Perioperative outcomes of robot-assisted laparoscopic radical prostatectomy (RALRP) and LRP in patients with prostate cancer based on risk groups

ABSTRACT Objective To compare the perioperative and pathological outcomes between robot-assisted laparoscopic radical prostatectomy (RALRP) and LRP based on the patient’s risk. Patients and methods The medical records of 588 patients with prostate cancer who underwent RP, using minimally invasive surgery (MIS) techniques (240 LRP and 348 RALRP) by a single surgeon during January 2008 to June 2018 at the Ramathibodi Hospital, were retrospectively reviewed. The patient’s risk was classified according to the National Comprehensive Cancer Network (NCCN) Guideline, 2018. The demographic, perioperative, and pathological data of patients were collected. The differences in perioperative and pathological outcomes between LRP and RALRP in each risk classification were assessed using chi-square, Fisher’s exact tests and logistic regression, as appropriate. Results In terms of positive margins, RALRP had significant advantages in high-risk patients when compared to LRP (adjusted odds ratio 0.46, 95% confidence interval 0.26–0.84), while there were no differences in the low- and intermediate-risk patients. Overall, the patients who underwent RALRP had significant advantages over those who underwent LRP in terms of operative time, estimate blood loss, and blood transfusion rate. While, adjacent organ injury rate and length of hospital stay were similar for both techniques in all subgroup analyses. Conclusion MIS techniques appear to be safe, especially RALRP, which has significantly better perioperative outcomes in all subgroups of patient risk classification, and in the high-risk patient group it seems to have better pathological outcomes when compared to LRP. Abbreviations EBL: estimated blood loss; LOS: length of hospital stay; PSM: positive surgical margin; (L)(O)(RAL)RP: (laparoscopic) (open) (robot-assisted laparoscopic) radical prostatectomy; MIS: minimally invasive surgery


Introduction
Prostate cancer is the fourth most common cancer affecting Thai men [1] and is a major health concern worldwide, being the second most common neoplasm and sixth cause of cancer-related deaths in the world [2]. Today, robot-assisted laparoscopic radical prostatectomy (RALRP) is the mainstay for the local control of the disease. However, the procedure is associated with significant morbidity and decline in quality of life because of the sequelae of the surgical procedure (urinary incontinence and erectile dysfunction) [3]. In the past decade, especially in Asia, the RP technique has shown a significant trend toward minimally invasive surgery (MIS) techniques [4][5][6]. Many studies show that MIS techniques (e.g. RALRP and LRP) have some advantages over a traditional open RP (ORP) technique in terms of the operative field and perioperative outcomes (operative time, estimated blood loss [EBL], blood transfusion rate, adjacent organ injury, oncological outcomes, and functional outcomes) [4][5][6][7]. It is evident that MIS techniques are safe and have significantly better perioperative outcomes with comparable pathological/oncological outcomes compared to ORP.
There are plenty of study data from high-volume centres that have reported on LRP and RALRP as MIS techniques worldwide. A meta-analysis [8] showed that RALRP, with its advantages of few postoperative complications and well-preserved urinary continence and erectile function, was an effective, and safe option for prostate cancer [8,9]. In Thailand, the cost of treatment with RALRP and LRP differs [10], so the total cost is a factor that determines the choice of treatment in shared doctor-patient decisions. Advocates claim greater precision in dissection and suturing, and accelerated attainment of surgical competency of the more costly RALRP over the standard LRP. There are few studies that compare the two techniques (LRP and RALRP) in depth in terms of subgroup analysis, especially in the high-risk patient group to reveal the advantages of one over the other.

Population and surgical techniques
This study was a single-centre single-surgeon retrospective observational study. In total, 716 patients with prostate cancer treated with RP by a single surgeon at the Ramathibodi Hospital, Mahidol University, Bangkok, Thailand, between January 2008 and June 2018, were reviewed. There were 588 patients who underwent MIS techniques (240 RALRP and 348 LRP) and 128 patients who underwent ORP. All patients who underwent MIS techniques were included in the study analysis. The LRPs were performed in an extraperitoneal fashion and the RALRPs were performed using the da Vinci® Surgical system Si (Intuitive Surgical Inc., Sunnyvale, CA, USA). The operation was selected for each patient depending on the shared decision of the patient and the doctor.

