Management of pregnancy and delivery in women with history of complex urologic reconstruction is challenging. There is a scarcity in the literature  to guide clinical practice. The aim of this study was to report pregnancy and delivery outcomes in patients with urologic reconstruction.

We conducted a retrospective national study in 16 expert centers, including all women who had at least one successful pregnancy after a major urological reconstructive surgery. Our main objective was to compare outcomes and morbidity of vaginal delivery and C-section in patients with a history of augmentation cystoplasty, artificial urinary sphincter and/or continent cutaneous urinary diversion.

We enrolled 68 women who presented 96 deliveries between 1998 and January 2019. The underlying diseases were as follows: 26 had spinal dysraphism, 7 sacral agenesis, 20 spinal cord injuries, 10 classical bladder exstrophy and 2 multiple sclerosis. Sixty-three underwent augmentation cystoplasty, 31 a continent cutaneous urinary diversion (Mitrofanoff or Monti’s principle) and 6 artificial urinary sphincter implantation before pregnancy. Twenty-two also previously underwent one another urological surgical procedure, mainly on bladder neck (Young Dees, Goebel Stoeckell, aponevrotic sling …). Of the 10 women with bladder exstrophy, one had a bicornuated uterus and one have had a complex vaginoplasty. Forty-seven women presented only one delivery, 14 had 2 deliveries and 7 had 3. 45 women (55.5%) were under antibioprophylaxis during pregnancy. The rate of febrile urinary tract infection during pregnancy was of 34.8% (16) in the antibioprophylaxis group compared to 31.4 (11) in the non-antibioprophylaxis group (p=0.751). We observed 13.5% renal colic (13) and 14.6% (14) required urinary diversion by nephrostomy or ureteral stent during pregnancy. Ten percent reported difficulties in self-catheterization during pregnancy and in case of cutaneous urinary diversion, the rate  was higher when the stoma was located in the right iliac fossae. 13.5% (13) of the included women reported de novo or augmented urinary leakage during pregnancy. No cystoplasty perforation was observed. No in utero fetal death was reported. Concerning delivery mode, 51% gave birth by a planned C-section, 31% by eutocic vaginal delivery, 5% by instrumental vaginal delivery, 10% by emergency C-section for spontaneous labor before planned-C-section, and 3% by emergency C-section after attempted vaginal delivery and for fetal suffering. Regarding anesthesia during labor and delivery, 19 (20%) gave birth with classical analgesics alone. Thirty-four had epidural or rachi-anesthesia (35%). Thirty gave birth by general anesthesia (31%). 35.3% of the newborns were preterms, with a median gestational time of 36 weeks of amenorrhea [21-41]. Median fetal weight was 2600g [429-4090g]. Of the 96 newborns, 3 died at birth. One third of the planned C-section got complicated, mainly by injury of the urological reconstruction, despite more than half of the procedures being performed in the presence of an expert urologist. 
Concerning functional outcomes, urinary continence one year after delivery was unchanged for 93.5% of the women. The mode of delivery did not seem to impact these results. 8.3% of them required surgical management for secondary urinary incontinence (mainly sub-urethral slings). 85% of the women were continent 6 months after delivery, compared to 91% before pregnancy. 6 % reported persistent de novo urinary incontinence after delivery, with no statistical difference between C-section and vaginal delivery. Four women required surgical management of secondary pelvic organ prolapse. Among them, one had a bladder exstrophy and gave birth by C-section, and the 3 others had spina bifida (one delivered vaginally and the two others by C-section).

To our knowledge, the current study is the biggest cohort reporting pregnancy and delivery issues in women with complex urinary reconstruction. Our results did not find any impact of antibiotic prophylaxis on febrile urinary tract infection episodes during pregnancy, nor on fetal weight or on term of delivery (p> 0.05). The literature reports between 45-100% of symptomatic lower urinary tract infections and about 30% of febrile UTI. In our study, 32% of the women presented at least one pyelonephritis during their pregnancy, which corresponds to the usual frequency encountered in the literature. In our serie, 1/10 of the patients reported difficulties in self-catheterization at late stage in pregnancy, which is less than the 30% usually reported in the literature. The literature reports between 10 and 40% of emergency diversion of the upper urinary tract during pregnancy, for severe acute renal colic or for pyelonephritis resisting the initial conservative management, with an average of 11% requiring urinary diversion during pregnancy. In the existing literature, planned C-section still represents the predominant mode of delivery in this population. The overall complication rate of planned C-section in our cohort is of 1/3. These complications are mainly represented by urological complications, followed by digestive complications related to ileal or colic injury during laparotomy. Urinary continence after delivery is mainly unmodified, with an average of 84.7% of the considered women being continent 6 months after delivery, compared to 91% before pregnancy. Delivery mode did not impact these results (p=0.293), nor did multi-parity (p=0.572). Systematic planned C-section to prevent any deterioration of the urological reconstruction is not justified by the existing literature.

Pregnancy and delivery in women with urinary reconstruction is challenging and must be managed in a reference center. Delivery mode does not impact urinary continence and C-section is at risk of urinary reconstruction injury. We strongly believe that all women who had undergone complex urinary tract reconstruction should be proposed vaginal delivery as the first option, if there is no obstetrical or neurological contra-indication, in order to prevent urological reconstruction injury.