Definition and Diagnosis of Urinary Leak

The definition of what constitutes a urinary leak following partial nephrectomy varies in published literature. Two main definitions exist. The first is clinical and requires a persistent drain output varying from 2 to 7 days whose chemical analysis is consistent with urine, defined as an elevated drainage creatinine level greater than that of the simultaneous serum level. While there is no specific cutoff for how much higher the drainage creatinine must be when compared to serum, any elevation should arouse suspicion in postoperative patients; however, a level of twofold or more is most suggestive. Care must be taken when interpreting these results in patients with severe impairment of GFR and CKD 4 and 5. The second definition of a urine leak following a partial nephrectomy is radiographic and based on the appearance of perioperative fluid collection on postoperative imaging. In this case, the differential diagnosis includes a serous, lymph, or urine collection, and the diagnosis must be confirmed by aspiration and creatinine analysis.

In cases of elevated or persistent drain output, a drain fluid creatinine level can be checked based on clinical suspicion (complex partial nephrectomy, prolonged ischemic time, etc.) or radiographic findings of contrast extravasation on CT, MR, or retrograde pyelogram. The timing of a urinary leak is typically within the first 14 days post partial nephrectomy. Care should be taken to avoid a false sense of security with low drain outputs after resection of complex tumors with prolonged ischemia whereas acute tubular necrosis may prevent the kidney from making urine within the first several days. Therefore as the kidney recovers and urine production increases, the drainage may pick up as the leak evolves. If a serum creatinine enables the use of iodinated contrast, a CT urogram is the single best test to perform with particular attention to the excretory phase to evaluate the renorraphy, collecting system, and the entire ureter for sites of urinary leak.

Symptoms of Urinary Leak

In the majority of series in the published literature, most patients with a urinary leak are asymptomatic and present with increased drain output that is then confirmed to be consistent with urine. In a study by Potretzke et al. evaluating urinary leak following robotic partial nephrectomy, the authors only tabulated symptomatic undrained urinary leaks, whereas all drains were removed prior to discharge. Patients presented at a median postoperative day (POD) 13 with symptomatic urinary leak with fever (14%), gastrointestinal complaints including ileus (29%), and pain (36%).

Incidence and Differential of Urinary Leak

The incidence of urine leak from renorraphy sites ranges from 1.5% to 18.5% after open partial nephrectomy (OPN), 1.5–16.5% after laparoscopic partial nephrectomy (LPN), and 0.78–3.7% after robotic partial nephrectomy (RAPN). The incidences vary with respect to the definition of urinary leak used (clinical, biochemical, and/or radiographic). Additionally, the literature is replete with inaccurate information regarding the timing of postoperative urinary leaks vis-à-vis drain removal. In one RAPN series (n = 1791) where all patients’ Jackson-Pratt (JP) drains were removed prior to hospital discharge, the authors report an absolute-symptomatic leak rate of 0.78%. In a combined OPN/LPN series consisting of 1118 cases, an overall leak rate of 4.4% was noted with delayed leak after drains were removed occurring in 0.4%. It is certainly likely that not all urine leaks following partial nephrectomy are clinically symptomatic, and the true incidence is therefore unquantifiable. According to the American Urological Association (AUA) guidelines on the management of the stage 1 renal masses, the incidence of a urinary leak following partial nephrectomy is 4.1–4.4% while that of symptomatic leaks requiring drainage is closer to 0.78%.

When a urinary leak is diagnosed following partial nephrectomy, a clinical incidence of suspicion must remain for the possibility of a ureteral injury or necrosis as its cause. Ureteral injuries are reported to occur in 0.3% after LPN and 0.6% after RAPN. Idiopathic ureteral injuries are usually identified intraoperatively and do not result in a urinary leak if repaired and properly drained. However, unidentified ureteral injuries can occur and need to be considered if a persistent urinary leak is encountered. Additionally, ureteral obstruction must always be considered and excluded, particularly for high-output urinary leaks following PN. While new-onset ureteropelvic junction obstruction (UPJO) is described in 1% of patients following PN, most cases of ureteral injury or obstruction occur as a result of iatrogenic crush, thermal or transection injury, and/or ischemia due to over-dissection. Downstream obstruction from the site of leakage can increase the persistence and amount of a urinary leak from the renorrhaphy site ( Fig. 13.1 ).

