Outcomes of Surgery of Female Urethral Diverticula Classified Using Magnetic Resonance Imaging

european urology 51 (2007) 1664 1670 available at www.sciencedirect.com journal homepage: www.europeanurology.com Female Urology Incontinence Outcomes of Surgery of Female Urethral Diverticula Classified Using Magnetic Resonance Imaging Deok Hyun Han a, Yong Sang Jeong a, Myung-Soo Choo b, Kyu-Sung Lee a, * a Department of Urology, Samsung Medical Center, Sungkyunkwan University School of Medicine, Seoul, South Korea b Department of Urology, Asan Medical Center, University of Ulsan College of Medicine, Seoul, South Korea Article info Article history: Accepted January 22, 2007 Published online ahead of print on February 6, 2007 Keywords: Diverticulum Magnetic resonance imaging Urethra Abstract Objectives: We evaluated the surgical outcomes of transvaginal diverticulectomies classified using magnetic resonance imaging (MRI). Methods: We evaluated 30 women (mean age: 46.4 yr; range: 31 73 yr) who were followed up for at least 12 mo after urethral diverticulectomies (mean follow-up: 29 mo; range: 12 93 mo). Characteristics of urethral diverticula were confirmed before surgery by MRI. Diverticula were classified as simple, U-shaped, or circumferential according to MRI features. Transvaginal excisions of urethral diverticula were performed using vaginal flaps and three-layer closures. Cure was defined as the absence of a diverticulum and symptoms. Results: Seventeen cases (57%) had simple diverticula, three (10%) had U-shaped diverticula, and 10 (33%) had circumferential diverticula. After the first operation, 23 cases (77%) were cured. None of the simple diverticula recurred, but 33% of the U-shaped and 60% of the circumferential diverticula did recur. Of the seven recurrent cases, three did not require a second operation because their symptoms resolved. Of the four cases that underwent a second operation, three were cured and one was cured after two additional operations. The success rate for circumferential diverticula after initial diverticulectomies was less than that of simple or U-shaped diverticula ( p < 0.05). Location, size, and multiplicity of urethral diverticula did not affect the surgical outcome ( p > 0.05). Conclusions: Transvaginal diverticulectomy is effective for treatment of female urethral diverticula. For circumferential urethral diverticula, however, surgical procedures should be adapted to achieve complete resections of the diverticulum. # 2007 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Urology, Samsung Medical Center, Sungkyunkwan University School of Medicine, 50 Irwon-dong, Gangnam-gu, Seoul 135-710, South Korea. Tel. +822 3410 3554; Fax: +822 3410 3027. E-mail address: LKS@smc.samsung.co.kr (K.-S. Lee). 0302-2838/$ see back matter # 2007 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2007.01.103

european urology 51 (2007) 1664 1670 1665 1. Introduction Female urethral diverticulum (UD) is uncommon and is present in 0.6 6% of the population [1,2]. The majority of cases present during the third to fifth decade of life, but cases of neonates and young female subjects have been reported [3]. It is difficult to diagnose female UDs because of the wide range of nonspecific symptoms and because 3 20% of patients are asymptomatic [1,4,5]. Several treatment methods have been described for symptomatic UDs. Although transvaginal diverticulectomy is an effective procedure [3,4,6], it is generally considered difficult because of the normal anatomic fusion of the urethra to the anterior vaginal wall, the lack of cleavage planes, and the involvement of the UD with the urethral sphincter mechanism. Consequently, this treatment is sometimes unsuccessful, resulting in a persistent symptomatic UD or other complications such as urethrovesical fistulae or stress urinary incontinence (SUI) [7]. We evaluated the surgical outcomes of transvaginal diverticulectomy according to the classification using MRI and identified preoperative factors that affect the success of the procedure. 2. Methods Five years prior to this study, 35 patients underwent urethral diverticulectomies at our institution. We conducted a review of the medical records of 30 women who were followed up for at least 12 mo after surgery. Preoperative evaluation included a thorough medical history, physical examination, urinalysis, and cystourethroscopy. Radiological imaging, including transvaginal ultrasonography (USG) and pelvis magnetic resonance imaging (MRI) using a surface coil, was performed to confirm Fig. 1 Anatomical configuration of three types of female urethral diverticulum according to T2-weighted axial magnetic resonance images (a, simple; b, U-shaped; c, circumferential).

