Surgical Margin Status After Robot Assisted Radical Cystectomy: Results From the International Robotic Cystectomy Consortium

Results 

Of 513 patients 35 (6.8%) had a positive surgical margin. Surgical margin status was significantly associated with increasing 10-year age group, higher pathological T stage and positive lymph nodes (table 1). Patients 70 to 80 years old and those 81 years old or older were 6.41 and 8.93 times more likely to have a positive margin on pathological review than patients 60 years old or younger (p = 0.014 and 0.004, respectively). Those with pT3 and pT4 disease were 6.28 and 41.4 times more likely to have positive surgical margins than those with pT2 or less disease (p = 0.001 and <0.001, respectively). Those with disease metastatic to lymph nodes were 6.63 times more likely to have a positive margin than those with negative lymph nodes (p <0.001).

Table 1. Patients with positive surgical margins at RC

		No. Margin Neg	No. Margin Pos	Total No.	OR (95% CI)	p Value
	Age group:
	60 or Less	126	2	128	Referent 1.0
	61–70	145	9	154	3.91 (0.83–18.43)	0.085
	71–80	167	17	184	6.41 (1.46–28.26)	0.014⁎
	81 or Greater	40	7	47	8.93 (2.20–55.2)	0.004⁎
	Gender:
	M	379	28	407	Referent 1.0	0.920
	F	99	7	106	0.96 (0.41–2.26)
	Stage:
	T2 or less	322	5	327	Referent 1.0
	T3	123	12	135	6.28 (2.16–18.20)	0.001⁎
	T4	28	18	46	41.4 (14.29–119.90)	<0.001⁎
	Nodal status:
	Neg	398	15	413	Referent 1.0
	Pos	80	20	100	6.63 (3.26–13.51)	<0.001⁎
	Sequential case No.:
	10 or Less	112	12	124	Referent 1.0
	11–20	98	9	107	0.93 (0.38–2.31)	0.882
	21–40	123	6	129	0.50 (0.18–1.36)	0.174
	41 or Greater	135	8	143	0.60 (0.24–1.52)	0.284
	No. institutional cases:
	50 or Less	236	19	255	Referent 1.0
	Greater than 50	242	16	258	0.82 (0.41–1.64)	0.575
	Totals	478	35	513

⁎ Statistical significance was achieved when the 95% CI did not exceed 1.0.

On multivariate analysis age group and pathological T stage remained significantly associated with a positive margin. Tumor T stage most significantly correlated with a positive margin. With all other variables remaining constant for each increase in pathological T stage above pT2 there was a 5 times higher chance (95% CI 3.00–9.57) of a positive surgical margin (p <0.001). With all other variables remaining constant each 10-year age increase above age 61 years also carried a 1.81 times higher chance (95% CI 1.15–2.85) of a positive surgical margin (p = 0.010).

Gender, sequential case number and total institutional volume were not significantly associated with a positive surgical margin. The predicted probability of a positive surgical margin remained constant at approximately 10% independent of overall institutional volume when using actual and linear fit models (see figure).

View Large Image Download to PowerPoint

Probability of positive surgical margin based on number of cases per institution using logistic regression for actual values and linear fit of logistic regression for linear fit correlation.

The proportion of patients in each stage group was similar in the early and later sequential case groups in terms of tumor stage at RARC (table 2). The incidence of high stage (pT3 or greater) disease was 30% to 38% independent of sequential case volume.

Table 2. Stage based on surgeon sequential case number

	Sequential Case No.	No. T2 or Less (%)	No. T3 (%)	No. T4 (%)	Total No.
	10 or Less	84 (62)	39 (29)	13 (9)	136
	11–20	67 (62)	23 (21)	18 (17)	108
	21–40	90 (70)	32 (25)	7 (5)	129
	41 or Greater	92 (64)	42 (29)	9 (7)	143
	Totals	333 (63)	136 (27)	47 (10)	516

Discussion 

Positive soft tissue surgical margins at RC are associated with progression to metastatic disease and dismal disease specific survival in patients with invasive urothelial carcinoma.3, 8, 9 The data show an overall 7% positive soft tissue surgical margin rate, comparable to the 1% to 10% in large open cystectomy series.3, 4, 5 The data are also in line with small laparoscopic and robotic cystectomy series showing a 1% to 6% positive margin rate.10, 11, 12 However, previously reported minimally invasive studies were low volume (typically less than 50 patients), single institutional experiences. In this large, multi-institutional setting the overall positive surgical margin rate approximated those in open series in a multi-institutional setting.

