Computed tomography staging of colon cancer: improved patient selection for neoadjuvant therapy with combined radiologic tumor and nodal staging.

BACKGROUND: Patient selection for neoadjuvant therapy in colon cancer (CC) needs to be improved as utilizing computed tomography (CT) tumor (T) staging alone is associated with overstaging and overtreatment. Therefore, we sought to identify specific nodal imaging features that can be combined with radiologic T staging to improve patient selection. METHODS: Pre-operative CTs of stage I-III CC patients treated with upfront resection (2018–2023) were assessed by an expert abdominal radiologist blinded to the histopathologic staging. Radiologic T and node (N) staging based on five imaging features (single lymph node (LN) > 1 cm, ≥ 3 prominent LNs, irregular borders, heterogenous enhancement, and extramural venous invasion) were compared to the pathologic staging, grouped by proficient or deficient mismatch repair status (pMMR vs. dMMR). RESULTS: Of the 177 patients, 147 and 30 had pMMR and dMMR CC, respectively. The overstaging rate for pathologic T staging was 6.1% pMMR vs. 6.7% dMMR CC (p = 0.91). The overstaging rate for pathologic N disease by any imaging feature was 17.0% pMMR vs. 33.3% dMMR (p = 0.04) but improved to 4.8% pMMR vs. 6.7% dMMR (p = 0.67) with irregular borders and/or heterogenous enhancement alone. Compared to radiologic T staging only, the addition of N staging (any feature) to T staging resulted in decreased overstaging rates of stage I/low-risk stage II to high-risk stage II/III CC from 25.2% to 15.6% for pMMR CC and 60.0% to 33.3% for dMMR CC. Overstaging rates further decreased to 4.8% for pMMR and 16.7% for dMMR CC with the combination of radiologic T and N staging by irregular borders and/or heterogenous enhancement specifically. CONCLUSION: Combining radiologic T and N stage based on any of the five imaging features resulted in lower overstaging rates than radiologic T staging alone, but overstaging rates continued to improve when only irregular borders and heterogenous enhancement were used to assess nodal disease. These results can be used as an initial framework for combining CT T and N staging for CC and improve patient selection for neoadjuvant treatment.