Patient stratification based on urea cycle metabolism for exploration of combination immunotherapy in colon cancer

Abstract

Background

Owing to the low ratio of patients benefitting from immunotherapy, patient stratification becomes necessary. An accurate patient stratification contributes to therapy for different tumor types. Therefore, this study aimed to subdivide colon cancer patients for improved combination immunotherapy.

Methods

We characterized the patients based on urea cycle metabolism, performed a consensus clustering analysis and constructed a risk model in the cancer genome atlas cohort. Colon cancer patients were further categorized into two tags: clusters, and risk groups, for the exploration of combination immunotherapy. In addition to external validation in the Gene Expression Omnibus datasets, several images of immunohistochemistry were used for further validation.

Results

Patient characterization based on urea cycle metabolism was related to immune infiltration. An analysis of consensus clustering and immune infiltration generated a cluster distribution and identified patients in cluster 1 with high immune infiltration levels as hot tumors for immunotherapy. A risk model of seven genes was constructed to subdivide the patients into low- and high-risk groups. Validation was performed using a cohort of 731 colon cancer patients. Patients in cluster 1 had a higher immunophenoscore (IPS) in immune checkpoint inhibitor therapy, and those other risk groups displayed varying sensitivities to potential combination immunotherapeutic agents. Finally, we subdivided the colon cancer patients into four groups to explore combination immunotherapy. Immunohistochemistry analysis showed that protein expression of two genes were upregulated while that of other two genes were downregulated or undetected in cancerous colon tissues.

Conclusion

Using subdivision to combine chemotherapy with immunotherapy would not only change the dilemma of immunotherapy in not hot tumors, but also promote the proposition of more rational personalized therapy strategies in future.