The reliability of cysto-urethrographic signs in the diagnosis of detrusor instability in children

Authors


Dr Batista Fundacion Puigvert, Cartagena 340, 08025 Barcelona, Spain.

Abstract

Objective

 To assess whether the signs associated with detrusor instability (DI), as assessed by video-urodynamic studies, can be evaluated by conventional voiding cysto-urethrography (VCUG).

Patients and methods

 Fifty-nine children who underwent cystometry and VCUG were reviewed and divided into two groups; group 1 comprised 51 neurologically normal children who had DI (47 girls and four boys, mean age 7.9 years, range 4–14), with no malformations or previous surgery. Most had mixed symptoms, including urinary tract infections (44) and nocturnal enuresis with daytime symptoms (20); group 2 (control) comprised eight children (seven girls, mean age 9.7 years, range 6–15) with a stable bladder and the same clinical presentation. The interval between VCUG and cystometry ranged from 1 day to 5 months (mean 47 days) and was similar in both groups. Signs considered suspicious of DI were sought in the findings of VCUG.

Results

 In group 1, VCUG was normal or showed no specific signs (only vesico-ureteric reflux or vaginal voiding) in 25 (49%) of patients, whereas 26 (51%) had some signs suspicious of DI. Suspicious signs were urethral ballooning (in 11), bladder trabeculation or a constricting ring (eight), a ‘spinning-top’ urethra (three), urethral notching (two) and Mercier’s bar (one). In group 2, five patients had no abnormal findings on VCUG and three had suspicious signs. The positive predictive value of VCUG was high (0.89) but the diagnostic yield of suspicious signs was low, because the sensitivity (0.5), specificity (0.62) and overall accuracy (0.52) were low.

Conclusion

 Radiological signs suspicious of DI cannot be regarded as such in conventional VCUG, as although they were detected in half the patients with DI, they also occurred in three of eight children with a stable bladder.

Introduction

Detrusor instability (DI) is defined as the presence of phasic, involuntary detrusor contractions during cystometry in the absence of neurological diseases [ 1]. DI has been detected in children with enuresis with daytime symptoms, recurrent UTIs and VUR with varying incidences [ 2[3][4]–5]. Some of these children undergo radiological studies before urodynamics and several authors have described the radiological signs associated with DI [ 3, 6[7]–8] either during modified voiding cysto-urethrography (VCUG) (with an indirect measurement of bladder pressure) or during video-urodynamic studies (VUD). As VUD are not available in many hospitals, the diagnostic yield of serial films of a conventional VCUG needs to be defined. Thus the aim of the present study was to assess whether these signs are present in VCUG and if they are of any value in diagnosing DI.

Patients and methods

The study included 59 children who underwent cystometry and VCUG within 6 months; on reviewing the records they were divided into two groups. Group 1 comprised 51 children with DI (47 girls and four boys, mean age 7.9 years, range 4–14). Their initial complaint was UTIs alone (23) or with other symptoms (21), nocturnal enuresis (20, 16 with recurrent infections and four with daytime urgency) and storage symptoms in two. DI was defined as the presence of phasic uninhibited detrusor contractions during water cystometry; from published criteria, the DI was considered pathological if the detrusor pressure was >15 cmH2O [ 5, 9[10]–11]. This group represented 40% of the children in whom DI was found in a 5-year period and who were referred for the evaluation of storage symptoms (urgency, frequency or urge incontinence) with or without enuresis, or recurrent UTIs. Exclusion criteria for the present series were; concomitant neurological diseases, malformations (e.g. PUV), previous urological surgery and patients not having VCUG, or with a short follow-up.

 During the same period, eight children fulfilled the criteria of group 1 but were found to have a stable bladder with an adequate age-adjusted bladder capacity and normal compliance (>40 mL/cmH2O) [ 12]. They comprised group 2 (six girls and two boys, mean age 9.7 years, range 6–15). Their initial complaints were UTIs in four, enuresis with daytime symptoms in three and storage symptoms in one.