Urinary leak after partial nephrectomy in a solitary kidney of a patient with a highly complex tumor (nephrometry score 2+3+3+P+3 = 11P) with proximal ureteral stricture. Ureteral stent was placed on postoperative day 4, leading to resolution of urinary leak in 35 days.

Risk Factors of Urinary Leak

Most studies evaluating risk factors of urinary leak compare those with and without a leak without adjusting for other clinicopathologic characteristics of a patient. These studies indicate that central/endophytic tumor location, large tumor size, collecting system entry, RENAL nephrometry complexity, downstream obstruction (UPJO, ureteral stricture, infundibular stricture, BPH with high-pressure voiding), intrarenal pelvis >50%, increased estimated blood loss, preoperative CKD stage ≥III status, increased warm ischemia time, increased operative time, surgeon’s learning curve, and surgical approach (increased with OPN vs RPN) are all risks and more highly associated with urinary leak. Other studies indicate that surgical approach, tumor multifocality, operative time, or EBL are not predictive of urinary leaks.

Two studies evaluated the risk factors for urinary leak following PN using multivariable models. One study suggested that a surgeon’s early operative experience (odds ratio [OR] 7.8), tumor proximity to collecting system (OR 9.2), and preoperative CKD stage ≥ III (OR 3.1) are indicative of urinary leak after adjusting for other factors including surgical technique, operative time, warm ischemia time, race, BMI, individual RENAL score attributes, and use of intraoperative ureteral catheterization. Tomaszewski and Uzzo et al. demonstrated that an intrarenal pelvis (OR 24.8), endophytic tumor (OR 4.5), and collecting system entry (6.1) are most predictive after adjusting for age, gender, race, BMI, Charlson comorbidity index, surgical approach, ECOG score, individual RENAL attributes, ischemia time, and operative time. To measure intrarenal pelvis anatomy these authors developed the renal pelvic score (RPS) as measured on CT or MRI by drawing a line connecting the two polar lines (upper and lower) and quantifying the proportion of renal pelvis volume within the kidney parenchyma. An RPS >50% is considered intrarenal and at a higher risk of urinary leak with prolonged time to closure in cases of complex PN.

Prevention of Urinary Leak

The knowledge of risk factors for urinary leak is paramount in efforts to avoid this complication. Techniques used include the placement of a preoperative ureteral catheter for localization, injection of methylene blue to identify collecting system entries, and “priming” with intravenous indigo carmine prior to arterial clamping. Early series of partial nephrectomy routinely place ureteral catheters and inject saline or methylene blue to help identify intraoperative urinary leakage during excision and reconstruction. However, recent studies question the utility of this since no difference in postoperative urinary leak was seen in those with and without the ureteral stent despite a higher incidence of collecting system entry detection intraoperatively.

In our practice, we are highly selective regarding in whom we place preoperative ureteral stents based primarily on complex nephrometry score, prior surgery, and intraabdominal or retroperitoneal anatomy. In these cases we place a 5Fr ureteral stent using a flexible cystoscopy prior to patient positioning and leave it for 24–48 hours as a preventive measure to decrease renal pelvis pressure in select patients with intrarenal pelvis, endophytic tumors, or tumors abutting the collecting system. Occasionally we will use it to instill retrograde saline with 1 mL of methylene blue to identify the collecting system entry. As these are the patients with highest odds of developing a urinary leak, we feel that this may contribute to identifying and preventing a urinary leak. This is, however, difficult to test in a prospective study. Additionally robotic or loupe magnification during excision and renorrhaphy is considered very helpful. The complexity of a collecting system entry should be assessed prior to and during excision, with care taken to avoid complete infundibular transection or extensive unroofing, which would make reconstruction complicated. Moreover, suture occlusion during fastidious repair should be carefully avoided.

