bs_bs_banner

International Journal of Urology (2014) 21, 781–785

doi: 10.1111/iju.12447

Original Article: Clinical Investigation

Robot-assisted radical prostatectomy in the Korean population: A 5-year propensity-score matched comparative analysis versus open radical prostatectomy Kyo Chul Koo, Patrick Tuliao, Young Eun Yoon, Byung Ha Chung, Sung Joon Hong, Seung Choul Yang and Koon Ho Rha Department of Urology and Urological Science Institute, Yonsei University College of Medicine, Seoul, Korea

Abbreviations & Acronyms BCR = biochemical recurrence BCRFS = biochemical recurrence-free survival CSS = cancer-specific survival LN = lymph node LND = lymph node dissection ORP = open radical prostatectomy OS = overall survival PCa = prostate cancer PLND = pelvic lymph node dissection PSA = prostate-specific antigen PSM = positive surgical margin RALP = robot-assisted laparoscopic radical prostatectomy RP = radical prostatectomy Correspondence: Koon Ho Rha M.D., Ph.D., Department of Urology and Urological Science Institute, Yonsei University College of Medicine, 134 Shinchon-dong, Seodaemun-gu, Seoul 120-752, Korea. Email: [email protected] Received 2 November 2013; accepted 18 February 2014. Online publication 24 March 2014

© 2014 The Japanese Urological Association

Objectives: To compare the oncological outcomes of robot-assisted laparoscopic radical prostatectomy with those of open radical prostatectomy in contemporary Korean prostate cancer patients. Methods: From a group of 1172 patients consisting of 592 (50.5%) robot-assisted laparoscopic radical prostatectomy and 580 (49.5%) open radical prostatectomy cases carried out between 1992 and 2008, 175 robot-assisted laparoscopic radical prostatectomy cases were matched with an equal number of open radical prostatectomy cases by propensity scoring based on patient age, preoperative prostate-specific antigen, biopsy Gleason score and clinical tumor stage. Competing-risks survival analyses were used to evaluate oncological outcomes, including rates of positive surgical margin, biochemical-recurrence, adjuvant therapy, cancer-specific survival, overall survival and metastasis-free survival during the mean follow up of 58.4 months. Results: Positive surgical margin rates were comparable between robot-assisted laparoscopic radical prostatectomy and open radical prostatectomy cohorts (19.4% vs 21.8%), with comparable rates for all pathological stages and risk subgroups. Positive surgical margin rates according to location were comparable, with the apical margin being the most common location. Robot-assisted laparoscopic radical prostatectomy recovered higher lymph node yields compared with open radical prostatectomy (12.5 vs 3.8; P < 0.001). The robot-assisted laparoscopic radical prostatectomy and the open radical prostatectomy groups showed equal oncological outcomes regarding 5-year biochemical recurrence-free survival (log-rank P = 0.651), metastasis-free survival (log-rank P = 0.876), cancer-specific survival (log-rank P = 0.076) and overall survival (log-rank P = 0.648), respectively. Between groups, there was no difference in the rate of adjuvant therapy, time to first adjuvant therapy failure or in the rate of subsequent secondary treatment. Conclusions: Robot-assisted laparoscopic radical prostatectomy represents an effective surgical approach for the treatment of prostate cancer in the Korean population, as it provides equivalent oncological outcomes to open radical prostatectomy.

Key words: prostate cancer, prostatectomy, robotics, treatment outcome.

