[Risk factors for local recurrence of rectal carcinoma]

The impact of distal resection margins and the mode of operation on pelvic recurrence rate was assessed in 331 cases following abdomino-perineal resection (APR; n = 134), and anterior resection (AR; n = 197) for rectal cancer. Local recurrence was observed in 55 cases (16.6%) after a median interval of 16 months. Only 20 of 212 patients (9.4%) without positive lymph nodes developed a local recurrence, but 35 of 119 (29.4%) with nodal involvement. The recurrence rate was 33% (10/30) in cases with local spread to adjacent structures (T4), and 15% in cases with less extensive penetration. Of 59 poorly differentiated cancers 15 developed pelvic failure (25%), as opposed to 40 of 272 (14.7%) well or moderately differentiated tumours. 17.7% developed local recurrence after AR, and 14.9% after APR. The recurrence rate following AR was 30% in 33 cases with resection margins less than or equal to 10 mm and 17.4% in 115 cases with margins from 11 to 40 mm, and 10% in 49 cases with wider margins. Cox's multiple proportional hazards regression revealed that nodal involvement (p = 0.0003), local invasion (p = 0.0055), poor differentiation (p = 0.066), and AR vs. APR (p = 0.099) were independent risk factors for pelvic failure. For the AR cases the factors were nodal involvement (p less than 0.0001), local invasion (p = 0.0043), and a resection margin less than or equal to 25 mm on the fixed specimen (p = 0.0039). For patients with negative lymph nodes local invasion was the only independent risk factor, whereas the variables "anterior resection" and narrow resection margin were significant only in node positive cases.(ABSTRACT TRUNCATED AT 250 WORDS)