Presentation,Endoscopic Management,and Significance of Hemostatic Clip Migration into the Lower Urinary Tract Following Radical Prostatectomy

Introduction

Radical prostatectomy (RP) remains an accepted and widely used treatment modality for clinically localized adenocarcinoma of the prostate. Due to the desire for athermal hemostasis near the neurovascular bundles and urethral rhabdosphincter, hemostatic clips are often used during RP. Although these clips provide reliable and efficient hemostasis, there have been reported cases of clip migration into the bladder and the rectum that may present years after RP.^1–8

Hemostatic clip migration (HCM) into the lower urinary tract is a relatively rare complication of RP that has only been described in case reports and very small series. To date, the number of cases of HCM after RP reported in the literature is fewer than 20 patients. As such, HCM after RP remains poorly described and anecdotal in nature. Furthermore, although HCM has been associated with bladder neck contracture (BNC), most BNCs occur without findings of HCM.^7,9

Due to its rarity, little is known about the presenting symptoms, preferred treatment, and subsequent functional outcomes patients with HCM experience after management. The purpose of this study was to evaluate initial presentation, possible risk factors, management modalities, and functional outcomes of treatment for HCM occurring after RP.

Patients and Methods

After institutional board review, we reviewed our institutional management of HCM after RP from 1977 to 2012. All patients who underwent prostatectomy at our institution or were referred to us for voiding symptoms after prostatectomy elsewhere and found to have HCM were included. Patient characteristics, RP pathology, and perioperative complications of RP were assessed. Additionally, operative reports were reviewed where available to assess estimated blood loss, RP duration, and comments on intraoperative complexity to act as surrogates for RP difficulty.

Multiple surgeons performed RP over the study interval using standard operative techniques that have been described previously.^10,11 All robot-assisted RP (RARP) procedures were performed with a transperitoneal approach. Although impacted by intraoperative findings and need for hemostasis, clips were generally placed on the vas deferens, vessels near the seminal vesicles, and over the lateral pedicles of the prostate.

For patients undergoing endoscopic clip removal, all procedures were performed transurethrally with the patient under general anesthetic. Clips were removed with grasping forceps. In the case of difficulty in removing the clip, a holmium laser was used to free the clip from the surrounding tissue. HCM management and outcomes were reviewed. Statistical analysis was performed using standard commercial software.

Results

In all, we identified 17 patients with HCM, 15 of whom underwent RP at our institution and 2 referred after RP performed elsewhere. Patient characteristics are summarized in Table 1. Median patient age was 66 (range, 48–79) years. Nine patients (53%) had undergone prior RARP, and 8 (47%) had received radical retropubic prostatectomy. Operative difficulty was noted in 3 cases (19%). One patient had undergone a prior attempted RARP, which was aborted due to the large size of the prostate; he subsequently underwent completion RARP at our institution, and significant scarring was noted at the time of surgery. One patient was noted to have a very large prostate intraoperatively, which increased the technical difficulty of RP, and another had extremely scarred posterior planes for which the seminal vesicles were unable to be taken easily. Median RP operative time was 192 (range, 135–240) minutes, and estimated blood loss was 237 (range, 150–900) mL. Postoperative anastomotic leak occurred in 2 patients. Of 14 patients with available pathology reports, final prostate cancer pathologic stage was T2 in 10 and T3 in 4. Positive margins were noted in 3 patients. Three patients (18%) had undergone external beam radiation therapy (EBRT) prior to HCM diagnosis.

Presenting symptoms are reported in Table 2. Sixteen (94%) patients with HCM were symptomatic at initial presentation. One asymptomatic patient was found to have evidence of HCM when a catheter could not be placed into the bladder. Fourteen (82%) had predominantly voiding symptoms, including obstructive or irritative voiding and dysuria. Three (18%) had perineal pain, 2 (12%) had hematuria, and 2 (12%) had urinary tract infection. Five (29%) patients had concomitant BNC. The median time from RP to diagnosis of HCM was 8 (range, 1–252) months.

Management and outcomes are summarized in Table 3. Fifteen patients (88%) underwent endoscopic clip removal for HCM, with 2 electing to forego intervention (Figs. 1 and 2). Of the 5 patients with concomitant BNC, 3 underwent dilation, 1 received a transurethral incision, and 1 underwent transurethral resection at the time of clip removal. Hem-o-lok^® (Weck Surgical Instruments, Teleflex Medical, Research Triangle Park, NC) clips were identified in 9 (53%) patients versus metal clips in 8 (47%).

OPEN IN VIEWER

FIG. 1.

Clip erosion into the vesicourethral anastomosis following robot-assisted radical prostatectomy.

OPEN IN VIEWER

FIG. 2.

Open vesicourethral anastomosis following successful transurethral clip removal.

At a median follow-up of 1.6 years, 13 (87%) patients experienced complete symptom resolution following endoscopic removal of clips. Two (13%) patients required repeat endoscopic removal for additional clip migration into the lower urinary tract. Two of 5 patients with concomitant BNC had recurrence of BNC during follow-up and required urethral dilation. Worsening urinary incontinence and ongoing urinary tract infections were reported by 1 patient each following clip removal. At latest follow-up, no patient had undergone artificial genitourinary sphincter placement for stress urinary incontinence.

Discussion

HCM is a long-term potential complication of both open and robotic RP. In this study, we present what we believe to be the largest experience of HCM into the lower urinary tract following RP. Overall, our experience suggests that most patients are symptomatic at presentation and that complete endoscopic removal of eroded clips will alleviate symptoms in a vast majority of patients with minimal impact on functional outcome.

