Apical Surgical Margins Status in Robot-Assisted Laparoscopic Radical Prostatectomy Does Not Depend on Disease Characteristics

Abstract

Background and Purpose:

Positive surgical margins (PSM) during robot-assisted laparoscopic radical prostatectomy (RALP) are generally considered an adverse event. We attempted to identify the factors associated with PSM and their location.

Patients and Methods:

Records of patients undergoing RALP between 2003 and 2009 were retrospectively reviewed. We collected demographic (age, race, body mass index [BMI]), cumulative surgical experience (years from RALP introduction at our center), clinical (prostate-specific antigen [PSA] levels, and biopsy Gleason sums), nerve-sparing technique (yes/no), and pathological variables, including stage (organ-confined vs. non), Gleason sums, prostate weight, status, and location of the surgical margins. Multivariate regression models were constructed to identify the factors associated with PSM at prostate apex, periphery, proximal, and all locations.

Results:

A total of 560 patients were analyzed. Median age was 60.1 (interquartile range [IQR] 55.1–64.7), 19% were African-Americans, median BMI was 28.1 (25.8–30.8 kg/m²), PSA levels were 5.3 (3.9–7.1 ng/mL), and prostate weight was 45.2 (36.8–57.0 g). Gleason sums were as follows: ≤6 in 42.5%, 7 in 53.4%, and >7 in 3.1%. Overall, PSM were reported in 130 (23.2%), including 58 (44.6%) apical, 81 (62.3%) peripheral, and 20 (15.4%) proximal. The overall rate of PSM was associated with surgical experience, PSA, prostate weight, and Gleason sums. Apical PSM were independently associated only with surgical experience. Peripheral PSM were associated with PSA, stage, Gleason sums, and prostate weight. Finally, proximal margin status showed an association with PSA levels only.

Conclusions:

While peripheral, proximal, and overall PSM are largely associated with inherent disease biology (grade, PSA levels, etc.), apical margin status is independently associated only with cumulative surgical experience. These results suggest that a lower rates of positive apical margins may be obtained as the cumulative center experience grows, suggesting a potential role of a “teaching learning curve,” independently from disease characteristics.

Introduction

R obot-assisted laparoscopic radical prostatectomy (RALP) currently represents a very popular treatment choice for clinically localized prostate cancer (PCa). Surgical margin status is among the most important determinants of oncologic success of radical surgery¹ and the location of the positive margin has been reported to have prognostic significance.²

Reducing the rates of positive surgical margins (PSM) may improve the oncological outcomes of patients undergoing RALP for localized PCa. In order to achieve this goal, the risk factors for this adverse event should be understood, and the knowledge derived should be applied to current surgical practice. However, currently available series analyzing risk factors for positive margins in RALP are mostly single surgeon experiences^3
–5 and may not be applicable to teaching centers where multiple learning curves intersect with the mentors' experience.

In this study, we assessed the potential risk factors for PSM, specifically addressing the location of positive margin, in a large tertiary teaching center RALP experience that includes a dedicated residency and fellowship program.

Patients and Methods

After approval by the Institutional Review Board, we performed a retrospective search of medical records in our prospectively maintained registry to identify patients undergoing RALP between 2003 and 2009 within our structured teaching program involving three chief residents and one fellow each year, under one mentor (D.M.A.).

The following variables were collected: demographics including age and race, body mass index (BMI), surgical experience with RALP (years since the technique was introduced in our center, that is, time between the first case performed and the date of surgery of any given patient, in years), preoperative prostate-specific antigen (PSA) levels, nerve-sparing technique (yes/no), biopsy and final pathology Gleason sums, prostate weight as determined on final specimen analysis, pathological stage according to the American Joint Committee on Cancer (AJCC) 6th edition classification,⁶ and status and location (apical, peripheral, and proximal) of surgical margins, considered independently from disease staging. Records with missing data on any of the aforementioned parameters were excluded. Outside biopsy slides were reviewed by uropathologists at our institution.

