The use of a collagen button plug to treat PVF was first reported by Gonsalves et al., with healing observed in 4/7 (57 %) of ileal pouch-vaginal fistulas at 16 weeks [32]. The technique involves securing the button portion of the collagen plug on the pouch side of the fistula with four dissolvable sutures. The button of the plug detaches within 4 weeks with the collagen matrix left in situ. Disappointingly, these results were not maintained long-term with 0/11 PVF successfully healed at 2 years [33]. Early success probably related to the persistence of the collagen plug within the tract, but failure of local tissue in-growth coupled with the relatively short length of PVF led to long-term failure. Given these results, the use of biological tissue plugs cannot be recommended for the management of PVF. Tsujinaka et al. [31] reported the instillation of fibrin glue in the fistula tract with complete healing in 1 patient with a minimally symptomatic fistula and failure in 2/3 symptomatic patients who eventually required pouch advancement and a redo pouch.

An ileal pouch advancement is essentially a variation of the mucosal advancement flap used for a high perianal fistula. A flap of mucosa and submucosa is mobilized from the ileal pouch, the internal opening is excised, and the flap is advanced and sutured beyond the internal fistula opening. Mallick et al. [10] reported healing rates of 42 % (20/48) when advancement flap was performed as a primary procedure and 66 % (4/6) when performed secondarily after a different procedure. Similar results have been reported by others. Tsujinaka et al. [31] showed healing rates of 60 % (6/10), while Shah et al. [18] and Ozuner et al. [34] reported success rates of 44 % (17/39) and 45 % (15/24), respectively. Lee et al. [23] had a slightly higher success rate of 50 % (10/20), with the rate increasing to 83 % (10/12) when excluding patients with CD. Wexner et al. [15] reported successful fistula healing in 8/16 patients with this approach in a survey of North American colorectal units, whereas Groom et al. [14] reported only one success in 10 attempts. Advantages of the ileal pouch advancement flap include the relative simplicity of the procedure and that the flap has more distal mobility [9]. The disadvantages of this approach include the suboptimal exposure, the risk of damage to the sphincters in patients with borderline incontinence, and the fact that the flap lies on the high pressure side of the PVF. Circumferential advancement of the pouch is both technically easier and ensures more mobilization than does anterior or anterolateral flap advancement.

Sagar et al. [35] reported the results of transvaginal repair for PVF in 11 patients, each of whom had previously undergone an attempt to close the fistula with a collagen button plug. Nine (81 %) were successful at a median follow-up of 14 (6–56) months and the remaining two patients described symptomatic improvement. Burke et al. [36] published the St. Mark’s Hospital experience with transvaginal repair for PVF in 14 patients. They reported total success in 11/14 patients (78 %), although 8 required multiple attempts to achieve long-term success. The largest series of transvaginal repair of PVF reported by Mallick et al. [10] from the Cleveland Clinic described a 55 % healing rate (15/27) when repair was performed as a primary procedure and 40 % (2/5) when performed secondarily after a different procedure. O’Kelly et al. [13] reported successful repair in 5/7 patients (71 %) with this approach, and once again some patients in this series required more than one attempt before complete healing was achieved. Others have reported success rates of 0 % (0/1) [18, 31], 27 % (3/11) [15], and 100 % (1/1) [12, 14]. The repair can also be augmented by placement of a collagen patch between the pouch and the vagina.

Advantages of the transvaginal approach include better exposure than the transanal approach, decreased risk of damage to the anal sphincters, and decreased tension. The procedure can be repeated if necessary and yields satisfactory results with relatively less morbidity. Possible complications include dyspareunia, although none of the patients reported dyspareunia in the series from St. Mark’s [36], and hematoma because of the vascularity of the vagina. However, this risk can be minimized with meticulous technique, drainage, and use of a vaginal pack [13, 18]

There are five small published series reporting on the utility of the gracilis muscle interposition flap specifically for the treatment of PVF. Gorenstein et al. [37] reported successful repair in two women with PVF. Previous attempts at local repair had failed in both patients and a simultaneous diverting loop ileostomy was constructed. Anterior sphincteroplasty was performed in one patient for associated incontinence. Wexner et al. [15] reported results of a multicenter study including treatment of PVF in 26 patients, 4 of whom underwent gracilis interposition flap with a 50 % success rate. In a later publication, Wexner et al. [38] published results of gracilis flap in 53 patients, two of whom for the indication of PVF. One patient had complete healing and the patient who did not heal was eventually diagnosed with CD and opted to have a permanent ileostomy. Zmora et al. [39] published their experience with the gracilis interposition flap in 9 patients. Only one patient had a PVF and the fistula ultimately completely healed. Another report by Tsujinaka et al. [31] described one patient with a failed gracilis interposition. In general, interposition flaps are particularly useful after previous failed repairs as well as when abdominal procedures are contraindicated. The expected perioperative morbidity is 33–50 % and includes perineal wound infection, urethral stricture, fever, urinary retention, and perineal bleeding [38, 40]. Perhaps because of the technical challenge, the procedure seems to have been underused. This procedure should be preceded by fecal diversion. At present, the low reported numbers and the relative complexity of the procedure prevent it from being strongly recommended as a first-line treatment. Another form of flap used for treating rectovaginal fistulas is the martius flap; however results with treating PVF have not been published.

The technique of transanal disconnection of the ileal pouch from the IPAA, advancement of the pouch, and re-suture at the dentate line can be employed in patients with PVF, especially in slimmer patients with demonstrable mobility of the pouch above the level of the anastomosis. As noted above, advantage of this procedure is that it allows healthy, full thickness tissue to be delivered to the perineum. This operation should be offered after stoma creation. Both Fazio et al. [41] and Heriot et al. [19] showed that this procedure was successful in 1/2 of their patients.