Predictive Factors for Percutaneous Nephrolithotomy Outcomes in Neurogenic Bladder Population

Introduction

P atients with neurogenic voiding dysfunction with or without urinary diversion are at an increased risk of urolithiasis, recurrent stone disease, and surgical morbidity. ^{1 –6} Risk factors for urolithiasis include recurrent urinary tract infections, urinary stasis, and hypercalciuria associated with prolonged immobility. Urinary collecting system access is also difficult in this population because of altered body habitus, such as severe kyphoscoliosis and contractures, as well as urinary diversion. We present a single center retrospective review that to our knowledge is the largest series to date.

Patients and Methods

After Institutional Review Board approval, we reviewed our percutaneous nephrolithotomy (PCNL) experience from January 1, 2002 to December 31, 2009. Forty-seven patients were identified with neurogenic bladder who underwent PCNL. The medical records were reviewed retrospectively for data relating to their nephrolithiasis and outcomes. Age, sex, race, laterality, stone composition, stone size, level of lesion, method of bladder management (including urinary diversion), preoperative urine culture results, preoperative antibiotics, and access were recorded. In addition, length of stay (LOS), intensive care unit (ICU) stay, complications, initial and final stone-free rate, and follow-up were collected.

One surgeon performed PCNLs at our institution until 2009, when an additional surgeon was added. All patients received preoperative upper tract imaging with either abdominal plain radiography, intravenous urography, or noncontrast CT. A urine culture was obtained at the preoperative visit, and appropriate culture-specific antibiotics were instituted 1 week before the procedure. If the culture showed organisms without sensitivities to oral antibiotics, a peripherally inserted central catheter was placed, and intravenous antibiotics were administered. Our practice has been to admit patients with neurogenic bladder patients and administer broad-spectrum intravenous antibiotics or antifungal agents when appropriate.

Before placement of the percutaneous tube, with the patient under sedation, a 7F ureteral catheter was advanced to the renal pelvis, and retrograde pyelography was performed to visualize the collecting system. Patients with urinary diversion that did not allow for retrograde access to the collecting system did not have a ureteral catheter placed. A 16F coudé catheter was placed and secured to the ureteral catheter. The desired tract or tracts were marked for the radiologist. If the desired tract was unable to be accessed, the interventional radiologist chose another option after consulting with the urologist. The choice of tract was determined by stone location, stone burden, and presence of spinal deformity.

Until 2009, all patients had access provided by the interventional radiologist. In total, 92% (60/66) of procedures had access placed by the radiologist (5 urology, 1 unknown). Percutaneous puncture of the tract was performed by the radiologist in the standard fashion. For the five procedures in which access was performed by the urologist, the patient was placed under endotracheal anesthesia in the supine position. A 7F ureteral catheter was then advanced to the renal pelvis and retrograde pyelography performed. The patient was then placed in the prone position, and access was placed in the standard fashion.

For dilation and stone removal, the patient was placed under general endotracheal intubation in the supine position. After successful intubation, the patient was placed in the prone position. For patients who were unable to be placed in the prone position because of severe kyphoscoliosis or contractures, a semiprone position was used. No patient underwent PCNL in the supine position. After standard skin preparation and draping, a super stiff guidewire (0.038 inch, polytetrafluoroethylene-coated [Amplatz Extra Stiff Wire, Cook Urological, Spencer, IN]) was positioned through the catheter into the ureter. A safety guidewire was placed, and the tract was dilated to a 30F diameter using a dilating balloon and sheath (NephroMax, Boston Scientific Microvasive, Natick, MA).

A 26F rigid nephroscope was used with an ultrasonic lithotripter for fracturing the stone. Large fragments were removed with graspers. All calices were inspected for retained stone fragments with a 26F rigid nephroscope and a 16F flexible cystoscope. In addition, the 16F flexible cystoscope and flexible ureteroscope were used to evaluate the ureter for fragments. Stone fragments found with flexible endoscopy were treated with a holmium laser or removed with flexible graspers/baskets.

