Genitourinary Imaging
Original Research
MRI Evaluation of Urethral Diverticula and Differential Diagnosis in Symptomatic Women
Audio Available | 
Share
OBJECTIVE. The purpose of this study was to evaluate the role of MRI in the diagnosis and differential diagnosis of urethral diverticula in symptomatic women.
MATERIALS AND METHODS. Women referred for MRI at a single institution because of suspicion of urethral diverticula were included. All MRI examinations were independently evaluated by two radiologists and compared with patients' follow-up data. Sensitivity and specificity of MRI for urethral diverticula were calculated using surgery and clinical confirmation as the reference standards. Image quality of the urethra and periurethral region performed with the endoluminal coil was compared with the pelvic phased-array coil.
RESULTS. From a study group of 60 patients (mean age, 44 years), 20 patients (33%) had urethral diverticula and 28 (47%) had an alternative diagnosis, of which 13 (46%) were visualized with MRI. In the remaining 12 patients (20%) no abnormalities were found. For urethral diverticula, MRI had both sensitivity and specificity of 100%. Twenty patients had a total of 27 diverticula; these were mostly locally round (n = 12) with sharp margins (n = 25) and high (n = 19) homogeneous (n = 16) signal intensity on T2-weighted sequences. The ostium of urethral diverticula was identified in 23 diverticula (85%) by both readers. Agreement was 93% with κ = 0.72. Endoluminal coil placement in the vagina showed the best image quality of the urethra and periurethral region.
CONCLUSION. Dedicated MRI is an excellent imaging modality for urethral diverticula; furthermore, MRI will show the alternative diagnosis in almost one half of the remaining patients.
Keywords: differential diagnosis, lower urinary tract symptoms, MRI, urethral diverticulum
Female urethral diverticulum is often overlooked and frequently misdiagnosed because of un awareness of the condition. Urethral diverticula are estimated to occur in 1–6% of women. Although usually diagnosed between the third and fifth decade of life, it can affect all age groups [1–3]. Usually, an array of nonspecific genitourinary symptoms predominate. The most frequent symptoms described are frequency and urgency (40–100%), dysuria (30–70%), recurrent urinary tract infection (30–50%), postmicturition dribble (10–30%), dyspareunia (10–25%), and hematuria (10–25%) [4, 5]. It may also present with a tender mass (35%), urinary incontinence (32%), stones (1–10%), discharge of pus from the urethra (12%), and retention (4%) [6]. This condition should always be considered in women with unexplained lower urinary tract symptoms. In addition to interstitial cystitis, urethral syndrome, and urgency-frequency syndrome, the clinician should include urethral diverticula in the differential diagnosis [7].
Appropriate investigations play an important role in the diagnosis of urethral diverticula and ideally should provide the surgeon with information regarding location, number, size, configuration, and communication of the urethral diverticula [8]. MRI has now become the imaging study of choice [8].
The purpose of this study was to evaluate the role of MRI in the diagnosis of urethral diverticula and differential diagnosis for this condition in symptomatic women.
| Materials and Methods |   |
|---|
This study was performed without financial support, and the authors had exclusive control of the data and information presented in this article.
Institutional review board approval was obtained for this retrospective study, and informed consent was waived. Women referred for MRI at a single institution because of clinical suspicion of urethral diverticula were included in the study. Patients were electronically identified using a department database. All patients had two or more lower urinary tract symptoms, including pain, urinary incontinence, dyspareunia, and frequency or urgency. Patients were excluded when no convincing information for suspicion of urethral diverticula was presented. From hospital- and department-based digital information systems, a search was performed to assess the specific diagnosis for the complaints for which the patient was referred to the MRI unit.
The reference standard was surgery or, in case no surgery was performed, confirmation of MRI findings with a second imaging modality, including video urodynamic studies and voiding cystourethrography and clinical examination.
