A New Postoperative Predictor of Time to Urinary Continence after Laparoscopic Radical Prostatectomy: The Urine Loss Ratio

Transcription

1 european urology 52 (2007) available at journal homepage: Laparoscopy A New Postoperative Predictor of Time to Urinary Continence after Laparoscopic Radical Prostatectomy: The Urine Loss Ratio Mutlu Ates a, Dogu Teber a, Ali Serdar Gozen a, Ahmet Tefekli a, Marcel Hruza a, Marto Sugiono a, Sarper Erdogan b, Jens Rassweiler a, * a Department of Urology, SLK Klinikum Heilbronn, University of Heidelberg, Germany b Kocaeli University, Faculty of Medicine, Department of Public Health, Kocaeli, Turkey Article info Article history: Accepted December 13, 2006 Published online ahead of print on December 21, 2006 Keywords: Prostate cancer Laparoscopy Radical prostatectomy Continence Urine loss Urine loss ratio Abbreviations: ULR: Urine loss ratio LRP: Laparoscopic radical prostatectomy MV: Micturition volumes UL: Weight of urine loss QoL: Quality of life Abstract Objectives: To assess the correlation of a newly defined parameter, urine loss ratio (ULR), with the time to continence and probability for incontinence after laparoscopic radical prostatectomy (LRP). Materials and methods: A standardized micturition protocol that uses 24-h pad testing to objectively quantify urine loss after removal of the catheter was obtained from 939 patients who underwent LRP and were provided complete follow-up regarding continence. ULR was defined as the weight of urine loss in the pad divided by daily micturition volume, distinguishing between ULR on the first day after catheter removal and the last day of hospital stay. The time to continence was classified as early (0 3 mo), midterm (4 12 mo), and late continence (13 24 mo). Results: Early continence was attained in 69.8% (n = 655) of patients, midterm continence in 18.4% (n = 173), and late continence in 3.5% (n = 33). Of 939 patients in whom first-day ULR was quantified, 495 patients were not discharged immediately and their last-day ULR was quantified (2.3 d following catheter removal). There was a linear correlation between time to continence and ULR, which was more significant for last- than first-day ULR ( p < 0.001). A cutoff point of more than 15% of urine loss indicates a high risk of incontinence (ie, 8-fold for first-day ULR, 55-fold for last-day ULR). Conclusion: ULR predicts the time to continence and may be used to select patients for specific rehabilitation programs and early adjuvant medical therapy, particularly when urine loss exceeds 15%. # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Urology, SLK Klinikum Heilbronn, Am Gesundbrunnen 20, D Heilbronn, Germany. Tel ; Fax: address: jens.rassweiler@slk-kliniken.de (J. Rassweiler) /$ see back matter # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: /j.eururo

2 european urology 52 (2007) Introduction Urinary incontinence is one of the most feared complications of radical prostatectomy. Reported rates range between 2.5% to 87%, and differ considerably according to its definition, follow-up, and surgical technique [1 3]. In a recent metaanalysis, continence rates at 12 mo after laparoscopic radical prostatectomy (LRP) were shown to range between 60% to 97% [1]. Incontinence has a great impact on quality of life (QoL) and causes more concern than the loss of erectile function, especially in older patients. Furthermore, the risk of postoperative incontinence influences some patients to seek alternative treatment options [4,5]. The risk of urinary incontinence following radical prostatectomy has declined in the past decade because of advances in surgical techniques [2]. Despite this improvement, the prevalence of postprostatectomy incontinence has risen, paralleling the increase in the number of surgical procedures performed annually [6]. Therefore, postprostatectomy incontinence has become an important issue in urologic practice, demanding the best efforts for prevention, diagnosis, and treatment. Although the time to continence after removal of a urethral catheter is one of the most frequently asked questions, we still do not have enough predictive parameters. Therefore, we investigated the correlation of urine loss ratios (ULRs) with the time to continence to give a probability for this complication. all patients 24 h after removal of the catheter. The micturition volumes (MVs) and weight of urine loss (UL) in the pads for each void were measured separately every day until discharge. The ULR was calculated by dividing UL by MV (UL/MV). First-day ULR was defined as the ULR on the day of catheter removal. After removal of the catheter, 495 