Anorectal reconstruction after abdominoperineal resection. Experience with double-wrap graciloplasty supported by low-frequency electrostimulation.

PURPOSE: The aims of the study contained herein were to analyze the efficacy and safety of a chronically electrostimulated double-wrap graciloplasty for restoration of continence after a curative abdominoperineal resection for rectal carcinoma and to evaluate late results of a stimulation protocol that was begun early.
METHODS: During the last six years, 31 consecutive patients underwent this procedure: in 24 patients, electrostimulated double-wrap graciloplasty was performed simultaneously with abdominoperineal resection for lower rectal cancer, 7 strictly selected patients underwent conversion to an abdominal stoma following previous abdominoperineal resection (mean length of time from stoma creation, 71.4 months) Anorectal reconstruction was performed following a surgical scheme already standarized since 1985 in 102 patients: after abdominoperineal resection, the distal colon was pulled through to the perineum and surrounded by both gracilis muscles following an "alfa and new-sling" configuration; using platinumiridium electrodes, both muscles were then connected to pulse generator, which was implanted subcutaneously in the abdomen. All surgical steps were performed during the same surgical session to allow early postoperative stimulation of the transposed muscles. A contemporary covering stoma was abandoned as a standard procedure; the distal colon was left closed for a few postoperative days, then it was resected and sutured to the perineum under local anesthesia. Eighteen patients underwent preoperative or postoperative radiotherapy or both, without any significant adverse outcome. To increase gracilis resistance to prolonged "tonic" contraction, patients underwent a chronic, low-frequency stimulation protocol. In the last 11 patients, a new "over-the-nerve and intramuscular" implant was adopted to optimize fiber recruitment and to reduce electrostimulation thresholds. At regular intervals, all patients were evaluated using continence scores and questionnaires, electromanometry, endoluminal ultrasound study, and defecography.
RESULTS: Twenty-six of 31 patients were evaluated for continence, with a mean length of follow-up of 37.8 (range, 4-68) months; 3 patients died because of cancer recurrence, 1 underwent conversion to an abdominal stoma, and 1 is waiting for stoma closure. Continence to liquid and solid stools was achieved in 22 patients (85 percent), and electromanometry findings confirmed a good muscular contraction postoperatively and during follow-up intervals. No postoperative mortality (40 days) was observed; the postoperative complication rate was high 22-percent), but early treatment (drainage and temporary diversion in 7 patients) led to favorable outcomes (4 resolutions, 3 partial muscular impairments). Four stimulators had to be temporarily explanted because of late complications, and two stimulators had to be replaced because of battery exhaustion after three years of use with high stimulation parameters. A significant difference was observed comparing full-contracting threshold after intramuscular (14 patients) and the new over-the-nerve and intramuscular implant technique.
CONCLUSIONS: The study contained herein confirms the efficacy of the surgical scheme we have adopted since 1985 to reconstruct sphincteric apparatus after abdominoperineal resection of the rectum. The "one-step" timing of surgical and electrostimulation-related procedures and the early start of stimulation did not show a significant increase in the complication rate and did not produce noticeable muscular or nerve damage. Adoption of chronic electrostimulation protocols using implantable devices increased the rate of fully continent patients; nevertheless, the overall cost for devices and medical staff duties was high, and a small increase of late morbidity was observed. Finally, the preliminary experience with our new technique of electrode implants encourages further application.