Abstract
Background/Aim: To investigate whether surgical margin (SM) status would affect the biochemical recurrence (BCR) after robot-associated RP (RARP). Patients and Methods: We evaluated BCR after RARP and the association between pre- and postoperative predictive factors and BCR. Results: Positive SM (PSM) was observed in 97 out of 365 enrolled patients. On multivariate analysis, preoperative prostate specific antigen, biopsy Gleason score (GS), clinical stage, GS ≥7 at the PSM and pathological GS ≥7 were predictive factors for BCR. The 5-year BCR-free survival rate was 84.1% in the negative SM (NSM), 87.4% when GS=6 at the PSM, and 47.6% when GS ≥7 at the PSM. There was no statistically significant difference in BCR-free survival between the NSM group and GS=6 at the PSM group (p=0.966). Conclusion: It would be desirable to evaluate GS at PSM when PSM is present in a specimen removed by RP.
Robot-associated radical prostatectomy (RARP) is well established as a minimally invasive surgery for local prostate cancer (PC). Recent reports have indicated that RARP may reduce the rate of biochemical recurrence (BCR) compared to open surgery and laparoscopic surgery (1, 2). Several studies concluded that surgical margin (SM) status is an independent factor of BCR in patients with PC after RP (3). However, according to other studies, not all patients with positive SM (PSM) show BCR (4). Several others have reported the importance of considering the details of the PSM, such as the PSM length, its location, the PSM focality in organ confined disease, and the Gleason score (GS) at the PSM (5-8). Whether these factors are predictive or not has been under debate. In this study, we evaluated the GS at the PSM and examined whether it was associated with postoperative BCR in the age of robotic surgery. Moreover, we examined the predictive factors for BCR, especially SM status, after RARP in our hospital.
Patients and Methods
Four hundred and fifty-four patients with clinically localized PC who underwent RARP at our hospital from April 2009 to November 2019 were enrolled. Patients with no recorded prostate specific antigen (PSA) values postoperatively (n=1) or patients who received neoadjuvant hormonal therapy (n=88) were excluded. In total, we retrospectively analyzed 365 patients who underwent RARP for localized PC. The observational period was from the date of surgery to the date of the final PSA measurement. The date of BCR was defined as the day when the PSA level increased to ≥0.2 ng/ml for the first time after RARP or when salvage treatment was initiated, if the patients who received RARP had a PSA level of <0.2 ng/ml. In cases where the PSA level never dropped to <0.2 ng/ml after RARP, the date of the surgery was defined as the date of BCR. We comprehensively evaluated clinical T stage by digital rectal examination, transrectal ultrasonography, and magnetic resonance imaging (MRI). We performed risk assessment before treatment in accordance with the D’Amico classification (9). We performed nerve-sparing on the side without cancer in patients with low- or intermediate-risk PC, according to the patients’ choice. Ninety-seven patients with PSM were evaluated the GS at PSM. All these tissues were examined by one pathologist. Microscopic examination of the neoplastic areas in contact with the ink was evaluated as SM positivity. PSMs were diagnosed based on Gleason classification. Since there are multiple Gleason grades (GGs), we assigned the final GS of the tumor at the margin for that case based on the entire tumor at the margin in different anatomic locations. If there were multiple SMs, all were integrated and the ratio of each GG was calculated to evaluate the GS. In cases of PSM in the absence of extra-prostatic extension anywhere in the gland, a stage denotation was given as pT2+. We grouped them according to their SM status, and the groups with PSM were further grouped by their GS at PSM and PSM length. Subsequently, we compared the BCR rates of each group.
The Kaplan–Meier method was used to estimate the BCR-free survival rate after RARP and the Log-rank test was performed to compare the BCR rate between different D’Amico classification groups. Univariate and multivariate Cox proportional hazards regression analyses were used to evaluate the association of pre- and postoperative predictive factors with BCR. Covariates included preoperative parameters, such as age, preoperative PSA, biopsy GS and clinical stage as well as postoperative parameters, such as nerve-sparing status, removed prostate weight, SMs, pathological GS and pathological stage. Covariates were tested for significance in the univariate model and included in the multivariate model if the p-Value was <0.05. All statistical analyses were performed using SPSS (IBM SPSS Statistics ver.25).
Results
The characteristics of all patients are listed in Table I. The median follow-up period after surgery was 32 months (range=1-123 months). Of the 365 patients, 53 (14.5%) experienced BCR during the follow-up period and the BCR-free survival rate at 5 years was 79.3% (Figure 1A). Eighty-nine patients (24.4%) were in the low-risk group, 152 patients (41.6%) were in the intermediate risk and 124 patients (34.0%) were in the high-risk group. BCR-free survival rate according to the different risk groups was evaluated (Figure 1B). The 5-year BCR-free survival rate was 88.1% in the low-risk group, 81.9% in the intermediate-risk group, and 79.9% in the high-risk group. There was a statistically significant difference in BCR-free survival between the low-risk group and high-risk group (p=0.002), and intermediate vs. high-risk group (p=0.012).
