICI has been considered a standard therapy by the FDA, including the use of programmed death receptor 1 (PD-1) monoclonal antibodies and CTL-associated protein 4 (CTLA-4) monoclonal antibodies (1–3). However, the vast majority of CRC cases (approximately 85%) are characterized by proficient mismatch repair (pMMR)/microsatellite stability (MSS) tumors that do not respond to ICIs (4). Recent studies suggest that a subgroup (approximately 10-20%) of MSS-CRC patients might benefit from combination regimens of ICIs (5–7). Therefore, predictive biomarkers for screening these patients are urgently needed.

Current clinical and investigational studies of screening MSS-CRC patients who would benefit from ICIs treatment are limited. PD-L1 expression is a classic biomarker, but the Keynote-028 study demonstrated that PD-L1⁺ MSS-CRC patients could not benefit from ICI therapy (8). POLD1/POLE mutations are predictive but occur in only 1% of MSS-CRC patients (9). Biomarkers such as tumor mutation burden (TMB), tumor-infiltrating lymphocytes (TIL), neo-antigen load (NAL), and immune-regulatory gene expression profiling (iGEP) may allow the selection of clinical patients for ICIs. However, the lack of uniform detection methods and validated cutoffs limit the use of these methods (5, 10–12). Several emerging biomarkers, such as gut microbiota and T-cell-receptor (TCR) sequencing, have also shown predictive value, although they are not yet clinically applicable (7, 13). DNA damage response (DDR) gene mutations may induce a hypermutational phenotype (14), and recent studies have shown that patients with MSS-CRC and mutations in the DDR system have better immune responses and outcomes following ICI therapy (15, 16). However, the pathogenicity of different DDR gene mutations in MSS-CRC remains unclear, and their incidence is significantly lower than in endometrial, ovarian, or biliary tract cancers (17).

Epigenomic alterations can affect tumor immunogenicity and anti-tumor responses by regulating genome stability and chromatin accessibility (18). Additionally, several epigenetic-related gene mutations have been shown to exhibit predictive functions in ICI therapy for multiple types of tumors. ARID1A, an AT-rich interactive domain-containing protein 1A, is a component of the switching defective/sucrose non-fermenting (SWI/SNF) complex that plays a role in chromatin remodeling (19), and increasing evidence suggests that ARID1A alterations are correlated with better outcomes after ICI therapy for bladder cancer, nonsmall-cell lung cancer (NSCLC), and gastric cancer (20, 21). ARID1A mutation is defined as an immunologically active subgroup in MSS-CRC patients with abundant intra-tumoral T-cell infiltration (22). Lysine methyltransferase 2 (KMT2) family members facilitate transcription and gene accessibility by methylating lysine 4 on histone H3 (H3k4) (23), and KMT2 family mutations have also been linked to a favorable response to ICIs in multiple cancers (24). Furthermore, as identified using clustered regularly interspaced short palindromic repeats (CRISPR), KMT2D mutant tumors exhibit an increased mutation burden, IFN-γ-stimulated antigen presentation, and a higher sensitivity to ICIs. Moreover, disruption of DNA methylation signatures has been identified as a marker of anti-PD-1 therapy efficacy in NSCLC (25), and TET1, a DNA demethylase, enhances the immunotherapeutic effect (26). Although this evidence points to the role of epigenetic regulation in anti-tumor immune responses, there is no clinical data on the association between comprehensive epigenetic-related gene mutations (mutations in genes that are involved in epigenetic modifications) and the clinical benefit of ICIs in MSS-CRC.

Given the proposed role of epigenetic regulation impairment in predicting the response to ICIs, we hypothesize that epigenetic-related gene mutations in MSS-CRC cause hypermutation and improve the expression of immune response gene sets. As a result, we conducted this study to clarify the value of epigenetic-related gene mutations as an indicator of immunotherapy efficacy in patients with MSS-CRC. For this purpose, we analyzed whole-exome sequencing (WES) data from TCGA to study TMB, frameshift mutation (FS-mutation), and immune characteristics of Epigenetic_Mut and Epigenetic_Wt groups of MSS-CRC samples. Additionally, in a Chinese clinical MSS-CRC cohort of 89 patients who received PD-1-based treatment, we found that Epigenetic_Mut was associated with favorable clinical outcomes. Here, we report the relationships between epigenetic-related gene mutations and TMB, FS-mutation, immunomodulatory mRNA expression signature, and ICI therapy efficacy in patients with MSS-CRC.