Treatment of Urinary Leak (see Fig. 13.2 )

More than 80% of urinary leaks following PN are diagnosed in otherwise asymptomatic patients with prolonged output per JP drain, while the remainder is diagnosed after drains have been removed. Given that the majority of patients presents asymptomatically, we routinely place JP drains in those who have significant risk factors for urinary leak (large endophytic lesion, collecting system entry, or intrarenal pelvis). Avoidance of JP drains in many of these patients without risk factors has been shown to be safe without an increase in operative interventions for urinary leak drainage.

Treatment algorithm of urinary leak after partial nephrectomy.

Fortunately, the majority of patients who develop a urinary leak can be managed with a combination of prolonged JP drain and selective use of Foley catheterization without any other operative intervention. If, however, prolonged JP drain does not decrease the outputs, a ureteral stent should be considered. Ureteral stents have been employed in 20–60% of cases of prolonged urinary leak with successful resolution of the leak in the vast majority of cases. The stents are usually placed at a median time of 2–3 weeks after partial nephrectomy but could be placed sooner in the case of excessively high drain outputs. A gentle retrograde pyelogram should be done at the time of stent placement to assess patency of the ureter and good positioning of the stent. A leak does not need to be demonstrated on retrograde if there is clinical evidence of a leak. The JP drains are kept until output significantly declines. The stents may be kept in place for 6–12 weeks, and a retrograde pyelogram may be done to evaluate for leak before they are removed; however, our practice is to omit retrograde unless clinically indicated. The combination of a ureteral stent with intraabdominal JP drain may result in over 90% resolution of urinary leak. The median duration of a urinary leak is between 20 and 64 days. It is expected that unobstructed patients will heal within a reasonable time period of 1 to 4 weeks. Some experts advocate that the drain be placed to straight drainage as opposed to bulb suction as this may prolong urinary leakage.

The use of Foley catheters has been advocated in cases of prolonged or high-output drainage, but we reserve this primarily for men with high-pressure voiding. Alpha-blockers are often used to minimize bladder pressures, which may induce reflux up the stent. Care should be taken to maintain good hygiene of the drain site. Drainage may reduce gradually or occasionally quite precipitously. If this occurs, an ultrasound is often helpful to assess for retroperitoneal fluid reaccumulation or hydronephrosis. Drains are usually removed before the stent, which should be maintained for several additional weeks after the leak has clinically resolved. In cases of persistent high-output leakage despite maximal drainage, ureteral or stent obstruction, unrecognized additional injuries, or necrosis should be considered. Although not scientifically evaluated, desmopressin administration has been suggested to reduce the duration of urinary leak.

Whereas the majority of urinary leaks resolve on their own following PN, patience is essential, and individuals should be counseled that their stent and drain may occasionally need to stay in place for weeks or, rarely, months. In these cases some have successfully accelerated resolution with dual ureteral stents or a fashioned 14Fr to 16Fr ureteral catheter. An additional urinoma drain or a nephrostomy tube may rarely be necessary and has been utilized along with stents on occasion.

The astute clinician should be very hesitant to ever reoperate on a urinary leak following PN as long as obstruction is ruled out and drainage is maximized. Very rare instances of persistent leakage may require attempted renal or ureteral reconstructive or completion nephrectomy. However, prior to undergoing nephrectomy, a careful pyelo­gram is needed to assess and to rule out ureteral strictures, infundibular stenosis, or ureteropelvic junction obstruction, all of which may prolong the urinary leak. If infundibular stenosis or strictures are identified, endoscopic dilations may be attempted and have been shown to be successful in case reports. Additionally, there have been case reports of fibrin sealants that can be successfully injected in the offending calyx either percutaneously or by retrograde ureteroscopy to resolve any further urinary leakage. However, any salvage endoscopic measures may require multiple operative procedures, and patients need to be counseled appropriately. Reconstruction may be preferable to nephrectomy depending on clinical circumstances.