All examinations were performed using a 1.5-T MR imager (Gyroscan NT Intera 1.5, Philips Healthcare) with application of a rigid endoluminal coil placed in the vagina (n = 17) or in the anus (n = 2) or with a pelvic phased-array coil (n = 6).
The endoluminal coil is commercially available and consists of a fixed, rectangular, 60-mm-long rigid receiver coil with a width of 16 mm. The coil is contained within an 80-mm-long cylindric coil holder with a diameter of 19 mm. Before the introduction of the coil into the vagina or anus, a condom was placed over the coil and ultrasound gel was used as a lubricant. It should be noted that this rigid endoluminal coil is not the same as the inflatable rectal coil but is of smaller diameter and is designed to be placed in the vagina or anus.
The phased-array coil is a commercially available four-channel coil (Synergy, Philips Healthcare) placed around the pelvis. A standardized imaging protocol of T2-weighted turbo spin-echo and gradient-echo sequences with and without fat saturation was performed in all cases, with high-resolution slices in three planes (axial, coronal, and sagittal) through the urethra and periurethral region. The FOV included the bladder base and perianal region. At endoluminal MRI, image volume encompassed the entire sensitive region of coil.
For the endoluminal coil, parameters were transverse T2-weighted fast-field echo imaging (TR/TE, 23/14; matrix, 205 × 256; flip angle, 60°; FOV, 140 mm; thickness, 2 mm with no gaps; and two signals acquired). Transverse T2-weighted fast spinecho MRI was performed with and without fat saturation (TR/TE, 5086/100; matrix, 186 × 256; flip angle, 90°; FOV, 120 mm; section thickness, 4 mm with a 0.4-mm gap; and three signals acquired). Coronal and sagittal T2-weighted fast spin-echo MRI was performed without fat saturation (TR/TE, 2454/100; matrix, 186 × 256; flip angle, 90°; FOV, 120 mm; section thickness, 4 mm with a 0.4-mm gap; and four signals acquired).
For the pelvic phased-array coils, parameters were transverse, coronal, and sagittal T2-weighted fast spin-echo sequences without fat suppression (TR/TE, 2500/70; matrix, 512 × 256; flip angle, 60°; FOV, 300 mm; section thickness and gap, 3 mm and 0.3 mm, respectively; and two signals acquired) and transverse T2-weighted fast spin-echo with fat suppression (TR/TE, 4000/85; matrix, 512 × 256; flip angle, 60°; FOV, 300 mm; section thickness and gap, 3 mm and 0.3 mm, respectively; and two signals acquired).
Transverse fat-saturated T1-weighted fast spinecho imaging after IV gadolinium chelate administration (0.2 mL/kg of 0.5 mmol/mL of gadopentetate dimeglumine [Magnevist, Schering]) was performed in few (n = 6) cases (TR/TE, 3.61/1.4; matrix, 512 × 256; flip angle, 15°; FOV, 300 mm; section thickness and gap, 4 mm and 0.4 mm, respectively; and one signal acquired).
All MRI examinations were loaded in a PACS viewing station from digital storage facilities and were examined by two observers with experience in reading approximately 2500 and 500 pelvic MR examinations, respectively. The observers were blinded to the patient history, findings in the clinical workup, or other imaging studies and to each other's results.
The following items were evaluated separately by both observers: MRI examination technique using the pelvic phased-array or endoluminal coil placed in the vagina or in the anal canal; number of urethral diverticula; size of urethral diverticula on three axes: anteroposterior, transverse, and craniocaudal measured in mm; shape of urethral diverticula (locally round, oval shaped, circumferential, or elongated); identification of the ostium of urethral diverticula in the urethra; localization of the ostium with “uncertain” added as a separate category; distance of the ostium to the bladder neck measured in mm; signal intensity of urethral diverticula on T2-weighted images; borders of urethral diverticula (well-defined or ill-defined); important remarks, such as stones or suggestion of other solid components within urethral diverticula; added value of T1-weighted sequences with IV gadolinium chelate administration; and alternative diagnoses.
