Imaging guidelines for urinary tract infection in childhood; time for change?

D espite the frequency of urinary tract infection (UTI) in childhood, and the numerous contributions on the subject in the literature, there is surprisingly little consensus on urinary tract imaging requirements, perhaps reflecting the paucity of high quality intervention studies and data on long term outcome. Diagnostic imaging following UTI in childhood has been accepted practice for nearly 40 years since the original studies showed a high prevalence of abnormalities, and specifically a link between renal scarring and vesicoureteric reflux (VUR). Inherent in this strategy was the assumption that identification of these abnormalities would influence outcome. In the intervening decades, much has been learned about additional risk factors for post-infection renal scarring, such as obstructive uropathy, recurrent febrile UTI, particularly in the infant and young child, diagnostic delay, inadequate treatment, dysfunctional voiding, the host inflammatory reaction, as well as factors specific to the infecting bacterium. In addition, clear gender specific differences in renal scarring have been established, with global parenchymal loss occurring more often in boys with dilated VUR, and focal scarring in the absence of VUR being seen more often in girls, particularly after recurrent febrile UTI. These findings support the contention that high grade (4 and 5) VUR is associated with renal dysplasia that occurs during fetal life with UTI as a risk factor for postnatal renal functional decline. Fortunately the great majority of children with UTI have an excellent prognosis, and although more long term follow up data are necessary, it appears that the risk of renal functional impairment and/or hypertension is much lower than previously thought, 7 and likely to be limited mainly to those with renal dysplasia. Intervention studies in patients with VUR have shown no benefit of antireflux surgery over medical treatment in UTI recurrence, renal function, or new scar formation. However, low dose antibiotic prophylaxis is the commonest intervention used to prevent UTI, but a recent review highlighted the fact that this intervention has not been appropriately studied in patients with VUR. Initial imaging guidelines for febrile children up to the age of 2 years by the American Academy of Pediatrics, involve a combination of ultrasound (US), contrast (MCUG), or isotope cystography. UK guidelines 12 are similar for the infant and young child, but also include renal cortical scintigraphy with technetium-99m dimercaptosuccinic acid (DMSA). Guidelines for the older child reflect differing professional opinion, predominantly on the need for scintigraphy. In the light of the increasing use of prenatal US, Zamir and colleagues, in this issue, have tested the value of routine diagnostic imaging and specifically questioned whether US is necessary. The study involved a prospective assessment of the value of US in 255 children under the age of 5 years with a first diagnosed uncomplicated febrile UTI. Abnormalities on US were found in 14.1% of patients, but in none did this influence management. MCUG revealed VUR in 47 (18.4%) of patients, grade 1–3 in all but one patient. No scintigraphy was carried out. This report and a similar one published early last year, confirm the previously recognised low sensitivity of US for VUR. The authors suggest that in the patient who has a normal late prenatal US, further US should be limited to those who have a complicated course and MCUG should be used as the sole screening test. However, apart from the invasive nature of MCUG and the reluctance of many radiologists to undertake this study beyond infancy, this protocol is predicated on the unproven assumption of the value of prophylaxis in patients with VUR. In addition, VUR is of low grade in the majority of patients and associated with a low risk of parenchymal scarring. Both studies referred to above involved infants and young children with febrile UTI, and whether the same recommendations would apply to the older child remains an open question. Despite the availability and recognised high sensitivity of renal cortical scintigraphy, the use of this modality in imaging practice is variable. Some advocate DMSA scanning as a first line investigation in lieu of US because of the greater sensitivity in identifying acute pyelonephritis (APN), although it is difficult to justify this approach unless confirmation of APN is shown to alter initial management and subsequent outcome. There is nevertheless some logic in adopting an imaging strategy beyond the acute phase, that defines the presence or absence of renal scarring and/or dysplasia as this is associated with long term risk, albeit small. However, as renal scintigraphy is invasive there is a need for a controlled application dictated by factors such as age, recurrence, and the presence of fever/systemic upset, accepting that some of the abnormalities found may have no long term significance. One must exercise some caution before dispensing with US in the child with uncomplicated UTI. Prenatal anomaly scanning is by no means universal in the UK and for maximum sensitivity for the detection of a dilated fetal urinary tract the scan would have to be carried at or greater than 20 weeks gestation. In addition, US is of value in demonstrating bladder wall morphology and voiding efficiency in the older child with UTI and day wetting. In view of the link between referral and imaging, one would have to ensure that other important aspects of the consultation, particularly simple preventive advice, are not lost merely because imaging may be considered unnecessary. Finally, the debate on imaging has to a significant extent diverted attention from the importance of early and accurate diagnosis and treatment and care must be taken to avoid minimising this aspect of management. In conclusion, the existing encouraging long term outcome data, particularly if confirmed, along with the advent of controlled trials on prophylactic antibiotic therapy and universal prenatal anomaly scanning, should encourage the development of imaging guidelines that are both targeted and less invasive.