Congenital cervicovaginal agenesis is due to defective development in the Müllerian ducts and sinovaginal bulbs. The failure of distal Müllerian structures to develop may give rise to agenesis of the cervix and proximal vagina. Failure of the development of the sinovaginal bulbs and fusion with the Müllerian duct can cause agenesis of the lower part of the vagina with an intact cervix and a blind-ending upper part of the vagina [8]. The uterus, tubes, ovaries and external genitalia are normal in this clinical condition, unlike Mayer–Rokitansky–Kuster–Hauser syndrome, where there are varying degrees of uterine hypoplasia ranging from rudimentary uterus to uterine agenesis. Congenital agenesis of the uterine cervix and vagina in the presence of a functional endometrium is an extremely rare Müllerian anomaly. The incidence of congenital vaginal atresia is reportedly 1 in 4,000 to 1 in 5,000 female births, with normal development of the uterus observed in less than 10% of these subjects [9]. The American Society of Reproductive Medicine (ASRM) system of classification categorises vaginal agenesis as 1a and cervicovaginal atresia as 1b [9, 10]. ASRM classification is widely used. The other classification systems for Müllerian anomalies are the embryological–clinical classification [11], the Vagina Cervix Uterus Adnexa-associated Malformation classification [12] and the European Society of Human Reproduction and Embryology/European Society for Gynaecological Endoscopy classification [13].

Patients with cervicovaginal agenesis present with primary amenorrhea and cyclical pain in the abdomen as they have a functioning endometrium. Generally, these anomalies go undiagnosed until sometime after menarche, as these patients have age-appropriate female secondary sexual characteristics, and the external genitalia may look normal [2]. Pain is due to obstructed menstrual flow or endometriosis. The primary goals of treatment in these patients are to construct a conduit for a free flow of menstrual blood, thereby relieving cyclical pain and preserving the uterus; a neovagina for sexual intercourse and, if possible, to maintain fertility. Historically, these groups of patients were managed using different procedures such as hysterectomy and neovaginoplasty, cervical canalisation combined with vaginoplasty or uterovaginal/uterovestibular anastomosis [13,14,15,16,17,18]. There are limited data available on the safety and success of various reconstructive surgical procedures [18]. Rock et al. [14] and Fujimoto et al. [2], in their review, have reported high restenosis rates with recanalisation procedures. A failed procedure with inadequate menstrual drainage would cause pyometra [7].

The patients in our series have a poor socio-economic status and a rural background where adequate medical facilities are not always available. The average duration of symptoms in our series was 4.5 years, which indicates misdiagnosis and improper guidance or referral. Approximately half (53%) of our patients underwent failed surgeries before reaching our centre, resulting in scarring. Apart from relieving pain, preservation of the uterus and menstruation is also a social and cultural affirmation of gender role identity in this set of women. The absence of a uterus in a woman and the resultant infertility would have a huge negative influence on the chances of the woman being accepted in an arranged marriage practice prevalent in a conservative Indian rural background. The ability to afford the increased cost of multiple/staged procedures becomes difficult for these patients. With this background, we decided on restoration of uterovaginal continuity by utero-colo-neovaginoplasty/sigmoid colon vaginoplasty as the ideal procedure to fulfil the treatment aims of pain relief, preservation of the uterus, restoration of menstrual flow and the ability to have sexual intercourse. Although Kannaiyan et al. [7] described this surgery as a “compromise” in view of unphysiological anastomosis between the colon and uterus at the cost of compromised fertility, we still feel that, with a follow-up of 2 to 12 years (average 9 years) without much morbidity, this procedure can be considered in this set of patients with health care limitations.

There were four (25%) complications in our series in the initial couple of patients. Two (12.5%) developed stenosis of the neovaginal orifice. One (6.25%) developed stenosis of the colo-uterine anastomosis and 1 more had mucosal prolapse at the neovaginal orifice. None had an anastomotic leak. All the patients eventually had resumption of painless cyclical menstrual flow.

Similar rates of neovaginal stenosis were seen in the literature. Kannaiyan et al. [7] reported 1 (9%) stenosis in 11 patients, whereas Kisku et al. [13] reported 2 (10%) in 20 patients. Kannaiyan et al. [7] proposed a wide myomectomy by excising a 3- to 4-cm disc of the posterior wall of the uterus to avoid stenosis and obstruction at the colo-uterine anastomosis. We have adopted the same in our later cases and found it to be useful. The distal end of the sigmoid colon conduit was anastomosed to the vaginal vestibular edge in a circular end-to-end fashion initially. This was later modified as a cruciate anastomosis to the V-shaped mucosal flaps raised in the vestibule, which we believe has helped us to avoid neovaginal stenosis. Tension in vestibular anastomosis can be avoided by rotating the sigmoid conduit by 180° for anti-peristaltic anastomosis to the vestibule. Prolapse of the neovagina and excessive mucosal discharge from the sigmoid conduit can be minimised by keeping the length of the conduit optimum, as a redundant sigmoid conduit can be responsible for these complications. Tailoring of the lower segment of the sigmoid conduit by excising a linear strip of conduit in its ante-mesenteric border is done if the diameter of the conduit is quite wide. The advantage of the sigmoid conduit is that it is lubricated, lined with epithelium and resistant to local trauma. Five of our patients have male partners and are sexually active. When the couples were interviewed, they did not report any feelings of inadequacy during any stage of the sexual act.

Minimal access surgery has further improved the outlook for these complex anomalies. Alborzi et al. reported a series of 7 cases of cervicovaginal atresia in whom the sigmoid colon vaginoplasty was done laparoscopically, with good outcomes [19].

Fertility in cervicovaginal or vaginal agenesis depends on the presence or absence of the cervix. Pregnancy should be possible in the presence of a competent cervix if there are no other fertility issues. Four patients in this series with distal vaginal agenesis having a normal cervix and upper vaginal pouch were counselled to have a pregnancy but not vaginal delivery. Obstetricians should be aware of the altered anatomy and be prepared for caesarean section.

Patients with cervicovaginal agenesis have a lower chance of fertility [5, 20]. Associated uterine anomalies, endometriosis and previous attempts at recanalisation further reduce the fertility [5, 9, 13, 20]. Few successful pregnancies have been reported in the literature [3, 5, 16, 21] in the case of cervical atresia following various recanalisation procedures such as cervical drilling [15], stenting [3, 16] and skin graft [18]. Recanalisation procedures carry a higher risk of stenosis and pyometra. Additionally, compared with isolated cervical atresia, these recanalisation procedures may not be possible for cervicovaginal agenesis. The utero-colo-neovaginoplasty is a wide patent anastomosis allowing free drainage of uterine content, but, at the same time, it is associated with a high chance of pregnancy loss or related complications. We are not aware of any reports of successful pregnancy in patients with cervicovaginal agenesis who have undergone sigmoid colon vaginoplasty. Thirteen patients with cervicovaginal agenesis in our series were counselled not to conceive and opt for surrogacy. In surrogacy, the retrieval of oocytes is performed regularly via a transvaginal posterior fornix approach. This may not be possible in view of the altered anatomy in these patients, and the best possible approach will be laparoscopic transabdominal ultrasound-guided oocyte aspiration [22]. One of our patients was counselled for the same by our infertility experts.

Given the rarity of the condition, our study is limited by the number of cases. This could be addressed by a multicentre review or a systematic review with meta-analysis.