The possibility of partially restoring anatomy should lead to a more physiologic evacuation in these patients [26, 27]. Although ABS is actually more expensive than graciloplasty, it is easier to implant and more easily accepted by patients because of less difficult training. The ABS does not need the substitution of a pacemaker battery. Moreover, the results of the TAR with graciloplasty both in terms of complications and faecal continence are quite controversial. Both early and late complications have been reported. Among these, graciloplasty stenosis, fibrosis and necrosis of the muscle, perineal ptosis and perineal infection have been frequently described [28-32].
Patient selection for ABS implantation is mandatory. We believe the following conditions should be considered as exclusion criteria:
• severe cardiovascular and respiratory diseases;
• perineal Crohn's disease;
• advanced neoplastic disease (T3-T4, involvement of perirectal fatty tissue or perirectal lymph nodes);
• poorly differentiated tumours and anaplastic forms (because of high risk of local or systemic recurrences);
• patients requiring post-operative radiation therapy.
As for rectal cancer, patients with tumour staging T1-2N0 and early involvement of the sphincter can represent a good indication for this technique. Moreover, patients must be well motivated and both physically and psychologically skilled to manage the device.
This procedure can be performed as a synchronous or delayed reconstruction. In the first case the per-ineal colostomy is performed at the same time as the Miles operation and a sizer is placed around the colostomy. In the delayed procedure the perineal colostomy with the placement of the sizer is performed at least 2 years after the APR. In both the synchronous and delayed procedures, after two or three months, with a small perineal incision the sizer can be removed and easily replaced with the cuff (deferred procedure). The other components of the ABS are then implanted. The goal of this deferred procedure is both to allow a careful selection of indications on the basis of the definitive pathological report (advanced stages are excluded) and to prevent erosion of the colon with subsequent infection. In fact, the sizer previously placed around the perineal colostomy will elicit fibrosis and a barrier between the implant and the colon wall will result. An additional advantage is economically related, as an ABS is not wasted should any infection occur in the interval between the placement of the sizer and the definitive implant.
All the patients must be followed up by manomet-ric and radiological evaluations. Manometry is the most reliable method to achieve an objective evaluation of ABS effectiveness. Three manometric parameters must be evaluated:
• basal pressure with the cuff inflated, a post-operative significant increase of this value contributes to faecal continence;
• basal pressure with the cuff deflated, a low value of this parameter implies a wide neo-anal opening and easy defecation, whereas high pressure lead to develop symptoms of obstructed defecation;
• the time required to inflate the cuff again after the opening of the artificial sphincter to evacuate. Sufficient time is necessary to completely empty the rectum as some patients complained of impaired defecation because of a closure of the cuff quicker than the seven minutes normally required.
As for defecography, a series of X-rays allows the filling and the emptying of the cuff to be checked, as well as the correct sphincter function.
Recently, an Italian multicentre study reported disappointing long-term results after ABS implant for faecal incontinence . The same complications may occur in patients who undergo TAR, that is:
• cuff deplacement
• mechanical impairment of device
• obstructed defecation
Constipation occurred in three cases of our series. The loss of sensitive receptors in the levator and sphincter muscles surgically ablated inevitably impairs the ability to be aware of the presence of faecal contents in the neorectum and thus activate evacuation. Clinical experience with TAR and electros-timulated graciloplasty has provided clear evidence of the constant occurrence of this complication, so that ingenious surgical solutions have been proposed to overcome the problem . After any type of TAR patients must be trained to evacuate the neorectum at definite intervals of time with the help of enemas and suppositories.
Most Authors reported a high rate of infections, cuff erosions and reoperations for ABS previously implanted for faecal incontinence [33, 35]. Although in our series we reported three cuff explantations for skin erosion respectively 7, 10 and 21 months after the operation, the rate of infections was significantly lower. Attention to some technical details at opera tion such as location of the device far from the skin and loose around the bowel, absolute sterility and suture of a finger glove to the neo-anus that allows a finger to be inserted in the bowel without an accidental passage of faeces  were of the utmost importance. A further improvement in the complication rate might be explained by the presence of an ileostomy and the use of the sizer.
As far as radiotherapy is concerned, only one patient in our series received radiotherapy because of the more advanced pathological stage; this patient had the complete removal of the device 40 days after the operation. Pre-operative radiotherapy has been recently reported to significantly reduce the local recurrence rate of rectal cancer although a survival benefit remains to be proven [36, 37] and its use for early stages is also questioned. The complication rate seems to be higher when compared with surgery alone with particular reference to leakage rate and sphincter function. Miles reconstruction with the use of a sphincteral substitute is reserved for T1-2 cancer with sphincter involvement and does not require a standard anastomosis, so that neoadjuvant treatment should not be considered an absolute contraindication.
As compared with the pre-operative condition, QoL was significantly improved in patients treated with the delayed procedure. Similar good scores were also reported after the ABS implant in patients operated on with the synchronous procedure (Tables 5, 6). A careful evaluation of patients' psychological habitus is important to achieve good results.
The ABS is a valid option for reconstruction of selected patients previously treated with an APR. Nevertheless, a long-term follow-up shows that the results of the TAR performed using an ABS may deteriorate with time and may be worse than patient expectations, so that the patient should always be correctly informed and aware of the possibility of failure.
Was this article helpful?