 Cystometry was performed with water infused at 20–50 mL/min through an 8–10 F urethral catheter. A rectal balloon was used to subtract abdominal pressure. Bladder compliance was defined as the volume at capacity divided by detrusor pressure at that volume, and calculated at the lowest detrusor pressure closest to the bladder capacity. As all patients had involuntary contractions, this was calculated at the lowest pressure between contractions. Low compliance values were considered severe if <10 mL/cmH2O [ 13, 14]. All studies were reviewed by the same urologists (J.E.B. and P.A.) and performed by the same technician. Methods, definitions and units conform to the standards defined by the ICS, except where specifically noted [ 1].

 VCUG was performed by catheterizing the bladder with a 7–10 F Nelaton catheter; 35% meglumine-iothalamate (Conray 35, Chemie Linz AG, Austria) was instilled by drip infusion into the bladder through a 10 F catheter with brief, intermittent fluoroscopic observation, until the child had the sense of fullness or the bladder was seen to be fully distended. The catheter was then withdrawn and the child encouraged to void. During voiding, at least one film was exposed; a mean (range) of 4 (3–6) films were exposed at each VCUG. Lateral views were available in 13 patients. As described in other series [ 15], the mean (range) interval between the VCUG and cystometry was 47 days (1 day to 5 months) and was similar in both groups. No change in symptoms or medication were recorded during the interval. All VCUG films were reviewed simultaneously by two observers (J.E.B. and J.C.), unaware of the radiologist’s report. Previously reported signs considered suspicious of DI were examined (as films or graphs from each of the original papers) and summarized ( Table 1). Means were compared between groups using Student’s t-test and the incidences using the chi-square test.

Table 1.  The incidence in the present patients of radiological signs described previously for detrusor instability. Some patients showed more than one feature Thumbnail image of

Results

The interval between the initial symptoms and the urodynamic study ranged from 1 to 95 months (mean 26) and was similar in both groups. In group 1, VCUG was not useful in 25 children (49%), showed no abnormality in 16 and showed no specific signs in nine (eight VUR and one vaginal voiding). In 26 (51%) there were suspicious signs, either alone or with VUR ( Table 1). More than one radiological sign was detected on reviewing the films of 12 children (i.e. 23% of all and about a third of those with changes). All children with more than one sign had at least one suspicious sign. Urethral ballooning and bladder trabeculation (Fig. 1a) were the most frequent, followed by a constricting ring at the base of the bladder and the ‘spinning-top’ urethra (STU) (Fig. 1b). All boys had abnormalities on VCUG, two with various findings and two with VUR. Signs on VCUG were equally distributed among all ages, with no significant differences in age among the groups ( Table 2). Patients with a single radiological sign had a longer interval between symptom onset and diagnosis ( Table 2). This diagnostic interval was not significant compared with other radiological patterns except when patients with a single nonspecific radiological sign were compared with patients with a single suspicious sign (P=0.05, unpaired t-test; ( Table 2).

Figure 1.

Figure 1.

a, Urethral ballooning (arrow) and bladder trabeculation (arrowheads). b, the ‘spinning-top’ urethra (arrow) and trabeculation (arrowheads).

Table 2.  Age and time to diagnosis (months) for different radiological findings in the group 1 (unstable) Thumbnail image of

 In the control group, four patients had a normal VCUG; a constricting ring was seen in two (including the boy) and urethral ballooning in one. All had stable bladders with a normal compliance and bladder capacity (at least 80% of the age-adjusted capacity), and therefore, there were no urodynamic differences between patients with and without signs.

 The diagnostic yield of radiological signs in group 1 (DI) was low; both the sensitivity (0.50) and specificity (0.62) were low. The positive predictive value was high (0.89) but the negative predictive value was unacceptable (0.1), resulting in a overall accuracy of 0.52.

Discussion

Some of the current concepts in uroradiology and neurourology have been defined recently, and the relationship between functional and anatomical abnormalities is constantly evolving [ 16, 17]. Involuntary detrusor contractions on cystometry are frequently found in neurological patients (detrusor hyper-reflexia) and early reports of radiological changes related to voiding dysfunction occurred in this group [ 18, 19]. The use of VCUG as a routine investigation in children with recurrent UTI has been questioned [ 20] and we only perform it when there is a UTI with fever, if there is resistance to conventional suppressive treatment or if associated storage symptoms are recalled.

