The management of metastatic colorectal cancer is usually based on the assumption that metastases within a given patient are homogeneous. However, current treatments, despite notable surgical and systemic advances, do not prevent a risk of recurrence and mortality exceeding 50% after metastasis resection. Classical clinico-pathological prognostic factors, such as tumor regression grade (TRG), have proven insufficient to accurately predict patient outcomes. In this context, analysis of the immune reaction within the tumor microenvironment has emerged as a leading prognostic avenue.
This work is based on a comprehensive analysis of all resected metastases in the patients concerned, namely 338 metastases from 153 stage IV patients who underwent complete metastatic resection with curative intent. Automated whole-slide quantification made it possible to map the spatial distribution of immune cells: total (CD3), cytotoxic (CD8), memory (CD45RO) and regulatory (FOXP3) T cells, as well as B cells (CD20). Infiltration proved to be non-uniform, with generally higher densities at the invasive margin. Most notably, metastases within a given patient were heterogeneously infiltrated and responded differently to treatment, calling into question the assumption of homogeneity of metastatic disease.
The degree of immune infiltration, but also its variability across the different metastases of a given patient, proved decisive for survival. The least-infiltrated metastasis turned out to be particularly important, presumably because it most effectively escapes immune elimination and can promote progression. The Immunoscore (CD3 and CD8) and the TB score (CD8 and CD20), assessed on a random metastatic biopsy or averaged across all metastases, were associated with relapse and survival; their assessment within the least-infiltrated metastasis most accurately predicted long-term survival. In multivariate analysis incorporating the relevant clinico-pathological factors, the Immunoscore remained the only statistically significant parameter for disease-free survival and overall survival, outperforming histopathological characteristics including TRG.
Adaptive immune cell infiltration also proved more informative than tumor regression and pathological response. Notably, a dense adaptive infiltrate prolonged survival independently of treatment response, whereas patients with a good pathological response but low infiltration had short survival. Available as a certified in vitro diagnostic test, the Immunoscore identifies patients at risk of recurrence and death, from early to metastatic stages. This work underscores its clinical utility and reinforces the role of the immune response within the tumor microenvironment, now incorporated as an essential diagnostic criterion in the international classification of digestive tumors.