Gentamicin resistant E. coli as a cause of urinary tract infections in children

Urinary tract infections (UTIs) are common in childhood, constituting 5–7% of cases of fever of unknown origin in infants under 2 years of age.^1,2 The early institution of adequate treatment is crucial in order to prevent kidney lesions.³Escherichia coli is the main uropathogen in childhood (70–90% of cases),^1,2,4 and it is thus important to understand its local resistance patterns in order to select the empirical antibiotic treatment.⁵

Various paediatric protocols consider gentamicin to be a first-line treatment for UTIs requiring parenteral antibiotic therapy.^1,2,6,7 We recently performed a retrospective review of UTI cases diagnosed at the children's emergency department of one tertiary hospital between January and December 2014. We included significant growths in urine collected using suprapubic aspiration (any amount), catheterisation (>10⁴ colony-forming units [CFUs]), spontaneous urination (>10⁵ CFUs) or an adhesive bag (>10⁵ CFUs of the same germ in at least two bags) in children under 14 years of age, excluding untreated asymptomatic bacteriuria. In cases caused by gentamicin-resistant E. coli, we recorded the existence of factors that could favour this resistance,^6–8 such as a history of prematurity, nephro-urological malformations or other chronic diseases, recent hospital admissions, stays in intensive care, previous UTIs or receiving antibiotic prophylaxis. 78% of the isolates were caused by E. coli (201/258), 15% of which were gentamicin-resistant (30/201). 25 were considered UTIs with significant growth, 15 by spontaneous urination (60%), nine by catheterisation (36%) and one in two urine collection bags (4%). The 25 episodes occurred in 21 children, with a median age of 12.4 months (interquartile range: 3.6–23.4), requiring hospitalisation on 13 occasions (52%). Resistance risk factors were observed in 13 patients (61%): 8 (38%) with nephro-urological disease, 7 (33%) had had previous UTIs, 5 (23%) were receiving antibiotic prophylaxis, 5 (23%) had been hospitalised in the six months prior, 3 (14%) were in intensive care and 5 (23%) had other diseases (encephalopathy, Down's syndrome, congenital CMV infection and heart disease). 57% (12) presented more than two risk factors, and 28% (6) presented more than three. 44% of the isolates combined resistances to amoxicillin/clavulanic acid and cefuroxime (11), 36% to cefotaxime (9) and 32% to ciprofloxacin (8). 28% (7) produced extended-spectrum beta-lactamases (ESBL) and 8% (2, from the same patient) VIM-type carbapenemases. 90% of the resistant strains subjected to a susceptibility analysis were sensitive to amikacin (9/10).

If this trend were confirmed in children with risk factors, gentamicin might no longer be the empirical treatment of choice in UTI cases among these patients as, in order to select an empirical treatment, we must consider that potential aetiological agents may not present resistances of over 10–20%.⁵

We noted that these strains combined resistances to drugs such as cephalosporins and ciprofloxacin, as mentioned previously in other studies.^6–9 The increase in cephalosporin resistances is particularly relevant as they constitute the routine outpatient treatment and the treatment administered to children with kidney failure.²

Rising rates of ESBL-producing strains have recently been published in community UTIs.^9–11 As in our series, these resistances are observed in patients with risk factors.^6–8,12 Similarly, other authors have observed high rates of gentamicin resistance in community-acquired UTIs caused by E. coli.⁶

However, hospitals in our geographical area have published high rates of gentamicin sensitivity in E. coli strains that cause community UTIs in children (95.6%).² As our hospital is a referral centre for certain paediatric diseases, the patients treated at our emergency department often present with comorbidities, histories of long hospital stays and exposure to broad-spectrum antibiotics. In view of our results and those presented by other groups, it is becoming increasingly important to adapt antibiotic strategies according to these risk factors.

In our study, we observed that amikacin could be an excellent empirical UTI treatment option in these patients within our setting. Other authors have already proposed the use of amikacin in this population, given its adequate coverage for other common uropathogens and its excellent diffusion to the renal parenchyma.^7,9

In conclusion, we consider it essential to continue the epidemiological surveillance of UTI-causing strains in risk populations. These factors must be assessed on establishing an empirical treatment to avoid complications and treatment failures.