1666 european urology 51 (2007) 1664 1670 the presence of the UDs and to evaluate their locations, sizes, multiplicities, and shapes. UDs were classified into three types according to preoperative pelvic MRI. We defined UDs that were round or oval and located laterally or posterior to the urethra as simple, UDs that extended partially around the urethra as U-shaped, and UDs that extended completely around the urethra as circumferential (Fig. 1). For treatment of UDs, we performed transvaginal excisions of the diverticular sac using a vaginal flap technique as described previously [2]. After the patient was placed in the dorsal lithotomy position and draped in a sterile manner, a 16-Fr Foley catheter was inserted through the urethra. A Scott ring retractor with hooks and a weighted vaginal speculum were used for retraction. After the anterior vaginal wall was exposed, an inverted U-shape incision was made on the anterior vaginal wall over the UD with the apex of the U midway between the UD and the urethral meatus. A saline solution with 1:800,000 epinephrine was injected into the anterior vaginal wall to facilitate dissection and haemostasis. This anterior vaginal wall flap was dissected with Metzenbaum scissors inferiorly and laterally up to the level of the bladder neck to expose the periurethral fascia. We identified the location of the UD by palpation and confirmed the existence of the UD by aspirating fluid from the diverticular sac using a syringe. In instances in which the UD could not be palpated after exposing the periurethral fascia, we reviewed the MRI carefully and located the UD. We then confirmed its location by aspirating fluid from it. After the location of the UD was identified, a transverse incision was made into the periurethral fascia overlying the UD. The diverticular sac was dissected, mobilized off the periurethral fascia, and excised completely after identifying its communication with the urethra. In some cases, the UD was opened and dissected at the level of the bladder neck. A three-layer closure consisting of the urethral wall, the periurethral fascia, and the anterior vaginal wall was made using nonoverlapping sutures. We used a Martius flap interposition between the urethra and the vaginal wall when there was insufficient periurethral fascia or when the operation was performed on recurrent cases. Concomitant anti-incontinence surgery was applied when the patient had severe or urodynamically demonstrable SUI. After repairing the UD, we filled the bladder with normal saline and confirmed that there was no leakage from the urethra or bladder. The Foley catheter was routinely removed 3 d after the operation. However, if the urethral opening was large or the bladder was injured during surgery, the Foley catheter was maintained for 7 d and, if necessary, cystography was performed before removing the Foley catheter. All women were followed up at 3 and 12 mo postoperatively with clinical examinations, urinalysis plus urine culture, and transvaginal USG or pelvic MRI. Cure was defined as the absence of a UD and disappearance of symptoms. Statistical analyses were performed using Fisher s exact test for categorical variables and the Wilcoxon rank sum test for continuous variables to test the predictive factor of surgical failure. Multiple logistic regression was used to detect independent prognostic factors for the outcome of diverticulectomy. A value of p < 0.05 was considered statistically significant. 3. Results The mean age of patients who underwent diverticulectomy was 46.4 yr (range: 31 73 yr). The mean duration of symptoms at the time of presentation was 26.6 mo (range: 1 mo to 10 yr). The presentations were diverse. The most common symptom was dysuria (11 patients, 37%). We observed palpable vaginal masses in seven patients (23%), postvoid dribbling in six patients (20%), SUI in six patients (20%), pus discharge from the urethra in five patients (17%), and perineal pain in five patients (17%). Three women (10%) had histories of recurrent urinary tract infections. On physical examination, an anterior vaginal mass was palpable in 25 patients (83%). In 15 women (50%), purulent material could be extruded from the urethral meatus when the UD was compressed. Preoperative MRI was used to delineate the diverticular anatomies listed in Table 1. Seventeen cases (57%) were simple UDs, three were U-shaped (10%), and 10 were circumferential (33%). Twentythree cases (77%) were located at the proximal urethra and the remaining seven cases (23%) were located at the distal urethra. The UD was single in 26 cases and multiple in four (Fig. 2). The mean maximal diameter of the UDs was 2.4 cm (range: 0.5 5.0 