The clinical relevance and prediction efficacy from therapy of tumor microenvironment related signature score in colorectal cancer.

Introduction: As the top 3 cancer in terms of incidence and mortality, the first-line treatment for CRC includes FOLFOX, FOLFIRI, Cetuximab or immunotherapy. However, the drug sensitivity of patients to regimens is different. There has been increasing evidence that immune components of TME can affect the sensitivity of patients to drugs. Therefore, it is necessary to define novo molecular subtypes of CRC based on TME immune components, and screen patients who are sensitive to the treatments, to make personalized therapy possible.

Methods: We analyzed the expression profiles and 197 TME-related signatures of 1775 patients using ssGSEA, univariate Cox proportional risk model and LASSO-Cox regression model, and defined a novo molecular subtype (TMERSS) of CRC. Simultaneously, we compared the clinicopathological factors, antitumor immune activity, immune cell abundance and differences of cell states in different TMERSS subtypes. In addition, patients sensitive to the therapy were screened out by correlation analysis between TMERSS subtypes and drug responses.

Results: Compared with low TMERSS subtype, high TMERSS subtype has a better outcome, which may be associated to higher abundance of antitumor immune cell in high TMERSS subtype. Our findings suggested that the high TMERSS subtype may have a higher proportion of respondents to Cetuximab agent and immunotherapy, while the low TMERSS subtype may be more suitable for treatment with FOLFOX and FOLFIRI regimens.

Discussion: In conclusion, the TMERSS model may provide a partial reference for the prognosis evaluation of patients, the prediction of drug sensitivity, and the implementation of clinical decision-making.