 Some signs have been considered suspicious of DI, e.g. urethral ballooning. In 1986, Fotter et al. [ 6] described a modified video-cystography with simultaneous monitoring of bladder pressure, in which they were able to assess the opening of the bladder neck during involuntary detrusor contractions. Opening of the bladder neck during cystography had a sensitivity of 93% in predicting DI in children referred for enuresis or recurrent UTI. This rate is higher than in the present study, as we did not monitor bladder pressure during cystography. However, this is a less accurate assessment of DI than obtained during VUD, as neither the pattern of DI nor compliance can be assessed. A lower sensitivity was found by the same authors in a latter series of 102 girls [ 7]. There is some overlap in the terminology of these signs. If ‘opening of the bladder neck’ as defined by Fotter et al. [ 6] ( Fig. 1) is compared with the ‘ballooning of the posterior urethra’ as described by Kondo et al. [ 3] they represent virtually the same pattern.

 The STU is defined as widening of the muscular segment of the urethra or ‘urethral widening’ on lateral views [ 8, 21]. Some authors report this finding predominantly during the filling phase, when involuntary contractions produce filling of the posterior urethra and leakage is prevented by tightening of the sphincter [ 8]. STU can be seen as ‘ballooning’ in the anteroposterior views, but as it was defined on lateral projections, we have applied this concept. For other authors, the sign represents a normal variant [ 22] and this might explain the presence of ballooning in one of the stable patients. Saxton et al. [ 8] studied 30 girls with STU using VUD; 28 had instability, 21 a congenital wide bladder neck anomaly (CWBNA, defined as opening of the bladder neck with no increase in detrusor pressure) and 20 had both. These authors concluded that STU was always a sign of DI or CWBNA. They reported similar findings in boys with DI, probably caused by unstable contractions filling the posterior urethra against the external sphincter contracting to avoid leakage [ 21]. Previous reports found a 16% incidence of CWBNA, but no DI [ 23].

 Urethral notching can be considered a multiple dilatation of the urethra and bladder neck. The earliest report in neurologically normal children with probable DI dates from 1969, in two children with UTI, and one with enuresis and diurnal symptoms [ 24].

 We consider trabeculation in the absence of obvious anatomical obstruction (i.e. PUV) or neurological disease a highly suspicious sign of DI. Mercier’s bar (inter-ureteric ridge) is probably a variant of the same feature and both can be present in neurological patients [ 25]. The presence of a constricting ring was more frequent in the control group than in those with DI, but the sample was too small to draw conclusions.

 Other signs can be considered nonspecific, as their link to DI is not as strong; DI has been reported with varying rates in children with VUR [ 3, 5, 7, 9]. The low incidence of VUR in children with DI, as found in the present series (25%) could be caused by a bias in selection; at our institution, children with VUR undergo cystometry if conservative treatment fails, if re- implantation surgery is planned or if they have storage symptoms [ 26]. In the present series, five of the 13 patients with reflux had some suspicious sign of instability. Vaginal voiding can be considered an artefact during normal VCUG, but can also be found in several types of voiding dysfunction [ 22, 27].

 Kondo [ 3] found signs in 40% of 200 children with instability studied with VUD, providing a comprehensive review of all signs previously described. The overall incidence of suspicious signs in the present study was higher (51%); this may be caused by bias in the selection of this group with a long history of symptoms (mean 25 months). In a study of 251 children with incontinence or storage symptoms, opening of the bladder neck was a common finding and 22% of children with that sign had DI [ 28]. Those authors considered ballooning and STU to be equivalent. The presence of these signs does not seem to be related to the age of patients or severity of DI. Patients with a single suspicious sign had a longer time to diagnosis than those with a single nonspecific sign ( Table 2).

 A major disadvantage in the present study was the small size of the control group that was not comparable with the DI group. This precludes an appropriate assessment of the sensitivity and specificity of these signs. Many patients, especially children, will not undergo invasive studies if the initial evaluation is negative. Therefore, the patients in our unit consist mainly of children with a long history of voiding dysfunction and the overall incidence of DI in children is very high. Although VCUG has a lower diagnostic yield than modified cystography, half the patients with DI had suggestive signs, a rate similar to that from VUD. It is common that patients are referred to the urologist with a previous VCUG in which these signs may go unreported. It is our current practice to perform urodynamics in patients with these signs; to date, DI was detected in all seven patients so evaluated. Further experience will enable us to assess the sensitivity and specificity of these signs.

Ancillary