In the treatment of very low rectal cancer, a distal resection margin of more than 1 cm can be obtained by partial internal sphincteric resection, allowing a sphincter preserving surgery. Thus, intersphincteric resection (ISR) has been proposed as an alternative to abdominoperineal resection for selected low rectal cancer. The aim of our study was to assess the morbidity, mortality, and the long-term oncologic and functional results of ISR. Charts of patients who had ISR between 1992 and 2004 were reviewed. Cancer-related survival and locoregional recurrence rates were calculated using the Kaplan-Meier method. Functional outcome was assessed by using a standardized gastrointestinal functional questionnaire. Incontinence was assessed by the continence score of Wexner. Ninety patients (59 males, 31 females) with a tumor at a median distance of 35 mm (range, 22-52) from the anal verge had an ISR. Thirty-seven patients (41%) had preoperative radiotherapy. Histologically complete remission after neoadjuvant radiotherapy (ypT0) was observed in 7 patients (8%), 12 patients (13%) were pT1, 35 patients (39%) pT2, 32 patients (36%) pT3, and 4 patients (4%) pT4. Five patients (5.5%) had synchronous liver metastases. R0 resection was obtained in 85 patients (94.4%). The median distal resection margin on the fixed specimen was 12 mm (range, 5-35) and was positive in 1 case. The circumferential margin was positive (< or =1 mm) in 4 patients (4.4%). There was no mortality. Complication rate was 18.8%: anastomotic leakage occurred in 8 patients (8.8%) and 1 patient had an anovaginal fistula. Five patients (5.6%) underwent secondary abdominoperineal resection: 1 for positive distal margin, 1 for colonic J-pouch necrosis, and 3 for local recurrence. ONCOLOGIC RESULTS: After a median follow-up of 56.2 months (range, 13.3-168.4), local, distant, and combined recurrence occurred in 6 (6.6%), 8 (8.8%), and 2 patients, respectively. Thirteen patients (14.4%) died of cancer recurrence. Five-year overall and disease-free survival was 82% (80-97) and 75% (64-86), respectively. In univariate analysis, overall survival was significantly influenced by pTNM stage and T stage (pT 1-2 vs. 3-4: P = 0.008 and stage I-II vs. III-IV: P = 0.03). In multivariate analysis, we did not find any impact on local recurrence-free survival for the investigated prognostic variables. For a total of 83 patients the mean stool frequency was 2.3 +/- 1.3 per 24 hours. Forty-one percent of patients had stool fragmentation, one-third nocturnal defecation, 19% fecal urgency, and 36% followed low fiber diet. Thirty-four patients (41%) were fully continent, 29 patients (35%) had minor continence problems, and 20 patients (24%) were incontinent. After adjustment for age, gender, tumor level, and pTNM stage, preoperative radiotherapy was the only factor associated with a risk of fecal incontinence [OR (IC 95%) = 3.1 (1.0-9.0), P = 0.04]. In selected patients, ISR is a safe operation with good oncologic results. It achieves good functional results in 76% of patients. Functional results are significantly altered by preoperative radiotherapy.