Definition and Diagnosis of Urinary Leak

There is no consensus on a definition regarding urinary leak following radical prostatectomy. Similar to urinary leak after partial nephrectomy, two main definitions exist, one based on the persistence of prolonged drain output with chemical analysis consistent with urine and the second based on radiographic findings on cystogram.

In the majority of historical prostatectomy series, cystogram was routinely performed before catheter removal. It is diagnostic for a urinary leak with the identification of contrast extravasation. Several groups have attempted to classify and quantify cystogram findings. Fenig et al. classified these into no extravasation, mild leak (wisp of extravasation), moderate leak (extravasation 25% or less of instilled contrast medium), and major leak (>25% extravasation). The purpose of the classification was to determine the need for a repeat cystogram. Patients with mild leak were recommended to undergo removal of Foley catheter without a repeat study. Patil et al. classified these into grade I (extraperitoneal within 6 cm of the urethrovesical anastomosis), grade II (extraperitoneal extending to the side-wall) >6 cm, and grade III (intraperitoneal). Here, the purpose was to report on clinically significant versus nonsignificant leakage where patients with higher grades tended to require intervention. Both of these classification schemes illustrate the importance of a uniform definition of a urinary leakage.

Several other modalities including CT cystogram have been described to evaluate the urethrovesical anastomosis following radical prostatectomy. Although TRUS has been described, we do not recommend its use in the perioperative setting due to patient discomfort and less precise image resolution compared to CT. CT cystogram has been shown to be more sensitive and accurate than cystogram. Indications for CT cystogram as opposed to standard cystogram remain unclear. We recommend consideration of CT cystogram if the patient has suffered a prolonged or complicated course of urinary leakage with secondary complications such as bleeding, uroperitoneum, sepsis, or abscess. More extensive imaging may identify drainable collections or additional concurrent pathology. In these cases it is more critical that a leak is not present when catheter is removed.

Recently, many have questioned the overutilization of cystogram and attempted to stratify patients in whom it may be omitted. A study on open prostatectomy suggested that cystogram could be omitted in the absence of red or pink urine on POD 8 with a negative predictive value of 98.6%. Another study suggested that an intraoperative test for a watertight anastomosis is predictive of no leakage. Ischia et al. recommended that over 97% of patients might avoid cystogram on POD 7 if a negative intraoperative leak test was confirmed. With the advent of robotics and improved visualization and degrees of freedom during suturing, many have questioned the need in robotic procedures and perform a cystogram only in patients with a high clinical suspicion or risk of a leak such as in salvage prostatectomies following radiotherapy. Guru et al. prospectively analyzed the use of cystogram after robotic prostatectomy and illustrated that it can be safely avoided.

Symptoms of Urinary Leak

In the majority of published series, patients with urine leaks following prostatectomies are asymptomatic and present with increased drain output that is then confirmed to be consistent with urine or on a routine postoperative cystography. In Guillonneau et al.’s study, while 8.1% were diagnosed as having immediate postoperative urinary leak (persistent drain output for more than 6 days), 1.9% presented with secondary anastomotic leakage days after catheter removal. These patients were diagnosed with acute pain, acute urinary retention, and peritoneal irritation symptoms. The study also illustrated that acute retention develops as a function of duration of catheterization where the incidence may be 5–25% at POD 2–3 and around 1% by POD 6. This may be due to resolution of anastomotic swelling. Removal of the catheter earlier in the postoperative course is, indeed, a risk factor for urinary retention. While not all patients with retention develop a secondary urinary leak, retention appears to be a risk factor for anastomotic breakdown and leakage, particularly if the catheter is replaced traumatically.

