The ureter can be well protected by surrounding tissue in the retroperitoneal space and is not susceptible to injury. Iatrogenic injury is the most common causative factor of ureteral injury (1), which may result in ureteral stenosis or atresia, hydronephrosis, and even progressive impairment of renal function.

Short stenosis of the proximal and middle ureter can be managed by endoscopic procedures, pyeloureteroplasty, or end-to-end ureteroureterostomy (1). However, the management of long ureteral stenosis (>2 cm) is a challenge for urologists. Although bowel interposition and renal autotransplantation have been applied for long ureteral stenosis, bowel and vascular complications are not uncommon and cannot be ignored (2, 3). Oral mucosal graft (OMG) is well-suited for the urinary tract, and several studies have reported the application of oral mucosal grafts in the treatment of ureteral strictures (4–6). However, to our knowledge, no study has reported the application of lower-lip mucosal graft (LLMG) in ureteroplasty. Based on our experience of urethroplasty with LLMG and the application of the da Vinci Xi surgical system, we performed robotic ureteroplasty with LLMG for proximal and middle ureteral stenosis. We presented our initial experience as follows.

The demographic characteristics and preoperative data are summarized in Table 1. The etiologies of ureteral stenosis included ureteral Holmium laser lithotripsy, percutaneous nephrolithotomy, open ureterolithotomy and laparoscopic uretero-ureterostomy. A total of seven male and six female patients met the criteria with a mean age of 45.9 ± 13.6 years. Proximal ureteral stenosis was observed in ten patients, and middle ureteral stenosis was detected in three patients.

All surgical procedures were completed without conversion to open surgery or other reconstructive surgeries. No intraoperative complications were observed. The median stenosis length was 3.5 cm (ranged, 3.0–4.5 cm). The mean length and width of the LLMG were 3.81 cm and 1.27 cm, respectively. The mean operative time was 212.31 ± 23.06 min, and the mean anastomosis procedure of the ureter and LLMG was 36.54 ± 6.58 min. The median blood loss was 35 ml (range, 20–60 ml), and none of the patients needed blood transfusion. The median postoperative hospitalization duration was 7 days (range, 5–9 days). All patients underwent removal of the double-J stent at 8 weeks after the operation. The clinical outcomes are presented in Table 2.

A total of three patients suffered postoperative complications. One patient suffered from fever (Grade I) and was treated with antipyretics. Urinary tract infection requiring antibiotics (Grade II) was observed in one patient. Complications at the lower lip donor site, including prolonged discomfort, pain, and numbness, were not observed in all patients.

During the follow-up period (range, 6–18 months), all thirteen patients underwent ultrasonography and intravenous urography. Furthermore, computed tomography urography was performed in ten patients and another three patients underwent magnetic resonance urography. Hydronephrosis improvement was observed in 12 (92.3%) patients, and the radiological results indicated good ureteral patency (Figures 4A–D). Preoperative flank or abdominal pain was relieved. Pre- and postoperative ⁹⁹mTc-mercaptoacetyltriglycine renography were performed in seven patients. The preoperative and postoperative glomerular filtration rates of the operative side kidney were 19.57 ± 4.43 ml/min and 26.57 ± 4.92 ml/min, respectively. Seven patients underwent ureteroscopy. One patient experienced renal colic and fever 8 weeks after removal of the double-J stent, after which nephrostomy tube insertion and ureteroscopy were performed. Anastomotic edema and short stenosis (Grade III) were observed, and a 7 F double-J stent was inserted again for 3 months; however, the patient failed to have the double-J stent removed again and was treated with a chronic double-J stent.

Iatrogenic factors are the most common etiology of ureteral injury (9). Ureteral injury may lead to ureteral stenosis, atresia, and ureteral fistula. Endoscopic surgeries and ureteroureterostomy are suitable for short ureteral stenosis. Long-segment stenosis remains an enormous challenge for urologists. Repeated replacement of the ureteral stent and permanent nephrostomy are compromise treatments; however, these treatments may result in poor quality of life and recurrent urinary tract infections. Ureteral reconstructive surgery is still considered the gold standard treatment for long-segmental ureteral stenosis (10). Bowel interposition (appendix, ileum), the bladder flap technique, OMG ureteroplasty, and renal auto-transplantation are optional therapies for long ureteral stenosis.

The OMG has special histological characteristics such as hairless, thick, and non-keratinized epithelium, compatibility with a wet environment, and rich vascularity of the lamina propria (11), and it has been demonstrated to be a good substitute for urethroplasty with good outcomes (12). Based on similar histological characteristics of the urethral mucosa and ureter mucosa, urologists have performed ureteroplasty with OMG for long and complex ureteral stenosis (4–6, 13).

Unlike the bowel interposition and bladder flap, the freed OMG is sutured at the target site of the ureter and acts as part of the ureteral wall. Both the normal ureter distal to the stenosis and the vesicoureteric junction are preserved, so the physiologic peristalsis and anti-reflux functions of the ureter are not affected. Moreover, the OMG does not require the simultaneous transfer of blood supply vessels, however, ileal ureteric replacement requires a segment of the intestine with mesentery, and the harvested bladder flap needs to be attached to the bladder for blood supply.

