Journal of Pediatric Urology (2008) 4, 260e264

Unilateral vesicoureteral reflux: Does endoscopic

injection based on the cystoscopic appearance
of the ureteral orifice decrease the incidence
of de-novo contralateral reflux?
Jonathan C. Routh a,*, Brant A. Inman a, Richard A. Ashley a,
David R. Vandersteen b, Yuri Reinberg b, James J. Wolpert b,
Stephen A. Kramer a, Douglas A. Husmann a

a
Department of Urology, Mayo Clinic, 200 1st Street SW, Rochester, MN 55901, USA
b
Division of Urology, Pediatric Surgical Associates, Minneapolis, MN, USA

Received 3 November 2007; accepted 17 December 2007

Available online 6 March 2008

KEYWORDS Abstract Objective: In patients with unilateral vesicoureteral reflux (VUR), it has been sug-
Vesicoureteral reflux; gested that injection of a non-refluxing but cystoscopically abnormal contralateral ureteral or-
Dextranomer; ifice (UO) with dextranomer/hyaluronic acid (Dx/HA) should be performed to prevent the
Ureter; development of de-novo contralateral VUR. We evaluate the effectiveness of this practice.
Endoscopy; Patients and methods: Patients with primary unilateral VUR undergoing injection of Dx/HA
Cystoscopy from 2002 to 2005 at two institutions were eligible. Patients with unilateral VUR with cystos-
copically abnormal contralateral UOs were injected with Dx/HA, while patients with normal
appearing UOs received no treatment. Multivariate logistic regression models were used to es-
timate the impact of prophylactic injection on the development of de-novo contralateral VUR.
Results: In total, 101 patients with unilateral VUR and an abnormal appearing contralateral UO
underwent prophylactic injection of Dx/HA while 45 patients with a normal appearing contra-
lateral UO were untreated. In patients receiving prophylactic Dx/HA, 9% (9/101) of the previ-
ously non-refluxing ureters developed de-novo VUR. Similarly, 13% (6/45) of patients with
a normal appearing UO treated by observation alone developed de-novo VUR (P Z 0.55).
The overall incidence of 10% (15/146) de-novo contralateral VUR matches published results
where this protocol was not followed.
Conclusions: Our findings suggest that cystoscopic assessment and prophylactic treatment of
an abnormal appearing, non-refluxing contralateral UO with Dx/HA is of little clinical benefit
and should be abandoned.
ª 2008 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.

* Corresponding author. Tel.: þ1 507 284 2511; fax: þ1 507 284 4951.
E-mail address: routh.jonathan@mayo.edu (J.C. Routh).