cm). The mean operative time was 87.8 min (range: 20 210 min). The mean postoperative catheterization period and hospital stay were 4.9 3.1 d and 4.1 3.1 d, respectively. A concomitant surgery tension-free vaginal tape (TVT) procedure was performed on one patient who had severe SUI. Five patients with SUI had minimal symptoms and did not want antiincontinence surgery. The initial operative cure rate for UD was 77% (23 of 30 cases); cure rates according to MRI features are shown in Fig. 3. Concomitant SUI was cured in the patient who underwent the TVT procedure. None of the five patients with mild SUI complained of Table 1 Preoperative variation in the anatomy of female urethral diverticula Diverticulum No. of patients (%) Location proximal urethra 23 77 distal urethra 7 23 Number single 26 87 multiple 4 13 Configuration simple 17 57 U-shaped 3 10 circumferential 10 33

european urology 51 (2007) 1664 1670 1667 Fig. 2 A multiple urethral diverticulum showing multiple separate lesions and a discontinuous outline (a, T2-weighted axial image; b, T1-weighted axial image with enhancement). aggravated SUI postoperatively. Of the seven recurrent cases of UD, three who displayed no symptoms of UD were diagnosed using transvaginal USG or pelvic MRI during routine follow-ups. However, four patients complained of symptoms after the operation, and imaging tests confirmed the recurrence of UD. The interval before diagnosis of recurrence ranged from 3 to 12 mo. Three patients in whom recurrent UD was detected only from imaging tests did not require surgery because their symptoms had resolved and the sizes of their UDs had diminished. Four patients who had recurrent UD plus symptoms underwent surgery. Three of four patients were cured after a second operation but one patient was cured after three additional transvaginal diverticulectomies with Martius flap interpositions. Preoperative factors (location, multiplicity, and size of the UD) did not affect the outcome of the operation. However, in cases of circumferential UDs, the success rate of diverticulectomy was significantly lower than that of simple or U-shaped UDs ( p = 0.002; Table 2). Multiple logistic regression tests demonstrated that circumferential UD ( p = 0.03) was an independent factor for a lower cure rate. Complications were bladder injury in two cases (7%) and de novo SUI in two cases (7%). In patients with bladder injury, the bladder was repaired without problems. Of the two patients with de novo SUI, one underwent TVT and was cured and the other did not require anti-incontinence surgery because of minimal symptoms. There were no surgery-related transfusion or wound problems such as urethrovaginal fistulas. 4. Discussion All 30 cases of UD were cured by urethral diverticulectomy. The cure rate after the first operation Fig. 3 The outcomes of surgery of female urethral diverticula according to diverticular configuration.

1668 european urology 51 (2007) 1664 1670 Table 2 Preoperative factors affecting the outcomes of urethral diverticulectomies Variable Not cured (n = 7) Cured (n = 23) Univariate Multivariate Patient s age (yr) * 51.0 8.5 45.0 10.6 0.077 0.208 Symptom duration (mo) * 30.1 23.8 25.6 38.2 0.362 0.618 Size of diverticulum (ml) * 2.8 0.7 2.2 1.2 0.116 0.179 Location Proximal 6 17 1.000 0.985 Distal 1 6 Multiplicity Single 5 21 0.225 0.465 Multiple 2 2 Configuration Simple or U-shaped 1 19 0.002 0.030 Circumferential 6 4 * Mean SD. was 77% and depended on the configuration of the UD as defined by MRI. Female UD is an uncommon surgical entity. It was first described by Hey in 1805 [8] and was rarely reported until the introduction of positive-pressure urethrography in 1956 [9,10]. Female UD can be diagnosed using a combination of the patient s history, physical examination, cystourethroscopy, and imaging. However, diagnosis of UD is often delayed because most patients present with nonspecific lower urinary tract symptoms. The presentations of our UD patients were diverse. More female UDs are now diagnosed than previously because of a high index of clinical suspicion and improved diagnostic techniques [3]. Endorectal and surface coil MRIs have been used to evaluate female UDs [11]. Although MRI is expensive, it is an excellent modality for revealing periurethral and diverticular anatomy because of its multiplanar capabilities and excellent soft-tissue contrast [12,13]. The UD often extends partially or completely around the urethra, which presents a challenging problem to the surgeon because access to the UD may be limited. MRI clearly identifies urethral pathology and provides superior information for surgical planning by accurately delineating the extent of