Clinical nodal findings and indocyanine green-guided sentinel-node mapping in non-metastatic colon cancer: an exploratory prospective cohort stratified by postoperative mismatch-repair status

Abstract Background Nodal enlargement can be difficult to interpret in mismatch-repair-deficient (dMMR) colon cancer, and indocyanine green-guided sentinel lymph node biopsy (ICG-SLNB) can miss nodal metastases. We described clinical nodal findings and ICG-SLNB results by postoperative mismatch-repair (MMR) status. Methods This exploratory, non-consecutive single-centre cohort included 40 patients undergoing elective colectomy for clinically non-metastatic colon cancer (May 2024 to March 2026). The primary question, how clinical nodal enlargement and ICG-SLNB differ across MMR strata, was addressed by a binary ≥ 5-mm nodal-enlargement marker (CT-LAP) on retrospective CT reread blinded to final pathological nodal status (pN) and MMR status, compared across strata. Sentinel-node detection and false-negative results were described within each stratum. The overall detection rate, capture of pN-positive disease, ultrastaging and pN classification by CT-LAP were exploratory. Results MMR was available for 38 patients (14 dMMR; 24 mismatch-repair-proficient [pMMR]). CT-LAP was present in 14/14 dMMR and 7/24 pMMR tumours (difference + 70.8 percentage points, 95% CI 41.4 to 85.1). Sentinel nodes were detected in 13/14 dMMR and 23/24 pMMR tumours, and in 38/40 overall (95.0%, 95% CI 83.5 to 98.6). Four of six false-negative sentinel results were dMMR. In exploratory analyses, pN-positive disease occurred in 7/14 dMMR and 6/24 pMMR patients (difference + 25.0 points, 95% CI − 5.6 to + 51.6), and CT-LAP specificity was 0/7 in dMMR and 15/18 in pMMR tumours. Among 12 pN-positive patients with successful detection, sentinel nodes captured six (sensitivity 50.0%, 95% CI 25.4 to 74.6). One further pN-positive dMMR patient had non-detection. Ultrastaging upstaged none of 24 pN-negative patients with invasive tumours and successful detection (upper 95% limit 13.8%). Conclusions In this selected cohort, enlarged regional nodes (≥ 5 mm) were far more frequent with dMMR than pMMR status. This points to a dMMR tendency towards nodal enlargement, which in exploratory analyses did not distinguish pN status within dMMR. Sentinel nodes were detected in almost every patient, but four of six false negatives after detection were dMMR. ICG-SLNB reliability in dMMR tumours therefore remains open. These hypothesis-generating findings require confirmation in a consecutive, preoperatively MMR-stratified cohort with the same ultrastaging of sentinel and non-sentinel nodes. Trial registration Not registered.