Abstract
PURPOSE: The efficacy of postoperative adjuvant chemotherapy in patients with stage II colorectal cancer has been a subject of debate. This study aimed to evaluate the prognostic and predictive significance of DNA ploidy and stroma ratio in patients diagnosed with stage II colorectal cancer (CRC). METHODS: Clinical data and tumor tissues from 179 patients with stage II CRC were collected retrospectively. DNA ploidy (P) and stroma (S) were assessed using automatic image analysis tools powered by machine learning. RESULTS: Patients were categorized into three risk groups: PS-low (diploid and low stroma, PS-L), PS-intermediate (non-diploid or high stroma, PS-M), and PS-high (non-diploid and high-stroma, PS-H). According to the univariable model, the PS-H group exhibited significantly poorer 5-year overall survival rates at 73.0% compared to 87.8%, with a hazard ratio (HR) of 2.281 (95% CI: 0.946-5.502, P = 0.066), as well as lower 5-year disease-free survival rates at 69.4% versus 86.6%, HR = 2.323 (95% CI: 1.016-5.308, P = 0.046) among stage II colorectal cancer patients. Notably, chemotherapy was associated with improved overall survival [HR = 83.460 (95% CI: 0.179-38925.833), P = 0.003] and disease-free survival [HR = 8.628 (95% CI: 1.059-70.265), P = 0 .044] in individuals within the PS-high group. CONCLUSION: While ploidy and stroma alone do not possess predictive power regarding survival outcomes for stage II colorectal cancer patients, those receiving chemotherapy within the PS-H group demonstrated enhanced survival rates. Therefore, combining assessments of ploidy and stroma may serve as an adjunctive tool in clinical decision-making processes to guide chemotherapy treatment strategies for patients diagnosed with stage II colorectal cancer.