Surgical transposition of gluteus muscles, one of the oldest surgical treatments for fecal incontinence, is experiencing a modest renewal of interest. Initially reported by Chetwood [1] in 1902, the use of the gluteus muscle has been suggested for a variety of reasons. First, it is a well-vascularized muscle supplied by the inferior gluteal artery. Second, it is larger and stronger than the gracilis muscle as it provides more bulk to help buttress the anal canal. Third, it is activated during walking, which allows it to function as a significant adjunct to the external sphincter. Finally, mobilization of the gluteus muscle does not impair gait or pelvic stability. Although a variety of techniques have been used, the most studied method involves division of both gluteus muscles from the sacrum, splitting one muscle and subcutaneously tunneling and wrapping one strip anteriorly and one strip posteriorly to the rectum. These muscle strips are first sutured together and then to the ipsilateral mobilized muscle [2]. In a cadaveric study, Pak-art and colleagues [3] proposed a modification of this procedure when they found that increased muscle length and thus decreased tension could be achieved by using a proximally based as opposed to a distally based gluteal flap.

Similar to the gluteus muscle, the gracilis muscle is a superficial adductor not essential for locomotion that is easily accessible and has a constant proximal neurovascular supply from the obturator nerve and profunda femoris artery. Gracilis muscle transposition has been used extensively in the treatment of fecal incontinence. The use of gracilis muscle transposition was first described by Pickrell et al. [4] in 1952 for neurogenic incontinence in children. The gracilis muscle seemed to be an excellent choice owing to its proximity to the anal canal, easy mobilization, proximal blood supply, and innervations; however, improvement in continence was only marginal in these children because of the rapid fatigability of the gracilis muscle. The gracilis was unable to sustain contraction because it is a fatigue-prone type II (fast-twitch) muscle. Double-gracilis anorectal neosphincter is a technique for reconstruction after necrosis of the distal part of the muscles. This technique seemed to permit the optimal use of the contractile potential of the residual gracilis muscles, which did not show evidence of defunction-or ischemia-related fibrosis.

In 1981, Salmons and Henriksson [5] found that electrical stimulation produces profound changes in the morphological, physiological, and biochemical character of the skeletal muscle. In their work, x inserted electrical stimulation converted the gracilis muscle into a slow-twitch (type I) muscle, allowing its x function as a sphincter. The first dynamic graciloplasty was reported in 1991 by Baeten et al. [6].

After placing the patient in the lithotomy position, either one long or two to three short incisions are made on the inner aspect of the thigh. The muscle is identified and detached from its insertion on the tibial tuberosity. Care is taken to preserve the proximally based neurovascular bundle. The muscle is then transposed around the anal canal through two lateral incisions and sutured to the contralateral tibial tuberosity. Six to 8 weeks later electrodes are implanted through a subcutaneous tunnel into the muscle and a neurostimulator is implanted in a pocket in the abdominal wall. Stimulation of the gracilis muscle involves progressively increasing activation of the generator. After 8 weeks the nerve stimulator is left on continuously. The patient is given a magnet to turn off the nerve stimulator to defecate [7].

In a study by Rongen et al. [8], the long term results after 24 months in patients without a stoma showed that 16 % were continent 100 % of the time, 43 % had >50 % improvement in continence from baseline, 11 % had <50 % improvement in continence, 7 % had a stoma created, and 23 % discontinued the study. In patients who had a preexisting stoma before graciloplasty, 33 % had 100 % continence, 17 %had >50 % improvement in continence, 22 % had <50 % improvement in continence, 6 % had a stoma re-created, and 22 % exited the study. Long term results in patients without stoma showed reduced incontinence, both solid and liquid, and a decreased use of pads after 24 months; however, these results were not statistically significant.

This procedure is not without complications. According to a report by Matzel and colleagues [9], 211 adverse events were noted in 121 patients. Eighty-nine of these complications were classified as severe, requiring hospitalization or even surgery. Fifteen percent experienced major infectious complications that required reoperation. Nine patients had numbness or significant pain. There was also a number of other complications, including minor infections, thromboembolic complications, lead dislodgment, constipation, and complications associated with stoma closure. Although this procedure is associated with a high morbidity rate, most complications can be treated successfully, and although continence after the procedure is not perfect, patients do receive a significant improvement in their quality of life.

The obturator internus can function as an anal neosphincter as well. Early electromyographical data of this muscle, autotransplanted around the anus, manifested that it contracts during straining [10]. The muscle is reliably approachable through a perineal approach with division of the obturator tendon for transposition. The technique was adopted after a series of technical articles about the subject by Skácel and Laichman [11–15]. This approach has also been adopted for use in patients suffering from anismus in the creation of an anal dilating mechanism, with satisfactory results in half the cases involved in a small study of 20 patients [16].

Less frequently used alternatives to the gracilis muscle include the semitendinosus and biceps femoris muscles. In a study by Rab et al. [17], 30 semitendinosus muscles and 15 long heads of the biceps femoris were dissected from human cadavers. The long head of the biceps femoris muscle received its dominant vascular supply from the first and second perforating arteries and its nerve supply from a motor branch of the sciatic nerve. The proximal muscle usually receives a branch from the inferior gluteal artery with anastomosing vascular loops between the internal iliac, external iliac, femoral, and profunda femoris vessels [18]. Although most cases there is one nerve supply and the available muscle length measures about half of the thigh, there has been variable use of this muscle because of the complex intramuscular anastomoses within the long head of the muscle [19]. The semitendinosus muscle showed variations in vascular supply, receiving dominant vascular pedicles from the medial circumflex femoral artery close to the ischial tuberosity and from the second perforating artery. The nerve supply consisted of two motor branches the sciatic nerve. As a potential neosphincter, the semitendinosus was more advantageous because of its vascular and neural topography. This same approach has been used in the sartorius muscle in dogs for electrically stimulated anal transposition [20].

A number of animal models use autologous smooth-muscle cuffs to improve continence in both the urethral and anal sphincters [21, 22] and also to create semicontinent colostomies utilizing free smooth grafts that are either stimulated or unstimulated [23, 24]. In children with high congenital anorectal anomalies, a smooth-muscle half cylinder cuff has been fashioned immediately after the pull-through procedure as well as a part of a secondary support in infants who have repeat pull-through operations, levatorplasty, postoperative rectal prolapse, and megacolon or as a conventional supplement to posterior sagittal anorectoplasty [25].