Predictive relevance of AI-quantified tumor cell density in LARC patients receiving neoadjuvant CRT: post hoc study of the phase 3 ARISTOTLE trial

Tumor cells and tumor-associated stroma are key components of the tumor microenvironment, and their interaction impacts disease progression and treatment resistance in rectal cancer. Post hoc analysis of the phase 3 ARISTOTLE trial evaluated the prognostic relevance of tumor cell density (TCD) quantified by artificial intelligence in patients receiving neoadjuvant chemoradiotherapy (nCRT). The data of ARISTOTLE participants with available digitized pre-treatment biopsies were analyzed. TCD was defined as the proportion of tumor cells within the tumor and stroma tissues, and patients were stratified as TCD-high or TCD-low. TCD status was combined with treatment arms (CRT: capecitabine plus radiotherapy, or IrCRT: capecitabine plus irinotecan plus radiotherapy). A total of 188 (45%) of 414 patients were classified as TCD-high and 226 (55%) as TCD-low. The figure shows disease-free survival (DFS, left panels) and overall survival (OS, right panels). In the TCD-high subgroup, use of IrCRT was associated with significantly longer DFS and OS, along with higher pCR rate, compared to CRT. Howeer, in the TCD-low subgroup, no significant difference in DFS was observed between the treatment arms, while patients receiving IrCRT exhibited a trend towards worse OS. The authors conclude that in this post hoc, hypothesis-generating analysis of ARISTOTLE, higher TCD was associated with differential outcomes after irinotecan-intensified nCRT compared with standard nCRT.