UTIs are primarily caused by Gram-negative bacilli (71%), with Gram-positive pathogens and fungi accounting for the remainder of microorganisms [6]. Escherichia coli is by far the most common cause of community-acquired and nosocomially acquired UTI. Most UTIs arise from ascending infection by enteric organisms that colonize the perineum and distal urethra. Specific clones of E. coli have evolved that readily colonize the uroepithelium and cause UTI. These clones possess the requisite set of virulence genes needed to successfully attach, survive, and invade the urinary tract in nonimmunocompromised patients with anatomically normal genitourinary (GU) tracts [10].

An essential characteristic of uropathogenic E. coli is its ability to adhere to uroepithelial membranes. Urinary isolates of E. coli possess an array of adhesions including type I (common pili), S pili, FIC pili, and P pili. These bacterial surface structures facilitate attachment to epithelial surfaces. Type I pili bind to mannose-containing polysaccharides on the cell surface of epithelial membranes. This allows the organism to attach and persist within the urinary tract and avoid elimination during micturition [11].

Another important adhesin of uropathogenic E. coli is the expression of P pili on the bacteria’s outer membrane [12]. P pili bind to α-D-galactose 1 → 4 β-D-galactose (Gal-Gal) containing disaccharides of the globoseries of glycolipids. These glycolipids are found primarily on the epithelial surfaces of the upper urinary tract, enterocytes, and erythrocytes. The ability of E. coli to express P pili is particularly important in the establishment of upper UTIs where Gal-Gal disaccharide-containing glycolipids are found in large concentration. Recent genetic analysis reveals that bacterial pathogens cluster their virulence factors in discreet loci along the chromosome known as pathogenicity-associated islands (PAIs). These genetic elements contain a large number of genes associated with virulence and distinguish uropathogenic strains from nonpathogenic colonizing strains [13].

Other genera of the Enterobacteriaceae, including Citrobacter, Klebsiella, Enterobacter, Serratia, Proteus, Morganella, and Providencia spp, become more common causes of UTI when patients receive antibiotics or have anatomic or functional abnormalities in urine flow [14]. The microbiology of UTI after short-term urinary catheterization is similar to that observed in the noncatheterized patient. However, long-term (> 30 days) catheterization generates an environment that supports a complex and often polymicrobial microflora. An extensive extracellular array of microbial-derived polysaccharides surrounds bacterial microcolonies within the lumen of the long-term urinary catheter. This biofilm structure protects bacterial populations for immune, phagocytic, or antibacterial clearance [15]. Bacteria found in the urine in chronically catheterized patients differ from noncatheterized patients. Proteus, Providencia, Morganella, and Pseudomonas species become more common, whereas E. coli and Klebsiella species become less common (Fig. 82.1). Proteus species, some other Gram-negative enteric organisms, and Staphylococcus saprophyticus synthesize the enzyme urease, a known bacterial virulence factor in the urinary tract. The generation of ammonia from the breakdown of urea increases regional pH, favoring the generation of the “triple-phosphate crystals” struvite and apatite in urine. Struvite crystals can block urinary catheter flow and promote the formation of urinary calculi [16].

Gram-positive bacteria occasionally cause UTIs in critically ill patients. The isolation of S. aureus in the urine is significant as it often accompanies staphylococcal bacteremia. S. aureus isolation in urine cultures, particularly in noncatheterized patients, should prompt a search for extrarenal sources of staphylococcal infection. S. aureus may also colonize chronically catheterized patients. This is particularly true for methicillin-resistant S. aureus strains, which may thrive in hospital settings with many elderly, catheterized patients [8].

Enterococci are prevalent in the GU tract of elderly populations and in patients with long-term urinary catheters. The remarkable ability of this organism to resist antimicrobial agents, including β-lactam antibiotics, aminoglycosides, quinolones, and recently vancomycin, makes this organism a frequent urinary pathogen in hospitalized patients [6,7,8]. Candida species and other fungal organisms may colonize or infect the GU tract. Candiduria may be associated with hematogenous dissemination (“descending UTI”) or ascending UTIs from perineal surfaces. The unique problems associated with the isolation of Candida species of the urinary tract are considered in the final section of this chapter.

