| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Pediatric Imaging |
1 From the Departments of Radiology (H.J.P., L.A.D.), Urology (D.A.D., A.A.), and Biostatistics (D.Z.), Children’s Hospital, Harvard Medical School, 300 Longwood Ave, Boston, MA 02115. Received April 10, 2003; revision requested June 25; final revision received December 16; accepted January 29, 2004. Address correspondence to H.J.P. (e-mail: harriet.paltiel@childrens.harvard.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXREFERENCES |
|---|
MATERIALS AND METHODS: Sonograms obtained at mean intervals of 1.4 and 12 months after treatment were compared with cystograms obtained at mean intervals of 2.7 and 12 months. Sonograms were analyzed for chondrocyte mound contour and volume, changes in mound volume over time, and presence of hydroureteronephrosis. Mean differences in mound volume were detected with paired t tests in 14 patients with early and late sonograms available for review.
RESULTS: Early cystograms and sonograms were available for 25 of 32 patients (45 of 56 treated ureters). Reflux occurred in 16 of 45 ureters. In 16 ureters, chondrocyte mounds were absent in six, unilobed in seven, and multilobed in three. In 29 treated ureters without reflux, mounds were unilobed in 28 and multilobed in one. Three patients had mild hydronephrosis and one had distal hydroureter; these conditions resolved. Late cystograms and sonograms were available in 18 of 22 patients (30 of 38 treated ureters). Reflux occurred in seven of 30 ureters. In seven ureters, mounds were unilobed in five and multilobed in two. In 23 treated ureters without reflux, mounds were unilobed in 21 and multilobed in two. One patient had mild hydronephrosis that resolved. In 14 patients with early and late sonograms available for review, mean mound volume in late group (0.37 cm3± 0.25 [standard deviation]) was significantly smaller than that in early group (0.56 cm3± 0.39; P = .004, paired t test).
CONCLUSION: Absence of chondrocyte mound or presence of multilobed mound contour was associated with persistent reflux. Mean mound volume decreased over time. Treatment-induced hydroureteronephrosis was uncommon and self-limited.
© RSNA, 2004
Index terms: Bladder, abnormalities, 83.85 • Bladder, US, 83.1298 • Ultrasound (US), in infants and children, 82.1298, 83.1298 • Ureter, interventional procedures, 82.1269 • Ureter, reflux, 82.85
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXREFERENCES |
|---|
Surgical correction is generally recommended in children with severe reflux, children with recurrent febrile urinary tract infections despite antibiotic treatment, prepubertalgirls with persistent reflux, and children in whom there is concern in regard to compliance with a medical regimen and adequate follow-up. Endoscopic treatment of reflux is an attractive therapeutic option that can be used to avoid long-term antibiotic administration, as well as the potential sources of morbidity associated with open surgery, such as abdominal and bladder incisions, several hours of general anesthesia, and a hospital stay of several days. In children, the endoscopic procedure is performed on an outpatient basis in less than 15 minutes with a general anesthetic, with minimal postoperative morbidity and rapid recovery.
The proposed mechanisms by which endoscopic injection corrects reflux include improved fixation of the ureter to the trigone, coaptation of the distal ureter, and creation of a solid support behind the intravesical portion of the ureter (7). The injected material should be stable and should not migrate. Endoscopic treatment of reflux was reported by Matouschek (8) in 1981 when he performed subureteric injection of polytetrafluoroethylene paste in a patient in whom open antireflux surgery had failed. Subsequent experimental and clinical demonstration of polytetrafluoroethylene particulate migration to distant organs such as the lung and brain has precluded widespread acceptance of this substance (9–12), and injectable alternatives to polytetrafluoroethylene have been sought.
Many materials and endoscopic delivery systems have been proposed, including particulate silicone microimplants, autologous and nonautologous collagen, bioactive glass (made from calcium salts, phosphorus, sodium salts, and silicon) that bonds with living bone tissue and repairs and promotes bone regeneration (Bioglass; Department of Materials Science and Engineering, University of Florida, Gainesville), dextranomer microspheres, detachable balloons, bladder muscle, and autologous fat (13–20). An ideal method, however, has not yet been identified. Bovine collagen has been used with few complications, but progressive absorption of collagen and concomitant loss of the antireflux effect have been documented (21–24). In 1994, Atala et al (25) reported the successful use of autologous chondrocytes in minipigs without untoward side effects and with maintenance of the cartilage bead over time. The potential advantages include a sustained antireflux effect from viable chondrocytes, as well as an avoidance of biocompatibility risks because of their autologous nature.