Postprocedure drainage was maintained with a 22F Councill-tip catheter (Bard, Covington, GA) placed over a guidewire. Antegrade nephrostography was performed to visualize all calices and check for any extravasation before procedure conclusion. Many patients with spinal cord injury (SCI) and spina bifica (SB) have contractures or severe kyphoscoliosis making interpretation of postoperative chest fluoroscopy difficult. In addition, kyphoscoliosis can alter chest anatomy, resulting in increased risk of pleural injury during access and dilation of the collecting system. Therefore, all patients underwent postoperative chest radiography in the recovery room.

The patient was maintained in the hospital for at least 2 days after the procedure. Our postoperative protocol involves plain abdominal radiography on postoperative day 1. If no remaining stone burden is present, the wire is removed and the nephrostomy tube is taken off gravity drainage and clamped. If the patient tolerates clamping (ie, no nausea/vomiting, intractable pain, fever >101.5°F), the tube is removed the next day, and the patient is discharged. The patient is maintained on culture-specific antimicrobials while admitted and for 3 days after discharge. Daily serum chemistry results are checked. We routinely do not use second-look nephroscopy. Residual fragments, however, were treated with a single second procedure or a combination of second-look PCNL, extracorporeal shockwave lithotripsy, and ureteroscopy. Patients were instructed to follow up in 1 month with another plain abdominal radiography and renal ultrasonography or in certain cases, a noncontrast CT scan to evaluate for residual stone disease.

The relationships between dichotomous or categorical predictors and outcomes were studied using contingency tables and the chi-square statistic. Relationships between the dichotomous and categorical predictors and numeric outcomes were studied using the Wilcoxon rank sum test. For numeric predictors and outcomes, the Spearman correlation coefficient was used. For numeric predictors and dichotomous outcomes, we used logistic regression. P values<0.05 were considered statistically significant, while those between 0.05 and 0.10 were noted as being suggestive of a relationship. All analyses were performed using SAS (v9.2). All complications were graded based on the Clavien-Dindo system. ⁷

Results

A total of 66 PCNLs were performed on 47 patients between January 1, 2002 and December 31, 2009. There were 26 patients with SCI, 16 patients with SB, and 5 heterogenous causes for neurogenic bladder (2 with multiple sclerosis, 1 amytrophic lateral sclerosis, 1 sacral agenesis, 1 triad syndrome). The patient with triad syndrome performed clean intermittent catheterization (CIC). Mean age was 42.9 years (11–85 y). The method of bladder management was variable with the most common being diversion at 27.6%. Ten of the 13 (77%) patients with a diversion had an ileal conduit. Four patients with bladder augmentations performed CIC. Stone composition was mainly infectious in nature. Preoperative culture data were obtained for 59 procedures, and 55 patients showed colonization/infection. The most common organism was two or more (n=17). Six of the patients had a PCNL performed on a solitary kidney. Additional data on patient demographics can be seen in Table 1.

Access was performed by the interventional radiologist in 92% of cases (n=60). A urologist performed access in five cases, and one patient had access placed at an outside institution. For 11 procedures, the percutaneous nephrostomy tube was placed emergently for urinary sepsis with a mean time to stone removal of 29 days. Seven procedures needed multiple access tracts. Information on access can be seen in Table 2. The mean LOS was 5.3 days. Nine patients needed ICU care, and their mean LOS extended to 13.9 days. Three SCI patients needed ICU care—two with postoperative sepsis and one with prolonged intubation. Six patients with SB needed ICU care—secondary to sepsis in five. The remaining patient with SB was transferred to the pediatric ICU for close monitoring.

The initial stone-free rate, defined as no evidence of stone on imaging after the first procedure, was 60.6%. In 40 of the 66 procedures, plain abdominal radiography was used to determine the stone-free rate. Noncontrast CT scan was used in 13 procedures to determine the initial stone-free rate. The remainder of patients underwent either limited intravenous pyelography or antegrade nephrostography. The final stone-free rate, defined as no evidence of stone on imaging after multiple procedures for the same stone episode, was 69.7%. Many patients chose to follow up and receive adjunctive procedures with a referring urologist.