patients were hospitalized for an additional average of 2.3 d (range: 1 11) because of comorbidities; they received the protocol for 24 h prior to discharge from the hospital. The last-day ULR was described as the ULR at the last day of their hospital stay Definition of continence and time to continence In addition to a self-administered modified ICS questionnaire, continence status was evaluated by physical examination including Valsalva or cough stress test at 1 and 3 mo postoperatively, then at quarterly intervals within the first year and semiannually thereafter. The patient characteristics analyzed in this study are shown in Table 1. Patients with no urine leak during coughing or sneezing, as well as those who stayed totally dry, were considered. Patients who were consistently dry but used a safety pad occasionally during normal daily activity (ie, work, exercise, walking) were also considered as. Those who used more than one protective pad per day and/or had urine leak during coughing or sneezing or during the night were considered in. All patients were advised to perform Kegel s exercises after removal of the catheter. None of the patients received any surgical therapy for stress urinary incontinence within 24 mo of the evaluation. The time to continence was classified into three groups: early (within 3 mo), midterm (4 12 mo), and late continence (13 24 mo). 2. Materials and methods 2.1. Patient data Table 1 Patient characteristics All patients We performed 1514 LRPs using the Heilbronn technique [7] from March 1999 to April All patients were prospectively registered in a specific database (Microsoft Excel) accessible to only selected persons through use of a password. Eight hundred ninety patients were followed at least 24 mo; they completed a modified ICS/QoL questionnaire by mail. Another 62 patients were provided a minimal follow-up of 12 mo. Thus, excluding 13 patients with preoperative incontinence, a total of 939 patients with an entire follow-up regarding continence were evaluated. The urethral catheter was generally removed 5 7 d after the operation when there was no leak on cystogram. In case of a leak, patients without significant comorbidities were discharged with an indwelling catheter and underwent a repeat cystogram as an inpatient prior to removal of the catheter. The mean follow-up period was mo (range: 12 81) Micturition protocol and urine loss ratios A standard 24h Micturition Protocol, in which pad testing was used to objectively quantify urine loss, was obtained from Age (43 82) Preoperative PSA ( ) Gleason score (n = 928) <7 505 (54%) (37%) >7 75 (8%) Clinical stage 1a 9 (1%) 1b 20 (2%) 1c 137 (15%) (53%) (29%) Pathologic stage pt0 11 (1%) pt2 545 (58%) pt3 349 (37%) pt4 34 (4%) Nerve sparing No 686 (73%) Unilateral 116 (12%) Bilateral 137 (15%) Catheterization time (4 56) PSA = prostate-specific antigen.

3 180 european urology 52 (2007) day, and and for the last day. The correlation of these ratio groups with the time to continence was investigated by Mantel-Haenszel linear test. Thereafter, the cutoff value of first-day and last-day ULR was defined by determining the point with greatest sensitivity+specifity value for continence and incontinence on the receiver operator characteristic (ROC) curve drawn by the Statistical Package for the Social Sciences (SPSS Inc, Chicago, IL, USA) (Figs. 1 and 2). 3. Results 3.1. Continence Fig. 1 Percentage of cases according to first-day urine loss ratio Data evaluation and statistical analysis ULR parameters obtained for each first- and last-day were compared by Kruskal-Wallis test. The correlation of first-day ULR and last-day ULR with the time to continence was analyzed with the use of the Spearman correlation test. QoL was evaluated by Pearson chi square and Fisher exact tests. ULRs were stratified as , , and for the first The mean urethral catheterization time was d (range: 4 56). In 339 (36%) patients the catheter was kept indwelling for more than 7 d (ie, leak at cystogram). Early continence (0 3 mo) was observed in 655 (69.8%) patients. During followup, 173 (18.4%) patients became midterm (4 12 mo), and 33 (3.5%) became late (13 24 mo). Thirty-two patients who were using only one safety pad all day were considered. They were consistently dry, although they used one pad occasionally for moderate exercise activities. A total of 78 (8.3%) patients with no improvement in urinary control 24 mo after surgery were considered in (43 suffered from grade I, 32 from grade II, and 3 from grade III stress urinary continence). In addition, 38 patients were using two pads, and 40 were using more than two pads a