1477-5131/$34 ª 2008 Journal of Pediatric Urology Company. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jpurol.2007.12.008
De-novo contralateral VUR after deflux
Introduction
Vesicoureteral reflux (VUR) is found in approximately 40% of
children presenting with a urinary tract infection (UTI),
with roughly half of those diagnosed initially found to have
unilateral VUR [1e4]. Interestingly, with extended follow-
up, up to one third of patients initially thought to have uni-
lateral VUR will be found to have bilateral VUR on repeat
VCUG [1]. The development of contralateral VUR after an
initially negative study is believed to be due to either the
lack of sensitivity of the radiographic examination, the
presence of intermittent VUR, or a high-grade contralateral
refluxing unit serving as a ‘pop-off’ mechanism for a low-
grade ipsilateral refluxing unit [2,5,6].
The development of contralateral VUR is problematic
after unilateral surgical intervention, with published stud-
ies reporting that 7e20% of patients undergoing either
endoscopic injection therapy or ureteroneocystostomy will
develop de-novo contralateral VUR [6e11]. In an attempt to
prevent the development of de-novo contralateral VUR af-
ter endoscopic intervention for unilateral VUR, Elmore
et al. have recently recommended assessment of the con-
tralateral, non-refluxing ureteral orifice (UO) at the time
of cystoscopy. In patients with an abnormal appearing con-
tralateral UO, and in select circumstances in patients with
a normal UO, these authors recommend that endoscopic in-
jection of that ureter be concurrent with dextranomer/hy-
aluronic acid (Dx/HA) injection [11]. This recommendation
seems to be based on the hypothesis that an abnormal ap-
pearing UO is associated with a higher likelihood of devel-
oping de-novo contralateral VUR, while a normal
appearing UO has a decreased risk for de-novo reflux. By
following this protocol, it is hypothesized that the overall
risk of de-novo contralateral VUR (currently reported at
7e15% in most studies) that occurs following endoscopic
treatment of unilateral VUR could be significantly de-
creased [10e15] The purpose of this study is to report our
results with a policy of prophylactic injection of Dx/HA
into a non-refluxing, abnormal appearing UO, and its im-
pact on the onset of de-novo contralateral VUR.
Patients and methods
Patient selection and clinical features
After obtaining approval from the relevant institutional
review boards, we conducted a retrospective cohort study
of all patients undergoing Dx/HA injection for the treat-
ment of VUR at two institutions. The first patient was
enrolled in April 2002 and the last in December 2005.
Inclusion criteria were a minimum of one multi-cycle VCUG
or radionuclide cystogram (RNC). The families of all
children radiographically diagnosed with VUR at our in-
stitutions are routinely presented with a detailed explana-
tion of all currently accepted management options for VUR,
including observation, antibiotic prophylaxis, Dx/HA in-
jection and open surgical ureteroneocystostomy. Only
patients with primary, unilateral VUR undergoing their first
Dx/HA injection were included in this study. All patients
with secondary VUR or anatomic abnormalities such as
ureteral duplication and periureteral diverticula were
261
excluded. Injection of the contralateral, non-refluxing
ureter was performed based upon the attending surgeon’s
assessment of the UO. We elected to treat contralateral
non-refluxing UOs that had either stadium, horseshoe or
golf-hole configurations and that were associated with
a grade II or higher appearance of the UO upon hydro-
distension. Of note, no patients in this study were found to
have a golf-hole appearance to the non-refluxing-contra-
lateral UO [13,16]. The amount of injected Dx/HA was
based upon the resulting degree of coaptation of the UO,
i.e. the injection was continued until adequate coaptation
was achieved. The injection technique used was either sub-
ureteral (subureteric transurethral injection, STING) or in-
traureteral (hydrodistention-implantation technique, HIT).
The decision on which technique to use was based on the
discretion of the attending surgeon. Routine postoperative
radiologic follow-up evaluations were obtained 3 months
following the injection and consisted of either a cyclic
VCUG or RNC along with a renal ultrasound. Treatment suc-
cess was defined as the absence of VUR on the 3-month
postoperative cyclic radiographic study.
Statistical methods
Univariate logistic regression models were constructed to
predict treatment success at 3 months. Covariates analyzed
include: gender, age, VUR grade, presence of dysfunctional
voiding, amount of injected Dx/HA, injection technique
and the operating surgeon. Model residuals revealed no vi-
olation of regression assumptions. Multivariate modeling
was not possible because an inadequate number of out-
comes were observed in our sampledless than 10 per can-
didate covariatedand the resulting model would have been
overfitted. Statistical analyses were performed using SPSS
14.0 (SPSS Inc., Chicago, IL). All tests were two-sided and
P-values of 0.05 were considered statistically significant.
Results
One hundred and forty-six patients with unilateral VUR met
our inclusion criteria. Tables 1 and 2 summarize the charac-
teristics of the cohort. There is a statistically significant
variance between the two study groups regarding mode of
presentation. Specifically, patients undergoing prophylactic
bilateral Dx/HA injection were more likely to have pre-
sented with a UTI, while patients undergoing unilateral in-
jection were more likely to have their VUR diagnosed during
an evaluation for sibling reflux or antenatal hydronephrosis.
The two study groups were otherwise statistically similar.
The median patient age at the time of treatment for both
groups was approximately 6 years (range 1 month to
18 years), both cohorts were predominantly female (91%),
and the median number of preoperative cyclic radiographic
studies was 3 (range 1e6). There were no significant differ-
ences between the two cohorts in terms of number of pre-
operative cyclic VCUGs or RNCs.
In 101 patients (69%) the non-refluxing contralateral
ureter had an abnormal appearing UO and was prophylac-
tically injected with Dx/HA, while in 45 patients (31%) the
contralateral UO appeared cystoscopically normal and thus
no treatment of the non-refluxing ureter was rendered.
262 J.C. Routh et al.

