FEBRILE URINARY TRACT INFECTION AFTER RADICAL CYSTECTOMY WITH URINARY DIVERSION: DIFFERENT CHARACTERISTICS IN PATIENTS WITH ILEAL CONDUIT AND ORTHOTOPIC NEOBLADDER

J Mens Health Vol 16(3):e38–e46; August 27, 2020 This article is distributed under the terms of the Creative Commons Attribution-Non Commercial 4.0 International License. ©2020 Seungsoo Lee et al. Background and Objective To compare the incidence of febrile urinary tract infection (UTI) and bacterial identification between patients with orthotopic neobladder (ONB) and ileal conduit (IC). Materials and Methods Data of 164 patients who underwent radical cystectomy with ONB and IC for bladder cancer between January 2009 and January 2018 at our institution were analyzed. Febrile UTI observed was listed and subsequently compared. Incidence of febrile UTI, clinicopathological characteristics, and microorganisms identified were reported at 3 months interval; and preoperative predictors of febrile UTI were evaluated with Cox regression analysis. Patients were divided into ONB and IC. Results The study cohort included 52 patients with ONB and 112 patients with IC. Febrile UTI was diagnosed in 49 (29.9%) patients. Compared to IC group, ONB group had significantly higher incidence of young age (p=0.00), lower cancer stage (p=0.013), longer hospital stay (p=0.049), longer operation time (p=0.00), and higher incidence of febrile UTI within the first 3 months after surgery (p=0.006). On univariable and multivariable analysis, factors associated with significantly increased febrile UTI risk were diabetes (odds ratio [OR]: 4.52; p=0.002) and ONB diversion (OR: 1.27; p=0.031). Forty-four (89.8%) patients were culture positive. However, significant difference in microorganisms was not detected between patients who under-went ONB or IC diversion. Conclusion Diabetes and ONB diversion were associated with higher risk symptomatic UTI following radical cystectomy.


INTRODUCTION
Radical cystectomy with pelvic lymph node dissection followed by the formation of urinary diversion represents a standard of care for muscle invasive bladder cancer. 1,2 The most common diversion types are orthotopic neobladder (ONB) and ileal conduit (IC). 3 However, urinary diversion is associated with many complications, including bowel obstruction, prolonged ileus, urinary tract infections (UTIs), stoma problem, ureteral stricture, metabolic acidosis, electrolyte imbalance, urolithiasis, and renal dysfunction. 4 Improvements in surgical technique and modern perioperative care have diminished perioperative complications rate substantially.
However, febrile UTI occurs frequently in patients who have undergone radical cystectomy. [5][6][7] In most cases, managements are complicated due to accompanying conditions, thus affecting prognosis of patients.
The objective of this study was to analyze the incidence, timing, risk factors, and microorganisms of febrile UTI in patients with bladder cancer treated by radical cystectomy according to methods of urinary diversion after radical cystectomy.

