Introduction

The aim of this paper is to focus on urinary tract infection (UTI) associated with conditions causing urinary tract obstruction and stasis, other than urolithiasis and neuropathic bladder dysfunction. The prevalence and causes of urinary tract obstruction in patients with UTI vary considerably in different reports in the literature. The most common causes are benign prostatic hyperplasia (BPH) in up to 45% of cases, prostatic cancer in 2–8%, urethral stricture in 3–8%, and pelviureteric junction obstruction (PUJO) in 2% (Fig. 1) [1].

Fig. 1
figure 1

Excretory urogram shows bilateral hydro-ureteronephrosis with large postvoid residual urine volume due to bladder outlet obstruction

Full size image

Methods

An electronic literature search was performed on PubMed using the key words urinary tract infection (UTI) alone and in combination with the following terms: benign prostatic hyperplasia, BPH, hydronephrosis, obstruction, reflux, diverticulum, urethra, and stricture. Abstracts were reviewed, relevant articles were studied in full-length version, and further references were obtained from these articles. In total, 520 abstracts were reviewed, 210 articles were studied in detail, and 36 were included as references in this paper.

A more extensive version of this work was published as a chapter titled “Urinary tract infection in patients with underlying urological abnormalities” in [37].

Bladder outlet obstruction (BOO)

An experimental animal model in mice has shown that after transurethral inoculation with E coli, bacteria will be rapidly eliminated by urination in normal mice, but will cause bacteremia and pyelonephritis in mice with reversible urethral obstruction for 1–6 h. Urethral obstruction enhances the uropathogenicity of nonpathogenic E coli and causes more severe renal lesions in mice challenged with uropathogenic E coli [2].

A large postvoid residual (PVR) urine volume may be caused by BOO, but may also result from inadequate detrusor contraction, a large bladder diverticulum or even high-grade vesico-ureteric reflux. It seems logical that a large PVR should predispose to UTI, because bacteria are not mechanically washed out by normal urinary flow, and stasis allows more time for bacterial adherence and multiplication. However, there is little evidence in the literature that the occurrence of UTI in the aging male population is associated with either PVR or BOO. In elderly men, bacteriuria does not appear to correlate with age, PVR, or symptoms of BOO [3]. In elderly women, an increased PVR does not appear to be associated with bacteriuria. However, the incidence of bacteriuria after urodynamic studies is higher in men (36%) compared with women (15%), possibly due to larger PVR urine volumes, or to less effective washout of inoculated bacteria [4].

Benign prostatic hyperplasia (BPH)

A review of more than 200 studies published between 1954 and 1988 concluded that most men with UTI had a functional or anatomic abnormality of the genitourinary tract. BPH and genitourinary instrumentation were the major predispositions to UTI in men [1]. In older surgical series, UTI was the indication for surgical intervention in about 12% of men with BPH [5]. However, among placebo-treated patients in the Medical Therapy of Prostatic Symptoms (MTOPS) study, the cumulative incidence of UTI at 4 years was <1% [6].

Bacteriuria prior to prostatectomy in men with BPH has been reported in 28–54% of cases. The incidence is higher in those catheterized before operation (44–57%) than in those not catheterized (18%). Several studies have indicated a correlation between the incidence of preoperative bacteriuria and the duration of catheterization as well as PVR urine volume [7].

Bacteria can be cultured from prostatic tissue obtained at prostatectomy in 21–44% of cases (67% in catheterized and 28% in uncatheterized patients) [8]. Bacteremia after transurethral resection of the prostate (TURP) has been reported in 23–29% of patients despite antibiotic prophylaxis, in 54% of patients with preoperative bacteriuria and 8% without bacteriuria. In 70–81% of cases, an identical species was isolated from preoperative urine cultures, and in 54% the organism causing bacteremia was identical to that cultured from prostatic tissue. Bacteremia was significantly more common in patients with preoperative UTI [9].

Prostate cancer and prostatitis

In men with febrile UTI, the prostate is frequently involved, causing PSA elevation. The increased total PSA and decreased free/total PSA ratio in prostatitis may be falsely interpreted as a sign of cancer. The use of empiric antibiotic treatment to decrease the PSA and to avoid unnecessary prostate biopsy is controversial, but increases the risk of bacteremia caused by resistant bacteria after transrectal prostate biopsy [10].