Most complications were infectious or pulmonary. Complications are listed by the Clavien-Dindo classification in Table 3. Sepsis developed in eight patients (four SCI, four SB) postoperatively, and only one of those patients had a positive blood culture. Pulmonary complications included prolonged intubation, acute respiratory distress syndrome, pneumonia, and hemothorax necessitating video-assisted thorascopic surgery (VATS) and chest tube decompression. Acute renal failure developed in two patients; it resolved before discharge. Both of these patients had SB and a large bilateral stone burden. One of the patients in whom acute renal failure developed after right PCNL had a functional solitary renal unit with a Lasix renal scan revealing 24% function on the left. Miscellaneous complications were one seizure, one pressure wound development, and one death secondary to overwhelming sepsis. One patient underwent bilateral below-knee amputations as well as amputations of upper extremity digits secondary to extensive necrosis from high-dose vasopressors.

Five patients needed two procedures and five needed three procedures for the same stone episode. There were seven second-look PCNLs, four shockwave lithotripsies, and one ureteroscopy. Three patients eventually needed a nephrectomy during the time frame of the study. All nephrectomies were performed secondary to nonfunctioning kidneys after obtaining a Lasix renal scan.

Several factors predicted outcomes in this population. Multiple access was associated with adverse outcomes, including increased LOS (P=0.01), ICU stay (P<0.01), need for transfusion (P<0.01), and pulmonary complications (P=0.03). Upper-pole access was associated with decreased initial stone-free rate (P=0.04). There was a trend, however, toward increased initial stone-free rate with midpolar access (P=0.07), but this did not reach statistical significance. Midpolar access predicted increased final stone-free rate (P=0.04). Mean stone size was 3.31 cm and was predictive of an increased number of procedures (P=0.04). Larger stone size was also predictive of decreased initial stone-free rate (P=0.03) and final stone-free rate (P=0.05). Bladder management, including urinary diversion, was not predictive of adverse outcomes. In addition, stone composition had no effect on complications. There were no statistically significant differences between SCI and SB patients when evaluating LOS, ICU stay, and initial/final stone-free rate.

Discussion

The patient cohort examined in this series is at particular risk for nephrolithiasis. In addition, there are little data published in regard to outcomes of PCNL in this population. Known risk factors include urinary stasis, urinary tract infection or colonization, hypercalcemia secondary to immobilization, ^2,3 frequent instrumentation, and elevated bladder pressures. ⁸

Our initial stone-free rate was 60.6%, and final stone-free rate was 69.7%. Culkin and colleagues ⁹ reported that SCI patients had a lower stone-free rate compared with an ambulatory population. More recently, Rubenstein and associates ¹⁰ reported a 96% stone-free rate in 23 patients. It is important to note that all but two patients in this series underwent a second-look procedure. In addition, four patients needed a third-look procedure and two needed a fourth-look procedure. This could account for the increased success rate in this series. Donnellan and Bolton ¹¹ reported on 76 PCNLs in patients with SCI who had struvite stones with a clearance of 84% for mean of 1.3 procedures. Symons and coworkers ¹² described a 48.7% stone-free rate by first PCNL in a cohort with various causes for spinal neuropathy.

Our initial stone-free rate compares favorably with this series as well as that of Culkin and colleagues ⁶ who reported a 53.6% stone-free rate with a single procedure. ⁶ In addition, our rate may be elevated secondary to decreased sensitivity of plain radiography as postoperative imaging in 60% of cases. Our initial stone-free rate was decreased compared with the ambulatory population stone-free rate of 75% to 98%. ¹³ This emphasizes the challenge of stone clearance in patients with neurogenic bladder (NGB). There were no statistically significant differences between the patients with SCI and SB in terms of outcomes.

Complications are common in this patient population. Culkin and associates ⁶ reported 8.5% major complications after PCNL in patients with SCI including transient respiratory arrest in a quadriplegic patient, two perirenal abscesses, and one hydrothorax; 64% experienced fever. ⁶ In addition, Culkin and coworkers ⁹ found a higher complication rate—20% in SCI patients—in a comparative study with ambulatory patients. Symons and associates ¹² reported their experience in patients with spinal neuropathy, including SB. They found a 10.3% major complication rate and 18% of patients experiencing fever. Rubenstein and colleagues ¹⁰ described their experience in patients with spinal neuropathy and found a 7% complication rate; however, nearly a third of the patients experienced fever. In this study, 30% of procedures (n=20) resulted in fever, and a major complication rate of 13.6%. Major complications included hemothorax necessitating VATS, four patients with prolonged intubation, one seizure, one extremity necrosis, and one death. We found no statistically significant differences in complications between patients with SCI and patients with SB.

