The AJCC/UICC-TNM staging system provides useful but incomplete prognostic information for early-stage colon cancers. Classifications based solely on tumor cell characteristics offer only moderate accuracy, even though risk assessment is critical in deciding on adjuvant therapy. In so-called high-risk stage II patients (T4 tumors, insufficient number of lymph nodes sampled, poorly differentiated tumors, lymphatic, vascular, or perineural invasion, obstruction, or perforation), chemotherapy may be considered, but these criteria remain imperfect. In stage I, where chemotherapy is not usually recommended, approximately 10% of tumors recur despite surgical resection. The Immunoscore, a classification method quantifying the densities of CD3+ and cytotoxic CD8+ T lymphocytes at the tumor center and in its invasive margin, had already demonstrated its prognostic value in stages II/III, but its relevance remained to be established in stages I/II and in early-stage risk subgroups.
This international study, conducted under the auspices of the Society for Immunotherapy of Cancer, evaluated the consensus Immunoscore by digital pathology in 1,885 stage I/II patients from Canada and the United States (cohort 1) as well as from Europe and Asia (cohort 2). The primary endpoint was time to recurrence; secondary endpoints included disease-free survival, overall survival, and prognosis in the stage I, stage II, high-risk stage II, and microsatellite-stable (MSS) subgroups.
A high Immunoscore was associated with the lowest risk of recurrence in both cohorts. At five years, recurrence-free survival rates in stage I/II reached 78.4%, 88.1%, and 93.4% for low, intermediate, and high Immunoscores, respectively (HR high vs. low = 0.27; p < 0.0001). In multivariate Cox analysis, this association was independent of sex, T stage, tumor location, and microsatellite status (HR = 0.29; p < 0.0001). The Immunoscore significantly predicted survival in stage II, high-risk stage II, T4N0 tumors, and MSS patients, and remained associated with time to recurrence in stage I (HR = 0.07; p = 0.016). It represented the strongest contribution to risk (69.5% of the χ²). Among T4N0 tumors not treated with chemotherapy, patients with a high Immunoscore maintained a prolonged time to recurrence, with the Immunoscore then remaining the only significant parameter. The authors emphasize that the heterogeneity of a population drawn from thirteen countries constitutes a limitation, but that it also illustrates the robustness of the tool. In early stages, a low Immunoscore reliably identifies patients at risk of relapse, for whom enhanced surveillance or adjuvant therapy warrants consideration. The Immunoscore is now included in the WHO classification and in the ESMO guidelines.