The results of univariate and multivariate analyses for the various preoperative factors to predict BCR are demonstrated in Table II. Preoperative PSA, biopsy GS and clinical stage were found to be significant factors to predict BCR in the univariate and multivariate analyses. Among the postoperative factors, SM Gleason score ≥7, SM length >3 mm and pathological GS were found to be significant factors to predict BCR in the univariate analyses (Table III). The correlation coefficient between GS at PSM and PSM length was extremely high at 0.983 (p<0.001), and both GS at PSM and pathological GS were found to be significant factors to predict BCR in multivariate analyses. (Table III). When GS at PSM and PSM length were swapped and multivariate analysis was performed on PSM length and pathological GS, PSM length >3 mm (p=0.048), pathological GS >7 (p=0.014) and pathological GS >8 (p=0.001) was a significant predictor of BCR. We evaluated the impact of the PSM length and GS at the PSM for BCR free survival. There was no significant difference in the BCR rate among the groups with NSM and PSM lengths of ≤3 mm, and the BCR rate was significantly higher in the group with PSM length >3 mm compared with the groups with NSM and PSM lengths of ≤1 mm (Figure 2). The 5-year BCR-free survival rate was 84.1% in the NSM, 87.4% in the GS=6 at the PSM and 47.6% in the GS ≥7 at the PSM (Figure 3). There was a statistically significant difference in BCR-free survival between the NSM group and GS ≥7 at the PSM group (p<0.001), and between GS=6 at the PSM group and GS ≥7 at the PSM group (p=0.004). However, there was no statistically significant difference in BCR-free survival between the NSM group and GS=6 at the PSM group (p=0.966).
The relationship between pathological GS and GS at the PSM is presented in Table IV. Among the 97 cases, 34 (35.0%) had the same pathological GS and GS at the PSM, 43 (44.4%) had pathological GS higher than GS at the PSM, and 20 (20.6%) had pathological GS lower than GS at the PSM. The relationship between PSM pathology and PSM length was shown in Table IV. The PSM length was significantly shorter in the GS=6 at the PSM group than GS ≥7 (p=0.034). The effect of PSM length on the BCR rate in the group with GS=6 at the PSM was analyzed using the Kaplan-Meier method. The 5-year BCR free survival rates did not significantly differ; the group with PSM length ≤3 mm had a 90.5% 5-year BCR free survival (n=24) vs. 81.8% for the group with PSM length >3 mm (n=12) (p=0.513).
Discussion
It has been reported that RARP reduces BCR compared with conventional surgical methods. Some reviews showed that compared with the conventional surgical methods, RARP could not reduce the risk of PSM (1, 2); other reviews show that RARP decreases the risk of PSM in patients with T3 or high-risk groups (1, 2, 10). In this study, we aimed to investigate the relationship between the PSM and BCR in the robot age. In our study, the BCR-free survival rate at 5 years was 79.3%. Several articles have demonstrated that the BCR-free survival rate at 5 years was between 80.1-87.1% (11-13). The results in this study were comparable to the BCR-free rates reported in past reports. Surgeon volume, preoperative PSA, PSA density, biopsy GS, pathological stage, postoperative GS and SM have been reported to be predictors for BCR in multivariable analyses (11-13). In our study, preoperative PSA, biopsy GS, clinical stage, pathological GS, and SM status were predictors for BCR in multivariable analyses. In recent years, it has been reported that the identification of subclassifications of PSM, such as PSM length and GS at the PSM, were important for improving the BCR prediction (7, 14). Moreover, this information might be useful for patient counseling and the follow-up schedule. In previous studies, PSM length, PSM location and GG (or GS) at the PSM are often taken up as predictors. Therefore, we focused on these three factors in this study.
First, regarding the PSM length, Shikanov et al. reported the PSM length as a continuous and categorical variable (<1, 1-3, >3 mm), showed it was an independent predictor and found similar BCR rates in patients with PSM <1 mm and in patients with NSM (5). Preisser et al. showed that a PSM ≥3 mm was an independent predictor in multivariable models (7). In our study, there were no statistically significant differences between the NSM and PSM ≤3 mm; however, there was a statistically significant difference between the NSM and PSM >3 mm. Second, regarding PSM location, Gautier et al. reported that extensive apical PSMs significantly increased the risk of BCR (15). In contrast, Ploussard et al. reported that PSM subclassifications, including PSM location, do not improve the BCR prediction in organ-confined PC (6). Further, Stephenson et al. reported that location and extent of PSM did not improve the accuracy of predicting BCR (16). The definition of the locations of PSM, such as anterior, apex, base, lateral, or posterior of the prostate which did not have clear boundaries was ambiguous in all past studies. We did not select the location of PSM as a predictive factor in this study because of the difficulties of a clear definition of PSM location from our pathological findings. Finally, regarding GS at PSM, Iremashvili et al. showed that the presence of Gleason 4/5 pattern at the PSM was the most practical definition for high risk of BCR (17). Kates et al. demonstrated that higher margin GS (≥4+3) was associated with an increased risk of BCR, and that a lower GS at the margin than the final pathology GS was associated with a lower risk of BCR than if the final pathology and margin GS were the same (8). On the other hand, Preisser et al. reported that BCR free survival for Gleason pattern 3 at 72 months had a statistically significant difference vs. Gleason pattern 4 and higher. However, they could not find out whether the Gleason pattern at PSM is a predictor of BCR in a multivariate model (7). Evren et al. also could not find out whether the Gleason pattern at PSM is a predictor of BCR in a multivariate model (18). In our study, a GS equal to 7 or higher at the PSM had a significantly higher BCR rate compared with NSM and GS=6 at the PSM. In addition, there was no statistically significant difference between NSM and GS=6 at the PSM and the two groups showed a similar BCR free survival. In the present study, no less than 36.5% of the total PSM cases had a GS equal to 6 at the PSM.
