Pre- and Postoperative Predictors of Infection-Related Complications in Patients Undergoing Percutaneous Nephrolithotomy

Introduction

Since its initial description by Fernstrom and Johansson, ¹ percutaneous nephrolithotomy (PCNL) has emerged as the gold standard therapy for large renal calculi (>1.5 cm) and staghorn calculi due to decreased morbidity and high stone-free rates when compared with the alternative modalities of SWL, ureteroscopy, and open nephrolithotomy. While PCNL is regarded as a less morbid procedure than its open counterpart, PCNL is not without risks for significant complication with Clavien >II complications occurring in ∼9% of cases. ² One of the more dreaded complications occurring after PCNL is postoperative infection, ranging from fever to severe sepsis carrying with it a significant risk of morbidity and in rare circumstances even mortality. ³

The pathogenesis of infection after PCNL is likely multifactorial with large renal stones serving as foreign body to harbor colonized bacteria and making urine sterilization difficult even with antibiotic therapy. Technique may theoretically increase risk for infectious complication. Pyelovenous backflow and increased absorption of infected urine occur when an access sheath is not utilized due to renal pelvis pressure exceeding 30 mm Hg or when perforation of the collecting system occurs with extravasation of infected irrigant into an open system. ^{4 –7} The size of the access sheath may also play a role, as mini-PCNL has been associated with increasing renal pelvis pressure and risk of fever in one series. ⁸

While investigations into the pathophysiology of post-PCNL infection have yielded mixed results, prediction of post-PCNL infectious complications has been explored in an effort to prevent this potentially fatal complication. Many studies have investigated preoperative risk factors for postoperative fever, systemic inflammatory response syndrome (SIRS), and/or sepsis utilizing different definitions with varied results, including the presence of positive urine culture, female gender, preoperative nephrostomy tube, large stone burden, presence of residual fragments, staghorn calculus, pyuria, diabetes mellitus, urinary diversion, old age, and paraplegia. ^{9 –15} Postoperative risk factors for infectious complications remain less studied; however, there seem to be significant risks of postoperative complications with increased operative time, requirement of blood transfusion, positive stone culture, and multiple punctures. ^9,13,16 While there have been multiple studies investigating both pre- and postoperative risk factors for infectious complication after PCNL, there is a lack of data analyzing SIRS/sepsis and other infectious complications. Thus, the aim of this investigation is to describe pre- and postoperative predictors of all infection-related complications in individuals undergoing PCNL receiving extended perioperative antibiotics.

Patients and Methods

This study was approved by the Mayo Clinic Institutional IRB. Since September 2009, we have prospectively enrolled patients undergoing PCNL. All nonpregnant patients older than 18 years undergoing elective PCNL for symptomatic upper tract urinary stone disease were informed of our study and offered enrollment. For the current study, the 227 patients enrolled as of December 2013 were evaluated. CT images were available before the procedure to assess current stone burden. All patients with a preoperative urinary tract infection (UTI) or positive urine culture were treated with 2 weeks of culture-specific antibiotic therapy. All other patients with no growth on urine culture received 7 days of empiric therapy with an antibiotic of surgeon's choosing. If renal obstruction was noted before surgery, then a ureteral stent or percutaneous nephrostomy tube was placed preoperatively to decompress the collecting system before PCNL. During the perioperative period, patients received culture-specific IV antibiotics or if the urine culture demonstrated no growth as per AUA Best Practice Guidelines. Access was placed in either an upper or lower pole posterior renal calix to maximize stone removal and a 30F Amplatz access sheath was utilized in each case to decrease renal pelvis pressure and minimize irrigation absorption. One fragment of the removed stone was sent for culture, which includes fungal, anaerobic thioglycolate, blood agar, and EMB media, and the remaining for detailed structural and compositional analysis. Postoperative CTs were performed to assess stone-free status. Patients with residual stones were treated with either secondary PCNL or ureteroscopy in an attempt to render them stone free.

Infectious complications, including postoperative fever (defined as a temperature >38°C), were identified utilizing the following definitions ¹⁷ : SIRS, any two or more of the following: body temperature >38°C or <36°C, heart rate >90 bpm, respiratory rate >20 breaths/minute, white blood cell count >12 × 10⁹ or <4 × 10⁹ cells/L. Sepsis was defined as both SIRS and a positive postoperative blood or urine culture. Other infectious complications included in analysis were UTI and pyelonephritis defined on the basis of clinical signs and symptoms reported from the patient in the medical record, which included lower urinary tract symptoms in the presence of a positive urine culture for UTI and UTI plus fever without positive blood cultures for pyelonephritis. All infection-related complications were recorded up to 3 months postoperatively.

Univariate analyses were performed using Kruskal–Wallis, Wilcoxon rank sum, chi-squared, and/or Fisher's exact tests. The relative risk (odds ratio [OR]) was used in a multivariable logistic regression to determine risk factors for fever/SIRS/sepsis and overall risk for infection, including UTI and pyelonephritis. All reported p-values were two sided, with p < 0.05 considered to be statistically significant. Statistical analyses were performed using SAS software, version 9.3 (SAS Institute, www.sas.com).

Results

We identified 227 patients who underwent PCNL and were enrolled in the study during the time period. Of these, 98 (43%) were male with a median age of 58.9 years. Regarding choice of antibiotic, 114 (50.4%) of patients received nitrofurantoin, 31 (13.7%) trimethoprim/sulfamethoxazole, and 42 (18.6%) received a fluoroquinolone. Median duration of preoperative antibiotics was 7 days (interquartile range [IQR] 7–10) and postoperative was 7 days (IQR 7–14). A total of 64 (31.2%) of patients had a positive preoperative urine culture.