We attempted to identify any factors that might predict outcomes in this population. Multiple access was associated with increased LOS (P=0.01), need for ICU care (P<0.01), need for transfusion (P<0.01), and pulmonary complications (P=0.03). Hegarty and Desai ¹⁴ showed that multiple tracts in the ambulatory population had similar complication rates and blood loss to a single tract. ¹⁴ Netto and coworkers ¹⁵ reported an overall complication rate of 45.4% when multiple tracts were used in the ambulatory population. Complications included 13 patients with blood transfusion, one sepsis/bacteremia, and one hydro/hemothorax. Stoler and colleagues ¹⁶ showed multiple punctures were associated with need for blood transfusion. Upper-pole access was associated with decreased initial stone-free rate (P=0.04). Midpolar access predicted increased final stone-free rate (P=0.04). This is in contrast to data from Netto and coworkers ¹⁵ in which upper-pole access resulted in increased stone-free rates compared with multiple, midpolar/lower pole access. These data were in ambulatory patients, and a possible explanation for the findings in our cohort is altered body habitus of patients with SB and SCI.

Mean stone size was 3.31 cm and was predictive of an increased number of procedures (P=0.04) as well as decreased initial (P=0.03) and final stone-free rates (P=0.05). This is similar to data in the ambulatory population.

To our knowledge, no predictors of outcomes, other than LOS, have been published in this population. An abstract presented at the 2010 American Urological Association Annual Meeting, however, showed >24-hour nephrostomy tube placement decreased sepsis in patients with neuromuscular disorders. ¹⁷ Early nephrostomy tube drainage had no effect on outcomes in our series. Rubenstein and associates ¹⁰ reported increased LOS was associated with greater stone burden and multiple procedures. In addition, they indentified risk factors for stone recurrence, which included a high spinal cord lesion, indwelling urinary catheter, and ureterosigmoidostomy.

One interesting finding of this study is the prevention of sepsis in this population. Ninety-three percent of patients for whom data were available had a positive preoperative culture; however, there were only nine cases of sepsis postoperatively. These patients were aggressively treated with culture-specific antibiotics and, if necessary, intravenous antibiotics for 1 week preoperatively and were continued on antimicrobials throughout the hospital stay. In addition, all patients with NGB received intravenous antimicrobials in the hospital 24 hours before PNCL. No patients had stone or renal pelvic culture. Korets and colleagues ¹⁸ recently published a prospective series on factors predicting systemic inflammatory response syndrome (SIRS) after PCNL. The rate of SIRS was 9.8%, and multiple tracts and stone burden of ≥10 cm² were predictors of postoperative SIRS. The authors recommended collecting renal pelvic and stone culture to direct antimicrobial therapy.

Limitations of our study include its retrospective nature and small sample size. This is the largest series to our knowledge reported in this population, however. In addition, no adjustments for multiple tests were made in our statistical analysis, making confirmatory tests necessary for our findings.

Conclusion

Treatment of patients with spinal neuropathy and nephrolithiasis is challenging. These patients are at an increased risk of nephrolithiasis and complications during percutaneous nephrolithotomy. In addition, patient positioning and access is difficult secondary to contractures and kyphoscoliosis. Predictors of adverse outcomes in our analysis include stone size and multiple access. Multiple access was associated with increased LOS, need for ICU care, need for transfusion, and pulmonary complications. In addition, stone-free status was influenced by access site. Upper-pole access predicted decreased initial stone-free rate and midpolar access predicted increased final stone free-rate. This conflicts with data from the ambulatory population and could be because of altered body habitus. Method of bladder management, including diversion, and stone composition were not predictive of outcomes. There were no differences in outcomes between patients with SCI patients and those with SB.