Although stone formation around suture within the bladder had been previously reported, Miedema and Redman ¹² reported the first case of intravesical or intraurethral HCM following RP. They noted significant urinary symptoms in their patient prompting evaluation but did not comment on impact of endoscopic removal of the clip. In contrast to the mechanism of postoperative clip erosion through tissues, another early report of HCM into the lower urinary tract was described by Landrigan et al. ¹³ where inadvertent closure of the clips into the vesicourethral anastomosis was thought to be the cause. With the development of RARP and increased use of Hem-o-lok clips, additional reports surfaced describing a similar phenomenon of plastic clip migration into the urinary tract. ¹

Regardless of the proposed mechanism of clips entering the urinary tract following RP, most available reports suggest patients are symptomatic with voiding complaints or perineal pain upon presentation. Our results were consistent with this as all but 1 patient described symptoms prompting evaluation and the subsequent diagnosis of HCM. The single asymptomatic patient could not be catheterized for a procedure, and the clip was recognized at the time of cystoscopic catheter placement. Thus the clinician should keep a high degree of suspicion for HCM in the patient with intractable urinary symptoms following RP.

The influence of HCM on the development of BNC remains poorly defined. Prior studies have alluded to the possible causality of HCM with BNC. Moser and Narepalem³ reported on 2 patients with BNC found to have HCM at the vesicourethral anastomosis. Both cases occurred within their institution's first 50 cases of RARP, prompting increased attentiveness to the placement location and judicious limitation of the use of clips. Additionally, Blumenthal et al. ⁹ described their experience with BNC after review of their database of patients undergoing RARP. Of 524 consecutive patients, 4 patients were identified as having BNC, with 2 having concomitant HCM into the lower urinary tract. They concluded that erosion was possibly occurring secondary to clips placed at the midline when ligating the seminal vesicle and vasa arteries. After a practice change limiting clip placement to only the lateral prostatic pedicles, they reported no additional HCM. Yi et al. ⁷ reviewed their database of 641 patients undergoing RP to identify clip-related complications. Due to the low rate of BNC after RARP and 2 of 2 patients with BNC having concomitant HCM at the vesicourethral anastomosis, they concluded that any patient with BNC after RARP should be suspected of having HCM.

We did not specifically look into the overall incidence rate of BNC and HCM in our series. We did find that 5 patients (29%) had simultaneous BNC, further supporting the possible association of BNC in clip migration cases. Of the 5 patients, 1 had a postoperative urine leak leading to a longer indwelling catheter duration and prior EBRT, 1 was noted to have had a prior attempted prostatectomy, making RARP difficult, and prior EBRT, 1 had EBRT alone, and 2 had no identifiable risk factors for BNC. However, it is still unclear if HCM can be added to the list of risk factors for BNC or if shared risk factors such as EBRT, perioperative blood loss, and postoperative urine leak contribute to both independently. One could imagine the erosion of a foreign body into the urinary tract could cause an inflammatory reaction and subsequent scar formation as suggested by Blumenthal et al., ⁹ especially if tissue damage is exacerbated by EBRT. The overall consensus is to limit clip usage surrounding the vesicourethral anastomosis. Fortunately, although BNC recurred in 2 patients, no patients currently require self-intermittent catheterization or balloon dilation to maintain patency of the urethral lumen.

As HCM remains a rare postoperative event, it is difficult to determine actual causes of erosion or migration into the bladder or urethra. Surely pelvic radiation treatment could affect tissue healing and predispose to clip migration. Additionally, technical considerations should be considered, especially limiting clip placement during vasal artery and lateral prostatic pedicle ligation. Finally, the surgeon must take care not to incorporate a previously placed clip into the anastomosis when bringing the bladder neck to the urethra during vesicourethral anastomosis.

In most prior reports, HCM was treated with transurethral endoscopic removal. Previous studies have described operative ease ranging from simple high-flow washing of the urethra and subsequent dislodgement of the clip into the bladder to the need for assistance with a holmium laser for cystolitholapaxy and clip removal.^3,6 Most commonly, stone-grasping forceps are used. Although rarely reported in the literature, spontaneous passage of a clip is possible and was found to have occurred in 1 patient in our series. ² No patients needed laser excision of clips or stone material. The transurethral removal of symptomatic clips remains an excellent treatment modality with reasonable outcomes. In all, a vast majority of our patients described complete symptom resolution with clip removal. Of those who remained symptomatic, 1 patient experienced ongoing perineal pain, and another had recurrent urinary tract infections. Further evaluation in both demonstrated no additional evidence of HCM. After clip removal, 1 patient complained of worsening urinary incontinence but elected to forego further management. Overall, symptomatic patients can be counseled that transurethral removal of the clip will very likely alleviate symptoms and have little impact on functional concern.

We acknowledge the limitations of our study, especially being a single-institution experience with limited size and relatively short follow-up. Additionally, our practice often acts as a referral center, with many patients opting for local surveillance if offered. We do typically follow-up in person with patients at 3 months after RP, but often additional follow-up is defined based on travel and functional concerns. Another limitation is the lack of standardized assessment of functional outcomes using validated questionnaires in follow-up.

Conclusions

HCM after RP is often symptomatic with voiding symptoms, perineal pain, hematuria, infection, or BNC. Although it is difficult to definitively assign causality, HCM appears to be associated with BNC. Fortunately, clips can be removed endoscopically with expected symptom resolution in the vast majority of patients.