Management of Urologic Complications in Gynecologic Oncology Surgery

Ureteral Injuries and Management

The course of the ureter makes it a prime organ for injury and involvement during gynecologic surgical procedures. Although it is located posteriorly in the retroperitoneum, it is in close proximity to the uterus and upper vagina and their associated vessels ( Fig. 19.1 ). Ureteral injury can cause significant morbidity, prolonged hospital admission, repeat operative procedures, renal insufficiency, and reduced quality of life and can even result in legal action. In fact, “It is the most common non-obstetric complication leading to legal action against gynecologists.”

Seventy-five percent of the injuries to the ureter in general are iatrogenic, with 73% of these resulting from gynecologic surgical procedures, mostly involving the lower ureteral segments. Ureteral injury is reported to occur at an overall frequency of 1.6 per 1000 cases, with only 11.5% of these injuries recognized intraoperatively. It most commonly occurs during abdominal hysterectomies (0.04%–3.0%), laparoscopic or robotic hysterectomies (0.19%–6.0%), and vaginal hysterectomies (0.02%–0.47%).

If not recognized intraoperatively, the diagnosis of a ureteral injury is made postoperatively when high clinical suspicion arises. There are no specific or discreet symptoms for ureteral injury. Preventative measures such as intraoperative placement of temporary ureteral catheters or stents, routine use of cystoscopy, and near infrared (NIR) fluorescence imaging are sometimes used to identify the course of the ureter and to attempt to minimize the risk of ureteral injuries. The use of cystoscopy was studied in a meta-analysis by Gilmour and colleagues, who found that it revealed up to 90% of unsuspected ureteral injuries. This study focused mostly on benign gynecologic cases but still shows the usefulness of intraoperative cystoscopy in this setting. Indocyanine green (ICG) can be used for ureteral identification because it binds urothelial proteins. It has been shown to be effective in ureteral identification in robotic and laparoscopic procedures in which tactile feedback and two-dimensional viewing do not allow for ready identification of the ureter. In a study by Siddighi and colleagues of 10 patients, the authors were able to visualize the course of the ureter by using ICG and NIR laser imaging with the da Vinci robotic system (Intuitive Surgical, Sunnyvale, California). however, difficulty in ureteral identification was noted in morbidly obese patients. The patients were followed for 2 months postoperatively with no sequelae from the intravenous dye administration. Korb and colleagues performed a study involving the use of IRDye800CW, a renally excreted fluorescent dye with absorption and emission spectra that overlap those of ICG and therefore allow the same imaging modality to be used. This dye was tested in a laparoscopic setting in adult female pigs and is currently undergoing clinical trials. It showed that ureteral fluorescence was dependent on ambient light and blood supply. Fluorescent signal in the plasma peaked before signal in the urine; however, it still remained high and caused background fluorescence, making it hard to see the ureter at times. According to Korb and colleagues, the ureter was always identifiable. In an oncologic setting, however, in which normal vascularity is potentially altered, this could cause significant background noise and make it difficult to identify the true ureteral course.

Other possibilities for prevention include stents and lighted stents, as in laparoscopic or robotic surgical procedures, in which tactile feedback is not possible. Currently there is debate regarding whether the use of preoperative ureteral stents results in any added benefit in reducing ureteral injury versus added morbidity with placement. A study by Merritt and colleagues evaluated perioperative morbidity with preoperative stent placement in 315 patients who underwent gynecologic oncology surgical procedures. During this study the ureteral stents were placed by gynecologists, a process that took 5 to 8 minutes on average. Only one patient was found to have a ureteral injury, and this was found postoperatively when bilateral ureterovaginal fistulas occurred. Not surprisingly, this complication was felt to be a result of the extent of the pelvic procedure and the patient’s history of pelvic radiation, rather than stent placement. Other potential complications included urinary tract infections (UTIs) in 1.48% and acute renal failure in 0.6%. Merritt and colleagues concluded that preoperative ureteral stents did not cause increased morbidity with placement and were cost-effective.

