Long term prognosis of children with urinary tract infection having normal urinary system ultrasonography and pathologic DMSA scintigraphy findings

Aim: Urinary tract infections (UTI) might lead to renal scar formation and related complications. Dimercaptosuccinic acid (DMSA) scintigraphy is the gold standard for demonstration of renal scarring. However, DMSA scintigraphy is an invasive procedure and exposes children to radiation hazard. Children with UTI and abnormal urinary system USG are recommended to be evaluated by voiding cystourethrography (VCUG) and DMSA scintigraphy. Yet, children with normal urinary system by USG might still have abnormal DMSA scans. We aimed to compare the demographic, clinical, laboratory and VCUG findings of children with UTIs, normal USG and normal DMSA scan with those having normal USG but abnormal DMSA scan. Patients and Methods: Hospital files of children presenting to Dokuz Eylül University Medical Faculty Pediatric Nephrology Department due to UTI during the years 2009-2019 were evaluated retrospectively. Those with normal urinary system USG and evaluated by at least one DMSA scan were enrolled in the study. Patients were grouped as those having normal or pathologic DMSA scan according to the first DMSA scintigraphy results. Both groups were compared for age, gender, fever at presentation, body mass index (BMI), blood pressure, blood urea nitrogen (BUN), serum creatinine, C-reactive protein (CRP), erythrocyte sedimentation rate (ESR), white blood cell count, urinalysis, urine culture, vesicoureteral reflux (VUR), prophylactic antibiotic treatment and surgery requirement. In addition, patients with pathologic initial DMSA scans who were evaluated by a second DMSA scintigraphy were grouped as those with resolved or persistent pathologic DMSA findings. These groups were also compared for the same parameters. Statistical analyses of continuous variables were performed by student-t test and Mann Whitney U test with respect to case numbers in each group as ≥30 or <30, respectively. Univariate analyses was performed using the odds ratio (OR) and chi-square test to evaluate the association between binary clinical and laboratory findings and pathological DMSA scan. To evaluate the accuracy of significant continuous predictive variables, we used the area under the receiver operation characteristic curve (ROC area). Results: There were 147 patients [122 (83%) female 25 (17%) male; mean age 61.7±46.6 months (1-203)] fulfilling the inclusion criteria. Serum creatinine and BUN levels were within normal limits in all patients. Patients with pathologic initial DMSA scan (n=58) were older (72.0±46.7 vs 55.0±45.6 months, p=0.030), had higher BUN (11.4±3.7 vs 9.7±3.4 mg/dL, p=0.005), creatinine (0.43±0.14 vs 0.38±0.12 mg/dL, p=0.013) and ESR (28±13 vs 16±14 mm/h, p=0.025) compared to those with normal initial DMSA scans (n=89). CRP was also higher in pathologic DMSA group, but the difference was not significant (35.8±53.0 vs 20.0±47.0 mg/L, p=0.067). However, the rate of increased CRP (>5 mg/L) was higher in this group (31/55 vs 31/85, p=0.021; OR 2.25). VUR frequency was also higher in those children with pathologic DMSA scan (29/49 vs 19/50, p=0.035; OR 2.36). Presence of VUR was determined to be the most significant parameter predicting pathological DMSA scan by logistic regression analysis. Prophylactic antibiotic treatment (52/58 vs 67/89, p=0.030) and antireflux surgery requirement (20/58 vs 13/89, p=0.005) were higher in patients with pathologic DMA scan. Patients with persistent pathological findings in the control DMA scan (n=30) had lower BMI SDS (-0.29±1.60 vs 0.86±1.24, p=0.033), lower urine specific gravity (1013±6 vs 1018±5, p=0.025) and higher CRP levels (37.4±42.4 vs 14.0±34.2 mg/L, p=0.027) compared to those with resolved findings (n=9). The rate of patients with increased CRP (>5 mg/L) was higher in this group (21/28 vs 2/9, p=0.005; OR 10.5). Cut-off level of CRP for predicting persistent pathological DMSA findings was determined to be 5.2 mg/L by ROC analysis (sensitivity 75.0%, specificity 77.8%). VCUG was performed in 30 out of 46 children under 2 years of age and 19 of them (63%) had VUR. While DMSA scan was pathologic in 6 (32%) of those with VUR, only 1 of 11 (9%) children without VUR had pathological DMSA scan. Conclusion: High CRP and presence of VUR increase the probability of pathologic DMSA scan findings in children presenting with UTI and normal urinary system USG. Persistent scintigraphic pathology probability increases 10.5 times in those with increased CRP. These data indicate that ordering DMSA scintigraphy during follow up of children with UTI despite normal urinary system USG is feasible if CRP level is increased at presentation. On the other hand, two thirds of children under 2 years of age with normal urinary system USG had VUR and one third of these children has pathologic DMSA scan. This finding contradicts the American Academy of Pediatrics's suggestion that "if urinary system ultrasonography is normal in children under 2 years of age with UTI, no additional examination is required." Key words: Children, Dimercaptosuccinic acid scintigraphy, ultrasonography, urinary tract infection, vesicoureteral reflux