Introduction RP has been shown to provide favorable oncological control and prolonged survival for localized PCa by reducing the risk of metastasis and local tumor progression.1 There has been a rapid shift from the clinical use of conventional ORP to RALP, accounting for as high as 80% of RP in the USA in 2011.2 In line with a rapid increase of localized PCa incidence in Asia, a dramatic increase in the number of RALP has been seen; however, there is a paucity of data regarding long-term oncological outcomes of RALP carried out in an Asian cohort.3,4 Despite the advantages of the robotic system, there remains controversy about whether RALP is as effective as ORP in terms of oncological outcome. Indeed, the adoption of RALP into clinical practice has been too recent to confirm its performance regarding cancer control. Comparisons of oncological outcomes between ORP and RALP have been carried out by few studies, which are limited by insufficient outcome variables or relatively short follow-up periods.5,6 Although the ideal approach for comparing competing treatment modalities is a randomized controlled trial, such studies are mostly limited to observing long-term oncological outcomes. A feasible alternative is a propensity-score matched analysis, in which the possibility 781

KC KOO ET AL.

of the selection bias arising from the difference between the two groups can be excluded by adjusting for the multiple preoperative confounding variables that affect oncological outcomes. We aimed to carry out a propensity-based comparison of oncological outcomes between a contemporary cohort of Korean patients undergoing ORP and RALP in the robotic era. As the first report of an Asian cohort, the present findings imply that RALP is a feasible option, providing comparable oncological outcomes to those of ORP.

Methods Patient selection Among a group of 1172 consecutive patients who underwent radical prostatectomy between 1992 and 2008, 548 (46.8%) patients received ORP and 624 (53.2%) patients received RALP. We carried out the first RALP in the Department of Urology, Severance Hospital, Seoul, Korea, in 2005, and have been carrying it out routinely since then using a four-arm da Vinci Surgical System (Intuitive Surgical, Sunnyvale, CA, USA). The RALP and the ORP included in the present study were each carried out by two surgeons. Surgery was carried out for localized and locally advanced PCa patients without distant metastases who either desired surgical treatment or were determined to be reasonable surgical candidates because of otherwise favorable clinical characteristics. The decision on surgical modality was based on the surgeon’s discretion and on the patient’s preference. Men treated with neoadjuvant hormonal therapy or men with incomplete preoperative information including preoperative PSA, biopsy Gleason score and clinical tumor stage were excluded from the study. Furthermore, in an attempt to normalize surgeon learning curves, we excluded both RALP and ORP cases that were completed before each surgeon had carried out at least 100 cases per procedure. This resulted in an overall study population of 827 cases, consisting of 521 (63%) RALP and 306 (37%) ORP. To account for the inherent differences between patients undergoing ORP and RALP, a 1:1 propensitybased matching was carried out for patient age, preoperative PSA, biopsy Gleason score and clinical tumor stage, resulting with 175 patients in each group. All prostate specimens were serially sectioned, processed and confirmed by a single uropathologist, with a PSM being defined as cancer at the inked margin. The operating surgeon monitored all patients with PSA measurements every 3 months for the first year and every 3–6 months thereafter.

End-points We selected our primary end-point to be BCR, defined as the first of two or more consecutive increases in PSA >0.2 ng/mL at least 3 months after surgery. Secondary end-points were the rate of distant metastasis, CSS, OS and continence recovery. For all patients, the status of survival and cause of death were investigated based on the National Cancer Registry Database or by institutional electronic medical records. Urinary continence was assessed by self-completed questionnaires, defined as the absence of pads. The mean follow-up period of the overall cohort was 58.4 months. The study was carried out in lieu 782

of a formal ethics committee, followed by the principles of the Helsinki Declaration.

Statistical analysis Propensity scores were calculated for each participant using multivariable logistic regression based on the following covariates: patient age, preoperative PSA, biopsy Gleason score and clinical tumor stage. Propensity score-based matching was carried out to control for imbalances in confounding factors among distinct study cohorts. Continuous and categorical factors were combined to yield a propensity score for each individual in the study cohort. Individuals in each of the different study cohorts were matched to patients in the reference cohort based on their calculated propensity scores. The greatest advantage of this method is that variables are weighted by relative importance rather than assigned equal weights. Furthermore, cohort means and standard deviations related to covariates were equivalent when composite propensity scores were matched.7 Matching was carried out to select for the most similar propensity scores across each surgical group in a 1:1 ratio. SPSS version 18 (SPSS, Chicago, IL, USA) was used for all analyses. Appropriate comparative tests; that is, Student’s t-test, χ2-test and ANOVA, were used to compare oncological outcomes, which included the rate of PSM, BCR, adjuvant therapy, distant metastasis, and cancer-specific and other-cause mortalities. The time-dependent variables of BCRFS, CSS, OS, and continence recovery were estimated and compared using the Kaplan–Meier method and a log–rank test. All tests were twosided, with a statistical significance set at P < 0.05.

