Findings of Routine Apical Margin Biopsy During Robot-Assisted Radical Prostatectomy

Introduction

The variable anatomy of the prostatic apex and its proximity to the external urinary sphincter may affect the attainable urethral length during radical prostatectomy.^1,2 Furthermore, the desire of achieving maximal urethral length, a desirable goal for postoperative urinary continence, must be balanced against the goal of attaining a negative surgical margin.^1,2 Several studies have investigated the routine use of intraoperative frozen section (IFS) of the apical tissue as a means of assuring oncological control during radical prostatectomy.^3–6 Nonetheless, the benefit of apical IFS during radical prostatectomy remains controversial. ⁷

It remains to be determined if the surgical approach differentially influences the utility of an apical biopsy. It is not surprising that the apical margin is the area of the surgical specimen most likely to have benign glands detected on final pathology following open radical prostatectomy. ⁸ The presence of benign glands at the apex in these specimens may reflect the difficulty of performing an apical dissection due to the anatomical constraints and limited visualization afforded by the open technique. ⁸ Whether the magnified visibility of the robotic platform affects determination of apical anatomy during robot-assisted radical prostatectomy (RARP) is not known. To our knowledge, detailed characterization of the rates of benign glands or cancer at the apical margin during RARP has not been reported. The aim of this study was to determine the histological findings and test characteristics of routine apical tissue biopsy on IFS during RARP.

Patients and Methods

We reviewed our Institutional Review Board–approved Columbia University Robotic Database for patients who underwent RARP by a single surgeon from December 2007 to August 2011 (n=400). Cases in which apical tissue biopsies were obtained and processed for frozen and permanent section were identified. Additionally, we identified cases in which apical tissues were only submitted for permanent section.

Clinical and pathological data were recorded, including age, prostate-specific antigen (PSA) level at diagnosis, biopsy Gleason score, clinical stage, pathological Gleason score, pathological stage, lymph node involvement, and surgical margin status. Biopsy reports were reviewed to identify the individual core with highest Gleason score and percentage tumor volume.

The RARP technique used was similar to those in previous reports.^9,10 Following dissection of the apex, isolation and transection of the anterior urethra, and extirpation of the gland, an excisional soft tissue biopsy was performed from the distal urethral margin. Specimens were typically obtained from the anterior distal urethral margin. In some cases specimens were also obtained from the posterior distal urethral margin; no biopsy specimens were taken from the prostate gland itself. The decision to perform an additional biopsy from the posterior distal urethral margin was at the discretion of the surgeon, but this was typically performed in men with higher-grade disease and/or evidence of cancer at the apex at the time of transrectal ultrasound-guided biopsy. In addition to the apical biopsy, IFS of the neurovascular bundle or the base of the prostate was obtained in selected cases when there were visual concerns of tumor extension. Following biopsy, the tissues were routinely sent for frozen and permanent section analysis. Frozen section analysis was carried out, irrespective of nerve-sparing technique, and the results were communicated to the surgeon within 30 minutes. In cases where cancer was identified on IFS, additional resection in the area of interest was performed, and a re-biopsy of the tissue was performed to confirm absence of cancer. All pathologic specimens (frozen section, permanent section, or radical prostatectomy) were reviewed by a dedicated genitourinary pathologist in a standardized fashion.

Categorical variables were assessed with the chi-squared test, with continuous parametric data analyzed through Student's t test. We determined the sensitivity, specificity, and positive and negative predictive values, as well as overall accuracy of IFS, compared with permanent section results for identification of prostate cancer, as previously reported.^4,11 Univariate and multivariate Cox proportional hazard modeling were used to determine factors that predict the presence of prostatic tissue (benign glands or cancer) versus nonprostatic tissue on apical biopsy.

Results

Apical tissue biopsy were obtained and processed for frozen and permanent section in 335 cases. In 6 of 335 (1.8%) cases, apical tissues were only submitted for permanent section evaluation; these were included in our study. Overall, 329 patients had both IFS and permanent sections available.