An important subacute manifestation of urinary leak from the urethrovesical anastomosis includes osteomyelitis of pubic bone in the setting of a urinary tract infection. These patients can present with recurrent urinary tract infections, significant pubic and pelvic pain, and disturbance in gait. Often these patients are not diagnosed in time and treated with repeat courses of antibiotics, which fails to improve their symptoms. These can present as early as 1 month postoperatively or can present over the next several months to several years. Pelvic MRI remains the most sensitive and specific test to diagnose osteomyelitis of the pelvis. When osteomyelitis has progressed, pelvic CT may reveal extravasation of intravesicle contrast from the fistulous tract into the bone causing fractures of pubic rami.

While urinary leak may delay the return of continence following prostatectomy, it does not seem to worsen long-term continence rates at 1 year. Several studies have suggested that urinary leak may predispose to the development of bladder-neck contractures (BNC), while other studies indicate no significant relationship. Postoperative urinary leakage when the anastomotic apposition is good seems unlikely to lead to BNC. However, prolonged and more extensive urinary leakage, resulting from an anastomotic gap, may cause scarring due to healing by secondary intention.

Incidence and Differential of Urinary Leak

The incidence of urine leak from urethrovesical anastomoses ranges from 0.5% to 29% after open radical prostatectomy (ORP), 0.9%–22.3% after laparoscopic radical prostatectomy (LRP), and 1.1%–12.7% after robotic-assisted laparoscopic prostatectomy (RALP). The incidence varies with respect to the definition used with the radiographic definition, yielding higher incidences with a diagnosis of subclinical leak. The incidence of leak is also time-dependent. In one ORP series, the leak rate at POD 8 was as high as 78% and decreased to 38% and 20% on POD number 11 and 14, respectively. Similar data are available for LRP and RALP. However, due to improved continuous suturing with RALP, it is thought that RALP yields lower leak rates overall given improved tissue coaptation.

While there are several series with a high incidence of urinary leak overall, clinically significant anastomotic leaks are infrequent. Clinically significant anastomotic leaks can be defined as those that extended intraperitoneally, those that necessitated additional drainage, those that necessitated readmission due to ileus or dehydration, and those associated with fever and sepsis. The incidence of clinically significant anastomotic leaks ranges from 0.36% to 2.3% in multiple series.

Other sources of urinary leak following prostatectomy include bladder or ureteral injury not identified intraoperatively. The incidence of ureteral injury has been described to be 0.1% to 1.4% and that of bladder injury 0.2% to 1.6%. If bladder or ureteral injuries are suspected, a CT urogram or a cystogram is recommended.

Risk Factors of Urinary Leak

Risk factors for urethrovesical anastomotic leakage include previous TURP, difficult intraoperative anastomosis, increased EBL, pelvic hematoma, positive intraoperative flush test, postoperative UTI, ischemic heart disease, obesity (due to difficulty accessing the urethra during anastomosis), early learning curve, and salvage prostatectomy after radiation therapy. Additionally bladder neck reconstruction during RALP has also been shown to increase urinary leakage rates. Neither suture type, number of sutures, nor interrupted versus running anastomotic sutures has been shown to be associated with leakage. Barbed V-lock suture, rather than monofilament, has been shown to reduce the incidence of positive intraoperative leakage test, and the need for additional tightening of sutures. However, it has not been directly related to postoperative urinary leakage. There has been no difference in leakage rate with open versus minimally invasive approach or by extraperitoneal versus intraperitoneal surgical approach.

Prevention of Urinary Leak

Several preventative measures to minimize urinary leak during prostatectomy have been described in the literature. The use of a side-fenestrated catheter over a conventional Foley catheter has been advocated in one study. In side-fenestrated catheters, the fenestration is located in the catheter below the balloon at the site of the anastomosis. The aim is to improve drainage emanating from the healing anastomosis. In a prospective randomized study of 250 patients undergoing RALP, patients randomized to a fenestrated catheter had a significantly decreased incidence of urinary leak (4.6% vs 12.3%). Another preventive measure described is using fibrin sealant over the anastomosis to prevent leakage across the anastomosis. The maneuver may decrease overall intraabdominal drain output; however, the evidence is not strong for reducing the incidence of urinary leak.