Pathological processing of the specimens was previously described.⁷ Briefly, prostatectomy specimens were processed in a standard fashion: inked, weighed at fixed in formalin overnight. The apex and the bladder neck margins were shaved and sectioned radially to assess the margin status parallel to the urethra. The remainder of the specimen was step sectioned perpendicular to the rectal surface in 3 to 4 mm intervals. Paraffin-embedded stained sections were evaluated. Positive margins were defined as the presence of PCa cells at the inked margin of the prostatectomy specimen.

We generated descriptive statistics to define patient characteristics in this cohort and constructed a multivariate logistic regression model to identify potential risk factors for PSM at any apical, peripheral, and proximal locations. For logistic regression analyses, PSA levels and prostate weight values were log-transformed to approximate normal distribution. Pathological stage was dichotomized in organ-confined disease (pT2a-c) and non-organ confined (T3 or higher) disease.

Statistical analysis was performed using SPSS software v17 (SPSS, Inc., Chicago, IL). p-Values <0.05 were considered statistically significant. Data are presented as median (interquartile range [IQR]) and number (%) unless otherwise specified.

Results

A total of 705 records of patients undergoing RALP between 2003 and 2009 were identified. Of those, 145 were excluded due to missing data resulting in a final study cohort of 560 patients. The excluded cases did not differ significantly from the final cohort with regards to demographic and clinical variables (data not shown). Patient characteristics are reported in Table 1. The majority of the cohort was Caucasian, most presenting with biopsy Gleason ≤6, most procedures were performed with a nerve-sparing approach. On final pathology, T2c stage was assigned in 70% of specimens.

Table 1.

Patient Characteristics

Variable	Median/number	IQR/%
Age (years)	60.1	55.1–64.7
Race
African-American	108	19.3%
Caucasian	437	78.0%
Other	15	2.7%
BMI (kg/m²)	28.1	25.8–30.8
Preoperative PSA (ng/mL)	5.3	3.9–7.1
Biopsy GS
≤6	384	68.6%
7	151	27.0%
≥8	25	4.5%
Nerve-sparing technique
Yes	323	57.7%
No	237	42.3%
Pathological stage
Organ confined (pT2a-c)	470	83.9%
Locally advanced (pT3 or higher)	90	16.1%
Pathology GS
≤6	238	42.5%
7	299	53.4%
≥8	23	3.1%
Prostate weight (g)	45.2	36.8–57.0
Positive surgical margins
Any	130	23.2%
Apical	58	10.4%
Peripheral	81	14.5%
Proximal	20	3.6%
Surgical experience^a (years)	3.78	2.95–4.53

Surgical experience with robot-assisted laparoscopic radical prostatectomy was quantified as years since the technique had been introduced in our Center, that is, the time between the first case performed and the date of surgery of any given patient.

BMI=body mass index; GS=Gleason sums; IQR=interquartile range; PSA=prostate-specific antigen.

PSM were recorded in 130 (23.2%) cases. The distribution of PSM rates by location according to institutional experience in hundreds of cases is illustrated in Table 2. The incidence of positive margins by location was as follows: apical in 58 (10.4%), peripheral in 81 (14.5%), and proximal in 20 (3.6%). When stratified by pathological stage, the rate of PSM in organ-confined (pT2a-c) disease was 20.6% versus 36.7% (odds ratio [OR]=2.23, p=0.002) in locally advanced (pT3 or higher). Table 3 reports the distribution of surgical margins status and location between localized and locally advanced disease. Of note, the incidence of peripheral positive margins was significantly lower in localized disease (11.9% vs. 27.8%, p<0.001). There were no significant differences between the rates of PSM at other locations.

Table 2.

Distribution of Positive Surgical Margins by Location According to the Institutional Experience with Robot-Assisted Laparoscopic Radical Prostatectomy

Case number	1–100	101–200	201–300	301–400	401–560
Apical	17%	15%	12%	10%	2.5%
Peripheral	16%	24%	13%	9%	11.9%
Proximal	2%	5%	3%	3%	4.4%
Overall	28%	35%	22%	17%	17.5%

Table 3.