the UD [14]. We routinely performed MRI scans before surgery to determine the configuration of the UD. Several transurethral or transvaginal procedures have been described for the management of symptomatic female UDs [15 17]. Transvaginal diverticulectomy with a vaginal flap, first described by Sholem et al. [18] and improved by Leach et al. [19], has been performed by many surgeons. Currently, complete excision via a transvaginal approach is the most common surgical procedure for the treatment of UD and is considered effective [7]. However, various complications from urethral diverticulectomy have been reported: recurrence occurs in 1 12% of patients, urethrovaginal fistula in 0.9 8.3% of patients, and SUI in 1.7 12.5% of patients [3]. Because diverticulectomy is invasive and considered somewhat difficult, we examined the surgical outcomes and preoperative factors that affect the success of transvaginal diverticulectomy. To achieve satisfactory results, meticulous dissection to define the extent of the UD, complete excision of the diverticular sac and its communication to the urethra, and closure of urethral defects in layers without overlapping sutures or tension are essential [3,7,12]. Several other risk factors affecting surgical success have been described: purulent content, large size, proximal location, lateral or horseshoe shape, delayed diagnosis, and previous urethral surgery [7]. However, in our experience, the location, multiplicity, and size of the UD did not affect the outcome of transvaginal diverticulectomy, but the success rate of the circumferential type was significantly lower than that of the simple or U-shaped types. Diverticular sacs with circumferential configurations may not be completely removed by transvaginal diverticulectomy, especially parts adhering to the anterior portion, causing recurrence of the UD. In our series, 86% of recurrent cases (six of seven cases) were of the circumferential type. It has been suggested that total urethrolysis of circumferential UDs should be performed to identify the entire urethral circumference and facilitate isolation of the UD [7]. However, access to the anterior portion of the diverticular sac can be challenging. Gilbert et al. [20] reported a retropubic approach for treatment of anterior UDs and Clyne and Flood [21] described an approach for an anterior horseshoe-shaped UD using a suprameatal incision. However, a significant portion of the circumferential diverticular sac is

european urology 51 (2007) 1664 1670 1669 located ventrally, so complete excision of a diverticular sac using this approach would require a vaginal counter-incision to access the ventral portion of the diverticular sac. Rovner and Wein [22] reported a high success rate for diverticulectomies of eight circumferential UDs. They completely divided the urethra to access the dorsal wall of the UD with partial urethrectomy and restored urethral continuity by end-to-end urethroplasty in five patients and tabularized the dorsal (anterior) wall of the UD to construct a neourethral segment in three patients. All patients reported subjective relief of severe preoperative pain. We performed transvaginal diverticulectomies and recorded a low cure rate for urethral diverticulectomies of circumferential UDs. In one patient with a circumferential UD, four transvaginal diverticulectomies were necessary to effect a cure. Therefore, we suggest that in cases of circumferential UD, satisfactory results may not be achieved by transvaginal diverticulectomy alone and that adapted treatment methods such as additional incision or urethral reconstruction (diverticular sac urethroplasty or end-to-end urethroplasty) may be necessary. A large UD could affect the sphincter mechanism and result in SUI before or after the operation. However, we did not routinely apply the sling operation during transurethral diverticulectomy. In this study, one of the six patients with preoperative SUI underwent a concomitant TVT procedure because of severe symptoms; her incontinence was cured by the operation. In the five patients who did not undergo concomitant anti-incontinence surgery because of minimal symptoms, incontinence did not worsen after surgery. Only two patients developed de novo SUI. Our data indicate that concomitant anti-incontinence surgery should not be considered mandatory if incontinence is minimal and not urodynamically demonstrable. Although the UD in the patient who underwent TVT was cured after the operation, we were unable to determine whether the TVT influenced the outcome of the diverticulectomy. Our study had several limitations. The follow-up imaging modality was not uniform because it was a