Another technical modality to prevent a leak is to relieve tension exerted on the urethrovesical anastomosis. Patel et al. investigated the use of the posterior reconstruction of rhabdosphincter with the secondary outcome being leak rate. The free edge of the Denonvillier’s fascia is sutured to the posterior rhabdosphincter underneath the urethral plate. The intent is to minimize the downward mobilization of the urethra and reduce tension across the anastomosis. With this approach, the incidence of urinary leak was shown to reduce significantly (0.4% vs 2.1% in those without). Tewari et al. evaluated three different techniques of anastomosis – conventional anastomosis, anterior reconstruction, total anatomic restoration (anterior plus posterior reconstruction). The aim of the latter two techniques was to relieve the tension on the anastomosis. Clinically significant urinary leak rates were lowest with total anatomic restoration (0.3%) followed by anterior reconstruction (1%) and highest with conventional anastomosis (2.3%). Ultimately, the best prevention of urinary leak following radical prostatectomy is the achievement of a tension-free, watertight urethral anastomosis.

Treatment of Urinary Leak (see Fig. 13.3 )

Depending on the risk factors for urinary leak, an intraabdominal drain is considered at the time of prostatectomy. There are two divergent maxims associated with the pelvic drain: “when in doubt, drain” and “drains have only one role, they suck.” In between these two extremes exists the informed approach. The goal of the drain is to avoid the sequela of uroperitoneum with its associated risks of ileus and electrolyte abnormalities and to minimize the need for an interventional procedure in the postoperative period.

Treatment algorithm of urinary leak after radical prostatectomy.

Many utilize the intraoperative leak test as the main indicator as to whether or not to place a drain. Other factors taken under consideration include the technical difficulty with anastomosis (such as suture breakage, increased robotic arm collisions due to suture angles, tearing of urethra), prior TURP, or radiotherapy. In a prospective study on 846 patients undergoing prostatectomy, patients did not undergo drain placement if they had a negative leak test, adequate hemostasis, and no surrounding organ injury. This study illustrated no differences between those with and without drain placement with respect to leak rates.

When leakage is confirmed in the postoperative period either by drain output or on radiographic imaging, the Foley catheter should be examined to ensure that it is draining and that it is not kinked or clogged. A gentle flush may be needed to clear away the debris. If high-output leakage is seen, the drain can be taken off suction and allowed to drain passively. The drain may be withdrawn a few centimeters as well to avoid suctioning against the anastomosis. Fortunately, the majority of urinary leak can be managed with prolonged Foley and pelvic drain without the requirement for additional maneuvers. Less than 5% of patients may require additional percutaneous drainage of the uroperitoneum.

If the leakage remains high despite the above maneuvers, the Foley catheter can be placed on gentle traction against the anastomosis. The drain output subsequently may be dependent on the closeness between the ureteral orifices and urethrovesical anastomosis. If the ureteral orifices are situated far away from the urethrovesical line and there is profuse leakage, gentle traction or apposition of the catheter against the bladder neck can reduce leakage and external fixation may help. If the ureteral orifices are close to the anastomosis, the Foley balloon can oppose the ureteral orifices or lie above the ureteral orifices and impair urinary drainage, thereby worsening leakage. Traction against this will only increase the urinary leak. As such, in this case, gentle radiologic guided mobilization of the catheter with deflation of the balloon and proper external fixation of the catheter can be done to avoid dislocation and allow improved drainage via the indwelling catheter. Alternatively, 14-16Fr Cope loop nephrostomy tube can be placed as a urethral catheter to allow improved urinary drainage through several drainage holes.