The ostium of urethral diverticula was defined as a beaklike linear extension of the diverticular neck through the muscular and submucosal layers of the urethra with high signal intensity on T2-weighted images. Image quality of the urethra and periurethral region with the pelvic phased-array or endoluminal coil was evaluated by both observers. Every examination was scored independently by both observers using a 3-point scale (fair, satisfactory, and excellent), indicating the best detailed anatomy of the urethra and periurethral region in the perception of the observer.
Ninety-five percent confidence intervals were calculated for sensitivity and specificity. Comparisons of percentages between groups were performed using Fisher's exact test. The limit of significance was set at p = 0.05 (two-sided). Kappa statistics were used to quantify degrees of agreement between readers for identifying the ostium of the diverticulum in the urethra. The classification “uncertain” was included in the latter calculation as a separate category. A kappa value ≤ 0.20 was interpreted as slight agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; and ≥ 0.81, almost perfect agreement.
| Results | |
|---|
Between January 1995 and December 2006, 1541 women underwent MRI of the pelvis and pelvic floor for a variety of reasons. We identified 60 patients (all women; mean age, 44 years; age range, 18–80 years) who fulfilled the criteria for inclusion in the study population. In total, 20 patients had 27 urethral diverticula on 25 different MRI examinations. This included three patients with follow-up MRI examinations after surgery because of persistent complaints and proved residual or recurrent urethral diverticula. Two patients underwent control MRI before surgery because of a long delay of 19 and 26 months, respectively, between the first MRI examination and surgery. Two patients had two separate urethral diverticula with separate outlines and diverticular necks on a single MRI examination (Fig. 1).
 View larger version (208K) | Fig. 1—43-year-old-woman with multiple urethral diverticula. Axial T2-weighted image with fat saturation performed with pelvic phased-array coil shows urethral diverticulum in 9-o'clock position (short arrow), including diverticular neck (curved arrow). Ostium (not shown) was located in midline anteriorly (12-o'clock position). Separate urethral diverticulum is seen posteriorly (long arrow) with ostium in midline posteriorly (6-o'clock position). Both urethral diverticula, including diverticular neck and ostia, were confirmed surgically. U = urethra, V = vagina, A = anus. |
 View larger version (184K) | Fig. 2A—42-year-old-woman with elongated urethral diverticulum. A, Axial T2-weighted image shows subtle linear structure (arrow) with high signal intensity within submucosal layer of urethra posteriorly, indicating location of ostium of urethral diverticulum. Asterisk indicates endoluminal coil, A = anus. |
 View larger version (185K) | Fig. 2B—42-year-old-woman with elongated urethral diverticulum. B, Axial T2-weighted images show track of this elongated urethral diverticulum (arrow). Arrowhead in C indicates debris in apex of urethral diverticulum. Asterisk indicates endoluminal coil, R = rectum. |
 View larger version (193K) | Fig. 2C—42-year-old-woman with elongated urethral diverticulum. C, Axial T2-weighted images show track of this elongated urethral diverticulum (arrow). Arrowhead in C indicates debris in apex of urethral diverticulum. Asterisk indicates endoluminal coil, R = rectum. |
 View larger version (110K) | Fig. 2D—42-year-old-woman with elongated urethral diverticulum. D, Voiding cystoureterogram shows same urethral diverticulum (arrows) with good correlation with MRI appearance. To our knowledge, this type of urethral diverticula with its elongated nature has not been described before. Patient had no additional abnormalities of kidneys, ureters, and lower urinary tract. |
In 16 patients, urethral diverticula were surgically proven. Surgery was performed between 2 and 10 weeks after MRI. These patients had a total of 19 surgical procedures for urethral diverticula, including three patients with recurrent or residual urethral diverticula. In 17 cases, urethrocystoscopy was performed as a preoperative tool, with visualization of the ostium of urethral diverticula in eight cases. The location of the ostium described in the surgical report was in agreement with MRI. In the remaining nine surgical cases, identification and excision of urethral diverticula, including the diverticular neck, was documented. In these cases, MRI had good correlation concerning the orientation of urethral diverticula, including the diverticular neck, as was described in the surgical report. Four patients either refused surgery or preferred expectant management for urethral diverticula. In the patients with no surgery (n = 4), MRI findings of urethral diverticula were confirmed with follow-up video urodynamics within 3 months after MRI (n = 4), including visualization of urethral diverticula on an initial voiding cystourethrography (n = 2), which was performed 4 and 5 months before MRI (Figs. 2A, 2B, 2C, and 2D). In the clinical workup, which included palpation and massage of the urethra through the vagina, the diagnosis of urethral diverticula was considered confirmed and proved by the urologist as was documented in patients' records.