Others feel that ureteral catheterization reduces ureteral peristalsis and pliability and moves the ureter into an ectopic location, making it harder to identify. A large randomized trial by Chou and colleagues involving 3141 women undergoing major gynecologic surgical procedures showed no added benefit of prophylactic ureteral catheterization; however, severe ureteral injury was less common in patients who underwent preoperative stenting.

Some believe that prospective ureteral dissection and exposure can prevent future injury. This not only familiarizes the gynecologist with the ureteral anatomy, but in complex cases, such as oncologic procedures in which the anatomy is altered, it may provide an added benefit. However, one could argue that prospectively dissecting out the ureters routinely in all cases will prolong surgical time more than simple cystoscopy and ureteral catheter placement and could expose the ureter to potential damage, which is more likely than with simple placement of ureteral catheters.

Ureteral injury usually occurs lateral to the uterine vessels but can also occur at the uterovesical junction, near the infundibulopelvic ligament and the pelvic side wall, or where the ureter passes beneath the uterine artery ( Fig. 19.2 ). Mechanisms of injury include crushing, ligation with a suture, transection, ischemia due to overdissection, and electrocautery injury. If not recognized intraoperatively, this can lead to renal insufficiency, chemical peritonitis due to urine leak, ileus, prolonged hospital stay, and requirement for future procedures. The average time from original operation to postoperative diagnosis of ureteral injury (if not recognized intraoperatively) is 6 to 20 days.

Risk factors for ureteral injury include previous abdominal operation, history of pelvic radiation therapy, large uterus, large tumors, infection, endometriosis, and laparoscopic techniques owing to the higher learning curve. Sun and colleagues studied 378 patients with advanced cervical cancer (stage IIB–IVA) who were treated with concurrent chemoradiation therapy (CCRT) (186 patients) or chemoradiation therapy plus completion surgical therapy 10 to 12 weeks after CCRT. Of these patients, 51.6% underwent extrafascial hysterectomy and 42.2% underwent radical hysterectomy. The overall rate of urinary or bowel complications was 19.8%, with 5.3% experiencing ureteral strictures. No urologic injuries were recognized intraoperatively in this study. The researchers found a higher rate of postoperative complications with radical hysterectomy than with extrafascial hysterectomy and therefore recommended radical hysterectomy only in patients with macroscopic residual tumor after CCRT.

Several other studies have found an increased rate of urologic injury during consolidation surgery, with the rate of ureteral injury reported to be as high as 11%. Laparoscopic hysterectomy has also been associated with an increase in the risk of ureteral injury, especially during the initial learning curve. One study predicted that at least 50 laparoscopic hysterectomies with lymphadenectomies are required for a surgeon to obtain adequate skills. A study of 317 laparoscopic radical hysterectomies and lymphadenectomies for invasive cervical cancer found that ureteral injury occurred at a rate of 1.1%; all of these injuries were unrecognized intraoperatively. This rate is higher than in most studies dealing with benign tumors. Because of this, the authors recommended use of the harmonic scalpel to dissect the vesicocervical ligament because it enables more precise dissection and reduces lateral thermal damage. The “click maneuver” can be useful to identify the ureter; however, this requires a hand-assisted form of surgery or an open procedure because palpation is necessary. During this maneuver, the forefinger and a right-angle instrument or another finger grasp the bladder pedicle or pillar on both sides and massage it until a click is felt, identifying the ureter.

Once suspected, the diagnosis of a ureteral injury is relatively easy, although the signs and symptoms might not be site specific. Intraoperatively, proximal hydroureteronephrosis can be a sign that the ureter has been devascularized or has been tied off or obliterated distally. Postoperatively, flank pain, abdominal pain, nausea, vomiting, fever, ileus, and leukocytosis can also be associated with a ureteral injury and possible ureteral leak. Diagnosis of these injuries intraoperatively includes flushing the ureteral stents if present to assess for possible leak or obstruction, identifying peristalsis along the ureter, and administering dyes such as indigo carmine or methylene blue to watch for ureteral efflux if cystoscopy is used. Intraoperatively, an intravenous pyelogram (IVP) examination can be performed intraoperatively by giving the patient 1 mg of intravenous contrast agent per kilogram and waiting 15 minutes before performing radiography. This test has an estimated 80% to 100% sensitivity when performed correctly. Alternatively, if access to the urethra is not possible, a controlled cystotomy can be done and ureteral catheters can be passed up the ureters to assess for obstruction. This may be a viable alternative when the patient is in the supine position, and not in lithotomy position, in which cystoscopy and ureteral catheter placement can be performed. Postoperatively, a computed tomography (CT) scan with or without contrast with delayed images (CT urogram) will enable identification of hydronephrosis to the level of obstruction; if contrast material is used, then delayed films are useful to look for a ureteral leak within the peritoneum. A retrograde pyelogram not only is a very sensitive diagnostic study for ureteral injury but also allows for treatment at the same time; however, fluoroscopy and a lithotomy position are necessary for this procedure.

