In 1987, Christiansen and Lorentzen performed the first successful implantation of an artificial anal sphincter to treat patients with fecal incontinence [55]. A large multi-center trial noted clinical success superior to that of dynamic graciloplasty. This led to the use of the artificial bowel sphincter in TAR [56]. Encouraging results were reported by Romano et al. [57], Devesa et al. [21], and Ocares et al. [58]. However, explantation of the device can be as high as 47 % due to device-related complications [59], and long-term complications continue to increase with time [60].

The antegrade continence enema (ACE) as described by Malone et al. [61] in 1990 can be used to achieve a satisfactory pseudo-continent status with a perineal colostomy. With TAR, the most disabling symptom is constipation, and the addition of ACE can overcome this and provide better functional results in terms of both quality of continence and quality of life than retrograde enemas. Chiotasso first reported its use in conjunction with a perineal colostomy [62]. Quality of life was not significantly different between patients with an abdominal colostomy and those with an ACE and a perineal colostomy [63]. An advantage of using ACE with TAR is to render the patient pseudo-continent and to prevent disabling fecal impaction.

TAR attempts to restore the normal functions that have been lost: an acceptable reservoir with adequate capacity and ability to discharge, a discriminatory sensation, and a sphincter mechanism. The limitations of TAR, which in no way recreates the sensory apparatus and only partially recreates a functional reservoir and sphincter mechanism, prevent full continence in most patients.

All reports of TAR are retrospective analysis of a certain technique, with a small number of patients with relatively short follow-up. As such, only the morbidity and functional outcome can be evaluated. The impact of constipation, which is frequent after TAR, cannot be evaluated or compared between techniques. When compared with control subjects, quality of life appeared undiminished after TAR [64]. When comparing techniques, most studies fail to evaluate the role of a neo-rectum as discussed previously, and focus on the neo-sphincter.

Table 16.1 shows the outcomes of smooth muscle neo-sphincters especially when combined with ACE in TAR. As the quality of life is unchanged with ACE and a perineal colostomy [63], the addition of a neo-smooth muscle sphincter is questionable, but may act as a biological Thiersch cerclage of the colocutaneous anastomosis. The cuff has a higher-pressure zone and is able to maintain this increased pressure in most patients [31, 32], but Lasser pointed out that the lack of this tonicity did not correlate with a poor functional outcome. Early complications are related to dehiscence or necrosis of the perineal colocutaneous anastomosis, while late complications are related to mucosal prolapse and stricture of the neo-anus [31, 32]. The early complications were responsible for the conversion back to an abdominal stoma. Lack of irradiated tissue in the pelvis may also decrease the perineal complications [33]. The higher continence noted by Hirche et al. [65] may be attributed to the perineal and neo-sphincter training, external electrostimulation of the perineal cuff, biofeedback and colonic irrigation.

Table 16.2 shows the outcomes of dynamic and adynamic graciloplasty in TAR. The overall morbidity of TAR with dynamic graciloplasty is high with frequent complications including erosion, colonic perforation, perineal sepsis, neo-sphincter necrosis, and stenosis or necrosis of the neo-anus. Device explantation in dynamic graciloplasty was also high due to erosion and infection. Despite this morbidity, dynamic graciloplasty was associated with a high continence in several studies. The continence does improve over time [48, 51]. Surprisingly, the continence achieved with adynamic graciloplasty was comparable to the dynamic counterparts [44, 66, 67]. Both Violi [51] and Ho [48] noted similar continence in their dynamic graciloplasty patients when the stimulator was turned off. However, patients in another series were completely incontinent without dynamization [46]. Despite these contradicting results, dynamization may not have a clear benefit, and the graciloplasty may serve as cerclage akin to the neo-smooth muscle sphincter discussed before. Interestingly, addition of ACE to TAR with a neo-reservoir and dynamic graciloplasty showed complete continence to solid and liquid stool in only 50 % of the patients [68].

Table 16.3 shows the outcomes of the artificial bowel sphincter in TAR. The largest series by Romano et al. [57] was the largest cohort with no explants of the sphincter. The rest of the reports had either skin or colonic erosion that led to the explantation of the sphincter.