The human GU tract is remarkably resistant to UTI by mechanical, mucosal, and immunologic mechanisms. The flushing action of urinary flow itself is an important defense against UTI.
The frequent occurrence of UTI after obstruction or incomplete bladder emptying attests to the importance of micturition in clearing potential pathogens. Patients with neurogenic bladder or vesicoureteral reflux are highly susceptible to UTI and renal scarring. While urinary pathogens must possess a full complement of virulence factors to cause infection in the anatomically normal urinary tract, UTI in the obstructed urinary tract occurs with bacterial species devoid of special urinary virulence factors [10].

Urinary osmolarity, urea concentration, pH, and oxygen concentration limit the growth potential of many bacterial pathogens in the urinary tract. Continuous sloughing of uroepithelial cells, urinary mucosal glycocalyx (slime), and secretion of the Tamm–Horsfall protein assist in the mechanical removal of adherent bacteria that have entered the urinary tract [17].

The ability of bacteria to adhere to the mucosal surface of uroepitheilial cells is dependent on the mucopolysaccharide content of this surface and its chemical composition. Patients with high concentrations of Gal-Gal disaccharides on the cell surfaces in the urinary tract are predisposed to UTI from P-piliated E. coli [10]. Patients who are nonsecretors of blood group antigens have an increased risk of UTI [18]. Blood group antigens coat uroepithelial cells when secreted onto the mucosal surface. These antigens prevent attachment of bacteria to adhesin-receptor oligosaccharides on the surface of epithelial cells. Individuals who fail to secrete blood group antigens are rendered infection prone to UTI.

Although secretory immunoglobulin, neutrophils, and cell-mediated immunity contribute to the host defense against UTI, their roles are secondary to mechanical and physical barriers to infection. Uroepithelial cells produce the chemokine interleukin-8 (IL-8) in response to E. coli infection. IL-8 promotes neutrophil migration to the urinary tract which reduces the risk of disseminated infection. Patients differ in their level of expression of the IL-8 receptor CXCR1. Decreased CXCR1 expression in the urinary tract might contribute to increased susceptibility to pyelonephritis in some patients [19].

Adult women are much more likely to develop UTI than men. Women are more likely to develop pyelonephritis if they are sexually active, use spermicidal agents, experience urinary incontinence, have diabetes mellitus, or a family history of UTI [16]. The increased anatomic distance from the urethral orifice to the urinary bladder, the infrequent presence of Gram-negative bacteria around the male urethra, and the production of inhibitory prostatic secretions protect men from UTI until they become elderly [20]. Bladder neck obstruction from age-related benign prostatic hypertrophy causes urinary obstruction and UTI in elderly men.

Complicated UTIs often require radiologic methods to establish the correct diagnosis. Routine abdominal radiographs may assist in the diagnosis of complicated forms of UTI. The presence of radiopaque renal calculi can be readily detected on abdominal radiography. Emphysematous pyelonephritis appears as an abnormal collection of gas within the renal parenchyma. Gas is detectable in the urinary collecting system in many patients with pyonephrosis. Abnormal renal shadows and loss of psoas margins may suggest the presence of a perinephric abscess. Renal ultrasonography and computed tomography (CT) have replaced the intravenous pyelogram (or excretory urogram) as the principal radiographic technique in the detection of complicated UTI. The anatomic definition of the kidney and
perirenal tissues is superior with a contrast-enhanced abdominal CT scan and is generally the preferred imaging method for complicated UTI (see Table 82.1). Renal ultrasound provides another rapid method of detecting hydronephrosis and anatomic detail of the renal parenchyma. Ultrasonography can also determine the solid or cystic nature of a renal mass detected on abdominal CT. Ultrasound can study the kidney on any plane and may be performed urgently in the absence of intravenous contrast media. The CT scan or renal ultrasound is indispensable in the localization of inflammatory processes during diagnostic aspiration or percutaneous drainage procedures. Magnetic resonance imaging (MRI) provides detailed information about the renal structures and retroperitoneal space, but the CT has sufficient resolving power in most forms of renal inflammatory disease.