Baseline characteristic and preoperative parameter
All patients who underwent MIS techniques (RALRP and LRP) were categorised by risk stratification using the National Comprehensive Cancer Network (NCCN) Guideline, version 4 (2018), for subgroup analysis. Patients' demographic data including age, weight, body mass index (BMI), serum PSA level, clinical stage, and biopsy Gleason score, were collected from the patient medical records and laboratory reports.

Grouping
All patients who underwent MIS techniques (RALRP and LRP) were divided into two groups depending on the surgical approach. The patient risk was classified according to the NCCN Guidelines, version 4 published in 2018, for the subgroup analysis. Briefly, a patient with a PSA level <10 ng/mL and Gleason score ≤6 was categorised in the low-risk group; a patient with a PSA level of 10-20 ng/mL and Gleason score 7 was categorised in the intermediate-risk group; and a patient with a PSA level of >20 ng/mL and Gleason score ≥8 was categorised in the high-risk group.

Outcome of interest
The perioperative outcomes, including: operative time, EBL, blood transfusion, adjacent organ injury (bowel injury, unattended bladder injury, or vessel injury), and length of hospital stay (LOS), were retrospectively collected from medical records. The pathological outcomes, e.g. margin status, were retrieved from the pathological reports, which had been examined by experienced pathologists in the Division of Clinical Pathology. Other parameters from the pathological reports, e.g. prostatic weight, biopsy Gleason score, and pathological stage were also collected.

Statistical analysis
For descriptive statistics, means ± standard deviations (SDs), medians with interquartile ranges (IQRs) and proportions, were used to describe and summarise the collected study data as appropriate. The independent t-test, Wilcoxon rank-sum test or Kruskal-Wallis test were used to determine the differences of continuous parameters between surgical approaches or amongst the risk groups. For categorical data, chisquared or Fisher's exact test, as appropriate, was used to test the differences amongst groups. Univariate and multivariate logistic regressions were further used to identify and explore the factors associated with the study outcomes. The analysis was performed using STATA Statistical Software: release 14.0 (StataCorp., College Station, TX, USA) and a P < 0.05 was considered statistically significant.

Patients and prostate cancer stages and grades
The patients' demographic data, prostate cancer stages and grades are presented in Tables 1 and 2. The data were statistically different in terms of median age between LRP and RALRP only in the high-risk patient subgroup, at 70 (67-73) and 67 (63-72) years, respectively (P = 0.019). The median body weight in the highrisk RALRP patient subgroup was statistically significantly higher than that in the high-risk LRP subgroup, at 67 (59.5-75.2) and 64.2 (57.8-71.0) kg, respectively (P = 0.028). The median height and median BMI were not different between LRP and RALRP amongst the high-, intermediate-, and low-risk patient groups. Based on reported medical history (hypertension, diabetes mellitus, or dyslipidaemia) there was no difference between LRP and RALRP in all subgroups. The median PSA level in the LRP group was significantly higher than that in the RALRP group only in the highrisk patient group, at 32.3 (21.2-48.8) and 24 (12.4-44.1) ng/mL, respectively (P = 0.041). There were no differences in terms of preoperative PSA levels between LRP and RALRP patients in other groups (intermediate-and low-risk groups). The biopsy Gleason score in RALRP patients was significantly higher than that in LRP patients in the high-risk patient group, at 8 and 7, respectively (P = 0.025), but there were no differences between LRP and RALRP patients in the other groups.
The pathological stage was significantly different in each risk subgroup (P = 0.001). The overall median biopsy Gleason score was significantly higher in RALRP than in LRP. The prostate volume (measured by specimen weight) was not significantly different amongst all subgroups for LRP and RALRP.