Results Clinicopathological characteristics of the two groups for matched and unmatched variables are presented in Table 1. No significant differences existed between the two groups with respect to the variables used for propensity-score matching, confirmed by a comparable distribution of the D’Amico risk subgroups. Except for two patients in the ORP group, all patients received RP at a contemporary period. The mean follow-up duration was 57.6 months and 60.1 months for the RALP and the ORP cohorts, respectively. Examination of postoperative variables showed that the overall PSM rates were comparable between the RALP and ORP cohorts (29.1% vs 30.8%; P = 0.911), with comparable rates between the risk and tumor stage subgroups (Table 2). PSM rates according to location were also comparable between the two surgical groups (P = 0.521). The apical margin was the most common location, followed by the base and lateral margins (data not shown). In addition to analyzing PSM rates, we also examined at LN yields between groups (Table 1). RALP achieved a greater LN yield after PLND compared with ORP (12.5 vs 3.8; P < 0.001). However, the rates of metastatic nodal staging were comparable (P = 0.502). Between groups, there was no difference in the rate of extracapsular extension, the rate of seminal vesicle invasion, the distribution of the pathological Gleason score or tumor stage. Oncological outcomes stratified by the surgical approach are presented in Table 3 and Figure 1. No differences existed © 2014 The Japanese Urological Association

Feasibility of RALP

Table 1 Clinicopathological characteristics of matched ORP and RALP patients Surgical technique

Propensity-matched variables Age (years) PSA (ng/mL) Biopsy Gleason score (%) ≤6 7 8–9 Clinical T stage (%) cT1 cT2 cT3 cT4 Unmatched variables D’Amico risk criteria (%) Low Intermediate High Pathological characteristics Pathological Gleason score (%) ≤6 7 8–9 Pathological T stage (%) pT2 pT3 pT4 Tumor volume (cc) Extraprostatic extension (%) Seminal vesicle invasion (%) PSM (%) Total harvested lymph nodes (n) Positive pN stage (%)

ORP (n = 175)

RALP (n = 175)

65.9 (6.1) 12.2 (16.1)

65.6 (6.5) 12.3 (16.6)

50 (28.6%) 92 (52.6%) 33 (18.8%)

50 (28.6%) 92 (52.6%) 33 (18.8%)

31 (17.7%) 84 (48.0%) 59 (33.7%) 1 (0.6%)

28 (16.6%) 87 (49.7%) 58 (33.1%) 1 (0.6%)

38 (21.7%) 41 (23.4%) 96 (54.9%)

42 (24.0%) 38 (21.7%) 95 (54.3%)

75 (42.9%) 72 (41.1%) 28 (16.0%)

64 (36.6%) 80 (45.7%) 31 (17.7%)

71 (40.6%) 95 (54.3%) 9 (5.1%) 2.3 (3.3) 105 (60.0%) 18 (10.3%) 64 (36.6%) 3.8 (5.9) 2 (1.1%)

67 (38.3%) 95 (54.3%) 13 (7.4%) 2.3 (3.2) 109 (62.3%) 14 (8.0%) 62 (35.4%) 12.5 (7.8) 8 (4.5%)

P

0.589 0.948 1

0.818

Discussion 0.853

0.486

0.839

0.898 0.742 0.579 0.911

Robot-assisted radical prostatectomy in the Korean population: a 5-year propensity-score matched comparative analysis versus open radical prostatectomy.

To compare the oncological outcomes of robot-assisted laparoscopic radical prostatectomy with those of open radical prostatectomy in contemporary Kore...
225KB Sizes 0 Downloads 0 Views