We identified postoperative infectious complications in 37 (16%) patients: 11 (5%) UTI/pyelonephritis, 21 (9%) SIRS, 2 (0.9%) sepsis, 2 (0.9%) fever only, and 1 (0.4%) infected urinoma. There were no significant differences between those with and without infectious complication with regard to age, gender, stone size, presence of diabetes, or procedure duration (Table 1). We identified 67 patients with a positive stone culture. Of these, 15 (22%) developed an infectious complication. We identified 46 patients with an intercostal access, with 12 (26%) of these patients developing a postoperative complication. There were eight patients with >1 access site. Four of the patients with >1 access developed an infection-related complication. Those with infection-related complication were more likely to have a positive stone culture (p = 0.01), struvite stone composition (p < 0.01), staghorn calculi (p < 0.001), and multiple stones (p = 0.02) (Table 1). On univariate analysis, the presence of struvite stone, positive stone culture, and staghorn calculus were associated with an increased risk of infection-related complication (Table 2). Preoperatively, on multivariable analysis, only the presence of a staghorn calculus remained independently associated with increased risks of fever/SIRS/sepsis (OR 3.14; p = 0.02) and total infection-related complications (OR 2.53; p = 0.02) following PCNL. Although there was a positive trend, the presence of multiple stones did not predict infection-related complications (OR 4.2; p = 0.06) (Table 3). After controlling for pre- and post-PCNL risk factors, again, only staghorn calculi remained significantly associated with fever/SIRS/sepsis (OR 3.41; p = 0.01) and total infection-related complications (OR 2.91; p = 0.01) (Table 4).

Discussion

The current study investigated all infection-related complications, including fever, SIRS, sepsis, and their risk factors after PCNL, in a cohort of 227 patients. The overall rate of sepsis was low at 0.9% and SIRS was noted in only 9% of patients. The presence of a staghorn calculus was identified as the only independent risk factor for fever/SIRS/sepsis and overall infection.

Our findings are consistent with prior studies that have also concluded that a large stone burden or staghorn calculus was associated with infection-related complications after PCNL. Gonen and colleagues ¹³ prospectively investigated 61 patients undergoing PCNL to determine risks for postoperative fever. Of the 61 patients investigated, 10 developed postoperative fever, with one developing sepsis (1.6%). The authors found that stone size, length of operation, positive stone culture, and positive pelvic urine culture were all significantly associated with postoperative fever; however, a multivariate analysis was not utilized to determine an independent association with postoperative infection. In an even larger study investigating postoperative fever after PCNL, Gutierrez and colleagues ¹⁸ utilizing the CROES (Clinical Research Office of the Endourological Society) data found that 10.4% of the 5313 patients with pre- and postoperative data developed fever (>38.5°C). On logistic regression analysis, they determined that the presence of fever was independently associated with positive urine culture, diabetes, the presence of a staghorn calculus, and the preoperative use of a nephrostomy tube. ¹⁸ Unfortunately, these data only investigated postoperative fever, were from more than 96 different centers worldwide with different patient populations, and retrospectively collected. Due to the heterogeneity of the CROES study population, it is difficult to generalize the results of the study, thus limiting the power of the conclusions made by the authors.

In an investigation into post-PCNL SIRS, Chen and colleagues ¹⁶ retrospectively analyzed 209 patients and identified SIRS in 23.4% and sepsis in 3 (1.4%). Their analysis demonstrated an independent association in the development of SIRS post-PCNL with receipt of a blood transfusion (OR = 22.41, 95% confidence interval [CI] = 3.91, 128.53; p < 0.001), stone size (OR = 1.10, 95% CI = 1.05, 1.14; p < 0.001), and presence of pyelocaliectasis (OR = 3.35, 95% CI = 1.44, 7.81; p = 0.005). ¹⁶ While our investigation did find the presence of a staghorn calculus and a trend toward significance with stone number (likely corresponding with stone size in Chen and colleagues), receipt of a transfusion, while overall rates were similar (6.2% vs 6.2%), was not associated with infection significant risk for the postoperative infection and/or fever/SIRS/sepsis. Specifically, the rate of transfusion in staghorn calculi was 8%.

In a more recent study, Koras and colleagues investigated pre- and intraoperative risk factors leading to infection-related complications, SIRS, and sepsis post-PCNL. ¹¹ In their series, 83 of 303 patients were diagnosed with SIRS and 7.6% of patients developed postoperative sepsis. On multivariate logistic regression analysis, stone burden, recurrent UTI, and infected stone were all independently associated with development of SIRS and sepsis. Our data do appear to concur with some of the conclusions of the Koras study in that staghorn stones did predict complication; however, our rates of both SIRS (9%) and sepsis (0.9%) were significantly lower.

There are certain limitations to our study. The sample size is small compared with large multi-institutional studies; however, limiting to one institution ensures that surgical technique and treatment protocols were standardized. Furthermore, as a tertiary care center, the patient population is unique and the findings of the study may not be generalizable to the general practitioner.

Conclusion

In individuals undergoing PCNL on preoperative antibiotics, the risk of SIRS/sepsis was low. The presence of a staghorn calculus confers a greater than threefold increased risk of postoperative infection-related complications with multiple stones approaching a significant independent risk. These data can be used to appropriately counsel patients with large complex stone burdens of their infection-related risks.