Table 1 Study cohort characteristics

Characteristic Bilateral injection (n Z 101) Unilateral injection (n Z 45) P-value
n % n %
Gender
Female 94 (93%) 39 (87%) 0.221b
Male 7 (7%) 6 (13%)
Dysfunctional voiding
No 57 (56%) 31 (69%) 0.200b
Yes 44 (44%) 14 (31%)
History of UTI
No 19 (19%) 24 (53%) <0.001b
Yes 82 (81%) 21 (47%)
Laterality of refluxing ureter
Right 40 (40%) 17 (38%) 0.857b
Left 61 (60%) 28 (62%)
VUR grade, refluxing ureter
1 2 (2%) 2 (4%) 0.114b
2 69 (68%) 22 (49%)
3 27 (27%) 19 (42%)
4 3 (3%) 2 (4%)
Injection technique
Subureteral (STING) 54 (53%) 29 (64%) 0.278b
Intraureteral (HIT) 47 (47%) 16 (36%)
Previous bladder surgery
Yes 3 (3%) 3 (7%) 0.373b
No 98 (97%) 42 (93%)
De-novo VUR in non-refluxing ureter
Yes 9 (8%) 6 (13%) 0.364a
No 92 (92%) 39 (87%)
a
Wilcoxon’s rank-sum test.
b
Fisher’s exact test.

Contralateral orifice abnormalities in the injected cohort
were typically described as of a volcano or horseshoe
configuration with or without lateral displacement, and
were typically grade 2 or higher by hydrodistension. No
patient was reported to have a golf-hole configuration of
the contralateral UO.
When the results of postoperative imaging studies were
evaluated, we observed that de-novo VUR developed in the
previously non-refluxing contralateral ureter in 9% of the
patients (9/101) who underwent prophylactic bilateral
Dx/HA injection. Similarly, we found de-novo VUR devel-
oped in the previously non-refluxing contralateral ureter
in 13% of patients (6/45) in whom the non-refluxing UO ap-
peared normal and was managed conservatively. This dif-
ference was not statistically significant (P Z 0.55). We
were unable to identify any covariate factors that pre-
dicted the development of de-novo contralateral VUR on
univariate analysis, as summarized in Table 3.
Discussion
The development of de-novo contralateral VUR was brought
to popular attention in 1972, when Warren and associates

Table 2 Study cohort characteristics

Characteristic Bilateral injection (n Z 101) Unilateral injection (n Z 45) P-value
Median Range Median Range
Age (years) 5.9 [0.4e18.0] 6.2 [0.3e14.0] 0.849a
Number of preop. VCUGs 3 [1e5] 3 [2e6] 0.422a
Amount of Dx/HA injected (mL)
Refluxing ureter 0.8 [0.5e3.5] 0.8 [0.5e2.0] 0.533a
Non-refluxing ureter 0.8 [0.5e2.0] e e e
a
Wilcoxon’s rank-sum test.
De-novo contralateral VUR after deflux 263