Distribution of Surgical Margins Status and Location Between Localized (pT2) and Locally Advanced (pT3 or Higher) Disease

PSM	Localized	Locally advanced	p–Value
Overall, n (%)	97 (20.6)	33 (36.7)	0.002
Apical, n (%)	47 (10.0)	11 (12.2)	0.570
Peripheral, n (%)	56 (11.9)	25 (27.8)	<0.001
Proximal, n (%)	16 (3.4)	4 (4.4)	0.545

PSM=positive surgical margins.

Multivariate analyses to assess potential predictors of any PSM are detailed in Table 4. Detailed risk factor analyses specific for positive margin location (apical, peripheral, and proximal) are reported in Table 5. Demographic variables (age, race, and BMI) were not associated with the rates of PSM in any of the models. Overall institutional experience had an independent inverse association with overall, apical, and peripheral, but not proximal positive margins. In other words, growing experience was independently associated with lower PSM rates at any location, except proximal. The extent of the association can be estimated in almost 20% risk reduction (OR=0.804, p=002) of any positive margins for every year of robotic experience. For apical margins, every year of robotic experience was associated with more than 30% risk reduction (OR=0.699, p<0.001), and roughly 15% reduction for peripheral margins (OR=0.843, p=0.042).

Table 4.

Multivariate Regression Results in Assessing Potential Risk Factors for Positive Surgical Margins After Robot-Assisted Laparoscopic Radical Prostatectomy (R ²=0.169)

			95% CI
Variable	p-Value	OR	Lower	Upper
Age	0.448	1.013	0.980	1.046
Race
Non-AA (reference)	—	—	—	—
AA	0.464	1.221	0.716	2.080
BMI	0.918	0.998	0.952	1.045
Cumulative experience (years)	0.004	0.815	0.709	0.937
Nerve-sparing	0.092	0.692	0.452	1.062
PSA	0.001	7.709	2.353	25.254
Prostate weight	0.001	0.086	0.020	0.370
Biopsy GS	0.135	0.741	0.501	1.098
Pathology GS	0.001	1.877	1.287	2.737
Stage
Organ-confined (reference)	—	—	—	—
Locally advanced	0.115	1.565	0.897	2.730

AA=African-American; CI=confidence intervals; OR=odds ratio.

Table 5.

Multivariate Regression Results in Assessing Risk Factors for Positive Surgical Margins at Apical, Peripheral, and Proximal Locations After Robot-Assisted Laparoscopic Radical Prostatectomy

	Apical PSM				Peripheral PSM				Proximal PSM
			95% CI				95% CI				95% CI
Variable	p-Value	OR	Lower	Upper	p-Value	OR	Lower	Upper	p-Value	OR	Lower	Upper
Age	0.353	1.021	0.977	1.066	0.264	1.023	0.983	1.063	0.569	0.981	0.917	1.049
Race
Non-AA (reference)	—	—	—	—	—	—	—	—	—	—	—	—
AA	0.674	1.167	0.567	2.402	0.954	1.019	0.532	1.953	0.558	1.365	0.482	3.866
BMI	0.517	0.977	0.912	1.048	0.801	0.993	0.938	1.051	0.422	1.032	0.955	1.115
Cumulative experience (years)	<0.001	0.697	0.585	0.831	0.070	0.858	0.727	1.013	0.325	1.187	0.844	1.670
Nerve sparing	0.766	1.091	0.613	1.942	0.102	0.648	0.385	1.090	0.435	0.690	0.271	1.753
PSA	0.444	1.836	0.388	8.689	0.001	12.406	2.932	52.501	0.028	13.306	1.327	133.377
Prostate weight (g)	0.224	0.308	0.046	2.056	0.000	0.016	0.003	0.099	0.512	0.357	0.016	7.749
Biopsy GS	0.192	0.693	0.399	1.202	0.256	0.766	0.484	1.213	0.279	0.606	0.244	1.502
Pathology GS	0.070	1.553	0.965	2.502	0.018	1.709	1.098	2.662	0.092	1.960	0.896	4.286
Stage
Organ confined (reference)	—	—	—	—	—	—	—	—	—	—	—	—
Locally advanced	0.761	1.130	0.515	2.480	0.033	1.987	1.058	3.733	0.733	0.807	0.235	2.766