Our hospital participated in a clinical trial in which autologous chondrocytes were used to treat reflux (26,27). The purpose of the current study was to analyze the sonographic imaging features of the renal collecting systems, ureters, and bladder following transurethral subureteric injection of autologous chondrocytes for treatment of reflux and to determine whether any features are associated with persistent reflux.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXREFERENCES |
|---|
Surgical Procedure
A two-stage approach was used. In each stage, the procedure was performed by one of the authors (D.A.D.), a board-certified urologist with 17 years of experience, with the patient receiving a general anesthetic in an outpatient setting. In stage 1, cystoscopy was performed to rule out an anatomic contraindication to the endoscopic technique, such as a paraureteral diverticulum or an occult ureterocele. Two 0.5 x 0.5-cm samples of posterior auricular cartilage were harvested. Chondrocyte cells were grown in culture for 6 weeks until a concentration of 20 x 106 chondrocytes per cubic centimeter was achieved.
In stage 2, the chondrocytes were prepared in an alginate solution. Alginate is a hydrogel composed of glucuronic acid and manuronic acid that is designed to gel with the addition of calcium. Cells were cross-linked with calcium salts to produce an injectable pastelike substance immediately prior to injection with a 5-F needle through a pediatric cystoscope. The needle was placed beneath the lip of the ureter in the submucosal plane at the 6-o’clock position. A volume of chondrocyte suspension was injected that was sufficient to produce a nipple-shaped mound surmounted by a slitlike or inverted crescentic ureteral orifice, as described by O’Donnell and Puri (7) (Fig 1).
|
|
Reports obtained at pre- and postoperative fluoroscopic voiding cystourethrography and radionuclide cystography were retrospectively reviewed by the first author, and the degree of reflux was recorded (Appendix) (28). In addition, all of the voiding cystourethrograms and radionuclide cystograms were reviewed by one of the other authors (D.A.D.). Sonograms were retrospectively assessed by the first author for mound contour and volume and for the presence of hydroureteronephrosis. Review of the pre- and postoperative fluoroscopic cystourethrograms and radionuclide cystograms was performed (D.A.D.) with knowledge of the patient’s history and outcome. Review of the sonograms was performed by the first author prior to review of the medical history, which included injected chondrocyte volume and number of injections.
Patients ranged in age from 1.25 to 15 years at the time of treatment, with a mean age of 7.64 years ± 3.3 (standard deviation). The six boys ranged in age from 4 to 11 years, with a mean age of 7.2 years ± 2.6. The 26 girls ranged in age from 1.25 to 15 years, with a mean age of 7.70 years ± 3.4. Fifty-six ureters were treated, and 53 of them demonstrated reflux before surgery. Three ureters that did not demonstrate reflux at pretreatment cystography were treated with chondrocyte injection at the time of treatment of the contralateral ureter because of an abnormal appearance of the ureteral orifice at cystoscopy. In two ureters with grade I reflux prior to treatment, chondrocytes were injected at the same time as they were injected into the contralateral ureter, which in each case demonstrated a higher degree of reflux.
In accordance with the dictates of the clinical trial, each patient underwent early voiding cystourethrography or radionuclide cystography at a mean interval of 2.7 months (range, 1–4 months) after treatment and early sonography at a mean interval of 1.4 months (range, 1–6 months) after treatment. A late voiding cystourethrogram or radionuclide cystogram and a late sonogram were obtained at a mean interval of 12 months (range, 12–16 months) after treatment in 22 patients. Ten patients with postoperative reflux identified by using early voiding cystourethrography or radionuclide cystography received a second endoscopic treatment with autologous chondrocytes between 3 and 9 months (mean, 5 months) after the first treatment and therefore did not undergo late studies. The results of early and late voiding cystourethrography, radionuclide cystography, and sonography performed in patients following a single endoscopic procedure form the basis of the current study.
Early posttreatment voiding cystourethrograms and radionuclide cystograms were available for review in all 32 treated patients with 56 treated ureters. Early posttreatment sonograms were available for review in 25 (78%) of 32 patients and 45 (80%) of 56 ureters, since the images in seven patients were inadequate (the bladder was empty or hard-copy images were missing).
Late posttreatment voiding cystourethrograms and radionuclide cystograms were available for review in 22 patients with 38 treated ureters. Late posttreatment sonograms were available for review in 18 (82%) of 22 patients and 30 (79%) of 38 ureters. Sonograms in four patients were inadequate (the bladder was empty or hard-copy images were missing). Fourteen patients with 25 treated ureters and adequate early and late sonograms were included in both the early and late study group analyses. Three patients with early and late images were included in the early study group analysis but were excluded from the late study group analysis because of inadequate late sonograms. Four patients with early and late images were excluded from the early study group analysis because of inadequate early sonograms but were included in the late study group analysis. One patient with early and late images was excluded from both the early and late study group analyses because of inadequate early and late sonograms.