Although ICI-based combination therapies have shown certain effectiveness in pMMR/MSS CRC, especially in combination with antiangiogenic agents (Lenvatinib or Regorafenib) that resulted in an ORR of 20-30% (6, 33), most patients still cannot benefit from the combination therapy because of the high heterogeneity of pMMR/MSS CRC. Recently, the MAYA phase II trial (NCT03832621) showed that MSS-CRC patients with silenced MGMT could benefit from ICIs combined with temozolomide treatment (34). This trial showed 36% for 8-month PFS, 42% for ORR, and 18.4 months for the median OS. Therefore, screening the MSS CRC patients with active anti-tumor immune response may be the key to improving the efficacy of immunotherapy. However, the predictive biomarkers for ICI therapy in MSS-CRC patients are limited.

pMMR/MSS CRC is a cold tumor that contains few neoantigens and either no or inactive TILs (35). Meanwhile, CRC is a multilayered heterogeneous disease with specific treatment challenges and opportunities (36). Additionally, Previous studies have reported that epigenetic-related gene mutations affect both the tumor microenvironment and efficacy of ICIs (24–26). Mechanistically, epigenetic modification can reshape the tumor microenvironment by affecting genomic instability and enhancing the immunogenicity of tumor cells. First, epigenetic modification can affect the DNA damage repair response by regulating the accessibility of chromatin. Studies have shown that epigenetic-related gene mutations can lead to increased TMB in tumor cells, such as ARID1A and KMT2D. ARID1A specifically has a 6.7% mutation rate in MSS-CRC (22) and may increase the instability of the genome by adjusting the MMR pathway (21, 37). Mutations in the KMT2D gene are common in cancer patients, and their deficiency can increase the levels of genomic DNA damage and TMB, as well as increase transcription instability. Clinical studies have shown that individuals with mutations in genes from the KMT family are more likely to benefit from ICI therapy (24, 25). Furthermore, epigenetic-related gene mutation enhances the immunogenicity of tumor cells. Accounting for 5%-10% of genomic DNA sequences, human endogenous retroviruses (ERVs) are remnants of the evolution of germline integrations of exogenous infectious retroviruses (38, 39). These exogenous genes are not expressed in healthy tissues other than germ cells but are often abnormally expressed in tumors with epigenetic regulation defects. Here, neoantigen expression increases immunogenicity and triggers an innate immune response against tumors (40, 41). Recently, genome-wide technologies have revealed frequent mutations in epigenetic modifier genes, particularly in cancers (42). It is therefore necessary to analyze systematically the immune activity and the effect of immunotherapy in MSS-CRC patients with epigenetic regulation impairment.

In our study, we systematically analyzed 68 epigenetic-related genes from 13 pathways involved in chromatin regulatory processes in MSS-CRC samples. The mutation rate of epigenetic-related genes in the TCGA cohort was 18.35%. This mutation frequency was higher than that of any previous marker in the population, such as POLE or DDR mutations, and was closer to the potential benefit ratio in MSS-CRC clinical trials. ARID1A, KMT2C, RSF1, CHD9, PBRM1, and ATRX were the most mutated genes in the TCGA cohort, accounting for approximately 75% of the epigenetic-related gene-mutated MSS-CRC patients. This is consistent with previous reports, and ARID1A is thus a marker gene that should be investigated in clinical practice.

Using bioinformatics algorithms, we also assessed whether the MSS-CRC samples with epigenetic-related gene mutations from TCGA had better immune activity, including immune signatures, tumor-infiltrating lymphocytes, and expression of immune checkpoints and key genes. Furthermore, we validated our bioinformatic findings using immunohistochemical analyses of CD8⁺ and FOXP3⁺ cells from a cohort of MSS-CRC patients, and similar results were obtained at the histopathological level. In the Epigenetic _Mut group, CD8⁺ cells were higher and FOXP3⁺ cells were lower. The Epigenetic _Mut group also had a higher proportion of CD8/FOXP3 cells than the Epigenetic_Wt group. The VOLTAGE trial demonstrated that among MSS-CRC patients receiving ICIs as neoadjuvant treatment, patients with an elevated CD8/FOXP3 cell ratio were more likely to achieve pathologic complete response (pCR), suggesting that the CD8/FOXP3 cell ratio may be a predictor for ICI therapy efficacy (43).