The gallium-67–scan or indium-111–labeled leukocyte studies can occasionally be useful in the diagnosis of complicated UTI. These nuclear medicine studies assist in the differentiation between a renal neoplasm and a focal inflammatory process of the kidney. These studies are useful in the evaluation of patients with fever of unknown origin secondary to perinephric abscess or renal cortical abscess [27].

Intravenous pyelography (IVP) provides refined details of the calyces and ureters and remains an excellent diagnostic method for the diagnosis of papillary necrosis or small, radiolucent urinary calculi. The need for intravenous contrast media carries attendant risks of hypersensitivity reactions and radiocontrast-induced renal failure. The potential toxicity and limited resolution outside the urinary collecting system has relegated the IVP to an infrequently performed procedure in the workup of UTI in ICU patients [28].

Patients admitted to the ICU for management of UTI usually suffer from severe infections complicated by a systemic inflammatory response (sepsis) or suppurative complications of the GU tract. Medical management initially consists of stabilization of the patient’s hemodynamic parameters and supportive measures in the management of septic shock. After the completion of appropriate diagnostic studies, empiric antimicrobial therapy should be directed toward the most likely infecting urinary pathogen(s). A urinary Gram stain usually provides evidence of either a Gram-negative or Gram-positive bacterial pathogen. If this is unavailable or nondiagnostic, then broad-spectrum, empiric antimicrobial therapy is indicated.

In the septic patient with UTI, the initial use of a β-lactam antibiotic (assuming there is no history of allergic reactions to β-lactams) in combination with an aminoglycoside has been the traditional therapeutic regimen in hospitalized patients. The β-lactam/aminoglycoside combination supplies optimal therapy for systemic infections with enteric Gram-negative bacilli, enterococci, and nonfermentative, multiresistant, Gram-negative bacterial pathogens. Severely ill septic patients who are immunocompromised also warrant combination antimicrobial therapy [21]. Increasingly, the therapeutic trend in empiric therapy is away from aminoglycosides to monotherapy with β-lactams alone, β-lactam/β-lactamase inhibitors, and/or fluoroquinolones [29].

Community-acquired UTIs in nonimmunocompromised patients who have not received antimicrobial agents infrequently
harbor multiresistant Gram-negative bacilli or Pseudomonas sp. Should the urinary Gram stain exclude enterococci as a potential pathogen, then single therapy with a third-generation cephalosporin, extended-spectrum penicillin, carbapenem (e.g., imipenem or meropenem), β-lactam/β-lactamase inhibitor (e.g., piperacillin-tazobactam) trimethoprim-sulfamethoxazole, or a fluoroquinolone is acceptable therapy while awaiting culture results. Local susceptibility patterns of urinary pathogens should guide the selection of antimicrobial therapy until specific susceptibility data are available. There is no evidence that combination antimicrobial therapy is necessary for UTIs caused by Gram-negative bacilli unless Pseudomonas aeruginosa infection with neutropenia is present. A single antimicrobial agent known to be active against the infecting uropathogen should be employed once the causative organism is known. Parenteral therapy is generally administered until the patient has been rendered nontoxic and afebrile for 24 to 48 hours. Therapy may then be administered orally and should be given for a total of approximately 2 weeks [21,24]. Patients with obstructive lesions and complicated UTIs not amenable to corrective surgery may require prolonged courses of antimicrobial therapy, as indicated by their underlying urologic disorder. Common antimicrobial agents useful in the treatment of severe UTIs are listed in Table 82.2.

Standard therapy for severe enterococcal UTIs has been ampicillin and an aminoglycoside. Although this regimen remains active against most enterococcal isolates, progressive antimicrobial resistance to aminoglycosides, ampicillin, and other β-lactams and vancomycin has complicated the antimicrobial therapy for enterococcal infections [30]. Rare strains of β-lactamase–producing enterococci are susceptible to β-lactam inhibitors such as ampicillin/sulbactam or piperacillin/tazobactam. High-level aminoglycoside-resistant strains of enterococci are problematic, as the addition of an aminoglycoside no longer contributes to synergistic clearance of these infections. Aminoglycosides should not be used in this situation.