Cystography
Fluoroscopic voiding cystourethrography and radionuclide cystography were performed in a standardized fashion in all patients by multiple individuals with varying levels of experience. Fluoroscopic voiding cystourethrography was performed with the patient in the supine position. Following bladder catheterization, 17.2% iothalamate meglumine injection (Cysto-Conray II; Mallinckrodt, St Louis, Mo) was instilled by means of gravity drip through intravenous tubing. The bladder was intermittently imaged until it was full. Patients were turned in both oblique projections to visualize the ureterovesical junctions for reflux or other abnormalities. The urethra was observed during voiding in an oblique projection for boys and in the anteroposterior projection for girls. If reflux was present, the level and amount of distention were documented. Spot images of the bladder and kidneys were obtained at the end of voiding.
Radionuclide cystography was performed following bladder catheterization with injection of 2 mCi (74 MBq) of technetium 99m pertechnetate as a bolus into the closed hub of intravenous tubing connected to the catheter. The catheter, in turn, was attached to a bottle of normal saline that dripped into the bladder by means of gravity. With the patient lying in the supine position on the imaging table, a gamma camera was positioned to view the back of the patient. The renal and vesical areas were included within the field of view of the gamma camera. The radionuclide cystogram was recorded as a series of 10-second frames in a 128 x 128-matrix format for the duration of the filling and voiding phases of the study. Postvoiding images were routinely obtained. The degree of reflux was subsequently determined by using the classification outlined as follows.
Reflux documented with voiding cystourethrography or radionuclide cystography was graded according to the International Reflux Study Committee classification (Appendix) (28), which is based on the morphologic appearance of the collecting system and ureter at fluoroscopic voiding cystourethrography. A direct comparison of grading of reflux at voiding cystourethrography and at radionuclide cystography is difficult because of inherent technical differences between the two imaging methods.
It is possible, however, to recognize at least three degrees of reflux severity at radionuclide cystography. The mildest degree involves reflux limited to the ureter without reaching the renal pelvis; this degree corresponds to grade I reflux according to the International Reflux Study Committee classification. The second degree involves a small volume of reflux that reaches the renal pelvis with minimal or no visualization of the ureter. This appearance corresponds to international classification grades II and III, since it is not possible with the radionuclide technique to precisely assess ureteral diameter or pelvicalyceal system anatomy. The third degree involves a large volume of reflux within a dilated collecting system with appreciable dilatation and tortuosity of the ureter; this degree corresponds to international classification grades IV and V (29).
Sonographic Evaluation
Sonography was performed by multiple sonographic technologists with the supervision of board-certified radiologists (H.J.P.) with varying levels of experience. Gray-scale sonograms of the bladder and kidneys were obtained in transverse and longitudinal planes by using sector transducers with frequencies varying between 2.0 and 7.5 MHz. The transducer power output was set at less than 100 mW/cm2 in accordance with acoustic output guidelines for sonographic studies of the Food and Drug Administration. Studies were recorded on film.
Chondrocyte mounds were retrospectively classified as present or absent, and contours of the mounds were categorized as unilobed (Fig 2) or multilobed (Fig 3). Since the subjective impression of the image reviewer (H.J.P.) was that the mounds were ellipsoid, mound volume was determined with the formula for an ellipsoid as follows: L · W · T · 0.52, where L is length, W is width, and T is thickness. This formula was also employed by Frey et al (14) in their evaluation of subureteral collagenous implants.
|
|
|
Statistical Analysis
Continuous variables were checked for normality with the Kolmogorov-Smirnov test, and no skewness was detected (30). Therefore, data are expressed in terms of the mean ± standard deviation. A two-sample Student t test was used to compare differences in age between boys and girls, differences in injected chondrocyte volumes, and differences in calculated bladder mound volumes between ureters with and ureters without reflux in both the early and late study groups. The relationship between mound contour and the number of chondrocyte injections performed at the time of endoscopy was evaluated with the Pearson 
2 test of independence.
Chondrocyte bladder mound volume was compared between early and late sonograms with the paired t test. The Fisher exact test for binomial proportions was used to determine whether the proportion of ureters with posttreatment reflux was different between ureters without associated mounds compared with those with unilobed or multilobed mounds and to compare a multilobed mound contour with a unilobed mound contour between treated ureters with reflux and those without reflux (31). A power analysis indicated that a sample size of 25 ureters would provide 90% power (
= .05, ß = .1) to detect a mean difference of 0.20 cm3 in chondrocyte mound volume between early and late sonograms with a paired t test (nQuery Advisor, version 4.0; Statistical Solutions, Boston, Mass).
