The management of metastatic colorectal cancer has seen notable advances, yet the prediction of treatment response and patient outcomes remains imperfect. This study examined the immune landscape of metastases in order to assess to what extent the local immune infiltrate determines therapeutic response and survival. It is based on two cohorts comprising a total of 153 stage IV patients who underwent complete curative resection of their metastases, amounting to 441 lesions in all.
Immune densities were quantified in an automated manner on whole slides, in the center and at the invasive margin of each metastasis. Two indicators were analyzed: the Immunoscore, based on T lymphocytes (CD3 and CD8), and the T and B score (CD8 and CD20). Their value was related to radiological and pathological response, as well as to recurrence-free survival and overall survival, using multivariate Cox models. The spatial distribution of immune cells proved to be non-uniform: within a single patient, the different metastases displayed variable infiltrates and responded differently to treatment, reflecting the heterogeneity of the tumor microenvironment.
A favorable response to treatment was significantly associated with high immune densities. Among all lesions, it was the least-infiltrated metastasis that proved most informative for predicting survival. With a high Immunoscore (I 3-4), the five-year recurrence-free survival rate reached 27.9% versus 12.3% for a low score (HR = 0.45; P = 0.02), and overall survival 64.6% versus 32.5% (HR = 0.32; P = 0.001). The TB score yielded comparable results (overall survival of 63.7% versus 21.4%; HR = 0.25; P < 0.001). Patients combining high scores showed a median survival of 70.5 months, compared with 25.1 to 38.3 months for low scores. Notably, patients classified as non-responders but with a strong immune infiltrate also benefited from prolonged survival. In multivariate analysis, the immune parameters remained the only significant prognostic factors for recurrence-free survival and overall survival, whereas treatment response and pathological characteristics, including tumor regression grade, were not. The authors acknowledge as a limitation the heterogeneity of the administered treatments, even though no difference in infiltrate was observed between the regimens tested, and they emphasize the value of evaluating multiple metastases to account for lesional heterogeneity in immunotherapy trials. These data suggest that the survival of stage IV patients is largely governed by the state of the local adaptive immune response, particularly in the least-infiltrated lesion, and support the role of immunotherapy aimed at modulating pre-existing immunity.