Clinical and pathologic characteristics of the 329 patients are presented in Table 1. Overall, there was a 2.7% rate of cancerous glands (Group 1), 38% rate of benign glands (Group 2), and 59.3% rate of other benign histologies (Group 3) based on the permanent pathologic specimen. The benign histological findings in Group 3 included fibrovascular tissue, fibromuscular tissue, nerve tissue, and skeletal muscle. Median age and PSA level were similar among all groups. Preoperative Gleason score distribution varied significantly among groups, with a higher proportion of Gleason 6 tumors noted in Group 1 (P=.03). Pathologic Gleason score also varied significantly among groups (P=.04). Although 33% of the men in Group 1 had extracapsular extension compared with 17% and 19% of the men in Groups 2 and 3, respectively, this difference was not significant. There were no differences among groups in the percentage of biopsy-positive cores, highest Gleason score, or percentage tumor volume. Patients with cancerous glands in the apical tissue biopsy had a median of four biopsy specimens submitted for IFS or permanent section compared with a median of two biopsy specimens for men without cancer glands (P=.02).

Compared with the histological findings on permanent section, IFS of apical tissue had a 67% sensitivity, 99% specificity, 67% positive predictive value, 99% negative predictive value, and 98% overall accuracy for detecting prostate cancer in the apical tissue biopsy specimen (Table 2).

Table 3 compares the results of routine apical tissue biopsy with those performed for a suspicious bladder neck or neurovascular bundle margin. Relative to routine apical biopsies, those for a suspicious bladder neck or neurovascular bundle revealed the same rate of cancer (3%) but a significantly lower detection rate for benign glands (6% and 13%, respectively, versus 41%; P<.001). Consequently, the rate of non-gland benign histology was significantly higher in bladder neck and neurovascular bundle samples compared with apical samples (84% and 91%, respectively, versus 56%; P<.001). Permanent section analysis of suspicious bladder neck and neurovascular bundle biopsy specimens revealed findings similar to the corresponding IFS results. Nonetheless, whereas the rate of cancer on biopsy and IFS of suspicious neurovascular bundle margin was 3%, on permanent section the rate increased to 9% (P=.50).

On multivariate Cox proportional hazards regression model, biopsy Gleason score, number of apical biopsy specimens, and pathologic Gleason score were statistically significant predictors of the presence of prostatic tissue (benign glands and cancer) on IFS of the apical tissue (Table 4).

Discussion

Nearly 30 years since the landmark report of Walsh and Donker, ¹² there continues to be a collective effort by urological surgeons and investigators to optimize effective surgical treatment while preserving the quality of life of patients with prostate cancer.^1,2 Preoperative and postoperative membranous urethral length, including distance from the perineal membrane to the prostatic apex and the urethral stump length, may significantly affect urinary continence rates following radical prostatectomy.^1,13,14 Maximal preservation of urethral length is therefore a highly desirable outcome at the time of surgery. However, given the variable anatomy of the prostatic apex, aggressive preservation of the urethra may compromise oncologic control and result in an increased rate of positive surgical margins. The capacity of the robot to assist the surgeon in addressing these potentially competing goals is not well characterized. In this study we sought to determine the histological findings of routine apical margin tissue biopsy at the time of RARP. Furthermore, we sought to determine the test characteristics of IFS analysis of apical tissue biopsy for the detection of prostate cancer in the biopsy specimen.