MATERIALS AND METHODS
This study was approved by the Institutional Review Board of our hospital (Approval No. 06-2014-006). Between January 2009 and January 2018, a total of 184 patients at our hospital underwent radical cystectomy with urinary diversion by a single surgeon (JKN), including 164 patients who had been observed at least 2 years postoperatively.
We excluded patients who didn't undergo a postoperative follow-up of less than 2 years (n = 10), patients who died within 2 years (n = 6), and patients who had undergone another major surgery (n = 4).
These 164 patients were used as subjects of this study. Indications for radical cystectomy included muscle invasive bladder cancer without distant metastasis, recurrent multifocal super� cial tumor refractory to repeated transurethral resection, and bacillus Calmette-Guerin resistant carcinoma in situ. Radical cystectomy with urinary diversion was performed using standard technique. ONB was performed as Studer's ONB technique. Patients underwent urinary catheterization with Foley catheter and neobladder irrigation for 2 weeks in general. In the IC surgery, ureters were spatulated and anastomosed independently to the ileal segment. In both surgeries, ureteral stents were inserted for 8 weeks in general.
Preoperative urine culture was obtained from all patients, and if positive, we underwent radical cystectomy with urinary diversion after treatment. In all patients, low residual diet was performed from 2 days before surgery and bowel preparation was performed in the night before surgery. Follow-up was conducted every 3 months during the � rst postoperative year and every 6 months thereafter. We performed computed tomography or magnetic resonance imaging to recurrence, laboratory study including urine culture, and a medical examination by interview. Febrile UTI was d� ned as a pyuria in the presence of fever (≥38 °C), chills with or without associate�� ank or abdominal pain.
Febrile UTI observed was prospectively listed and subsequently compared. Baseline classi� cation 4 complication underwent intensive care unit care due to severe UTI. All those cases were recovered from the complications, and there were no recurrence UTI in 3 months. Compared to IC group, ONB group had signi�cantly higher incidence of young age (p < 0.001), dominant male patients (p = 0.019), lower cancer stage (p = 0.013), longer hospital stay (p = 0.049), and longer operation time (p < 0.001). During the � rst 3 months after surgery, 26.9% of the patients with ONB and 8% of the patients with IC had a febrile UTI episodes (p = 0.006). Figure 1 compares the total percentage of infections at 3 months interval during the � rst 2 years after surgery for patients with ONB and IC diversion.
We investigated on the basis of � rst febrile UTI within 3 months.  Table 2). Forty-four (89.8%) patients were culture positive. However, signi� cant difference in microorganisms was not detected between patients who underwent ONB or IC diversion (Table 3). Table 3 details the common pathogens involved in febrile UTI episodes strati-� ed by diversion type.
Those who were aged ≥65 years were considered as elderly patients. Body mass index was dichotomized as obese (≥25 kg/m 2 ) versus non-obese (<25 kg/m 2 ) according to WHO proposed cut points for adult Asians. 8 Pathological stage was based on 2010 American Joint Committee on cancer TNM staging system for bladder cancer. 9 Urinary diversion was classi� ed as two groups: ONB and IC. Incidence of febrile UTI, clinicopathological characteristics, and growth of microorganism were compared between the two diversion method groups. Mann-Whitney U test and Fisher's exact test were used to compare categorical variables.
Estimated probability of febrile UTI was calculated with Kaplan-Meier method, and predictors of febrile UTI were evaluated using univariable and multivariable Cox regression models. Any p-value of less than 0.05 was considered statistically sig-ni� cant. Statistical analyses were performed using SPSS software version 18.0.

RESULTS
A total of 164 patients were eligible for this comparative analysis. The mean age of the patients was 68 years (range, 40-84) with average follow-up in patients who are alive at 62.8 months (range, 24-132). Of these individuals, 52 received ONB diversion, and 112 received IC diversion. A total of 49 (29.9%) patients were diagnosed with febrile UTI. Clinicopathological characteristics of these subjects are summarized in Table 1.
All the patients with Clavien-Dindo classi� cation 3 complication underwent insertion of percutaneous nephrostomy or ureteral stent for UTI. In some cases, temporary anastomosis site stricture caused UTI. In addition, the patients with Clavien-Dindo