Finally, we validated the predictive power of epigenetic-related gene mutations in the HMUCH cohort of 89 MSS CRC patients who received immunotherapy and discovered that patients with epigenetic mutations were more likely to benefit from ICI-based combination therapy and had better clinical outcomes. These preliminary results demonstrate that epigenetic-related gene mutations can predict the response to ICIs in MSS-CRC patients.

This study has several limitations, including the validation cohort coming from a single-center, the small size of the cohort, and the lack of validation in other populations. This is because ICI-based regimens have not been recommended by any clinical guidelines for MSS-CRC. Numerous patients included in this study experienced the failure of standard treatment, and the treatment compliance and completeness of the clinical information in many of these patients, were not ideal. Additionally, since the genetic information in the HMUCH cohort was obtained from clinical testing, transcriptomic data were lacking. Thus, our TCGA cohort findings could not be validated. Instead, we performed immunohistochemical staining analysis of pathological sections to validate the immune activation status of the Epigenetic _Mut group, but a larger-scale validation remains necessary. Furthermore, the application of ICIs in MSS-CRC has not been standardized, and most patients enrolled in our study were patients who had experienced multiple failed lines of treatment, bringing considerable heterogeneity to the population of this study. Therefore, future prospective studies with larger cohort studies are needed.

In conclusion, our data suggest that identifying epigenetic-related gene mutations might help select the right immunotherapy for MSS-CRC patients and can be used as a biomarker to predict ICI therapy effectiveness. Importantly, the status of epigenetic-related gene mutations is highly accessible from clinical genetic testing, although it is often overlooked by clinicians. Further exploration of the molecular mechanisms underlying the increased effectiveness in specific MSS-CRC patients and prospective clinical trials are therefore warranted.

The data presented in the study are deposited in the Genome Sequence Archive in National Genomics Data Center repository (https://ngdc.cncb.ac.cn/gsa-human), accession number GSA-Human: HRA003408.

The study was approved by the Ethics Committee of Harbin Medical University Cancer Hospital (Harbin, China). Informed consents were obtained from all patients before surgery and during the experimental procedures.

CL: Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Software, Validation, Visualization, Writing - original draft, Writing - review & editing. HX: Data curation, Formal analysis, Investigation, Methodology, Software, Validation, Visualization, Writing - original draft, Writing - review & editing. LC: Investigation, Methodology, Validation, Writing - original draft, Writing - review & editing. LF: Investigation, Methodology, Writing - review & editing. SH: Investigation, Writing - review & editing. YR: Investigation, Writing - review & editing. WZ: Conceptualization, Project administration, Resources, Supervision, Writing - review & editing. YZ: Conceptualization, Funding acquisition, Project administration, Resources, Supervision, Writing - review & editing. The work reported in the paper has been performed by the authors, unless clearly specified in the text. All authors contributed to the article and approved the submitted version.

This work was supported by grants from the National Natural Science Foundation of China (No. 82102858), the Natural Science Funding of Heilongjiang (No. YQ2022H017), nN10 Excellent Discipline Construction Program (No. Hepatobiliary and Pancreatic Tumor 2017); Basic scientific research projects of universities in Heilongjiang Province (No. 2020-KYYWF-1465); Top Young Talents Project of HMUCH (No. BJQN2021-01); Haiyan Research Fund of Harbin Medical University Cancer Hospital (No. JJMS2022-04), and Beijing Science and Technology Innovation Medical Development Foundation (No. KC2021-JX-0186-102, KC2021-JX-0186-69).

The results of the pan-cancer analysis were based on data generated by TCGA Research Network: https://www.cancer.gov/tcga. The authors thank AiMi Academic Services (www.aimieditor.com) for English language editing services.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.