We searched for the clinically confirmed final diagnoses in the rest of the study group (Fig. 3). Twenty-eight of 60 patients (47%) had an alternative diagnosis. Thirteen (46%) were correctly diagnosed with MRI. Periurethral scarring, urethrovaginal fistula, infected cyst of Skene, cystocele, and endometriosis of the vaginal vault were confirmed by surgery. Periurethritis and inflammatory pseudotumor of the urethra were proved by biopsy. Dysfunctional voiding and interstitial cystitis were confirmed by clinical evaluation, including video urodynamics, with no abnormalities found on MRI. Urethritis and vaginitis by Candida species was proved by specific laboratory findings and response to specific therapy, with no abnormalities found on MRI. In the remaining 12 patients (20%), no abnormalities were found on MRI and in the clinical follow-up. The follow-up period was at least 18 months and ranged from 1.5 to 8 years (mean, 5 years).
 View larger version (25K) | Fig. 3—Flow diagram summarizes patient sampling. |
MRI showed the final diagnosis (urethral diverticula or alternative diagnosis) in 33 of 60 patients (55%). On the basis of MRI; clinical evaluation, including video urodynamics; and clinical data, 80% of patients were diagnosed with an abnormality that was responsible for their complaints.
For the diagnosis of urethral diverticula, MRI had sensitivity of 100% (20 of 20 patients; 95% CI, 83–100%) and specificity of 100% (40 of 40 patients; 95% CI, 91–100%).
Of 25 MRI examinations with findings of urethral diverticula, 19 were performed with the endoluminal coil placed in the vagina (n = 17) or anal canal (n = 2), and 6 examinations were performed with the pelvic phased-array coil (Table 1). Imaging with the endoluminal coil placed in the vagina had 17 of 17 graded excellent versus 0 of 6 with phased-array imaging (p < 0.001, Fisher's exact test). Too few patients (n = 2) were imaged with the endoluminal coil placed in the anus to evaluate the results statistically. Both observers fully agreed in each of the gradings displayed in Table 1 that visualization of the best image quality of the urethra and periurethral region occurred with the endoluminal coil placed in the vagina.
Urethral diverticula had different shapes (Table 2). The most frequent shape was locally round or oval. When considering the borders of the urethral diverticula, most cases showed well-defined margins (n = 25). Two urethral diverticula showed partly sharp and partly ill-defined margins. This was caused by perilesional inflammation, which was proved by histology (Figs. 4A and 4B).
TABLE 1: Quality of Images of the Urethra and Periurethral Region Using Different MRI Coils
TABLE 2: Shape of 27 Urethral Diverticula on 25 MRI Examinations in 20 Women
The ostia of urethral diverticula could be identified in 23 of 27 cases (85%) by both observers. Observer 2 could not identify the ostium in two diverticula and, in another two diverticula, considered this to be uncertain. Observer 1 scored the same four diverticula as uncertain. The agreement was 93% (25/27) with κ = 0.72. Twenty-three identified ostia of urethral diverticula by both observers were located in the midline anterior (12-o'clock position) (n = 1); left lower quadrant (between 3-o'clock and 6-o'clock positions) (n = 11); midline posterior (6-o'clock position) (n = 4); and right lower quadrant (between 6-o'clock and 9-o'clock positions) (n = 7) with the patient in the supine position. The mean distance of the ostia to the bladder neck was 12 mm (range, 7–36 mm) (Figs. 5A, 5B, and 5C).
