Ismail Gögenur, Tobias F. Justesen, Line Schmidt Tarpgaard, Mustafa Bulut, Torben Hansen, Lars H. Jensen, Hans B. Rahr, Mikail Gögenur, Tove Kirkegaard, Lukas Balsevicius, Michael Chieng, Hans Raskov, Mia K.S. Sørensen, Peter C. Petersen, Jens R. Eriksen, Søren Salomon, Anne-Marie K. Fiehn, Maria Dalgaard Mikkelsen, Søren Brandsborg, Kåre A. Gotschalck, Katrine J. Emmertsen, Anne Ramlov, Pernille W. Born, Ole Thorlacius-Ussing, Michael B. Lauritzen, René K. Olesen, L P Poulsen, Jakob Lykke, Jakob Schou, Laura Buskov, Peter‐Martin Krarup, Claus L. Andersen, Niels Pallisgaard, Per Pfeiffer, Camilla Qvortrup
PURPOSE: Neoadjuvant immune checkpoint inhibition has shown promise in localized deficient mismatch repair (dMMR) colorectal cancer, yet the optimal treatment duration, regimen, and methods for response evaluation remain undefined. This trial aimed to investigate the efficacy and safety of neoadjuvant pembrolizumab for patients with localized dMMR colon cancer. METHODS: Participants received a single cycle of pembrolizumab (4 mg/kg, max 400 mg, every-6-weeks dosing regimen), followed by a preoperative endoscopy with biopsies and surgery 3-5 weeks later. The primary end point was pathologic complete response (pCR); secondary and exploratory end points included major pathologic response (MPR), safety, survival, and endoscopic response assessment. RESULTS: From February 2023 to March 2024, 85 patients (median age 74 years) were enrolled. All received pembrolizumab; one patient declined surgery. Among 84 patients, pCR was achieved in 44% (37 of 84; 95% CI, 33 to 55), and MPR in 57% (48 of 84; 95% CI, 46 to 68). Two (2%) patients died of complications within 30 days of surgery. At a median follow-up of 18.4 months (IQR, 16.3-21.1), one patient had a recurrence, resulting in overall and disease-free survival rates of 98% and 96%. Grade 3 adverse events occurred in 11% of patients (9 of 85; 95% CI, 3 to 16), with three treatment-related events. In 81 patients, the sensitivity, specificity, and accuracy of the biopsies for predicting pCR were 68%, 75%, and 72%, respectively. In 76 patients, the sensitivity, specificity, and accuracy of the endoscopic images for predicting pCR were 77%, 93%, and 86%, respectively. CONCLUSION: A single cycle of neoadjuvant pembrolizumab led to pCR in nearly half of patients undergoing surgery for localized dMMR colon cancer. The utility of endoscopic evaluation may inform future strategies for patient selection in nonoperative management pathways.