The Use of Bacterial Isolate Cytokine Analysis to Assess the Risk of Pyelonephritis in Children

Sunday, October 3, 2010: 8:00 AM
Yerba Buena Salon 9 (San Francisco Marriott Marquis)
Douglas W. Storm, M.D.1, Dennis Horvath2, Birong Li2, Stephen A. Koff, M.D., FAAP1 and Sheryl Justice, PhD2, (1)Section of Pediatric Urology, Nationwide Children's Hospital, Columbus, OH, (2)Center for Microbial Interface Biology, The Research Institute at Nationwide Children's Hospital, Columbus, OH

Purpose: Despite heightened awareness of conditions such as urinary reflux (VUR) and bladder dysfunction that influence the risk of pyelonephritis and nephropathy in children, little progress has been made to prevent their occurrence by characterizing host defenses and bacterial virulence. Urinary cytokine levels, markers of genitourinary inflammation in UTI, mirror the degree of bladder inflammation and reciprocally reflect bacterial virulence: virulent bacteria evade host defenses, minimize bladder inflammation and suppress cytokine production. We hypothesize that VUR, neurogenic bladder (NGB) and non-neurogenic bladder dysfunction (NNGBD) differently influence the bacterial virulence characteristics and risk of pyelonephritis. To assess the interplay between microbial virulence and clinical manifestations, we compared in-vitro cytokine responses of uropathogens in children with and without these conditions.

Methods: Bacterial isolates causing UTI in children were collected and stored along with the patients' clinical data and urological evaluation. In-vitro cytokine analysis (IL-6) was performed by ELISA on the culture supernatant from bladder epithelial cells infected with each bacterial isolate.

Results: 43 bacterial isolates were collected in 34 children aged 1-204 months (mean 86 months). Patient sub-populations include: VUR (17%), NGB  (23%), NNGBD (29%) and no underlying UTI etiology (31%). 14 (41%) children developed clinical signs of pyelonephritis. This was more common with VUR than with NGB or NNGBD. Mean cytokine level of strains with pyelonephritis (41.3 pg/ml) was less than with cystitis (75.0 pg/ml) (p=0.04). The mean IL-6 response of isolates from patients with NGB and NNGBD was significantly higher than isolates from VUR (Figure 1). There was no cytokine difference between NGB and NNGBD patients or between those with VUR and no etiology (Figure 1). IL-6 levels above 65 pg/ml identified patients with a high risk of altered bladder function (90%; 9/10) and a low risk of pyelonephritis (20%; 2/10).

Figure 1.

Conclusion: Children with VUR had the lowest cytokine response and the highest risk of pyelonephritis (83%) suggesting normal bladder function and a bladder microenvironment that was unfavorable to all but the most virulent bacteria that, adept at evading host defenses, elicited only a minimal bladder inflammatory response before gaining access to the kidney. In contrast, NGB and NNGBD, by altering bladder resistance, enabled less virulent bacteria to typically develop cystitis only (88%) where a robust bladder inflammatory reaction confined and limited their ascent. These findings suggest that 1) both NGB and NNGBD can modify the bladder microbial environment to predispose patients to UTI by low virulence organisms, and 2) UTI's caused by bacteria producing high levels of in-vitro IL-6 (>65 pg/ml) appear to identify patients with a low risk of pyelonephritis and a high risk of having altered bladder function.

See more of: Reflux/UTI
See more of: Section on Urology
Previous Abstract | Next Abstract >>