Preoperative PSA levels demonstrated a significant and independent association with the risk of overall incidence of PSM and at any of the given locations. A similar pattern was observed for prostate weight that was inversely associated with the rate of PSM at any location with the exception of proximal. Interestingly, while biopsy Gleason sums failed to demonstrate a significant association with positive margins, pathological Gleason scores were independently predictive of PSM in all models (with the exception of proximal location, p=0.089). Non-organ confined disease showed double the risk of positive peripheral margins compared with localized PCa on univariate analysis (OR=1.99, p=0.032). However, no significant association was found between disease stage and positive overall, apical, or proximal margins on multivariate analyses.

Discussion

Surgical margin status and their location after radical surgery for PCa are strongly and independently associated with the chance of biochemical recurrence.^1,2 In this study, we assessed the potential risk factors for the incidence of PSM and their location in RALP in a tertiary center with a structured teaching program.⁸ To the best of our knowledge, this is the first study reporting on risk factors for PSM locations in RALP, indicating different risk patterns for apical, peripheral, proximal, and overall PSM.

The rate of PSM in the present study was 23.2%. While some studies have reported a lower incidence,^3,4,9
–11 others have shown comparable rates of positive margins^12
–14 with the robot-assisted technique. However, this represents a series from an institution with a structured residency and fellowship teaching program,⁸ whereas other similar studies represent mostly single, experienced surgeon series. We found that the overall rate of PSM was lower in organ confined disease (20.6% vs. 36.7%, p=0.002). Although this finding is not surprising,^3,4,11 we corroborate these data by showing that the increase of positive margins in locally advanced disease is mainly due to peripheral margins (11.9% for pT2 vs. 27.8% in pT3 or higher PCa, p<0.001).

Our multivariate analysis demonstrates that demographic variables are not associated with the incidence of PSM when adjusted for other parameters. Our results indicate a different pattern of risk factors for PSM by location. In fact, on multivariate analysis, apical margin status was associated solely with cumulative surgical expertise, whereas demographic and biological parameters did not show statistically significant associations suggesting the apical margin status to be largely independent from disease characteristics.

On multivariate analysis, serum PSA levels were associated with overall, peripheral, and proximal PSM. This is in accordance with previous reports in pure laparoscopic¹⁵ and robot-assisted^4,16 series. Similarly, we found a significant association between pathological Gleason score and the incidence of overall and peripheral PSM when adjusted for other variables. Higher PSA levels and higher pathological Gleason scores may reflect more aggressive disease and thereby account for a higher rate of positive margins due to a heavier disease burden.

Prostate weight was associated with a decreased risk of overall and peripheral PSM on multivariate analysis, indicating that RALP for larger prostates is less likely to result in positive margins. Similar results were reported by Zorn et al,⁵ who found a higher incidence of overall PSM in smaller prostates with organ-confined disease. In pure laparoscopic series, several studies have suggested that larger prostates are at a lower risk of positive margins.^15,17,18 Small prostate size has been also associated with more aggressive PCa features even when adjusted for surgical margin status, indicating the importance of prostate size in the general characterization and management of PCa.¹⁹ While prostate weight was associated with overall and peripheral margin status, no significant association was found with apical and proximal locations, emphasizing the importance of technical aspects (surgical experience) as main factors for apical margins and biological features for proximal margin status.