Differences in injected chondrocyte volumes and differences in calculated bladder mound volumes between ureters with and ureters without reflux were compared in the subset of 14 patients who were included in both the early and late study group analyses. The Fisher exact test for binomial proportions was used to compare a multilobed mound contour with a unilobed mound contour between treated ureters with reflux and those without reflux in this subset of patients. Statistical analysis of the data was performed by using software designed for this purpose (SPSS, version 11.0; SPSS, Chicago, Ill). All reported P values are two tailed.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXREFERENCES |
|---|
Posttreatment images for the early group included 10 voiding cystourethrograms and 22 radionuclide cystograms, and those for the late group included 11 voiding cystourethrograms and 11 radionuclide cystograms. Table 1 shows a comparison of the number of ureters with reflux and reflux severity prior to and following treatment for both the early and late study groups. There was persistent posttreatment reflux in 22 (39%) of 56 treated ureters in the early study group and in 10 (26%) of 38 treated ureters in the late study group. In addition, two untreated ureters in the early study group and one untreated ureter in the late study group demonstrated new grade I reflux following chondrocyte injection of the contralateral ureter.
|
|
2 = 13.43, df = 2, P < .001). Absence of a chondrocyte mound was associated with persistent reflux (P < .001, Fisher exact test). A chondrocyte mound was detected sonographically in 10 of 16 treated ureters with reflux in the early group and in all seven treated ureters with reflux in the late group. A multilobed mound was present in five (29%) of these 17 ureters. In contrast, in 52 treated ureters without reflux (29 in the early group and 23 in the late group), a multilobed mound was present in only two (4%) ureters (P = .008, Fisher exact test).
Early posttreatment sonograms demonstrated mild unilateral renal pelvic dilatation in four patients; one of these patients had ipsilateral posttreatment grade II reflux and another had mild unilateral distal hydroureter without associated reflux. These sonographic abnormalities all resolved at follow-up. Late posttreatment sonograms revealed mild unilateral renal pelvic dilatation without associated reflux in one patient, and this dilatation resolved at follow-up 1 year later.
Table 3 provides a summary of the injected chondrocyte volume, calculated mound volume, and mound contour in the subset of 14 patients (25 ureters) whose images were included in both the early and the late group analyses.
|
A chondrocyte mound was detected sonographically in all five of the treated ureters with reflux in the early group and in all four of the treated ureters with reflux in the late group. A multilobed mound was present in two (22%) of these nine ureters. In contrast, in the 41 treated ureters without reflux (20 in the early group and 21 in the late group), a multilobed mound was present in only two (5%) ureters (P = .14, Fisher exact test). Therefore, there was a significant difference in the presence of multilobed mounds between treated ureters with reflux and those without reflux for the combined early and late groups.
Two treated ureters demonstrated reflux at both early and late studies; three treated ureters with reflux at early studies demonstrated resolution at late studies, and two treated ureters without reflux at early studies showed reflux at late studies. One patient with a multilobed mound contour had reflux at the early study that resolved at the late study, although the multilobed contour persisted. One patient with a unilobed mound and no reflux at the early study had ipsilateral reflux that was associated with a multilobed mound contour at the late study.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXREFERENCES |
|---|
The role of endoscopic surgery compared with ureteral reimplantation in the treatment of reflux remains controversial because of suboptimal success rates. Endoscopic therapy offers a viable alternative to long-term antibiotic prophylaxis and to open surgery, as long as treating physicians and parents are aware of its limitations. The method is straightforward and can be completed in an outpatient setting in less than 15 minutes, with low morbidity. Endoscopy may be performed more than once on the same ureter if necessary, and the procedure does not preclude future performance of open antireflux surgery.
Since 1984, when O’Donnell and Puri (33) demonstrated that primary reflux could be effectively treated endoscopically, pediatric urologists and surgeons have awaited the development of an optimal injectable substance for this purpose. The polytetrafluoroethylene paste originally used proved effective, with success rates of greater than 85% after two injections in the largest reported series of 12,251 ureters (34). Polytetrafluoroethylene particles, however, have been shown to migrate to distant organs and to provoke a large-cell granulomatous reaction in some patients (9–12). Since the overriding concern in children is safety, use of polytetrafluoroethylene has been avoided by most American practitioners (35).