 View larger version (240K) | Fig. 4A—39-year-old-woman with urethral diverticulum imaged with pelvic phased-array coil. A, Axial T2-weighted image shows circumferential urethral diverticulum (short arrows) with internal inhomogeneity as a result of debris, septa, and stone (long arrow). |
 View larger version (242K) | Fig. 4B—39-year-old-woman with urethral diverticulum imaged with pelvic phased-array coil. B, Axial T2-weighted image shows unsharp margins of diverticulum, best seen on left (arrowheads), which is caused by perilesional inflammation (pathology proven). Ostium of this complicated urethral diverticulum was considered by two observers as “not identified” and “uncertain,” respectively. |
Twenty-three identified ostia of urethral diverticula by both observers were imaged with the endoluminal coil placed in the vagina (17 of 18 diverticula), or anal canal (2 of 2 diverticula), or with the pelvic phased-array coil (4 of 7 diverticula). These findings tended toward significance regarding the identification of the ostium of urethral diverticula with the endoluminal coil placed in the vagina compared with the phased-array coil (p = 0.053, Fisher's exact test).
On T2-weighted images, urethral diverticula had high signal intensity similar (n = 19) or slightly lower (n = 8) than the signal intensity of urine in the bladder. Signal intensity on T2-weighted sequences was homogeneous (n = 16) (Figs. 5A, 5B, and 5C) or inhomogeneous (n = 11) (Figs. 4A, 4B, and 6A). Inhomogeneous signal intensity was found in diverticula containing debris (n = 9) (Fig. 6A) or debris, septa, and stone (n = 2) (Figs. 4A and 4B).
In six MRI examinations, gadolinium-enhanced T1-weighted series were performed in addition to the T2-weighted sequences. Both observers were of opinion that contrast-enhanced studies had no additional value for the diagnosis of urethral diverticula (Figs. 7A and 7B).
 View larger version (146K) | Fig. 5A—35-year-old-woman with urethral diverticulum. A, Axial T2-weighted image shows endoluminal coil (asterisk) placed in vagina. Arrow indicates locally rounded urethral diverticulum on left, A indicates anus. |
 View larger version (133K) | Fig. 5B—35-year-old-woman with urethral diverticulum. B, On next lower slice image, ostium of urethral diverticulum can be seen posteriorly as linear extension of diverticular neck through layered structure of urethra in midline (arrow). Arrowheads show puborectal muscles with thinning and fibrosis of left compared with normal muscle thickness on right side. Asterisk indicates endoluminal coil, A indicates anus. |
 View larger version (202K) | Fig. 5C—35-year-old-woman with urethral diverticulum. C, Sagittal T2-weighted image shows diverticulum consisting of two parts (straight arrows) and facilitates precise measurement of distance of diverticular neck (curved arrow) to bladder base. Asterisk indicates endoluminal coil placed in vagina, B = bladder, S = pubic symphysis, A = anus. |
| Discussion | |
|---|
In our study, urethral diverticula were diagnosed in 20 of 60 symptomatic women (33%) referred for MRI. Other studies reported urethral diverticula in 10% of patients examined with endorectal coil MRI or 74% of patients examined with a combination of transvaginal, transperineal, and urethral sonography (using a catheter-based transducer) [9, 10]. Previous MRI studies of symptomatic female urethral and periurethral diseases were based on the use of the pelvic phased-array coil or the inflatable endoluminal coil placed in the vagina or in the rectum. To our knowledge, our study is the first with application of the rigid endoluminal coil in these patients. The rigid endoluminal coil can be inserted into the vagina by the patient herself, and there is no need for administration of glucagon or butylbromide to prevent rectal spasm, which could be the case when using the inflatable endoluminal coil [9]. Although a limited number of patients in our study was examined with the endoluminal coil placed in the anal canal or with the pelvic phased-array coil, a qualitative comparison