IFS has previously been evaluated as a means to decrease positive surgical margins during radical prostatectomy.^3–7 In the current study, we found that 2.7% of routine distal apical margin biopsy specimens harbor cancerous cells, whereas 37.5% contain benign prostatic glands. The 2.7% cancer rate in this study compares favorably with previously reported rates of 5% and 7% in conventional laparoscopic and open radical prostatectomy series, respectively.^3,4 This improved result may be partially due to the fact that RARP can facilitate the usually difficult decision of where to transect the urethra in most patients. We also found a similarly low rate of cancer detection on IFS biopsy specimens of suspicious bladder neck tissue (3%) or neurovascular bundle tissue (3%). On permanent section, the rate of cancer detection in suspicious bladder neck tissue remained 3% but increased to 9% for neurovascular bundle biopsy specimens, although this difference was not statistically significant. It is interesting that patients with cancer detected on IFS actually had lower biopsy Gleason scores. This likely represents the inherent sampling error of transrectal ultrasound biopsy, as there was likely no true difference between the groups in terms of their mean biopsy Gleason scores. This is further suggested from the pathologic Gleason scores at time of prostatectomy, which were similar across all groups. In an analysis of 1340 patients undergoing RARP, Tewari et al. ¹⁵ demonstrated a significant decrease in positive margin rate in cases when there was a conscious effort to use visual cues to aid dissection. The visual cues they considered important included appreciation of periprostatic (lateral prostatic) fascial compartments, color and texture of the tissue, periprostatic veins as a landmark for athermal dissection, signs of inflammation, and a freely separating bloodless plane showing loose shiny areolar tissue. The observed decline in positive margin rate occurred despite an increased rate of patients with high-risk tumors. This suggests that the ability of the robotic platform to improve visualization can directly affect how well the console surgeon is able to achieve negative oncologic margins. Furthermore, a study of patients with Gleason 8 and 9 tumors who underwent RARP demonstrated 5-year biochemical recurrence rates of 36% and 47%, respectively, which is similar to open surgery outcomes. ¹⁶ This suggests that despite the lack of tactile feedback, oncologic outcomes are not compromised by RARP for high-grade tumors. ¹⁶ We now demonstrate that RARP may offer a benefit of low rate of cancer (2.7%) at the urethral margin.

The 37.5% rate of benign glands detected on IFS of apical tissue biopsy specimens in our study is within the previously reported range of 16%–54%.^3,4 Several factors other than surgical approach could contribute to this variability, including differences in patient population, institutional tissue processing, and biopsy or sampling technique. In our study, 2–3 mm excisional biopsy specimens were taken and submitted separately to the pathologist, in contrast to other studies that used a circumferential biopsy method; our decision to perform excisional biopsies was in effort to preserve urethral length.^3,4 Lastly, our cohort of 335 patients was larger than the two earlier studies (198 and 95 patients, respectively), which could represent a regression to the mean phenomenon. Of note, we found no correlation between prostate weight and presence of benign glands on apical tissue biopsy during RARP, unlike the study of Shah et al., ⁸ which found larger glands during open radical prostatectomy were more likely to have benign glands at the surgical margin. Whether this difference is technology, technique, or patient-related is not known.

Prostate cancer detection by IFS of apical tissue biopsy specimens during RARP had the following test characteristics: 67% sensitivity, 99% specificity, 67% positive predictive value, 99% negative predictive value, and an overall accuracy of 98%. Our study confirms the performance benchmark characteristics that may be expected of IFS when evaluating apical tissue biopsy during radical prostatectomy in the modern era. The high specificity (99%) favors IFS as an intraoperative clinical tool for ruling out cancer at the apex when the surgeon suspects apical involvement or extraprostatic extension.

Although we report a large contemporary series, there are several limitations of this study that must be considered, including its single institutional retrospective methodology and lack of outcome data on continence and biochemical recurrence. In addition, we recognize the difficulty in generalizing these results to non–high-volume institutions. Despite these limitations, we have shown a detailed characterization of IFS at the apical margin during RARP. Further data and longer follow-up are required to determine whether the finding of cancerous cells on apical tissue biopsy during RARP should influence the recommendation for adjuvant treatment, in addition to whether patients found to have cancer on IFS are more likely to experience a biochemical recurrence.

Conclusions

We determined the histological findings of routine apical tissue biopsy on frozen and permanent section during RARP. Overall, there was a 2.7% rate of cancer detected in the apical tissue. Therefore, for most men undergoing RARP, attempts to preserve urethral length do not compromise oncologic control. IFS of apical tissue margin may be most helpful for patients with apical tumors or those with a suspicion of extraprostatic extension at the apex.