The potential advantages of autologous chondrocyte injection include a sustained antireflux effect from viable chondrocytes, as well as prevention of bioincompatibility due to their autologous nature. In our study, reflux resolved after one chondrocyte injection in 34 (61%) of 56 ureters in the early follow-up group and in 28 (74%) of 38 ureters in the late follow-up group. These results are comparable to those of other authors who have used alternative injectable substances. The higher percentage of reflux resolution in the 1-year follow-up group is, at least in part, explained by the fact that 10 of the patients with persistent postoperative reflux were excluded from the late study group. In a recent publication by Oswald et al (36), the researchers compared the effectiveness of a single injection of polydimethylsiloxane (Macroplastique) with that of dextranomer–hyaluronic acid copolymer (Deflux), a new biodegradable substance. There was no evidence of reflux at 3 months in 76% of ureters treated with polydimethylsiloxane and in 63% of ureters treated with dextranomer–hyaluronic acid copolymer. At 1 year, there was a slight decline in the reflux resolution rates in both groups.
As reported by Caldamone and Diamond (27), patients in whom treatment with autologous chondrocytes failed had evidence of volume loss, misplaced material, and shifting of the subureteral mounds at cystoscopy. Shifting was typically toward the bladder neck, although in some cases the material had shifted medially or laterally from the 6-o’clock position at the orifice. In the four patients who underwent open ureteral reimplantation because of failed endoscopic implantation, histologic sectioning revealed calcified alginate, without evidence of viable chondrocytes in the excised specimens, and a varying granulomatous reaction.
The use of sonography in the evaluation of endoscopic therapy for reflux has been the subject of several reports (37–41). Sonography is an ideal technique with which to image the ureteral orifice and to exclude upper urinary tract obstruction. Although the urinary tract may also be successfully imaged with computed tomography (CT) or magnetic resonance imaging, use of these alternate modalities is expensive and impractical. In the case of CT, it would also entail unnecessary radiation exposure.
To the best of our knowledge, this report is the first to include an analysis of the sonographic imaging features of the urinary tract following transurethral subureteric injection of autologous chondrocytes and the first attempt to associate differences in subureteric mound contour with the presence or absence of posttreatment reflux. As noted by Gore et al (38), adequate bladder distention was necessary for satisfactory mound visualization. Mann et al (39) prospectively used sonography to demonstrate subureteric mounds, with a sensitivity of 100%. In our retrospective study, several patients were eliminated from consideration, since sonograms were obtained with poorly filled bladders, and the mounds could not be identified.
The lack of correlation between the volume of injected chondrocytes and the calculated volume of the cartilaginous mounds is in accordance with the prior observations of Blake and O’Connell (40) and Rypens et al (41) in regard to patients treated with polytetrafluoroethylene paste. These authors speculated that the size of the subureteric mound is probably determined by individual tissue reaction, and this may play a role in the size of the mound with chondrocyte injections as well. Polytetrafluoroethylene particles elicit a foreign-body giant-cell granulomatous reaction. Although Atala et al (25,42) reported no evidence of granuloma formation at histologic analysis of experimental chondrocyte injection sites in animals, the histologic analysis performed by Caldamone and Diamond (27) following a trial of chondrocyte injection in humans demonstrated the presence of a varying granulomatous reaction.
As previously noted by Gore et al (38) and Blake and O’Connell (40), we also determined that the size of the subureteric mound bore no direct relationship to the success of the procedure. Interestingly, in the series of patients included in the study of Blake and O’Connell, posttreatment reflux was documented in only one of 23 ureters in which a subureteric mound was absent, whereas in the series of patients in our study, reflux was present in all cases of an absent mound. Blake and O’Connell (40) suggested the possibility that soft-tissue distortion or transient soft-tissue swelling might be present despite the absence of a mound, with sufficient modification of the ureteral orifice to prevent reflux.
We demonstrated a significant association between mound contour and the number of chondrocyte injections performed and noted that every patient with a multilobed mound received multiple subureteric injections at the time of endoscopy.
The study was limited by its retrospective nature, with a lack of a complete set of sonograms for every patient in the clinical trial.
In conclusion, sonography is a useful modality with which to image the distal ureter and to exclude obstruction following treatment of reflux with subureteric injection of autologous chondrocytes. The absence of a subureteric chondrocyte mound or the presence of a multilobed mound contour is associated with persistent reflux. Mean mound volume decreases over time. Treatment-induced hydroureteronephrosis is uncommon and self-limited.
|
APPENDIX |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXREFERENCES |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONAPPENDIXREFERENCES |
|---|
inar B, Cotiè D, Oblak C. Possible cause of unsuccessful endoscopic collagen treatment of vesicoureteral reflux in children. Eur Urol 1999; 36:635-639.[CrossRef][Medline]
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