There are controversial data on the influence of a nerve-sparing approach on the incidence of PSM. While some authors reported an increased risk of PSM using nerve-sparing procedures⁴ and suggested differences between nerve-sparing techniques,¹⁵ others suggested that nerve sparing did not compromise margin status²⁰ and the technique did not show significant differences.²¹ Our results add to the controversy in the literature suggesting no significant association between the nerve-sparing approach and PSM at apical, peripheral, or proximal locations.

Institutional surgical experience demonstrated a strong association with the rates of PSM independently of pathological and clinical characteristics. While this may be of no surprise and seem a mere validation of previous studies analyzing the learning curves for RALP,^4,12,14 one has to keep in mind that this series includes the initial experience and the learning curves of multiple surgeons, including residents, fellows, and attending, whereas most of the published data refer to single surgeon series. The intersection of multiple learning curves, however, did not mask the effect of gained experience of the program. The results of the multivariate analysis suggest a substantial (20%–30%) reduction in the risk of PSM with each year of RALP experience indicating the overall progress in the teaching process as experience is accumulated, whereas the rates of PSM were shown to be comparable between trainees and mentors.⁸ The learning curve for beginning surgeons may be accompanied by a “teaching learning curve” that shows steady improvements in the overall rates of PSM with each year of mentoring. In other words, the efficiency of mentorship may increase with experience so that trainees reach proficiency faster improving the overall outcomes. Our findings suggest that the importance of cumulative institutional experience may be of greater importance in reducing the rates of apical margins, whereas the incidence of peripheral and proximal margins may depend, to a larger extent, on PCa biology and be less surgeon dependent.

There are limitations to our study that should be considered when interpreting the results. First, the retrospective design has inherent biases. Second, we did not perform a review of the slides, as all the specimens had been analyzed at our institution by experienced uropathologists. In addition, our multivariate analysis for predictors of proximal margins status was statistically underpowered due to the low incidence of proximal PSM. This may have prevented us from identifying some of the risk factors. Moreover, we do not discuss follow-up and long-term impact of PSM in different locations on oncological outcomes, as they are beyond the focus of the current study. Finally, we could not report the exact break down of the cases according to the surgeon performing the procedure to further elucidate the multiple learning curves including that of the mentor and the trainees and their interactions.

To the best of our knowledge, this is the first study that specifically addresses the risk factors for PSM by location in RALP. In the present study, we demonstrate that the incidence of apical PSM is independently associated only with surgical experience but not with biological features of PCa, whereas the risk factors for the overall incidence of PSM are comparable to those for peripheral and proximal PSM. These results may indicate apical margin status as an indicator of proficiency in RALP. Overall, peripheral, and proximal PSM seem do depend, to a larger extent, on disease biology (serum PSA, Gleason scores, and prostate weight) and are less related to surgical experience. While the biological features cannot be modified, if not by patient selection, surgical expertise may be acquired, especially by structured training programs. Surgical expertise of the mentor may have a role in improving the trainees' results (“teaching learning curve”) and, therefore, lead to better immediate oncological outcomes of RALP.

Conclusions

This study demonstrates that the predictors of PSM during RALP differ based on location. Apical surgical margins status is independent from biological characteristics of PCa but is associated with institutional experience. On the other hand, peripheral and proximal margins' status is strongly related to PCa biology. These results suggest that lower rates of positive apical margins may be obtained as the cumulative center experience grows, suggesting a potential role of a “teaching learning curve,” independently from disease characteristics.

Footnotes

Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Swindle

, Eastham

, Ohori

et al. Do margins matter? The prognostic significance of positive surgical margins in radical prostatectomy specimens. J Urol, 2008; 179,5 Suppl:S47–S51.

Eastham

, Kuroiwa

, Ohori

et al. Prognostic significance of location of positive margins in radical prostatectomy specimens. Urology, 2007; 70:965–969.

Smith

Jr. , Chan

, Chang

et al. A comparison of the incidence and location of positive surgical margins in robotic assisted laparoscopic radical prostatectomy and open retropubic radical prostatectomy. J Urol, 2007; 178:2385–2389discussion 2389–2390.