Perioperative and pathological outcomes
The perioperative outcomes showed that operative time was significantly shorter for RALRP than for LRP, at 190 and 210 min, respectively (P = 0.015). Patients who underwent RALRP had a significantly lower EBL and blood transfusion rate than those who underwent LRP in all subgroup analyses. There was no statistical difference between the MIS techniques, LRP and RALRP, in all subgroup analyses for adjacent organ injury and LOS (Table 3).
For the pathological outcomes, the positive surgical margin (PSM) rate was only significantly different in the high-risk patient group, at 41.1% and 59.3% for RALRP and LRP, respectively (P = 0.012). After adjusting for other variables (Table 4) using multivariate logistic regression, the results still showed that the high-risk patients who underwent RALRP had a significantly lower risk of PSMs when compared to those who underwent LRP (adjusted odds ratio [OR] 0.46, 95% CI 0.26-0.84). The multivariate analysis also showed that the high-risk patients with higher PSA levels and higher biopsy Gleason scores were associated with higher risk of PSMs (OR 1.01, 95% CI 1.00-1.02; and OR 1.54, 95% CI 1.16-2.07, respectively).

Discussion
Two recent studies [11,12] demonstrated that RALRP is at least equivalent to ORP or LRP in terms of PSM rates and suggested that RALRP provides certain advantages, especially regarding decreased risk of adverse events. Although patients with high-risk prostate cancer have increased incidence of biochemical recurrence and requirement for secondary therapy, there are many treatment options available for this patient group, e.g. RP, radiation in combination with androgen-deprivation therapy, or observation amongst others, with a tendency to select RP. A previous study showed that the pathological and oncological outcomes in patients with high-risk prostate cancer were   *Statistical significance at similar amongst RALRP, LRP and ORP [13]. In our present study, only in the high-risk patient group was RALRP superior to LRP for PSM rates. The RALRP allows for more precise instrument movement when dissecting and improved operative field visualisation compared with LRP. According to our present results, if the patient is classified as high risk, RALRP is preferred as it results in better oncological outcomes. To confirm this difference in PSM rates, more prospective, randomised controlled trials are needed.
At the Ramathibodi Hospital, RP is the standard treatment for localised and locally advanced prostate cancer. Since 2007, the surgical trend toward LRP for prostate cancer is obvious [14][15][16], with LRP being used as the standard approach. In 2013, RALRP was introduced at the Ramathibodi Hospital. In many studies in the past, RALRP has been shown to be more advantageous than LRP [17] in terms of EBL, operative time, blood transfusion rate, LOS, and adjacent organ injury rate. According to previous studies [11,18], there  are ample data from high-volume centres reporting that RALRP has the advantages of fewer postoperative complications compared with LRP. Similar to our present study, the operative time, EBL, and blood transfusion rate were significantly lower for RALRP than for LRP. When comparing the LOS, in our present study we found no statistically significant difference between LRP and RALRP (6 and 5 days, respectively). In fact, in our experience, the LOS can be affected by many other factors such as patient preference, socioeconomic status, and disease complications, and these may explain why there was no difference between RALRP and LRP in our present study. In regards to the adjacent organ injury rate, there was no statistically significant difference between LRP and RALRP in all subgroup analyses, which is probably explained by the fact that both RALRP and LRP are MIS techniques with good visualisation of the surrounding anatomy.
Our present study has some limitations. With the retrospective study design, both known and unknown factors between LRP and RALRP could not be controlled for, such as medication given prior to the surgery. Furthermore, the improvement of surgical performance overtime by the surgeon could be a confounding factor when evaluating the outcomes of the procedure, although different techniques adopted by the single surgeon might reduce the bias from surgical performance amongst surgeons.

Conclusion
The MIS techniques of LRP and RALRP appear to be promising techniques for the treatment of organconfined prostate cancer and some locally advanced prostate cancers. Comparing LRP and RALRP, RALRP seems to have a shorter operative time, lesser EBL, and lower blood transfusion rate than LRP, but there was no difference in the adjacent organ injury rate and LOS. In regard to the pathological outcome, it was found that RALRP was superior to LRP for the PSM rate, but only in the high-risk patient group. As RALRP was more beneficial in terms of pathological outcome than LRP in the high-risk patient group in the present study, we therefore highly recommend RALRP for the treatment of prostate cancer in highrisk patients.