Comparison 0f Ultrasound Findings with Voiding Cystourethrography in Detecting Vesicoureteral Reflux in Children

Background & Aims: Vesicoureteral Reflux (VUR) is a common urinary tract disorder among pediatric population and defined as the retrograde flow of urine from the bladder into the ureters and renal collecting systems due to a failure in the ureterovesical valve function (1). Identifying children with VUR at an early age provides an opportunity to prevent episodes of acute pyelonephritis and the consequent renal scarring (2) . Voiding cystourethrography (VCUG) is the modality of choice for diagnosis and grading of VUR (3). In VCUG, the child is catheterized and radiocontrast material is injected through the catheter to fill the bladder (3). Disadvantages of this procedure are catheterization discomfort, complications and considerable radiation exposure of the children who are relatively more vulnerable to the adverse effects of ionizing radiation. In addition, the risk of carcinogenesis is higher in children as they have a longer life expectancy following the procedure than adults (4). Considering the mentioned disadvantages of VCUG, many attempts have been made to find a non-invasive alternative modality with adequate accuracy to detect VUR. Ultrasound (US) is a proper modality for evaluation pediatric urinary tract system due to its accuracy, accessibility and non-invasiveness (5, 6). However, there is controversy among clinicians regarding the accuracy of US for diagnosis of VUR. VCUG allows grading of VUR using the five-level International Reflux Scale (IRS). Grade of VUR is strongly associated with the outcomes such as spontaneous resolution, recurrence of UTI and renal scarring (7). Recent guidelines recommend clinical decision-making based on the grade of VUR on VCUG, including observation (without medical therapy) for selected children with grade I or II VUR But higher grades of VUR need medical or surgical interventions (7, 8). The diagnostic accuracy of VCUG for diagnosing reflux is very high and nearly 100% for high-grades of reflux (grade III -V) (9). The objective of our study was to determine the accuracy of ultrasound for detecting VUR in comparison with VCUG in children with UTI. Methods: In this cross-sectional study, the total numbers of 540 children younger than  8 years old with UTI referred to Ali-Asghar children’s hospital, a pediatric center in Tehran, Iran were enrolled, from April 2017 to May 2019. All patients underwent US and VCUG to detect VUR. All patients with obstructive nephropathy were excluded from the study. US was performed immediately after diagnosis, and VCUG were performed after the resolution of fever and confirmation of a negative urine culture. US was performed by a single radiologist using a Philips Affiniti 50 ultrasound machine  with 4–7 MHz convex  and  7-10 MHZ linear transducers with the bladder being both full and empty. The most important ultrasonographic findings related to VUR, were dilatation of the renal pelvis or the ureter. Additionally, changes in kidneys size and cortical echogenicity, reduction in the thickness of renal parenchyma, irregularity of the kidneys margin, and increase of urothelial thickening were also noted. Grading system for VUR on VCUG is according to the International Reflux Study Committee: o   Grade I: reflux into the ureter; o   Grade II: reflux into the ureter and renal pelvis without dilatation; o   Grade III: reflux with mild dilatation; o   Grade IV: reflux with moderate dilatation, rounded fornices; o   Grade V: gross dilatation of the ureter, ureter tortuosity, papillary obliteration. Grades I and II were classified as low grade and grades III, IV, and V as high grade reflux (7). We used the mean and standard deviation and percent for reporting the descriptive statistics of quantitative and qualitative variables, respectively. Qualitative variables were compared using the Chi square test and one way analysis of variance (One-way ANOVA) was used to compare the mean of quantitative variables. We used the sensitivity, specificity, positive predictive value, negative predictive value, overall accuracy and kappa agreement coefficient to investigate the efficacy of US for prediction of VUR based on the actual presence or absence of VUR confirmed by VCUG. Data was analyzed using Stata software, version 12 (StataCorp, TX) and p-value <0.05 was considered as the level of significant. Results: Among 572 patients evaluated for VUR, 540 patients entered our study. A total of 269 (49.8%) were boys and 271(50.2%) were girls. All patients underwent VCUG and US. Mean age of children who had VUR was 2.5 years old. Fifty-three cases (37%) had low-grades and 90 cases (63%) had high-grades of VUR. US was abnormal in 97 of 143 patients (67.8%) with confirmed VUR on VCUG and in 163 of 397 cases (41%) without VUR on VCUG. The overall sensitivity and NPV of US for detecting VUR were 67.83% and 37.31%, respectively. Among 90 children with high-grade VUR (grade III–V) on VCUG, 72 (80%) had abnormal findings on US and sensitivity and NPV of US among these cases, were 88% and 93.6 % respectively. All cases of grade V and 83.9% of grade IV VURs had abnormal US findings. Conclusion: There is considerable interest in prompt and early detection of VUR as it is linked to recurrent UTIs, renal scarring and renal insufficiency. VCUG is the modality of choice for detecting VUR (10, 11). However, owing to some disadvantages such as bladder catheterization and pediatric radiation exposure, there is a growing interest in finding alternative and less invasive methods with acceptable accuracy to detect VUR (12). In this study, we evaluated the accuracy of US in predicting VUR among children hospitalized with UTI. Several studies have evaluated the efficacy of US in diagnosis of VUR and their results have been conflicting with some reporting unreliability of ultrasound in evaluation of VUR. Mehnat and colleagues showed that the sensitivity and specificity of US for detecting VUR were 40% and 76%, respectively and demonstrated that renal US was neither sensitive nor specific for detection of VUR in children with a first-time UTI (11). In another investigation, Adibi and colleagues demonstrated the sensitivity, specificity; NPV and PPV of US in diagnosis of VUR were 70.9%, 51.4%, 69.6% and 52.9% respectively. They suggested that US is a sensitive but not specific method in diagnosis of VUR (12). In a review article in 2016, Shaikh N. and colleagues concluded that US could not replace VCUG in detecting VUR. On the other hand, some studies reported that US is a reliable modality for evaluation of VUR (3). Hey-young Lee and colleagues demonstrated that 95.3% of high grade VUR cases could be detected by US . However, they also stated that the diagnosis of VUR by US had some limitations in cases of low-grad VUR and detection ratio of these cases was only 62.5% (13). Similar to other investigations, we found that sensitivity and specificity  of US to detect low-grade VURs are low (respectively 67.83% and 58.94%). However for high-grade VURs, the sensitivity (88%) and NPV (93.6 %) of US are acceptable. Regarding high frequency of spontaneous resolution of low grade VUR while children grow up, it can be recommended that VCUG be performed only in children with abnormal findings on US, avoiding many unnecessary VCUG procedures. Although ultrasound is not sufficiently accurate to detect all grades of VUR, but has enough sensitivity and NPV for ruling out high-grade VUR. So avoiding unnecessary VCUG in children with normal ultrasound finding is recommended.