day. Patients who could achieve early continence were significantly younger than the others (ie, 63.4 vs yr); however, this difference is clinically insignificant. Using multivariate analysis, we evaluated the impact of age, preoperative prostate-specific antigen, previous transurethral resection of the prostate, pathologic stage, surgeon, and surgical technique (nerve-sparing and bladder neck sparing procedure) as well as ULR. We were unable to identify any other parameters, except age and ULR, which affected time or ability to achieve continence. For age, first-day ULR and last-day ULR, adjusted odds ratio, and 95% confidence intervals were 1.831, 1.792, and 1.116, 0.984, 1.217, respectively. Patients evaluated for first-day and last-day ULR had similar risk for incontinence (mean ages and ratios of comorbidities were 63.7, 64.1 and 50.5%, 50.7% for the patients evaluated for first-day and last-day ULR, respectively) Urine loss ratio Fig. 2 Percentage of cases according to last-day urine loss ratio. First-day mean ULR values were 0.079, 0.164, 0.306, and in each continence status group,

4 european urology 52 (2007) Table 2 Urine loss ratios according to time groups and pad usage Early Time to continence Midterm Late Patients using 1 security pad In (patients using >1 security pad) p value Age <0.05 First-day ULR (n = 655) (69.8%) (n = 173) (18.4%) (n = 33) (3.5%) (n = 32) (3.4%) (n = 78) (8.3%) <0.001 Last-day ULR (n = 300) (60.6%) (n = 120) (24.2%) (n = 22) (4.4%) (n = 22) (4.4%) (n = 53) (10.7%) <0.001 ULR = urine loss ratio. Fig. 3 Correlation with the increase in first-day ULR and time to continence. respectively (Table 2, p < 0.001). There was a linear correlation (R square = 0.37) with the increase in firstday ULR and time to continence (Fig. 3). The mean hospitalization period of patients who were not discharged immediately after urethral catheter removal (n = 495) was 2.3 d (range: 1 11). In these patients, mean last-day ULR values were 0.034, 0.077, 0.184, and in each continence status group, respectively (Table 2, p < 0.001). There was also a linear correlation with the increase in last-day ULR (R square = 0.56) and time to continence (Fig. 4). The correlation of first-day ULR and last-day ULR with the time to continence was and 0.603, respectively, which is also similar for patients younger and older than 65 (Table 3, p < 0.001). In a comparison of ULR according to pad usage, patients using more than one pad had higher ULRs (Table 2, p < 0.001). Patients with a security pad had a ULR value below (ie, better than) the late continence group. Using the first-day ULR values of <5%, 5 15%, and >15%, we found that the probability of becoming early, midterm, late, or in differed significantly. Increase in first-day ULR decreases the probability of early continence, but increases the probability of later Table 3 The correlation coefficient of first- and last-day ULR with the time to continence according to age groups All groups Patients 65 yr Patients >65 yr First-day ULR Last-day ULR ULR = urine loss ratio.

5 182 european urology 52 (2007) Fig. 4 Correlation with the increase in last-day ULR and time to continence. Table 4 Probabilities of continence time using first-day urine loss ratio First-day urine loss ratio Continence Continence time (n = 339) (n = 313) (n = 287) Incontinence Early Midterm Late 303 (89.4%) 230 (73.5%) 122 (42.5%) 29 (8.5%) 59 (18.9%) 85 (29.6%) 3 (0.9%) 8 (2.5%) 22 (7.7%) continence (midterm or late) (Table 4, p < 0.001). Similarly, the probability of time to continence did differ significantly for last-day ULR values, ranging between 0 to and 0.05 to 1 (Table 5, p < 0.001). Increase in last-day ULR was also associated with a diminished probability of early continence. For an estimated risk of how many times the probability of later continence or even incontinence 4 (1.2%) 16 (5.1%) 58 (20.2%) increases, we found a first-day ULR cutoff value of This value represents the peak point of the sums of sensitivity and specificity values for continence using ROC curves (Figs. 1 and 2): If urine loss exceeds 15%, the probability of later continence/ incontinence increases 8-fold. Using the same cutoff value for last-day ULR, we found that the probability increased even 55-fold. Table 5 Probabilities of continence time using last-day urine loss ratio Last-day urine loss ratio Continence Continence time (n = 285) (n = 210) Incontinence Early Midterm Late 228 (80%) 72 (34.3%) 48 (16.9%) 72 (34.3%) 5 (1.7%) 17 (8.1%) 4 (1.4%) 49 (23.3%)