In the presence of high-output leakage where the majority of urine is coming from the perianastomosis drain rather than per the catheter, CT urogram should be performed to rule out an iatrogenic ureteral injury and to ensure that the catheter balloon remains in the bladder. In cases of ureteral injury, a decision to do ureteral reimplantation or nephrostomy with delayed ureteral reimplantation is left up to the discretion of the surgeon, and no complication differences are seen with early versus delayed reimplantation, although it is our practice to repair early. If no ureteral injury is thus identified and urinary leakage exceeds >1.5 L, revision of the anastomosis can be performed even up to POD 6–12. This option should be reserved for rare circumstances.

Rather than a repeat revision of the anastomosis, several other options have been described as potentially effective. One option is to continue to observe the patient ensuring that all leaked urine is diverted effectively. This may be accomplished by leaving the JP drain or placing an additional drain under radiographic guidance. Another approach described is a needle-vented suction of the Foley catheter. The Foley catheter is connected to low wall suction at 15 to 20 mm Hg. To avoid continuous suction against the bladder mucosa, a sump is created by placing an 18-gauge angiocatheter needle in the tubing connecting the Foley to the wall suction. This technique allows a gentle vacuum to allow urine to preferentially drain through the catheter rather than leaking through the anastomosis. Another approach described is to divert all of the urine away by placing single J internal–external ureteral stents bilaterally. This allows all urine to become diverted and will minimize all urinary leakage. A final approach described is to place either bilateral nephrostomy tubes to divert all urine or a unilateral nephroureteral stent, which can be placed on intermittent wall suction.

Ultimately, the drain output is monitored daily. When the output decreases significantly, another fluid creatinine can be done to rule out any residual leak. The drain is subsequently removed and after a repeat cystogram has ensured healing, the Foley catheter can safely be removed.

Definition and Diagnosis of Urinary Leak

Compared to leakage after radical prostatectomy or partial nephrectomy, urine leak after urinary diversions has not been studied as extensively in the literature as a primary end point. Richard et al. defines a urine leak as elevated drain fluid creatinine level beyond POD 7, while other studies define it as early as POD 1. A small amount of urinary extravasation in the early postoperative period seems to be acceptable and may resolve by POD 7 with adequate drainage.

The study of choice to confirm laterality or location of leakage is a combination of CT urogram and loopogram/pouchogram. However, while radiographic studies are more sensitive for a identifying a leak, they are generally done to localize the presence of a clinical leak. In a study by Ankem et al. on patients with continent urinary diversion, five of six leaks diagnosed with a pouchogram were clinically detectable. In a study by Brown et al., of 15 urinary leaks, three were localized by CT urogram, 10 by CT conduitogram, and the remaining two on nephrostogram done at the time of percutaneous nephrostomy tube insertion. Many patients with urine leak can also have an infected urinoma, and it is imperative to identify this early on CT imaging.

Symptoms of Urinary Leak

Two studies evaluated urinary leak as a primary end point and tabulated presenting symptoms. In 21 patients, Hensle et al. showed ileus to be a presenting symptom of a leak in four (19%), urinary drainage from wounds in seven (33.3%), elevated BUN without increased creatinine in 20 (95.2%), and sepsis in 10 (47.6%). In a study of 15 patients, Brown et al. indicated sepsis as a presenting symptom in eight (53.3%), increased drain output in six (40%), leakage from wound in five (33.3%), and decreased urine output in four (26.6%), with one asymptomatic patient diagnosed on a routine “stentogram” prior to removal of stent. In other series, urinary leak is also associated with prolonged ileus and prolonged length of hospitalization.