In our study patients were chosen based on an initial diagnosis of clinically nonmetastatic invasive urothelial carcinoma. No attempt was made to operate only in those with a minimal tumor burden. Of the 513 patients 181 (35%) had extravesical disease on final pathological examination, similar to the 30% rate of extravesical disease in larger open series.13

Pathological tumor stage and lymph node metastasis were significantly associated with positive surgical margins. Patients with pT3 or pT4 disease were 6 and 40 times more likely, respectively, to have a positive margin at RARC than those with pT2 or less disease. Likewise, those with lymph node metastasis were 7 times more likely to have a positive surgical margin. A recent series from Memorial Sloan-Kettering Cancer Center evaluating survival in the setting of positive surgical margins after RC also showed that pathological stage and lymph node metastasis are associated with positive margins.3 In this study of 1,589 patients the overall positive margin rate was 4.2%. Further analysis showed no positive margins in patients with organ confined disease and a 9% positive margin rate in those with extravesical disease. In this series 30 of 181 patients (16.6%) with extravesical (pT2 or greater) disease had a positive surgical margin, in line with the standards of the Bladder Cancer Collaborative Group, which cited a 15% or less positive margin rate as a goal at high volume centers.14 Upon further breakdown the positive margin rate was 8.8% in patients with pT3 disease (12 of 135) and 39% in those with pT4 disease (18 of 46). However, the margin rate in IRCC participants with extravesical (pT3 or greater) disease ranged from no positive margins in as many as 11 to 50%, representing 5 in a group in which 10 had pT3 or greater disease.

On a smaller scale advanced patient age was associated with an increased likelihood of a positive surgical margin. Those who were 71 to 80 years old and those who were 81 years old or older were 6 and 9 times more likely, respectively, to have a positive surgical margin than those 60 years old or younger. These data are similar to those in a prior collaborative group study showing that patient age correlated with surgical margin status and the number of lymph nodes retrieved at open RC.5 Increased positive margin rates in our series are explained largely by tumor stage. Of 231 patients 92 (43%) who were 71 years old or older presented with pT3 or higher disease while only 32 of 128 (27%) 60 years old or younger presented with extravesical disease (data not shown). Thus, while the positive margin rates were higher, older patients in our cohort often presented with more advanced disease.

Previous studies show that a learning curve may exist when incorporating robot assisted surgery into practice, specifically in regard to time and surgical margin rates for robot assisted radical prostatectomy.15, 16 Similarly incorporating RARC into practice may ultimately come with an unacceptable surgeon learning curve, particularly as it relates to surgical margin status. In this study sequential case volume served as a surrogate for a learning curve and no significant difference in surgical margin status was associated with cases 11 to 20, 21 to 40 or 41 and higher compared to the initial 10 cases. Institutional positive margin rates ranged from 0% in as many as 91 patients at 1 institution to 30% at 1 institution with 20 patients. When we looked at these 2 institutions, the institution with the 0% margin rate only had 3 patients (3%) with pathological T3 or T4 disease while the institution with the 30% positive margin rate had 15 (75%) with pT3 or T4 disease. However, overall surgical margin status was not significantly associated with institutional volume. Positive margin rates at centers with more than 50 cases were no different than those at centers with 50 or fewer.

Generally IRCC comprises a relatively experienced group of robotic surgeons with an average of 120 (range 0 to 400) robot assisted radical prostatectomies done before performing RARC. While prior robotic experience may have a role in the essentially equivocal positive margin rate, one must also evaluate large open cystectomy series in similar fashion. Most of the published literature has also been from high volume, experienced open surgeons.

IRCC surgeons did not appear to discriminate in terms of stage selection. About 30% to 40% of cases were high stage (pT3 or greater), comparable to the University of Southern California open cystectomy experience in which 29% were pT3 or greater.13 High stage cases were also relatively equally disbursed among the sequential case numbers. Of the 136 cases in the initial 10-case group 52 (38%) were high stage, similar to the 51 of 143 high stage cases (36%) in a group further along the learning curve (cases 41 or greater). Some may argue that simple cases should be selected early in the RARC experience but differences in the margin rate were not attributable to case number.

This study has several limitations. Pathological review of surgical specimens was not centralized, which may have introduced some variation in interpreting surgical margin status. Despite being what is to our knowledge the largest report of RARC experience to date the cohort comprised fewer patients than contemporary open series. Given more time, greater experience may the affect positive margin rate. Finally, as in any large database, there is the possibility of data entry miscoding. However, this potential error should be nonselective over the cohort analyzed and in effect negate overt bias.

Conclusions 

The positive surgical margin rate at RARC for advanced bladder cancer in a large, multi-institutional, prospective cohort was approximately 7%, similar to open cystectomy series. As in open series, increasing age, pathological T stage and lymph node metastasis appear to be associated with an increased likelihood of a positive margin. Case number, which is a surrogate for a learning curve and institutional volume, was not associated with a positive margin.