unilateral VUR has been reported to vary between 7% and

Table 3 Factors predicting the development of de-novo
14% [6,10,11]. Based on this relatively high rate of de-
VUR in a previously non-refluxing ureter
novo contralateral reflux, Elmore and associates recently
Characteristic Odds 95% Confidence P-value suggested consideration of bilateral Dx/HA injection when
ratio interval a patient undergoes endoscopic treatment of a unilaterally
Age 0.87 [0.71e1.05] 0.153 refluxing ureter [11]. We have heretofore agreed with this
Gender concept, and since 2002 our practice has been to treat con-
Female 1 [0.04e4.41] 0.802 tralateral non-refluxing ureters associated with an abnor-
Male 0.77 mal appearing UO when endoscopically treating unilateral
VUR. It should be noted that the problem with this treat-
UTI ment protocol is that the correlation between VUR and
No 1 [0.12e1.17] 0.089 the appearance of the UO is hardly perfect. Nevertheless,
Yes 0.38 at the current time published data would seem to suggest
Dysfunctional voiding that a golf-hole or laterally displaced horseshoe configura-
No 1 [0.52e4.90] 0.413 tion of the UO may be associated with VUR [12,14e16,21].
Yes 1.59 Other abnormalities of UO appearance, such as stadium or
non-displaced horseshoe configurations, or the more re-
Laterality of cently described assessment of UO appearance following
non-refluxing ureter hydrodistension, have a variable association with radio-
Right 1 [1.03e10.76] 0.045 graphically diagnosed VUR [12,14e16,21]. Our results
Left 3.15 would seem to suggest that endoscopic management of
VUR grade in the non-refluxing, contralateral UO should not be per-
refluxing ureter formed based on the cystoscopic appearance of the UO.
1&2 1 [0.30e3.19] 0.949 Even a normal appearance did not preclude the develop-
3&4 1.04 ment of contralateral de-novo VUR, with 13% (6/45) of
our patients with a cystoscopically normal UO developing
Prophylactic injection de-novo contralateral reflux despite having undergone
No 1 [0.23e2.41] 0.553 three cyclic radiographic cystograms prior to surgical inter-
Yes 0.70 vention. Similarly, the injection of a cystoscopically abnor-
Amount injected 1.26 [0.38e3.87] 0.696 mal UO was not protective against the development of de-
into non-refluxing novo contralateral VUR, as this occurred in 9% of patients
ureter (9/101) despite prophylactic bilateral injection. These find-
ings are discouraging. At best, we thought that we would
see a decrease in the incidence of de-novo VUR in the con-
tralateral renal units associated with a normal appearing
reported on a cohort of 50 patients undergoing unilateral UO; unfortunately. our results did not demonstrate this. In-
PolitanoeLeadbetter ureteroneocystostomy. In this publi- deed, our findings support the premise that the cystoscopic
cation, single-cycle preoperative cystograms were used, appearance of the contralateral UO is of little to no benefit
with the finding that 22% of patients (11/50) developed in the management algorithm for a contralateral non-re-
de-novo contralateral VUR on postoperative radiologic eval- fluxing ureter. One important caveat must be noted: none
uations [8]. These authors stated that the only factor they of our contralateral UOs were reported to have a golf-
found to be related to the development of de-novo contra- hole configuration [12,14e16,21]. We hypothesize that
lateral VUR was the gross appearance of the previously non- this particular group of UOs were identified by our routine
refluxing UO [8]. In the modern era, the true incidence of and serial use of cyclic studies. Published data suggest
de-novo contralateral VUR is difficult to estimate, in large that golf-hole UOs may be associated with persistent reflux,
part due to the highly variable sensitivity of the radio- and consideration could be given to prophylactically treat-
graphic studies used to make this diagnosis [14,17]. Current ing this subset. Unfortunately, our data can neither support
studies reveal that VCUG and RNC have similar sensitivities nor refute this claim, as none of the patients in this study
in diagnosing VUR, and that repeated filling cycles improve were noted to have this particular configuration.
the sensitivity of both techniques [2,14,17,19]. Unfortu- Our data do have limitations. As a retrospective study, it
nately, even with cyclic studies, both cystographic tech- is possible that we did not account for certain confounding
niques remain an imperfect diagnostic evaluation, with factors which may have influenced our analysis. Specifi-
sensitivities typically reported in the low 90% range cally, the lack of treatment randomization reduces our
[14,17e19]. Due to the fact that cystograms are not 100% ability to definitively conclude that no difference existed
sensitive, the literature regarding the de-novo onset of con- between the two patient cohorts. It is also noteworthy that
tralateral reflux is quite difficult to critically assess and in- up to one third of medically managed patients with
terpret. Specifically, comparison of published studies is unilateral VUR will develop contralateral reflux after an
often difficult due to failure to report the number and extended period of observation. Given the short follow-up
type (single versus cyclic) of cystographic studies obtained in this study, it is possible that our rates of de-novo contra-
