Intra-operative prostate examination: predictive value and effect on margin status

Authors


David E Rapp, Department of Surgery, Section of Urology, University of Chicago Pritzker School of Medicine, 5841 S. Maryland Avenue, MC 6038, Chicago, Illinois 60637, USA
e-mail: derapp@yahoo.com

Abstract

OBJECTIVE

To evaluate the ability of intra-operative prostate examination (IOPE) to predict extraprostatic extension (EPE) and its effect on margin status in the region of the neurovascular bundle (NVB) when combined with wide excision.

PATIENTS AND METHODS

We retrospectively reviewed 403 patients with clinical stage T1c prostate adenocarcinoma undergoing radical retropubic prostatectomy (RRP). All patients had IOPE during RRP, and those with palpable abnormalities in the region of the NVB underwent wide excision. Pathological outcomes were analysed.

RESULTS

Of 403 patients, 49 (12%) had a palpable abnormality in the region of the NVB. After wide excision, 18 (37%) of these 49 patients were found to have EPE at the site of the palpable abnormality; with wide excision of the NVB, only one of these 18 patients (6%) had a corresponding positive surgical margin (PSM). In 354 patients with a normal IOPE and who underwent bilateral NVB preservation, 30 were found to have EPE in the region of the NVB. The PSM rate in this subset was 23% (seven of 30). The positive predictive value of IOPE for detecting EPE was 37%.

CONCLUSION

IOPE detects abnormalities in 12% of patients with preoperative stage T1c prostate cancer. Although the predictive value of this test is limited, IOPE may decrease PSMs in a subset of patients with EPE in the region of the NVB. The present study reaffirms the value of IOPE for assessing the risk of extraprostatic disease, and for guiding surgical management.

Abbreviations
IOPE

intra-operative prostate examination

RRP

retropubic radical prostatectomy

PSM

positive surgical margin

EPE

extraprostatic extension

NVB

neurovascular bundle.

INTRODUCTION

Intra-operative prostate examination (IOPE) can guide surgical management during radical retropubic prostatectomy (RRP). However, the predictive value of IOPE and its potential role in decreasing positive surgical margin (PSM) rates are not clearly defined, and have become important issues in urological oncology with the advent of laparoscopic RP. We sought to evaluate the predictive value of IOPE for extraprostatic extension (EPE) and its benefit in minimizing PSM rates.

PATIENTS AND METHODS

We retrospectively reviewed the charts of consecutive patients with T1c prostate adenocarcinoma undergoing RRP by one surgeon (C.B.B.). In the absence of an abnormal DRE, no absolute finding on preoperative pathology or PSA testing (e.g. significantly elevated PSA level or percentage tumour involvement) was used to exclude patients. Data were analysed using Student's t-test and Fisher's exact test. Institutional review board approval was obtained.

Bilateral nerve-sparing prostatectomy was offered to all patients and performed as previously described, with recent technical modifications in the apical dissection [1]. The possibility of wide excision of the neurovascular bundle (NVB), depending on preoperative variables and IOPE, was discussed with each patient. More specifically, the risk of EPE based on preoperative variables including Gleason sum, high-volume disease, and number of cores involved, was discussed with all patients. However, the decision to perform wide excision was reserved until after IOPE, to better assess risk of extraprostatic disease.

Both the preoperative DRE and IOPE were performed by one surgeon (C.B.B.). In all patients with palpable intraoperative abnormalities encroaching on the NVB, the NVB was widely excised; the prostate and NVB were removed en bloc. The pathological examination was as previously described [2]; EPE was defined as tumour in contact with extraprostatic fat, and PSM defined as tumour in contact with the inked margin.