suggests better image quality of the urethra and periurethral region with the endoluminal coil placed in the vagina (Table 1 and Figs. 4A, 4B, 6A, and 7A). With intraanal placement of the endoluminal coil, the urethra falls slightly outside the sensitive region of the coil, whereas the perianal region is displayed in great detail (Fig. 6A). According to data in the literature and our experience, care should be taken to use the right scanning parameters. With the pelvic phased-array coil, a slice thickness of 3 mm and the use of axial T2-weighted turbo spin-echo sequences from the bladder base through the entire urethra has been recommended [11] with no need for T1-weighted sequences or contrast-enhanced series for the diagnosis of urethral diverticula [8] (Figs. 7A and 7B). With this dedicated imaging protocol, imaging times can be limited to an average of 15 minutes [8]. Contrast-enhanced studies, however, may improve visualization of granulation tissue or carcinoma within urethral diverticula and tumor spread to adjacent adipose tissue [12, 13]. Malignancy arising from a diverticulum can be visualized as enhancing soft tissue within the diverticulum [14]. Cancer within urethral diverticula that is still small may be overlooked with torso phased-array MRI including gadolinium-enhanced sequences [15]. In our study, among 19 resected and pathology-proven urethral diverticula, no carcinoma was found.
 View larger version (174K) | Fig. 6A—42-year-old-woman with circumferential urethral diverticulum. A, Axial T2-weighted image shows sharp borders (straight arrows) of urethral diverticulum with high inhomogeneous signal intensity containing debris on dependent side. Endoluminal coil (asterisk) is placed in anal canal. Curved arrow indicates urethra, and arrowheads indicate puborectal muscles. |
 View larger version (240K) | Fig. 6B—42-year-old-woman with circumferential urethral diverticulum. B, Axial T2-weighted image acquired almost 1.5 years after surgery shows recurrent or residual urethral diverticulum (arrows). Image was acquired with four-channel phased-array pelvic coil. For optimal imaging results of urethral diverticula, smaller FOV is recommended (see also Figures 5A, 5B, and 5C). |
 View larger version (137K) | Fig. 7A—47-year-old-woman with circumferential urethra diverticulum. A, Axial T2-weighted image shows endoluminal coil (asterisk) placement in vagina results in better image quality of urethra and periurethral region compared with Figures 2A, 2B, 2C, and 2D. Urethral diverticulum with sharp borders, homogeneous hyperintense signal intensity, and identification of ostium in right lower quadrant (arrow) are visualized. A indicates anus. |
 View larger version (147K) | Fig. 7B—47-year-old-woman with circumferential urethra diverticulum. B, Contrast-enhanced axial T1-weighted turbo spinecho image with fat saturation after IV gadolinium chelate administration image obtained at same level as A shows no enhancing components within urethral diverticula. Asterisk indicates endoluminal coil. |
Few studies have focused on the differential diagnosis of patients with urethral diverticula. In the differential diagnosis, one should also consider other regional cystic lesions. Viaualization of communication between the lesion and the urethra confirms urethral diverticula [16]. The differentiation between a communicating or noncommunicating cystic lesion with the urethra is important because the surgical approach for the two entities differs. Both lesions are resected; however, a diverticulum may require urethral reconstruction procedures [16–18]. To our knowledge, our study is the first to identify the ostium of urethral diverticula in the majority of cases (85%) on MRI with improved confidence using the rigid endoluminal coil rather than the pelvic phased-array coil. Kim et al. [19] could not identify the ostium of urethral diverticula on MRI with application of the body or pelvic coil. Blander et al. [20] identified the diverticular neck (ostium) in 11 of 27 patients (41%) on MRI with the inflatable endoluminal coil.