Liss

, Osann

, Ornstein

. Positive surgical margins during robotic radical prostatectomy: A contemporary analysis of risk factors. BJU Int, 2008; 102:603–608.

Zorn

, Orvieto

, Mikhail

et al. Effect of prostate weight on operative and postoperative outcomes of robotic-assisted laparoscopic prostatectomy. Urology, 2007; 69:300–305.

Grene

, Page

, Fleming

et al. AJCC Cancer Staging Manual, 6th. New York: Springer, 2002.

Freedland

, Banez

, Sun

, Fitzsimons

, Moul

. Obese men have higher-grade and larger tumors: An analysis of the duke prostate center database. Prostate Cancer Prostatic Dis, 2009; 12:259–263.

Schroeck

, de Sousa

, Kalman

et al. Trainees do not negatively impact the institutional learning curve for robotic prostatectomy as characterized by operative time, estimated blood loss, and positive surgical margin rate. Urology, 2008; 71:597–601.

Patel

, Shah

, Arend

. Histopathologic outcomes of robotic radical prostatectomy. ScientificWorldJournal, 2006; 6:2566–2572.

10.

Menon

, Shrivastava

, Kaul

et al. Vattikuti Institute prostatectomy: Contemporary technique and analysis of results. Eur Urol, 2007; 51:648–657; discussion 657–648.

11.

Yee

, Narula

, Amin

, Skarecky

, Ahlering

. Robot-assisted radical prostatectomy: Current evaluation of surgical margins in clinically low-, intermediate-, and high-risk prostate cancer. J Endourol, 2009; 23:1461–1465.

12.

Ploussard

, Xylinas

, Salomon

et al. Robot-assisted extraperitoneal laparoscopic radical prostatectomy: Experience in a high-volume laparoscopy reference centre. BJU Int, 2009[Epub ahead of print]10.1111/j.1464-410X.2009.09013.x.

13.

Ahlering

, Eichel

, Edwards

, Lee

, Skarecky

. Robotic radical prostatectomy: A technique to reduce pT2 positive margins. Urology, 2004; 64:1224–1228.

14.

Atug

, Castle

, Srivastav

, Burgess

, Thomas

, Davis

. Positive surgical margins in robotic-assisted radical prostatectomy: Impact of learning curve on oncologic outcomes. Eur Urol, 2006; 49:866–871discussion 871–862.

15.

Secin

, Serio

, Bianco

Jr.

et al. Preoperative and intraoperative risk factors for side-specific positive surgical margins in laparoscopic radical prostatectomy for prostate cancer. Eur Urol, 2007; 51:764–771.

16.

Chun

, Briganti

, Antebi

et al. Surgical volume is related to the rate of positive surgical margins at radical prostatectomy in European patients. BJU Int, 2006; 98:1204–1209.

17.

Levinson

, Ward

, Sulman

et al. The impact of prostate size on perioperative outcomes in a large laparoscopic radical prostatectomy series. J Endourol, 2009; 23:147–152.

18.

Frota

, Turna

, Santos

, Lin

, Gill

, Aron

. The effect of prostate weight on the outcomes of laparoscopic radical prostatectomy. BJU Int, 2008; 101:589–593.

19.

Freedland

, Isaacs

, Platz

et al. Prostate size and risk of high-grade, advanced prostate cancer and biochemical progression after radical prostatectomy: A search database study. J Clin Oncol, 2005; 23:7546–7554.

20.

Palisaar

, Noldus

, Graefen

, Erbersdobler

, Haese

, Huland

. Influence of nerve-sparing (NS) procedure during radical prostatectomy (RP) on margin status and biochemical failure. Eur Urol, 2005; 47:176–184.

21.

Shikanov

, Woo

, Al-Ahmadie

et al. Extrafascial versus interfascial nerve-sparing technique for robotic-assisted laparoscopic prostatectomy: Comparison of functional outcomes and positive surgical margins characteristics. Urology, 2009; 74:611–616.