6 european urology 52 (2007) Table 6 Evaluation of quality of life according to continence status based on self-administrated questionnaire Time to continence Continent In p value Early Midterm Late Do it again? Yes 634 (97%) 165 (95%) 25 (76%) 824 (96%) 64 (82%) <0.01 No 21 (3%) 8 (5%) 8 (24%) 37 (4%) 14 (18%) Life quality? Better/same 353 (54%) 91 (53%) 7 (21%) 451 (52%) 14 (18%) <0.01 Worse 302 (46%) 82 (47%) 26 (79%) 410 (48%) 64 (82%) Self-appraisal Continent 548 (84%) 90 (52%) 3 (9%) 641 (75%) 0 <0.01 Minimal incontinence 107 (16%) 83 (48%) 27 (82%) 217 (25%) 42 (54%) In (9%) 3 (0.3%) 36 (46%) Satisfaction Happy 602 (92%) 116 (67%) 12 (36%) 730 (85%) 8 (10%) <0.01 Unhappy 53 (8%) 57 (33%) 21 (64%) 131 (15%) 70 (90%) Total Quality of life data Our QoL results correlated significantly with the objective continence data (Table 6). Whereas overall QoL and the consequence to choose the procedure again were not influenced by the time to continence, patient satisfaction depended significantly on this fact. 4. Discussion Urinary continence after radical prostatectomy is a major concern with a great impact on QoL. Although about half of the patients achieve full continence within the first 3 mo, and most are dry at 1 yr, some patients may need up to 2 yr to recover [3]. According to the literature, 69.8% of our patients were 3 mo after surgery [3,8]. Moreover patient satisfaction is strongly correlated to early continence (Table 6). Laparoscopy may offer an improved identification of anatomic landmarks such as striated muscles and neurovascular bundles, resulting in minimized damage to the striated sphincter. All these efforts may have an impact on postoperative continence rates in addition to other functional and oncologic outcomes [1,9]. Following radical prostatectomy, only adequate function of the striated urethral sphincter guarantees continence [10]. Therefore, incontinence is inevitable after injuries to the distal urethral sphincter, resulting in internal sphincter deficiency. Since previous studies have shown that a combination of neurologic and anatomic factors play a role in postprostatectomy incontinence, we did not intend to stratify patients but we excluded patients with preoperative (mainly urge) incontinence [11]. According to Catalona [12], independently of the other investigated parameters, the main predisposing risk factor for postoperative incontinence appears to be age with a cutoff of 70 yr: On the basis of the logistic model including only age, return of continence was 2-fold more likely with each 10-yr decrease in age. There was no association between postoperative continence and the incidence of nerve-sparing surgery, tumor characteristics, history of transurethral resection of prostate, number of prior radical prostatectomies performed by the surgeon, or follow-up interval [12]. Apart from this finding, some authors consider that certain technical modifications appear to facilitate preservation of continence, independently of the approach [13 20]. Nevertheless, the surgical technique may still matter, and technical refinement has been shown to result in better continence rates [20]. Evidently the level of expertise significantly impacts functional results, particularly concerning early recovery of continence. This finding might be related to an improved apical and periprostatic anatomic dissection between the levator and prostatic fascia rather than to other specific technical steps [21]. Although the impact of any surgical modifications on postoperative continence is self-evident, it is very difficult to evaluate. According to Janetschek et al [22] who compared open and laparoscopic prostatectomy, the LRP group scored less well for early continence rates than the open group, but this phenomenon could be ascribed to the learning curve. As in our previously published report [23], we found no significant difference for 12-mo continence rates between our early and more recent 219 cases. Despite all these technical efforts, none of the parameters evaluated, except age, was able to give a probability for the time to continence. The importance of age is understandable in light of Oelrich s

