Neoadjuvant Immunotherapy Followed by Surgery Compared with Upfront Surgery Alone in Operable Colon Cancer with Deficient Mismatch Repair: Modeling Oncological Outcomes and Numbers Needed to Treat.

Abstract

Trials on neoadjuvant immunotherapy in operable colon cancer with deficient mismatch repair (dMMR) reported high pathological response rates in the surgical specimen, but long-term survival is not known. Neoadjuvant immunotherapy as a stand-alone therapy without subsequent radical surgery is currently not investigated in colon cancer. The aim of this study was to model outcomes between trial data and real-world patients after surgery. We conducted a comparative modeling study between prospective trial data (NICHE-1) and a prospective, population-derived, translational cohort study (ACROBATICC) of patients with operable colon cancer and microsatellite instability (MSI) status. Comparison was performed across immune-cell infiltrates, stages, MSI, and patient demographics for adverse events, reported oncological outcomes, and modeling numbers needed to treat (NNT) to prevent recurrence. Patient characteristics between the dMMR tumors in the NICHE-1 (n = 21) and ACROBATICC (n = 43) cohorts differed, with older patients and fewer hereditary dMMRs in the 'real-life' ACROBATICC cohort. Higher expression of CD8+ in dMMR tumors compared with proficient mismatch repair (pMMR) tumors was statistically significant across both cohorts. At long-term follow-up, 2/43 patients with dMMR had died from recurrent colon cancer in the ACROBATICC cohort. Assuming a curative effect of neoadjuvant immunotherapy in addition to surgery in dMMR tumors, the NNT would be >20 patients for any additional survivor. In unselected patients with colon cancer having dMMR, recurrence risk is very low after surgery. Assuming a curative effect of neoadjuvant immunotherapy beyond surgery alone, the NNT would be at least 20 patients to prevent one cancer death over surgery alone.