Identification of radiologic and clinicopathologic variables associated with tumor regression pattern and distribution of cancer cells after short-course radiotherapy and consolidation chemotherapy in patients with rectal cancer.

Background: Knowledge of the pattern of regression and distribution of residual tumor cells may assist in the selection of candidates for rectum-sparing strategies. Objective: To investigate and identify factors associated with tumor regression pattern and distribution of residual tumor cells. Methods: We conducted a prospective study of patients with T3/T4 N0/N+ adenocarcinoma of the middle and lower third of the rectum (≤10 cm) treated with radiotherapy (5×5 Gy) followed by 6 cycles of CAPOX chemotherapy. The pattern of tumor regression was classified as fragmented or solid. Microscopic intramural spread was measured. We used a model of distribution of residual tumor cells not yet applied to rectal cancer, defined as follows: type I (luminal), type II (invasive front), type III (concentric), and type IV (random). Results: Forty patients were included with a median age of 66 years; 23 (57.5%) were men. A fragmented pattern was identified in 18 patients (45.0%), and a solid pattern in 22 (55.0%). Microscopic intramural spread was identified in 25 patients (62.5%), extending from 1 to 18 mm (median, 4 mm). There were 14 cases (35.0%) of microscopic intramural spread ≥10 mm. All cases of fragmented regression pattern, except one, showed microscopic intramural spread. Within the fragmented pattern, microscopic intramural spread was 4-8 mm in 4 cases and ≥10 mm in the remaining cases. All cases of microscopic intramural spread ≥ 10 mm were within the fragmented pattern. Regarding the distribution pattern of residual tumor cells, 11 cases (31.5%) were classified as type I, 14 (40.0%) as type II, 10 (28.5%) as type III, and none as type IV. Carcinoembryonic antigen levels >5 ng/mL, downsizing <50%, residual mucosal abnormality >20 mm, and anatomopathologic lymph node involvement were significantly associated with the occurrence of fragmentation (P<0.05). Having received all 6 cycles of CAPOX chemotherapy and absence of microscopic intramural spread were significantly associated with the type I distribution pattern (P<0.05). Conclusion: The occurrence of a fragmented regression pattern is common, as is the presence of microscopic intramural spread. We could identify radiologic and clinicopathologic factors associated with the pattern of tumor regression and a type I distribution pattern.