retrospective study. We routinely followed up patients using transvaginal USG or pelvic MRI at 3 and 12 mo postoperatively to confirm the disappearance of UDs and to consult with the patients about surgical outcomes and prognoses. However, because not all patients were evaluated postoperatively with MRI, the recurrence rate could have been underestimated. Nevertheless, when patients complained of persistent symptoms after the operation, MRI was applied during follow-up and recurrence was evaluated. Because we did not use validated symptom or quality-of-life questionnaires, symptom improvement could not be quantitatively evaluated. Finally, multiple UDs were difficult to differentiate from single UDs with multiple septa using MRI. If the outline of the UD was smooth and not discontinuous, it was considered single with multiple septa (Fig. 1b, c); if the outline of UD was discontinuous and there were separate lesions, it was considered multiple (Fig. 2). 5. Conclusions The transvaginal approach for diverticulectomy is effective for treatment of female UD. However, the success rate of UDs with circumferential configurations according to MRI was significantly lower than that of simple or U-shaped UDs. This is most likely because complete resection of circumferential UDs is difficult. In cases of circumferential UDs, the surgical procedure must be adapted to achieve complete resection and layered closure. Conflicts of interest Authors disclose that they do not have any commercial relationship such as consultancies, stock ownership or other equity interests, patents received and/or pending, or any commercial relationship which might be in any way considered related to this submitted article. References [1] Andersen MJ. The incidence of diverticula in the female urethra. J Urol 1967;98:96 8. [2] Aspera AM, Rackley RR, Vasavada SP. Contemporary evaluation and management of the female urethral diverticulum. Urol Clin North Am 2002;29:617 24. [3] Ganabathi K, Leach GE, Zimmern PE, Dmochowski R. Experience with the management of urethral diverticulum in 63 women. J Urol 1994;152:1445 52. [4] Gómez Gallo A, Valdevenito Sepúlveda JP, San Martín Montes M. Giant lithiasis in a female urethral diverticulum. Eur Urol 2007;51:556 8. [5] Kim B, Hricak H, Tanagho EA. Diagnosis of urethral diverticula in women: value of MR imaging. AJR Am J Roentgenol 1993;161:809 15. [6] Leng WW, McGuire EJ. Management of female urethral diverticula: a new classification. J Urol 1998;160:1297 300. [7] Fortunato P, Schettini M, Gallucci M. Diagnosis and therapy of the female urethral diverticula. Int Urogynecol J Pelvic Floor Dysfunct 2001;12:51 7. [8] Hey W. Practical observations in surgery. Philadelphia: James Humphreys; 1805. p. 303 5.

1670 european urology 51 (2007) 1664 1670 [9] Davis HJ, Cian LG. Positive pressure urethrography: a new diagnostic method. J Urol 1956;75:753 7. [10] Davis HJ, Telinde RW. Urethral diverticula: an assay of 121 cases. J Urol 1958;80:34 9. [11] Blander DS, Rovner ES, Schnall MD, et al. Endoluminal magnetic resonance imaging in the evaluation of urethral diverticula in women. Urology 2001;57:660 5. [12] Romanzi LJ, Groutz A, Blaivas JG. Urethral diverticulum in women: diverse presentations resulting in diagnostic delay and mismanagement. J Urol 2000;164:428 33. [13] Neitlich JD, Foster Jr HE, Glickman MG, Smith RC. Detection of urethral diverticula in women: comparison of a high resolution fast spin echo technique with double balloon urethrography. J Urol 1998;159:408 10. [14] Siegelman ES, Banner MP, Ramchandani P, Schnall MD. Multicoil MR imaging of symptomatic female urethral and periurethral disease. Radiographics 1997;17:349 65. [15] Ellik M. Diverticulum of the female urethra: a new method of ablation. J Urol 1957;77:243 6. [16] Davis BL, Robinson DG. Diverticula of the female urethra: assay of 120 cases. J Urol 1970;104:850 3. [17] Lapides J. Transurethral treatment of urethral diverticula in women. J Urol 1979;121:736 8. [18] Sholem SL, Wechsler M, Roberts M. Management of the urethral diverticulum in women: a modified operative technique. J Urol 1974;112:485 6. [19] Leach GE, Schmidbauer CP, Hadley HR, Staskin DR, Zimmern P, Raz S. Surgical treatment of female urethral diverticulum. Semin Urol 1986;4:33 42. [20] Gilbert CR, Rivera Cintron FJ. Urethral diverticula in the female: review of the subject and introduction of a different surgical approach. Am J Obstet Gynecol 1954;67: 616 27. [21] Clyne OJ, Flood HD. Giant urethral diverticulum: a novel approach to repair. J Urol 2002;167:1796. [22] Rovner ES, Wein AJ. Diagnosis and reconstruction of the dorsal or circumferential urethral diverticulum. J Urol 2003;170:82 6.