The proportion of cases with a short resection margin was higher in GS=6 vs. GS ≥7 at the PSM. In the case of GS=6 at the PSM, there was no difference in the recurrence rate because of the difference in resection margin. The risk of resection length was relatively reduced when GS=6 at the PSM, which could be the one of the reasons why there was no significant difference in recurrence rate from cases with NSMs.
We have examined the details of the PSM for each factor, but in reality, each factor was thought to be closely related. Patients with worse GS at the PSM had longer PSM in this study; therefore, exacerbation factors may be related. PC with low volume (<0.5 cm3) and low grade (GS≤6) was classified as insignificant cancer according to the report of the Johns Hopkins group (19). Several large studies have reported that the rate of metastasis for GS=6 PC confirmed by RP specimen is close to zero (20). GS=6 PCs have a low risk of progression, and even if a few PCs remain in the PSM cases, they are very unlikely to be clinically problematic, which may be one of the reasons why there was no difference in BCR between NSM and GS=6 at PSM cases in this study.
So far, the pathological evaluation and report for GS at the PSM in RP specimen has not yet been established as standard pathological diagnosis. In fact, Iremashvili et al. reported that 85% to 88% of responders in a survey of the members of the International Society of Urological Pathology were not reporting the GS at the PSM (17). Improving the predictive accuracy of BCR has important clinical implications. To improve the accuracy of predicting BCR, it may be desirable to incorporate GS in PSM into the pathological evaluation of RP specimens, if a PSM is present. The American Urological Association/American Society for Radiation Oncology guidelines suggest that patients with PSM at RP should undergo adjuvant radiotherapy. However, radiation therapy is associated with an increase in the incidence of short- and long-term severe complications (21, 22). If accurate reports of the GS at the PSM become common, we can prevent patients with GS=6 at the PSM from experiencing the complications of unnecessary radiation. The nerve-sparing procedure was known to be associated with urinary incontinence prevention and preservation of sexual function (23); however, nerve-sparing was indicated to possibly increase the PSM rate (24). Some urologists may be strict in their indications for nerve-sparing to avoid getting a PSM. As shown in this study, that there was no difference in BCR rates between GS=6 at the PSM and NSM; the indication of RARP with nerve-sparing technique may be desirable in the GS=6 side, in addition to the non-cancer side of the prostate by systematic prostate biopsy, in order to obtain better urinary and sexual function.
This study had some limitations: its retrospective design, small sample size, short-term follow-up and lack of analysis of overall survival. In addition, the indication of lymph node dissection was ambiguous by the surgeon, which may affect the BCR rate. Moreover, location, extent and multifocality of PSM, which have the possibility of increasing the BCR rate, were not examined. The exact pathological stage could not be determined in all cases of our cohort because approximately 10% of the pathological specimens had positive SM in the absence of extra-prostatic extension (= pT2+); therefore, the risk of recurrence at the pathological stage could not be examined. However, finding no significant differences between GS=6 at the PSM and NSM, as shown in this report, would be useful in predicting BCR, and could have an impact on increasing the selection of patients for nerve preservation during RARP.
Conclusion
The 5-year BCR-free survival rate after RARP in this study was 79.3%, which is comparable to that of previous reports. Patients with PC, who underwent RARP, with GS=6 at the PSM had the same BCR rate as NSM in this study. There was a statistically significant difference between the NSM and PSM >3 mm. Therefore, it would be desirable to evaluate GS at the PSM when PSM was present in a specimen removed at the time of RP.
Footnotes
Authors’ Contributions
Hiroshi Kano: Data collection, Manuscript drafting. Yoshifumi Kadono: Conception and design, Data analysis, Manuscript revising. Suguru Kadomoto: Data collection, Manuscript revising. Hiroaki Iwamoto: Data collection, Manuscript revising. Hiroshi Yaegashi: Data collection, Manuscript revising. Masashi Iijima: Data collection, Manuscript drafting. Shohei Kawaguchi: Data analysis, Manuscript drafting. Takahiro Nohara: Conception and design, Manuscript revising. Kazuyoshi Shigehara: Data analysis, Manuscript revising. Kouji Izumi: Data analysis, Manuscript revising. Hiroko Ikeda: Histological evaluation. Atsushi Mizokami: Conception and design, Manuscript revising.
Conflicts of Interest
All Authors declare that they have no conflicts of interest.
- Received November 19, 2020.
- Revision received December 18, 2020.
- Accepted December 21, 2020.
- Copyright© 2021, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.