RESULTS

In all, 403 patients with T1c prostate adenocarcinoma had RRP between 1 July 1997 and 6 January 2004; the clinical and pathological characteristics of the patients are listed in Table 1. All 403 patients had an IOPE during RRP, and in 49 (12%) this revealed a palpable abnormality in the region of the NVB that had not been detected before RRP. Eleven of these patients had bilateral wide excision of the NVB. After wide excision of the corresponding NVB in these 49 patients, 18 (37%) were found to have EPE at the site of the palpable abnormality; only one of these had a corresponding PSM. One other patient had a PSM at the apex. Of the 354 patients whose IOPE was normal and who had bilateral NVB preservation, 43 (12%) were found to have EPE; 30 of these had EPE in the region of the NVB. The PSM rate in patients with EPE in the region of the NVB was 23% (seven of 30). Another six of the 43 patients (14%) in this group had a PSM in a different location (e.g. apex, base). The overall PSM rate in patients with an abnormal IOPE was 4% (two of 49), compared to 5% (18/354) in patients with a normal IOPE (P = 0.76).

Table 1. 
The characteristics of the patients
VariableValue
Total403
Median (sd):
 Age, years 59 (6.1)
 PSA level, ng/mL  5.6 (3.7)
 Gleason sum  6 (0.6)
Preoperative Gleason sum, n (%)
 3, 4  9 (2)
 5 32 (8)
 6303 (75)
 7 56 (14)
 8, 9  3 (1)
Final pathological stage, n (%)
 pT2335 (83)
 pT3a 61 (15)
 pT3b  7 (2)

We analysed differences in preoperative characteristics that might have predicted EPE (Table 2). Patients with an abnormal IOPE and confirmed EPE had significantly higher PSA levels than patients with a normal IOPE. (P < 0.01), but no other significant differences were identified between these groups.

Table 2. 
The characteristics of the patients according to the results of IOPE and the presence of EPE
VariableAbnormal IOPE and confirmed EPENormal IOPE
  • *

    P < 0.01.

N354 18
Mean (range):
 Follow-up, months21.6 (1.5–66)17.5 (1.5–48)
 Age, years59.2 (42–66)58.6 (41–74)
 PSA level, ng/mL 8.4 (4.7–23.6) 6.3 (3.1–19.6)*
Mean (sd) preop Gleason sum 6.2 (1.0) 6.0 (0.6)

To assess the predictive value of IOPE for detecting EPE, calculations were based on the number of sides revealing a palpable abnormality. In two patients undergoing bilateral wide excision of the NVB, wide excision of one side was performed despite a normal IOPE, because of other factors (e.g. patient preference based on discussion of preoperative risk assessment and preoperative erectile dysfunction), and these two sides were excluded from analysis. All other patients had wide excision after an abnormal IOPE. Accordingly, 58 sides were found to be abnormal on IOPE. The use of IOPE for detecting EPE in the region of the NVB had a positive predictive value of 37% and a negative predictive value of 92%.

DISCUSSION

Tactile feedback during surgery has been part of the surgical technique throughout the era of open surgery; such feedback during RRP facilitates parts of the surgical dissection (e.g. bladder neck), but the oncological benefit of IOPE is controversial. Specifically, there is no objective research to define how often IOPE detects palpable abnormalities in the region of the NVB, and whether the presence of palpable abnormalities accurately predicts EPE. Finally, it is not clear whether wide excision of the NVB based on intraoperative palpable abnormalities results in lower PSM rates.

In the present study, ≈ 12% of patients had palpable abnormalities in the region of the NVB that were not detected by preoperative DRE. As a result, all of these patients had wide excision of the ipsilateral NVB. Thus IOPE detects palpable abnormalities in a substantial number of patients and directly influences surgical management.

IOPE had a positive predictive value of 37% for corresponding EPE; thus many patients will undergo wide excision of the NVB unnecessarily. Two recent investigations support this conclusion, reporting finding tumour in only 18–33% of the resected NVBs [3,4]. Therefore, caution is needed when assessing the relative oncological benefit of IOPE and subsequent wide excision of the NVB, given the increased risk of erectile dysfunction.