In our study, urethrocystoscopy showed the ostium of urethral diverticula in 47% of cases. Other studies have reported 52% accuracy of urethrocystoscopy in identifying the ostium of urethral diverticula [20]. This warrants accurate preoperative imaging to assess the precise location of urethral diverticula, including the diverticular neck and ostium with additional information regarding size, extent, content, and complexity of the lesion. This practice may shorten the duration and improve the outcome of the surgical procedure.
The only study we found to clearly identify the ostium of urethral diverticula in the majority of cases used a combination of transvaginal, transperineal, and urethral sonography (using a catheter-based transducer) in 19 women with urethral symptoms [10]. The neck was precisely seen in 13 of 15 diverticula on sonography, especially via the transurethral approach. The disadvantages of this modality are that most diverticular necks were identified using the transurethral approach, which is not as convenient for the patient and is not as readily available as transvaginal and transperineal sonography [10]. Furthermore, transurethral sonography requires a dedicated intravascular sonographic unit as well as a catheter-based transducer [10].
In our study, MRI showed the final diagnosis (urethral diverticula or alternative diagnosis) in 33 of 60 patients (55%). On the basis of MRI; clinical workup, including functional urodynamic studies; and clinical data, 80% of patients were diagnosed with an abnormality that was responsible for their complaints. According to these findings, in the diagnostic workup of women with suspicion of urethral diverticula, both MRI and urodynamic studies can be considered. After clinical evaluation, MRI can be considered to assess a physical cause. In addition, urodynamic studies may be used to evaluate associated urinary incontinence and for planning additional incontinence surgery to obtain the best postoperative results for urethral diverticula [8, 18, 21]. Bennett et al. [22] advocated that addition of a sagittal dynamic sequence to the MRI protocol will permit detection of pelvic organ prolapse that may not be evident on static at-rest images and that may also go undetected at physical examination.
In 20% of our study group, no clear diagnosis could be established. No physical abnormalities were found on MRI and in the clinical follow-up. We believe that functional aspects of the pelvic floor and supporting ligaments of the bladder base and urethra may be a cause for the complaints in these patients. Follow-up studies, including dynamic MRI and functional urodynamic studies, may be needed to evaluate these patients thoroughly.
There were limitations to this study. First, our study was retrospective and included a relatively small number of urethral diverticula. Second, two control MRI examinations (because of a long delay of 19 and 26 months between the first MRI examination and surgery) were included for analysis of specific imaging features of urethral diverticula. This may have led to selection bias; however, we believe that the morphology of urethral diverticula may change in time depending on the state of filling and secondary inflammation. This may justify the inclusion of both examinations per patient. Third, we found that sensitivity and specificity of MRI for urethral diverticula were both 100%, using surgery and clinical confirmation as the reference standards. Not all 60 patients had the same additional evaluation for urethral diverticula, and thus the true sensitivity may not be 100%. Finally, there was no uniform reference standard. We consider surgery as the reference standard for urethral diverticula. Surgery was performed in the majority of patients (80%), with urethral diverticula proved. In the remaining patients, MRI findings of urethral diverticula were confirmed with a second imaging modality (video urodynamic studies and voiding cystourethrography) and considered proved by clinical examination and evaluation by the urologist. Furthermore, no reports were found in the hospital-based long-term follow-up data of these patients for whom the MRI findings of urethral diverticula were clinically in doubt. Nevertheless, we believe that adequate imaging and clinical follow-up information was available for all patients, and therefore we consider our conclusions valid.
In conclusion, there is an expanding role for dedicated MRI in women with suspicion of urethral diverticula. MRI will show the diagnosis in more than one half of all patients. On the basis of MRI, functional urodynamic studies, and clinical evaluation, 80% of patients will be diagnosed with an abnormality that is responsible for their complaints. Additional studies should follow to clarify the reason for complaints in the remaining 20% of patients.
| References | |
|---|
