Recent studies in mice suggest targeting COX-2 for treatment of recurrent urinary tract infections (RUTI) [1,2]. However, the role of COX-2 in human RUTI remains undefined. Our goal was to measure levels of PGE2, a product of the COX-2 enzymatic pathway, in patients with and without RUTI to determine if PGE2 has the potential to serve as urine biomarker for RUTI.

Following IRB approval and patient consent, urine was collected from post-menopausal women seen at a tertiary care urology clinic and cultured on CHROMagar. Patients had varying urine analysis (UA), history of urinary tract infections (UTI), and treatment regimens. Urinary PGE2 levels were measured in triplicate for each sample by an accurate and commercially available PGE2 ELISA (Enzo) that was normalized to urinary creatinine (Cr) by a quantitative creatinine assay (Sigma). PGE2/Cr levels were compared between groups of patients classified by several different criteria: clinical history of UTIs, clinical positive vs. negative UA, bacteriuria count (>10^4 CFU/ml), and urinary pH.

Urine samples of 65 postmenopausal women were analyzed. Women with a history of RUTI with a positive UA (+UA) (n=23) had statistically significant (p = 0.0303) increased levels of normalized urinary PGE2 (mean=1.957 pg/μg) compared to patients with no history of UTI with negative UA (-UA) (n=15, mean=1.281 pg/μg). In women with negative UA, the difference in normalized PGE2 levels between patients with a history of RUTI (n=27, mean=1.751 pg/μg) and those with no lifetime history of UTI was not significantly significant. No significant difference in normalized PGE2 levels was noted when patients were classified based on presence vs. absence of bacteriuria on CHROMagar or by urinary pH.

Recurrent UTIs may be explained by an interactive cycle of infection and inflammation. Host inflammation in response to initial bacterial pathogens include severe cystitis, exfoliation of urothelium, and COX-2 overexpression. This sets the stage for formation of quiescent intracellular bacterial communities, which is thought to cause recurrent infection [1]. Additionally, studies in mice have found urothelial remodeling after infection, in which the urothelial cells are smaller, have weakened junctions, and lack terminal differentiation [1,2]. This sensitized urothelium may make certain RUTI patients’ bladders more susceptible to infection and thus perpetuate the cycle of infection and inflammation. 
	The rationale for inhibiting COX-2 is to target the host inflammatory stage of the cycle. Decreasing inflammation may decrease the likelihood of urothelial sensitization and recurrent infections. This was the idea in Hannan et al’s study in which mice were given an oral COX-2 specific inhibitor [3]. These mice had reduced severity of pyuria, reduced incidence of chronic cystitis, and decreased bacterial titers when compared to control or COX-1 specific inhibitor [3]. Gross bladder inflammation was reduced, shown by less edema and urothelial erosion [3].
	The translational work to human RUTI is scarce. There have been randomized controlled trials of NSAID use in premenopausal women with uncomplicated UTIs showing resolution of infection with NSAIDs alone compared to antibiotics. In our cohort of post-menopausal RUTI women, we would potentially explore adjunct COX-2 inhibitor therapy in addition to antibiotic therapy. This may slow the progression of recurrence and decreases severity of each infection. 
Because of potential side effects of COX-2 specific inhibitors, much work is to be done to correlate COX-2, PGE2, and a patient’s degree of bladder urothelial sensitization  before implementing COX-2 inhibitor therapy. Urinary PGE2 levels could help as a guide or biomarker for women who are sensitized and thus potentially would benefit from COX-2 inhibition.

In this exploratory study on post-menopausal women, we found a statistically significant increase in levels of urinary PGE2/Cr in patients with a history of RUTI with +UA versus patients reporting never having a UTI. These new findings suggest that PGE2 may be a useful urine biomarker for RUTI to guide the administration of COX-2 inhibitor therapy.

O’Brien VP, Hannan TJ, Yu L, Linvy J, Roberson ED, Schwartz DJ, Souza S, Mendelsohn CL, Colonna M, Lewis AL, Hultgren SJ. A mucosal imprint left by prior Escherichia coli bladder infection sensitized to recurrent disease. Nat Microbiol. 2016 Oct 31;2:16196.
Sivick KE, Mobley HLT. Waging War against Uropathogenic Escherichia coli: Winning Back the Urinary Tract. Infection and Immunity. 2010 Jan; 78(2):568-585
Hannan TJ, Roberts PL, Riehl TE, van der Post S, Binkley JM, Schwartz DJ, Miyoshi H, Mack M, Schwendener RA, Hooton TM, Stappenbeck TS, Hansson GC, Stenson WF, Colonna M, Stapleton AE, Hultgren SJ. Inhibition of Cyclooxygenase-2 Prevents Chronic and Recurrent Cystitis. EBioMedicine. 2014 Oct 24;1(1):46-57.