7 184 european urology 52 (2007) [24] classic findings that as men age, the striated sphincter becomes progressively invaded by extensive vascular and connective tissue, producing a cavernous-like appearance. In the present study, patients in the early group were also significantly younger than the other patients; however, the age difference was clinically insignificant. Conclusively, the correlation between ULR and time to continence was similar between those younger than 65 yr and those older than 65. We therefore did not stratify our patients into groups according to age. Of course, it would be ideal to predict the development of continence prior to surgery, but according to the literature this seems to be impossible. Nevertheless, patients who have undergone a radical prostatectomy want to know when they have achieved full continence. In this scenario, we evaluated ULR as an independent postoperative parameter for continence after LRP. In light of a significant correlation between firstday ULR and the time to continence, Table 4 distinguishes certain probabilities for time to continence in patients whose first-day ULRs have been calculated: If the first-day ULR was less than 0.05, 89.4% achieved early continence. However the probability of early continence diminished to 73.5% and 42.5% with first-day ULRs of and , respectively. Although not available for all patients, last-day ULR provided an even better correlation with the time to continence (Table 5): If the last-day ULR was less than 0.05, 80% became within 3 mo and the risk of incontinence was only 1.4%. However, the probability of early decreased to 34% and the risk of incontinence increased to 23% if the last-day ULR ranged between 0.05 and 1. First-day and last-day ULRs also correlated significantly with pad usage. Patients using more than one pad had higher first-day and last-day ULRs (Table 2). This finding underlines that ULR may be useful for predicting high-risk patients (ie, using more than one pad). What are the reasons for the better predictive value of last-day ULR? The mean time interval between first- and last-day ULR measurements was 2.35 d in the present study. As mentioned above, sphincter weakness was assumed to be the most important factor for incontinence after radical prostatectomy. The recovery of the external sphincter after radical prostatectomy was shown to begin a few days after catheter removal. Furthermore, urodynamic studies demonstrated that maximum urethral closure pressure at rest and maximum voluntary sphincter contraction were significantly better in patients than in those with incontinence after radical prostatectomy [5,10]. For continence to be achieved, an initial period is needed until recovery of the external sphincter. This finding might be a partial explanation for lastday ULR correlating more significantly with time to continence than the first-day ULR. However, the mean interval of 2.35 d is definitely too short for the complete recovery of external sphincter muscles. Therefore, future studies should focus on ULR measured a few weeks (4 6 wk) after removal of catheter to determine the ideal predictive interval after surgery. Additionally, early continence on the first-day due to the edema at the anastomosis may give false-positive results (ie, first-day ULR). Patients evaluated for last-day ULR might have a higher risk for incontinence in spite of having statistically similar age and comorbidities. However, because correlation coefficients of ULR and time to continence were similar for young and old patients, ULR can be used independently for all old or young patients with or without comorbidities (Table 3). The hospitalization period, which is relatively long in the present series, can vary significantly according to different national health care and reimbursement systems. The German Health Care System actually supports patients being hospitalized until the catheter is removed. However, hospitalization is not required to receive the micturition protocol, because patients can easily measure urine output and the weight of their pads at home to assess the ULR. We currently offer all our patients the opportunity to perform this protocol at home. Interestingly there was no linear relationship between ULR and time to continence. The highest sensitivity and specificity of predicting postoperative incontinence using ROC curves was found at a level of 0.15 ULR. Above this level the risk of incontinence increased 8-fold (first-day ULR) and 55-fold (last-day ULR). It might be difficult, at least at this time, to use ULRs as selection criteria for any adjuvant therapy, but using ULRs could be useful in identifying patients who were unlikely to regain continence so that earlier interventions (collagen, sling, artificial sphincter) might be recommended to improve QoL parameters. On the other hand, our patients appreciate the information about their individual prognosis of continence based on the ULR data. Future studies may focus on long-term ULR measurements, the impact of ULR results on adjuvant treatment (medication, rehabilitation), and nomograms predicting postoperative continence including ULR together with other parameters (ie, age, operative technique, surgeon, tumor stage).

