In the case of cervicovaginal agenesis and functional uterus, surgery consists of three steps: 1- creation of neovagina, 2-creation of neocervix, and 3-maintaining the continuity between neovagina, neocervix, and uterus [6].
However, when there is no sufficient cervical tissue, the outcome of cervicovaginal reconstruction remains unclear and there is a high probability of cervical stenosis that may lead to hysterectomy [16].
Bowel vaginoplasty was traditionally used in transgender women or biologic women with failed skin graft vaginoplasty [17]. It is a kind of direct uterovaginal anastomosis which can be used for any kind of cervical anomaly (fibrotic band, fragmentation, complete absence) if performed by our technique, in which cervix is bypassed.
Baldwin did the first bowel vaginoplasty in 1907 with a segment of ileum and in 1911, Wallace performed sigmoid colon vaginoplasty [18]. Finally, Ohashi reported laparoscopic sigmoid vaginoplasty for the first time in 1996 [12].
Different segments of bowel have been used for vaginoplasty, such as sigmoid, ileum, cecum, and ascending colon. But the best alternative is sigmoid which has an appropriate diameter, produces lubricant mucosa, and due to its location, has the best accessibility for anastomosis as a neovaginal conduit [19].
Bowel vaginoplasty is a practical choice at any age because it creates a conduit with potential growth. Sigmoid neovagina is resistant to mucosal injury with a very low incidence of graft necrosis because of preservation of its vascularity [19].
In 2011, Kim et al. reported that 6 and 12 months post operation, sigmoid neovaginal pH was 6.5 and 6, and squamous transformation of columnar epithelium was observed in only one of the 12 patients [20].
Predisposing factors for bowel-related complications in sigmoid vaginoplasty are tension over vascularized pedicle, short mesentery, obesity, and masculinized pelvis. So, surgeons should be highly skilled [21].
We had a very small intraoperative complication only in one patient whose rectum was inadvertently opened which was diagnosed and repaired immediately. Risk of rectovaginal fistula in primary cervicovaginoplasty is about 0-4%. But we had no rectovaginal or ureterovaginal fistula [17].
In the literature, the risk of anastomosis leakage is 0-7.1% and we had no anastomosis leakage in our series [22].
We had no post-operative ileus and stenosis of anastomosis site.
Bouman et al. performed 42 transgender sigmoid vaginoplasties and reported one (2.4%) case of anastomosis leakage, six (14.6%) patients with introital stenosis, and one mucosal prolapse subject. We had no mucosal prolapse or introital stenosis in our study [17].
None of our patients needed blood transfusion, and they had no post-operative surgical site infection.
In the literature, neovaginal stenosis is divided to two subtypes: introital stenosis (0-55.6%) and diffuse stenosis [17]. One of our patients had only proximal part stenosis and an episode of PID and pyometra 14 months after the second surgery. Due to this episode of infection, there was a possibility of stenosis and incomplete drainage. However, she refused vaginoscopy and further investigations. But she has a regular period.
In 2009, Kannaiyan et al. reported a series of 11 subjects among whom seven patients had cervicovaginal agenesis and functional uterus and underwent sigmoid vaginoplasty by open method. The mean operation time was 239 min, they reported that no mold was required to prevent stenosis and only one (9%) patient had introital stenosis. They reported no other complications [23]. Our method is similar to theirs, but it was done by laparoscopy and the mean surgical time was lower in our study (207 min) despite the laparoscopic route.
The important point in these patients is to build a wide and patent neovagina-uterus anastomosis, to completely drain the uterus and diminish the risk of pyometra as a result of incomplete uterus drainage and the recurrence of dysmenorrhea and endometriosis [23]. In agreement with kannaiyan et al., in our study, resection of a healthy portion of myometrium was required to maintain an open standing anastomosis to reduce the risk of ascending infection [23]. By this method, we are able to make a non-competent free drainage passage. Therefore, if the patient desires future pregnancy, she needs abdominal cerclage before attempting to get pregnant. We did not put abdominal cerclage simultaneously with cervicovaginoplasty.
Additionally, we should caution the patients against natural vaginal delivery and emphasize the necessity of cesarean section [23].
Kisku et al. reported 20 patients with cervicovaginal atresia, functional uterus and open colovaginoplasty among whom the rate of anastomosis leakage and recurrence of dysmenorrhea and endometriosis was 5% and 10%. Mild stenosis of neovagina were detected in 10% of the patients [24].
Yang et al. reported 22 cases of laparoscopic assisted sigmoid vaginoplasty. But they performed hysterectomy for all patients with functional uterus (15 patients) due to cervical agenesis. In their study, the mean time of surgery was 279±30 min, the mean estimated blood loss was 334±71 cc and the mean neovaginal length was 11.3±1.2 cm [25].
The mean operation time in our study was 203.5 min, which is shorter than Kannaiyan series (laparotomy, 239 min) and Yang series (279 min, laparoscopic vaginoplasty and hysterectomy) and similar to Bouman series (210 min, laparoscopy in transgender women) despite the fact that we had an additional procedure of uteroneovaginal anastomosis in our patients [17, 23].
In our patients, the mean vaginal length was 8.9 cm (7.6-10.5) which is a very good length.
(Bouman: 16.3±1.5 cm, Cao: 12.5cm in and Yang: 11.3±1.2 cm) [17, 19, 25].
Given that cervicovaginal agenesis is mostly diagnosed and treated at peripubertal age, even after long follow-up periods, patients may not reach the age to attempt to get pregnant. Especially in our society, with cultural limitations, women may not have a sexually active life prior to matrimony.
In our study, just two patients were sexually active and both were satisfied with their sexual relationship. None of them had attempted to get pregnant prior to the research.
Due to rarity of this congenital malformation, we had a small sample size so the low rate of adverse events is not reliable in our study.
Although all the surgeries in our center were done by a gynecologist who is skilled in advance minimally invasive surgery (SA), it can also be done by multidisciplinary approach and collaboration of gynecologists and colorectal surgeons.