prior to surgical repair of the refluxing ureter [5e7,9,20]. lateral VUR would have been higher had we followed these
Taking this fact into consideration, the incidence of de- patients longer postoperatively. Likewise, it is unclear
novo contralateral VUR after endoscopic management of whether patients in the two cohorts would have fared
264
differently with increased follow up. On average three pre-
operative cyclic cystograms were obtained in this cohort,
implying that occult contralateral VUR is not likely to
have been a factor behind our de-novo rates.
Conclusion
Overall, these data raise serious concerns regarding the use
of contralateral Dx/HA injections in the presence of
unilateral VUR. The findings that Dx/HA injection into
abnormal ureteral orifices is of no benefit and that a normal
appearing contralateral UO is equally likely to develop de-
novo VUR call into question the effectiveness of the cysto-
scopic assessment of the contralateral non-refluxing UO
and the use of this treatment protocol. It is our opinion
that the ineffectiveness of this treatment protocol, coupled
with the added expense of contralateral Dx/HA injection,
strongly suggests that this practice be abandoned.
Conflict of interest statement
Dr. Douglas Husmann serves as a paid consultant to Q-Med
Scandinavia. No other author has any conflicts of interest to
disclose. There was no corporate involvement in this study.
No external funding sources were used for this study.
Acknowledgments
The authors gratefully thank Heidi Pfefferle, RN, and Sue
Rathbun, RN, for their assistance with this study.
References
[1] Sparr K, Balcom A, Mesrobian H. Incidence and natural history
of contralateral vesicoureteral reflux in patients presenting
with unilateral disease. J Urol 1998;160:1023.
[2] Unver T, Alpay H, Biyikli N, Ones T. Comparison of direct ra-
dionuclide cystography and voiding cystourethrography in de-
tecting vesicoureteral reflux. Pediatr Int 2006;48:287.
[3] Wein A, Schoenberg H. A review of 402 girls with recurrent uri-
nary tract infection. J Urol 1972;107:329.
[4] Elder J, Peters C, Arant Jr B, Ewalt D, Hawtrey C, Hurwitz R,
et al. Pediatric Vesicoureteral Reflux Guidelines Panel sum-
mary report on the management of primary vesicoureteral re-
flux in children. J Urol 1846;157:1997.
J.C. Routh et al.
[5] Diamond D, Rabinowitz R, Hoenig D, Caldamone A. The mech-
anism of new onset contralateral reflux following unilateral
ureteroneocystostomy. J Urol 1996;156:665.
[6] Kumar R, Puri P. Newly diagnosed contralateral reflux after
successful unilateral endoscopic correction: is it due to the
pop-off mechanism? J Urol 1997;158:1213.
[7] Minevich E, Wacksman J, Lewis A, Sheldon C. Incidence of
contralateral vesicoureteral reflux following unilateral extra-
vesical detrusorrhaphy (ureteroneocystostomy). J Urol 1998;
159:2126.
[8] Warren M, Kelalis P, Stickler G. Unilateral ureteroneocystos-
tomy: the fate of the contralateral ureter. J Urol 1972;107:466.
[9] Hoenig D, Diamond D, Rabinowitz R, Caldamone A. Contralat-
eral reflux after unilateral ureteral reimplantation. J Urol
1996;156:196.
[10] Routh J, Vandersteen D, Pfefferle H, Wolpert J, Reinberg Y.
Single-center experience with endoscopic management of
vesicoureteral reflux in children. J Urol 2006;175:1898.
[11] Elmore J, Kirsch A, Lyles R, Perez-Brayfield M, Scherz H. New
contralateral vesicoureteral reflux following dextranomer/-
hyaluronic Acid implantation: incidence and identification of
a high risk group. J Urol 2006;175:1097.
[12] Woodward J, Rushton H. Reflux uropathy. Pediatr Clin North
Am 1987;34:1349.
[13] Lyon RP, Marshall S, Tanagho EA. The ureteral orifice: its con-
figuration and competency. J Urol 1969;102:504.
[14] Dikshit M, Acharya V, Shikare S, Merchant S, Pardanani D.
Comparison of direct radionuclide cystography with micturat-
ing cystourethrography for the diagnosis of vesicoureteric re-
flux, and its correlation with cystoscopic appearances of the
ureteric orifices. Nephrol Dial Transplant 1993;8:600.
[15] Dewan P, Rao P. Predictive factors of resolution of primary vesi-
coureteric reflux: A multivariate analysis. Br J Urol 2006;98:915.
[16] Edmondson J, Maizels M, Alpert S, Kirsch A, Hanna M,
Weiser A, et al. Multi-institutional experience with PIC cystog-
raphyeincidence of occult vesicoureteral reflux in children
with febrile urinary tract infections. Urology 2006;67:608.
[17] Fettich J, Kenda B. Cyclic direct radionuclide voiding cystog-
raphy increasing reliability in detecting vesicoureteral reflux
in children. Pediatr Radiol 1992;22:337.
[18] Paltiel H, Rupich R, Kirulata H. Enhanced detection of vesi-
coureteral reflux in infants and children with use of cyclic
voiding cystourethrography. Radiology 1992;184:753.
[19] Novljian G, Kenig A, Rus R, Kenda B, Rajko B. Cyclic voiding
urosonography in detecting vesicoureteral reflux in children.
Pediatr Nephrol 2003;18:992.
[20] Ross J, Kay R, Nasrallah P. Contralateral reflux after unilateral
ureteral reimplantation in patients with a history of resolved
contralateral reflux. J Urol 1995;154:1171.
[21] Silva J, Diniz J, Lima E, Vergara R, Oliveira EA. Predictive fac-
tors of resolution of primary vesico-ureteric reflux: a multivar-
iate analysis. Br J Urol Int 2006;97:1063.