DISCUSSION
Febrile UTIs after radical cystectomy include � ank pain, nausea, vomiting, high fever, with or without costovertebral angle tenderness. 10 It is a common infectious complication following urinary diversion. In large series of patients undergoing radical cystectomy with urinary diversion, the rate of UTI after radical cystectomy ranges from 8.5 to 39%. [5][6][7][11][12][13][14] Its cumulative incidence during the � rst 3 months after surgery is about 10%. 15 However, diagnostic criteria for UTI utilized in these series   In a nationwide population study on complications following radical cystectomy in Sweden, patients with ONB have higher risk of UTI (OR: 1.21) than those with IC. 16 A single center series of 209 patients following robotic radical cystectomy with urinary diversion has also reported an increased complication risk of UTI among patients undergoing ONB. 6 In this study, risk factors of symptomatic UTI included the receiving of continent urinary diversion (OR: 2.563; 95% CI: 1.630-2.948; p = 0.031) and the presence of diabetes (OR: 4.52; 95% CI: 2.862-5.571; p = 0.002).
Mechanisms underlying these results are multifactorial. Some studies have examined infections after ONB diversion and suggested the following possible mechanisms. 15,17,18 First, bacteria can colonize the bowel epithelium more easily than the urothelium and then use intestine for urinary diversion, thus increasing UTI. Second, incomplete emptying of the bladder might promote infection. Some studies have shown signi� cant correlation various bacteria were not the same in our study. Our results showed that Gram-negative bacteria were present in 47.7% of cultures. Enterococcus species, E. coli, Pseudomonas aeruginosa, Candida species, and Enterobacter species were common pathogens in this study. This could be due to different perioperative antibiotic uses, epidemiologic reasons, and different clinical characteristics of patients who previously received multiple course of antibiotic therapy and surgical procedures such as cystoscopic biopsy and transurethral resection of bladder tumor. A positive urine culture is a common � nding in patients with urinary diversion. However, this does not mean symptomatic infection. 32 In general, treating an asymptomatic bacteriuria is not advocated for these patients. However, controversy exists regarding appropriate treatment of asymptomatic bacteriuria in urinary diversion patients. In a recent study, 78% of patients with ONB have pyuria and nitrates based on urinalysis. 33 In the subset of patients with ONB who had evidence of UTI on urinalysis, 50% of them developed symptomatic infection. 33 Studer et al. 34 have recommended antibiotics for patients with ONB and positive urine culture result. Clearly, UTI with symptoms such as high fever, chilling, nausea, vomiting, and � ank pain must be treated quickly with appropriate antibiotics. Moreover, if recurrent symptomatic UTI occurs, evaluation of ureteral obstruction or r� ux is necessary using � uoroscopy of the reservoir or conduit to determine whether there is poor emptying of the ONB and identify stricture or obstructed areas. 35 Early diagnosis and prompt drainage are required to prevent consequent renal functional loss with aggravated infectious complications.
This study has some limitations. First, it was an uncontrolled and retrospective study. Among the groups of patients, there were differences in baseline characteristics such as age, sex, and pathologic stage. This might have affected our results. However, in the selection of surgical method, there are several considerations including age, renal function, cancer stage, patient's socioeconomic status, comorbidities, and between post-void residual urine volume and UTI. In addition, serial urodynamic examinations in patients with neobladders presented that a significant increase in maximal � ow rate and bladder capacity concomitant to a decrease in residual urine volume were documented over a 6-month interval from surgery, possibly contributing to the decrease in UTI events. 19 Third, excessive mucus production by the bowel epithelium accompanied by an established infection can prevent effective clearance of microorganisms. However, after 3-6 months after surgery, atrophic changes in the intestinal mucosa lining the neobladder occur, and consequently, the neobladder is less susceptible to bacterial colonization and infection. 20,21 Finally, prolonged urinary catheterization is related to UTI. For these mechanisms, it can be suggested that UTI rate in patients with ONB decreases overtime especially after 3 months postoperatively.
UTIs in diabetic patients are common. These infections have more complex clinical course. 22,23 Increased prevalence of bacteriuria in diabetic patients might be due to differences in host responses between diabetic and nondiabetic patients or difference in the infectious bacterium itself. 22,23 Host-related mechanisms include the presence of urine glucose, functional change of leukocyte, and increased adhesion to urinary tract epithelium. 22,23 Some studies have reported increase of Klebsiella but decrease of E. coli in patients with diabetes. 14,24,25 However, recent studies have found no signi� cant difference of bacterial species in diabetic patients, 26,27 while results of other studies are different. 28 Our results showed that the prevalence of symptomatic UTI was increased in diabetic patients. However, such increase was not due to difference in bacterial species.
Regarding identities of microorganisms following radical cystectomy, some studies have demonstrated a predominance of Gram-negative bacteria (65-91%). 23,[29][30][31] In patients with neobladder-related UTIs, the most common implicated microorganism is E. coli which is responsible for 59% of monobacterial infections. 23 However, percentages of preference. In general, younger and healthy patients underwent ONB rather than IC; therefore, there were differences in characteristics between ONB and IC patients. Second, the number of patients in this study was relatively small. However, there were no signi� cant differences in culture results among the groups. Therefore, comparative analysis of subgroup and culture proven microorganism is limited. Our results were different from the results of some previous studies. 23,[29][30][31] Finally, we did not analyze preoperative and postoperative changes in voiding and renal function. Ureteral stricture, obstruction, estimated glomerular � ltration rate (eGFR), and voiding function are known to be associated with acute pyelonephritis. It was very hard to evaluate eGFR of the total cohort within the follow-up periods, because each patient had different health condition. Therefore, it is needed to determine long-term changes of the parameters in the future.

CONCLUSION
The results of this study revealed that at 3 months after radical cystectomy for bladder cancer, diabetes and ONB diversion signi� cantly increased the risk of febrile UTI. However, our study has many limitations, including its retrospective nature, few patients with positive cultures, and different baseline characteristics among groups. Therefore, further research is needed.