There appears to be a correlation between bacterial colonization or chronic inflammation of the prostate and the development of prostate cancer. It is postulated that during acute or chronic inflammation in the prostate, various cells are activated by chemokines via different chemotaxin receptors which then trigger processes in angiogenesis, cellular growth and neoplasia which may lead to prostate cancer [11].

Urethral stricture

In men with UTI, in some developing countries, the second most frequent underlying abnormality, after BPH, is urethral stricture. Men with urethral strictures present with UTI in up to 44% of cases [12].

Diverticula

Bladder diverticula are most often acquired (secondary to BOO or neuropathic bladder dysfunction) but may also be congenital. They are much more common in men than women and are most often associated with BPH, prostate cancer or urethral stricture, although obstruction is not present in all cases [13].

Congenital bladder diverticula most often present with UTI [14]. Urinary stasis in a bladder diverticulum may predispose to UTI and may impede the eradication of established infection. Relapsing or persistent UTI unresponsive to antibiotic therapy may be an indication for bladder diverticulectomy.

A study comparing patients with BPH plus a bladder diverticulum and patients with BPH only found that the group with BPH plus a diverticulum had a higher rate of acute urinary retention and UTI (22% vs. 3%), a larger mean PVR and greater urethral resistance [15]. In patients who underwent TURP with or without diverticulectomy, urethral resistance parameters and PVR decreased in both groups. Diverticulectomy showed an improvement of bladder contractility with longer detrusor contraction duration, supporting its use in cases of BPH-associated diverticula [15].

Urethral diverticula in women are most probably caused by infection of the periurethral glands, possibly by Neisseria gonorrhoeae, although the initial infection and especially subsequent reinfections may be caused by E coli and other coliform bacteria or vaginal flora. Recurrent infection of the periurethral glands leads to obstruction, suburethral abscess formation and subsequent rupture of the infected glands into the urethral lumen [16].

UTI is the mode of presentation in about 30% of women with urethral diverticula. E coli is the most common organism isolated, but other gram-negative enteric flora as well as N. gonorrhoeae, chlamydia, streptococci, and staphylococci are often present [17].

The management of female urethral diverticulum is surgical excision of the diverticulum and closure of the tract communicating with the urethra. In women with recurrent UTI and a large PVR, dilatation of the urethra to improve bladder emptying is often performed, although there is no evidence to support its efficacy [18].

Upper tract obstruction

It is generally accepted that urinary stasis provides the time and opportunity for bacteria to adhere to the urothelium, multiply, and infect the host [19]. A contributing factor may be that ureteric obstruction causes renal dysfunction, so the kidneys are less effective in concentrating antibiotics in the urine, making eradication of bacteria difficult and predisposing to bacterial resistance. The most important determinants of UTI caused by resistant strains are previous use of antibiotics and the presence of underlying urological diseases [18].

Hematogenous infection of the kidney is rare in normal individuals, but experimental data indicate that infection is enhanced when the kidney is obstructed [20]. In an animal model of hematogenous pyelonephritis, the kidney is relatively resistant to infection by organisms injected intravenously, but if a ureter is ligated, the obstructed kidney becomes infected [21].

In an experimental model in rats, transient ureteric obstruction (24 h) with inoculation of E coli in the bladder after release of ureteric obstruction predisposes to ascending pyelonephritis. It is postulated that acute ureteric obstruction induces foci of medullary necrosis in the papillae, which may provide a nidus for bacterial invasion, while altered urodynamics after the release of ureteric obstruction may promote ascending infection [22].

There are few clinical studies on the risk of UTI associated with upper tract obstruction. In a study of children prenatally diagnosed with severe hydronephrosis due to PUJO or ureterovesical junction obstruction and not on antibiotic prophylaxis, the incidence of UTI during the first 12 months postnatally was 36%. Most cases of UTI (93%) occurred in the first 6 months postnatally [23].