8 european urology 52 (2007) Conclusions Urine loss ratios, which can be easily assessed after removal of the urethral catheter, can be used to predict the time to continence. Both first-day and last-day ULRs that exceed 15% are significantly associated with an increased risk for incontinence, while lower ratios predict achievement of early continence after LRP. These newly defined parameters may be used for the selection of patients for early adjuvant therapy including rehabilitation programs or medical treatments (ie, off-label use of duloxetin). Conflicts of interest In this article there is no funding or any disclosure to companies, except from the fact, that the study was sponsored by the EUSP of the EAU. References [1] Rassweiler J, Schulze M, Teber D, et al. Laparoscopic radical prostatectomy: functional and oncological outcomes. Curr Opin Urol 2004;14: [2] Steiner MS, Morton RA, Walsch PC, et al. Impact of anatomical radical prostatectomy on urinary incontinence. J Urol 1991;145: [3] Hammerer P, Huland H. Uroynamic evalution on changes in urinary control after radical retropubic prostatectomy. J Urol 1997;157: [4] Feneley MR, Gillatt DA, Hehir M, et al. A review of radical prostatectomy from three centres in the UK: clinical presentation and outcome. BJU 1996;78: [5] Kundu SD, Roehl KA, Eggener SE, et al. Potency, continence and complications in 3,477 consecutive radical prostatectomies. J Urol 2004;172: [6] Lu-Yao GL, McLerran D, Wasson J, et al. An assessment of radical prostatectomy: Time trends, geographic variation and outcomes. The prostate patient outcomes research team. JAMA 1993;269: [7] Rassweiler J, Marrero R, Hammady A, et al. transperitoneal laparoscopic radical prostatectomy: ascending technique. J Endourol 2004;18: [8] Rassweiler J, Stolzenburg J, Sulser T, et al. Laparoscopic radical prostatectomy the experience of the German Laparoscopic Working Group. Eur Urol 2006;49: [9] Stolzenburg JU, Rabenalt R, Tannapfel A, et al. Intrafascial nerve-sparing endoscopic extraperitoneal radical prostatectomy. Urology 2006;67: [10] Majoros A, Bach D, Kesszthelyi A, et al. Urinary incontinence and voiding dysfunction after radical retropubic prostatectomy (prospective urudynamic study). Neurourol Urodyn 2006;25:2 7. [11] Giannantoni A, Mearini E, Stasi SM, et al. Assessment of bladder and urethral sphincter function before and after radical retropubic prostetectomy. J Urol 2004; 171: [12] Catalona WJ, Carvalhal GF, Mager DE, et al. Potency, continence and complication rates in 1,870 consecutive radical prostatectomies. J Urol 1999;162: [13] Klein EA. Early continence after radical prostatectomy. J Urol 1992;148:92 5. [14] Walsh P. Radical prostatectomy for localized prostate cancer provides durable cancer control with excellent qualitiy of life: a structured debate. J Urol 2000;163: [15] Heinzer H, Graefen M, Noldus J, et al. Early complications of anatomical radical retropubic prostatectomy: lessons from a single-center experience. Urol Int 1997;59:30 3. [16] Kaye KW, Creed KE, Wilson GJ, et al. Urinary continence after radical retropubic prostatectomy. Analysis and synthesis of contributing factors: a unified concept. BJU 1997;80: [17] Myers RP. Improving the exposure of the prostate in radical retropubic prostatectomy: longitudinal bunching of the deep venous plexus. J Urol 1989;142: [18] Stolzenburg J-U, Liatsikos EN, Rabenalt R, et al. Nerve sparing endoscopic radical prostatectomy effect of puboprostatic ligament preservation on early continence and positive margins. Eur Urol 2006;49: [19] Katz R, Salomon L, Hoznek A, et al. Positive surgical margins in laparoscopic radical prostatectomy: the impact of apical dissection, bladder neck remodelling and nerve preservation. J Urol 2003;169: [20] Touijer K, Guillonneau B. Laparoscopic radical prostatectomy: A critical analysis of surgical quality. Eur Urol 2006; 49: [21] Rassweiler J, Hruza M, Teber D, Su L-M. Laparoscopic and robotic assisted radical prostatectomy critical analysis of the results. Eur Urol 2006;49: [22] Janetschek G, Montorsi F. Open versus laparoscopic radical prostatectomy. Eur Urol Suppl 2006;5: [23] Rassweiler J, Seemann O, Schulze M, et al. Laparoscopic versus open radical prostatectomy: A comparative study at a single institution. J Urol 2003;168: [24] Oelrich TM. The urethral sphincter muscle in the male. Am J Anat 1980;158: