Although VCUG and radionuclide cystography are presently the examinations of choice for the evaluation of VUR, radiation exposure from both these techniques is still a cause of worry. Also, there is other disadvantages of radioisotope cystography include bladder catheterization, poor anatomical details of the urethra, and vesicoureteric junction. So, every effort should be made towards minimization of radiation exposure during evaluation of VUR in infants and children. In this regard, US is the most advantageous alternative because it is a safe, simple, and widely accessible modality with no radiation exposure to rule out residual hydronephrosis and examine the Deflux injection site. [16] [17]
In our study, US detected Deflux mounds at the base of the urinary bladder in 200 ureters (91%) with sensitivity of implant visualization on bladder US in predicting complete VUR resolution on postoperative VCUG 86.6%, specificity 54%, PPV 93%, NPV 35%, and accuracy 83% at last follow-up. Park et al [18] studied 56 ureters (36 children) and found Deflux mounds on bladder US in 38 ureters (68%). The sensitivity was 73%, specificity was 44%, PPV was 76%, and NPV was 39%. In comparison, our study showed higher sensitivity, specificity and PPV; this may be explained by the comparatively higher number of patients included in our study. NPV was found to be similar to that reported in their study.
Lee et al [19] examined 149 patients (220 ureters), and of them, 122 patients (82%) were injected with Deflux and 27 patients (18%) were injected with polydimethylsiloxane (Macroplastique). Bladder US detected injection implants in 152 ureters (69%). VUR was cured in 128 ureters (84%). The success rate in visualized injection implants group was 84.2%, while that in absent injection implants group was 48.5%. Bladder US had 79.5% sensitivity, 59% specificity, 84% PPV, 51.5% NPV, and 74% accuracy as a diagnostic tool for successful correction of VUR. Injection implants were detected by US in only 69% ureters compared to 91% ureters in our study. This may be explained by the use of Macroplastique and Deflux injections in their study, while all patients in our study were administered Deflux injections only. Also, US is an operator-dependent modality, and the results may differ depending on the operator or the machine used in the examination.
In our study, the number of US identified Deflux implants decreased from 91% at first follow-up to 82% at last follow-up. This finding is in line with Özcan et al [16] study, who described reduced rates of US detection of implants during follow-up from 80% in the first postoperative month to 45% at a mean follow-up of 2.2 years. He suggested that reduction in implant volume could be due to the replacement of dextranomer particles by collagen.
In our study, Bladder US could not detect Deflux mounds in 20 ureters (9%) at first follow-up; 12 (5.4%) of these ureters had corrected reflux and 8 ureters (3.6%) had persistent reflux. At last follow-up, the Deflux mounds were not detected in 40 ureters (18%); 26 (11.8%) of these ureters had corrected reflux and 14 ureters (6.4%) had persistent reflux. Our study showed low specificity and NPV for the identification of Deflux implant on bladder US and prediction of VUR resolution. So, the non-visualized Deflux implant could not be diagnostic option of persistent VUR. This finding has also been described in previous reports. Park et al stated that low specificity of non-visualized implants on bladder US for persistent VUR could be explained by ureteral orifice anti-reflux mechanism caused by tissue expansion resulting from endogenous collagen between microspheres. [18] Lee et al and Yucel et al attributed this to treatment‑associated factors such as surgeon skills, injection technique, injection mound location, injectable volume, US machines used, and operator-dependent factors. [19] [20]
Also our study is in line with Park et al [18] findings that stated VCUG may not be necessary if bladder US identified Deflux implants in the correct position and of good size. VCUG could be indicated when the implant is not identified on bladder US or is of small size.
Our study concluded that, Deflux injection cured reflux in 98% of ureters in low-grade VUR and 26.6% of ureters with high-grade reflux at last follow-up. Thus, postoperative US could be adjusted as a screening examination in the follow up of children treated for low grade VUR and can select patients with high grade VUR who will require subsequent VCUG to exclude persistent VUR especially if the implant is not clearly visualized and not has good volume. Our protocol was thus modified in the last few years to using US only at the follow-up of low-grade cases. Limitations of the present study include possible sampling bias due to retrospective nature of the study. Also, validation of the sonographic appearance and volume of the implant as a prognostic factor for treatment success in low-grade VUR cases may be necessary.