Management of ureteral injuries depends on the level of injury, whether it is complete or not, and the timing of identification. Management principles require a tension-free anastomosis, preservation of viable tissue, meticulous debridement of nonviable tissue, and establishment of a low-pressure drainage system. Per guidelines set forth by the European Association of Urology (EAU) and the American Urological Association (AUA), upper ureteral injury (injury above the pelvic brim) can be managed with ureteroureterostomy, transureteroureterostomy, or a ureterocalycostomy. Management of mid-ureteral injuries (injury within the bony pelvis), when limited to 2 to 3 cm, includes ureteroureterostomy, transureteroureterostomy, or Boari flap reimplantation; however, adequate bladder capacity is necessary for this last procedure, because a flap of bladder is tubularized to produce the distal aspect of the ureter. For distal ureteral injuries (injury below the inferior border of the bony pelvis), options include ureteral reimplantation (ureterocystostomy) with psoas hitch as necessary to minimize tension on the anastomosis. Complete loss of the ureter or long ureteral defects will require ileal interposition, autotransplantation of the kidney, or transureteroureterostomy; however, this is best suited to occur in a delayed fashion, and in the interim a nephrostomy tube can be placed and the ureter can be tied off intraoperatively to prevent a leak. In extreme cases, a nephrectomy might be needed, but this should be done very selectively ( Table 19.1 ). If the injury is noticed in a delayed fashion and it is a complete transection, then a stent placed in a retrograde manner will likely be unsuccessful. A nephrostomy tube can be placed, and if the urine is not adequately diverted, then placement of a periureteral drain or immediate open repair may need to occur, even in a semidelayed fashion. Usually after 1 to 2 weeks it is very difficult to reoperate; however, if the urine cannot be diverted appropriately, this may be the only course of action. If a ureteral injury is incomplete, then placement of a ureteral stent either cystoscopically or in an antegrade fashion through nephrostomy access placed by the interventional radiology team may be performed rather than open repair, especially if noticed in a delayed fashion—that is, after more than 1 week. If it is recognized within 1 week, then a primary repair may be performed as appropriate.

Primary Ureteral Repair

When a ureteroureterostomy or end-to-end ureteral reanastomosis is performed, the ureter must be spatulated with Potts scissors on both ends, on opposing sides, after all necrotic tissue has been debrided back to bleeding tissue ( Fig. 19.3 ). If the ends cannot be brought together without tension, then further dissection and release of the ureter and possibly the kidney (by dissecting the kidney from the adrenal gland cranially and mobilizing it away from the psoas muscle posteriorly and the abdominal wall laterally) will be needed to bring the ureteral ends together. A 4-0 or 5-0 absorbable suture is typically placed in a running or interrupted fashion by using the full thickness of the ureter to perform the anastomosis while handling the ureter in an atraumatic fashion. If the ureter is to be directly handled, Gerald forceps may be used for handling of the ureter, and a stay-suture of Vicryl or polydioxanone (PDS) may be placed at the distal tip of each ureteral end to allow for traction. A double- J stent is then placed according to ureteral length along the anastomosis just before closure and left in place for 4 to 6 weeks to avoid obstruction and extravasation. All anastomoses should be covered with peritoneum or omentum if possible to improve healing. A Jackson-Pratt (JP) drain may be placed near the anastomosis, but not directly next to it, to identify any urine leak. If a urine leak is suspected postoperatively, or before JP drain removal, a JP creatinine level can be assessed (and compared with serum creatinine).