In another study of children with prenatally detected renal pelvic dilatation, the estimated cumulative incidence of UTI was 8% at age 12 months, 13% at 24 months, and 21% at 36 months. Independent predictors of UTI during follow-up were female gender and the presence of vesico-ureteric reflux (VUR) or urinary tract obstruction [24].

In a study of children with prenatally diagnosed nonrefluxing hydronephrosis, UTI developed in 19% during 12 months of follow-up—in 84% of those cases within the first 6 months. UTI developed in 39% of children with and in 11% without obstructive uropathy. High-grade hydronephrosis was associated with an increased incidence of UTI: 40% in grade IV hydronephrosis, 33% in grade III, 14% in grade II, and 4% in grade I. The authors recommended antibiotic prophylaxis in neonates with obstructive uropathy, severe hydronephrosis or hydroureteronephrosis without VUR [25].

However, another study of children with antenatally diagnosed severe hydronephrosis secondary to obstructive megaureter or PUJO not maintained on prophylactic antibiotics reported an overall rate of UTI of only 4.3%. No statistically significant difference in the infection rate was noted according to sex, obstruction level, hydronephrosis grade, or circumcision status. The authors concluded that antibiotic prophylaxis is unlikely to benefit most children with hydronephrosis secondary to upper tract obstruction [26].

Despite a surprising lack of relevant clinical studies, it is generally accepted practice that in patients with UTI and upper urinary tract obstruction, empiric antibiotic treatment should be complemented with urgent intervention to drain the urinary tract (e.g., percutaneous nephrostomy or double-J ureteric stenting). This should be followed with targeted antibiotic treatment (according to urine culture), and definitive surgery to remove the cause of obstruction once infection is under control.

Renal papillary necrosis (RPN)

Although there are several factors that on their own may cause RPN, the coexistence of multiple factors (e.g., diabetes mellitus or obstruction and UTI) increases the risk of developing RPN [18]. Therefore, in ureteric obstruction, it is important to prevent UTI and, if UTI is present, to drain the kidney by ureteric catheter or percutaneous nephrostomy [27]. A necrotic renal papilla may cause obstruction, and the medullary cavity caused by RPN may cause urinary stasis, both conditions predisposing to UTI.

Renal calyceal diverticula

Renal calyceal diverticula are often asymptomatic, but may be associated with UTI and stone formation in up to 39% of cases (Fig. 2). The management is usually conservative, but indications for surgery include persistent pain, UTI, and nephrolithiasis. Open surgery has been largely replaced by percutaneous, uretero-renoscopic or laparoscopic approaches to obliterate the diverticulum or to dilate the connection between the collecting system and the diverticulum to improve drainage [28].

Fig. 2
figure 2

a and b Intravenous pyelogram showing calyceal diverticulum containing a stone (seen as a filling defect in the early pyelogram phase)

Full size image

Bacteriology

Urinary tract abnormalities may predispose to infection with organisms other than E coli, and long-term antibiotic therapy often leads to bacterial resistance or fungal superinfection with Candida albicans [29]. Urinary tract obstruction may permit ascending infection of E coli strains with lower adhesive ability. P-fimbriae are mannose-resistant adhesins of uropathogenic E coli which cause acute pyelonephritis (APN). The pap gene cluster (PapGI, -II, and -III) encodes the proteins required for P-fimbrial biogenesis [30]. The incidence of the PapG class II allele in E coli infecting adult patients with hydronephrosis is lower than in E coli infecting patients without urinary tract abnormalities (69% vs. 93%). This indicates that structural or functional abnormalities of the urinary tract may permit infection with E coli strains that would not be pathogenic in a normal urinary tract [31].

Special investigations

The aim of imaging in UTI is to detect conditions that must be corrected to avoid rapid deterioration of renal function or systemic sepsis and to prevent recurrent infections and long-term kidney damage. Indications for imaging in patients with UTI are shown in the Table 1 [18, 29, 32].

Table 1 Indications for imaging of patients with UTI [18, 29, 32]
Full size table

Intravenous urography (IVU) and ultrasound (US) have traditionally been used in the assessment of complicated UTI. US with abdominal radiography (AXR/KUB) is as accurate as IVU in detecting important urological abnormalities in men presenting with UTI.