There appears to be a bimodal temporal presentation of urinary leak. In Hensle et al.’s study, 8 of 21 (38%) presented within the first 24 hours and 13 of 21 (62%) presented after POD 6. Additionally, those who presented after POD 6 had worse outcomes. Brown et al. noted a median time to leak of 12 days, with 2 of 15 (13%) presenting within the first 6 days and 13 of 15 (87%) presenting after POD 6. Patients who presented within the first few days postoperatively may represent technical errors. In the Brown et al.’s study, for the two patients who presented with early leak the sources were “technical construction of ureteroenteric anastomosis” in one and a stoma catheter perforating staple line at the blind end of the conduit in another. On the other hand, those who presented after POD 6 may represent a manifestation of ischemic necrosis of the anastomosis. As such urinary leak is also thought to be a risk factor for development of ischemic ureteroenteric stricture.

Urinary leak has an adverse effect on mortality, which may be due to multiorgan failure from concomitant sepsis or may also be due to rapid cancer progression theorized to be a result of proinflammatory cytokines and suppressed antitumor inflammation response. Sepsis is one of the more common causes of mortality following radical cystectomy and urine leak has been diagnosed up to 40% of septic patients who died within 30 days. In the Hensle et al.’s study, alarmingly, 48% of patients with a urinary leak died perioperatively. Additionally, mortality risks are also associated with elevated serum BUN, an indirect marker of urinary leak, where rates are 5.8% if BUN <30 and increase to 33.3% with BUN >100.

Incidence and Differential of Urinary Leak

The incidence of urine leak after ileal conduit ranges from 1.87% to 16%. If limited to series with >500 patients, the incidence ranges from 1.87% to 2.57%. In more than 60% of cases, leakage occurred from the ureteroenteric anastomosis, but it can also occur anywhere along the loop including the butt-end of the conduit.

The incidence of urine leak after continent reservoir ranges from 2% to 12.8%. Leakage can occur from the ureteroenteric anastomosis or from the extensive suture lines.

The incidence of urine leak after neobladder ranges from 2.1% to 11.6%. If limited to series with >300 patients, the incidence ranges from 2.1% to 7.7%. Leakage can occur from either the ureteroenteric anastomosis, the neovesico-urethral anastomosis, or anywhere along the extensive suture lines of the neobladder. In a series by Hautmann et al., the incidence of leakage per ileourethral anastomosis was 6.6% and from the ileoureteral anastomosis was 1.1%.

Risk Factors of Urinary Leak

Evidence on the risk factors of urinary leak after urinary diversion is limited compared to the evidence of risk factors predisposing leakage after radical prostatectomy or partial nephrectomy. Studies describing risk factors are descriptive or achieve marginal statistical significance. No current multivariate model for urinary leakage exists following urinary diversion. Perhaps this is because urinary leak represents only a fraction of postoperative complications after cystectomy while it represents major complications after radical prostatectomy or partial nephrectomy. However, urinary leak after a diversion represents a greater concern as most are symptomatic and leakage can be a poor predictor of outcomes.

Risk factors for urinary leak after urinary diversion include palliative exenteration, BMI >30, prior pelvic radiation, and cardiovascular medical history. These factors may influence the vascular supply to the ureter or the ability to perform a high-quality diversion. Increasing medical comorbidities may also be associated with postoperative genitourinary complications, although no specific effect of comorbidities on urinary leakage is described.

Urinary leakage is more common on the left ureter than on the right, perhaps due to devascularization of left ureter with extensive mobilization or because the left anastomosis is more predisposed to be under tension. There do not appear to be any differences in urinary leak with interrupted or running suture anastomosis of the ureter. Stomal edema/obstruction may increase the intraconduit pressure and increase the risk of ureteroileal anastomosis disruption. Additionally, placement of a catheter postoperatively is the most common cause of iatrogenic injury to the stump of the conduit. Extraluminal hematoma/seroma may also increase the risk of urinary leak by disrupting the blood supply. Poor surgical technique (case reports of spermatic vein anastomosis to conduit) and early learning curve can also predispose to a urinary leak. Recurrence of malignancy can be a risk factor for late presentation of urinary leaks.