Despite the limited positive predictive value of IOPE, we feel that its importance is apparent in the pathological outcomes in the present patients. Of patients with palpable abnormalities, almost 40% had corresponding EPE. More strikingly, the rate of PSMs after wide excision was 6%, vs 23% in patients who had EPE in the region of the NVB that was not detected during RRP. This suggests that wide excision based on intraoperative palpable abnormalities can benefit this specific patient group. Given the negative consequence of PSM status on disease progression and survival [5,6] this benefit is of direct clinical significance. A longer follow-up is necessary to determine whether there will be a significant difference in the associated biochemical failure rates.

Although several studies reported that wide excision has little effect on overall PSM rates, many analyses are limited by their focus on PSM rates in all regions of the prostate [7,8]. To accurately assess the value of wide excision, statistical analysis must compare PSM rates only in the region of the NVB. This is supported by research showing that the apex is the most common site of PSMs after RRP [7–9] and by the finding that, despite a significant reduction in PSM rates in the region of the NVB, there was little difference in overall PSM rates between wide excision and nerve-sparing groups.

To our knowledge, only one other investigation has specifically evaluated intraoperative tumour staging in radical prostatectomy. Vaidya et al.[10] concluded that intraoperative tactile assessment of margin status in 100 patients did not accurately correlate with the pathological findings. However, they reported a positive finding on IOPE in only five (5%) patients, which might limit the statistical conclusions that can be drawn from that study.

Other preoperative factors might help to predict extracapsular disease. In addition to preoperative Gleason grade and PSA level, we analysed our data for the number of positive biopsies and the percentage core involvement. However, many of the pathology reports were from referral institutions and, although all slides were reviewed, inconsistencies made detailed analysis difficult. Although bilateral positive biopsies, the presence of two or more positive cores, percentage tumour involvement on needle biopsy, preoperative PSA elevation, Gleason score sum of ≥ 7, and the presence of perineural invasion on biopsy all correlate with the risk of EPE and/or PSM [5,11–16], none of these factors can be used individually to exclude a patient from nerve-sparing prostatectomy [17]. A recently reported nomogram by Tsuzuki et al.[18] attempted to predict more specifically for EPE in the NVB from preoperative clinical information and biopsy data. This model only allowed for limited risk stratification, as defined by NVB extension in <10% vs ≥10% of cases, highlighting the inability of preoperative variables to specifically predict NVB involvement. Thus, as reported by Walsh [19], the decision to widely excise the NVB must include an intraoperative assessment rather than using preoperative findings as an absolute indication for resection.

The present results show that previously undetected palpable abnormalities will be found in some patients, and this will affect the decision to preserve the NVB. When combined with wide excision, the PSM rate may be decreased in a subset of patients. Although our experience reaffirms the basic value of IOPE, recent reports suggest that laparoscopic techniques might give equivalent PSM rates to RRP, despite the loss of tactile feedback [20,21]. Further, many laparoscopic surgeons performing RP examine the surgical specimen immediately after removal. We would advocate this to help assess surgical margins, but it is unclear how best to approach the presence of palpable abnormalities after specimen removal, and whether further resection is beneficial in these cases. The importance of objective analysis of this issue is underscored by reports investigating the use of intraoperative frozen-section analysis during both laparoscopic and open RP, and a recent report by Ahlering et al.[22] described a modified surgical technique to decrease pT2 PSMs during robot-assisted laparoscopic prostatectomy [3,4,22].

The present data suggest several important conclusions. First, IOPE detects palpable abnormalities in many patients who had a normal preoperative DRE by the same surgeon. Second, IOPE predicted corresponding EPE in 37% of patients with palpable abnormalities. Finally, when combined with wide excision, IOPE resulted in a lower PSM rate in the region of the NVB in those patients who had an EPE corresponding to the palpable abnormality. These data, combined with the known disadvantage of wide excision on postoperative potency rates, might be used to optimize surgical decision-making and preoperative counselling about preserving the NVB. Most importantly, we think that the present study reaffirms the value of IOPE for assessing the risk of extraprostatic disease, and for guiding surgical management.

CONFLICT OF INTEREST

None declared.

Ancillary