Computerized tomography (CT) has become accepted as a more sensitive and specific modality. CT urography is increasingly performed as a comprehensive urinary tract imaging study. Magnetic resonance (MR) imaging is particularly useful in those with iodinated contrast allergies. Nuclear medicine has a limited role but is useful in the assessment of renal function prior to surgery [18]. Imaging with microbubble ultrasonographic contrast agents can overcome the limitations of conventional B-mode imaging and may become the modality of choice in the future [33].

It is generally accepted that urological evaluation should be carried out in men with febrile UTI, pyelonephritis or recurrent infection, or whenever a complicating factor is suspected. However, imaging may not be necessary in all men with febrile UTI. It has been suggested that in men younger than 45 years with a first acute UTI no radiologic, endoscopic or urodynamic investigation is required, provided a urethral stricture has been excluded [34].

In patients with complicated APN, urine cultures are positive in 90–98% of cases and bacteremia may occur in 21–42%, but in only a small minory (around 1%), the pathogens found in blood cultures are different from those in the urine [35]. Some authors have suggested that blood cultures should be reserved for patients with an uncertain diagnosis, those who are immunocompromised, and those who do not respond promptly to treatment. However, others recommend that blood cultures should be done in all patients with complicated APN, because bacteremia indicates severe disease, which is more likely to recur within 6 months in patients with non-E coli bacteremia and those with urolithiasis or hydronephrosis, especially men [35].

Antibiotic treatment

The treatment of UTI in the presence of urinary tract obstruction requires effective antibiotic therapy as well as appropriate Urological intervention to remove predisposing factors and to restore as far as possible the normal anatomy and function of the urinary tract in order to prevent septicemia and recurrent UTI [18].

Severely ill patients with possible urosepsis should be hospitalized, and empirical treatment may include intravenous ampicillin and gentamicin or alternatives such as ciprofloxacin, levofloxacin, ceftriaxone, aztreonam, ticarcillin-clavulanate or imipenem-cilastin. The choice of empirical antibiotic treatment should be according to local guidance or policy. Indiscriminate use of quinolones and cephalosporins is strongly discouraged because of increasing bacterial resistance. Every effort should be made to correct any underlying urinary tract abnormalities that may compromise treatment efficacy. Therapy is usually switched from parenteral to oral as soon as possible [18].

Empirical therapy of serious UTI should usually include an intravenous antipseudomonal agent. Targeted therapy should be initiated once susceptibility data are known. Agents commonly prescribed include aminoglycosides, beta-lactamase inhibitor combinations, imipenem, advanced-generation cephalosporins, and fluoroquinolones. Several pivotal clinical trials support the use of fluoroquinolones for serious UTIs, with ciprofloxacin the most frequently studied drug [36].

A 7-day course of therapy is recommended for women with symptoms of 1 week or more (i.e., complicated UTI), men (even those with apparently uncomplicated cystitis), and individuals with possible complicating factors. For patients with fever or more severe systemic infection, therapy for 10–14 days is recommended. Urine cultures should be performed prior to initiation of antibiotic therapy, during treatment if clinical response is unsatisfactory, and 7–14 days after cessation of therapy to determine whether treatment has been effective [1, 18, 34].

Conclusions

Despite the lack of evidence based on prospective randomized controlled clinical trials, logic dictates that in patients with UTI and obstruction, e.g., pyocystis or pyonephrosis, targeted antibiotic treatment should be complemented with emergency drainage by means of transurethral bladder catheterization, ureteric stenting or percutaneous nephrostomy, followed by definitive surgery to remove the cause of obstruction or stasis once infection is under control.

Further study is required to define the role, if any, for antibiotic prophylaxis in urinary tract obstruction, dilatation or stasis without infection, taking into consideration the risk of bacterial resistance and superinfection with highly resistant organisms. Further study is indicated to better define, in different populations and countries, the optimal type and duration of drainage of the infected, obstructed urinary tract. Given the increasing incidence of bacterial resistance due to indiscriminate use of antibiotics, further research is required to develop new antibacterial agents.