The donor sites of OMG include the buccal mucosa (inner cheek, lip) and lingual mucosa (tongue) (11). There is no consensus among urologists as to which site is preferred. However, to our knowledge, no study has reported the application of lower-lip mucosal graft (LLMG) in ureteroplasty. All patients in our study had OMG harvested from the lower lip for the following reasons. First, the donor site was easily exposed without the need for a mouth retractor even if the patient was placed in the special position described in our study. Second, orotracheal intubation could meet surgical requirements without the need for nasotracheal intubation. Third, salivatory duct damage can be avoided. Moreover, the postoperative condition of the donor site was easy to check. Kamp et al. compared donor-site complications of the lower lip with those of the inner cheek, and greater long-term morbidity was observed in the lower lip group (14). However, in our study, complications related to donor sites were not observed in any of the patients. Compared to the LLMG, the lingual mucosa could be harvested longer in length, but it was technically harder because a mouth opener was needed and the tongue needed to be pulled out. Complications such as difficulty with fine motor movements, numbness or slurred speech may occur after lingual mucosa harvesting.

The length and width of the graft were measured according to the characteristics of the intraoperative ureteral defect. Importantly, shrinkage of the graft size may have occurred during the growth time when we harvested the graft (15). Some details were emphasized in our study. The length of the LLMG graft was matched that of the defect, and we harvested LLMG with width of 1.0–1.5 cm, which were relatively wider than those of the ureteral defect. The boundary of the graft was marked before submucosal injection to avoid submucosal dilation affecting the true length and width. The bleeding point at the donor site was managed with continuous suturing and compression hemostasis rather than coagulation forceps to reduce thermal damage to blood vessels and nerves.

Comparing to open surgery, robotic ureteral reconstructive surgery has the advantages of less invasion, less bleeding, faster recovery, and shorter postoperative hospitalization (16, 17). Other advantages include image amplification, precise dissection and suturing. Moreover, the robotic surgical system has special superiorities including 3-D stereoscopic vision, multifold magnification of the image, multiple manipulator arms, and delicate operation, which could facilitate intracorporeal suture (18). These features make robotic approaches more suitable for ureteral reconstruction. Zhao et al. first described robot-assisted buccal mucosa graft ureteroplasty, and all four patients were successful during the follow up period (19). Lee et al. reported their multi-institutional experience with robotic ureter reconstruction by using BMG, and surgical success was observed in 87.0% (47/54) of the patients (6). Compared to previous reports, all patients in our study had BMGs harvested from the lower lip, and the success rate was comparable with the results of previous reports (6, 19).

Identification of the ureteral stenosis segment is the key step, and it is sometimes difficult in surgery, especially for patients with previous stenosis-related procedures. The hard and cicatricial tissue surrounding the ureter, which we call the “ureteral armor”, makes it difficult to distinguish the ureteral stenosis segment. Several methods can be used to solve this problem. First, preoperative radiological examinations could indicate the stenosis location; moreover, the gradual decrease in the degree of dilatation of the proximal ureter could help to identifu the stenosis site. Yang et al. applied real-time 3D imaging combined with a surgeon's cognitive fusion during the operation procedure to locate the stenosis and achieved satisfactory effects (5), but special imaging software was needed. Second, the intraoperative combination of ureteroscopy and laparoscopy is a simple and effective method. A modified supine lithotomy position was used, which is beneficial for ureteroscopic insertion. The ureteroscope is retrograded to the distal end of the stenosis, and then the luminous tip of the ureteroscope could be seen from the laparoscope so that we could identify the stenosis site. Second, intraurethral or intravascular ICG fluorescence imaging could facilitate the identification of the stenosis (18). A preexisting ureteral catheter or nephrostomy tube is needed for intraureteral ICG, and the ICG may spill out after the incision of the ureter; thus, the surgical field can be contaminated by ICG, which may influence the subsequent judgement of the ureter condition. However, intravascular ICG can avoid these limitations and was easy to use, and we preferred to use intravascular ICG during surgery.

The reconstructive principles of robotic ureteroplasty are consistent with those of open surgery, including tensionless, watertight, and mucosa-to-mucosa anastomosis, gentle manipulation of the ureter, and sufficient blood supply (17, 20, 21). In accordance with these principles, several technical difficulties must be considered. First, the “ureteral armour” must be completely excised until the ureter appears, and the scar tissues on the ureteral wall were excised until the pink healthy ureteral mucosa was exposed at both ends of the ureteral stenosis. The ureteral polyps in the ureter lumen must be completely excised. These details are important for preventing recurrent stenosis (4, 5, 22). Second, ischemia, infection, and localized adhesion constitute the major factors of failed ureteral reconstruction when the tissue transfer approach was used (23). The blood supply of the freed BMG was partly provided by the ureter. Extensive ureterolysis may disrupt the blood supply of the ureter, so precise positioning of the target segment is particularly important, which could provide maximum protection of the ureteral blood supply. Third, the surrounding tissues, such as perinephric fat and omentum, could provide extra blood supply for freed graft (24) which could improve the survival rate of the graft. Pedicled omentum flaps have been increasingly applied in ureteroplasty because of their its special characteristics, including vascular development and remodelling, anti-infection effects, and regeneration of tissue (25). We preferred to use an omentum flap as covering tissue because the harvested area was large enough for tension-free wrapping.

Except for one patient who experienced recurrent stenosis during the follow-up period, none of the remaining patients experienced recurrent stenosis or increased hydronephrosis. The patient underwent surgery at the early stage of the LLMG urteroplasty procedure, predominantly because of insufficient relevant experience. With the accumulation of experience and the continuous improvement in surgical techniques, our total success rate was 92.3%, which was comparable to what has been reported in previous OMG studies (4, 6).

The study summarized our initial experience and several limitations exist. The main limitation was that it was a retrospective study in a single center, and selection bias could not be avoided. Second, the control group was lacking and the total sample size was small. For further study, a multicenter randomized controlled study is recommended. Third, the follow-up periods of some patients were relatively short, and long-term follow-up is needed in future studies.

Robotic ureteroplasty with LLMG is a safe and feasible technique for treating ureteral stenosis longer than 2 cm.