Safety of Minimally Invasive Radical Prostatectomy in Patients with Prior Abdominopelvic or Inguinal Surgery

Abstract

Introduction: Despite the widespread use of minimally invasive radical prostatectomy (MIRP), there remain concerns regarding its safety in patients with a history of prior abdominopelvic or inguinal surgery.

Methods: A prospective database of 1165 MIRP procedures performed by a single surgeon at a high-volume tertiary care center from 2001 to 2013 was analyzed. After an initial period of transperitoneal MIRP (TP), an extraperitoneal (EP) approach was used preferentially beginning in 2005 (for both laparoscopic and robotic cases), and robotics were used preferentially beginning in 2010. Overall perioperative complications, major complications (Clavien-Dindo III or IV), and abdominal complications (e.g., ileus, bowel/organ injury, or vascular injury) were compared for patients with and without a prior surgical history. Uni- and multivariate logistic regression were used to control the impact of robotics, approach, operative time, estimated blood loss, case number, prostate weight, and primary Gleason on complications.

Results: Three hundred patients undergoing MIRP had prior abdominopelvic or inguinal surgery (25.8%). Of these, 102 (34%) underwent TP and 198 (66%) EP MIRP. Robotics was used in 286 cases (24.6%) and pure laparoscopy in 879 (75.4%). Complications occurred in 111 patients (9.5%) from the total cohort, with major complications in 32 (2.75%) and abdominal complications in 19 (1.63%). Prior surgery was not associated with overall, major, or abdominal complications. Of the controlling factors, only increasing operative time was associated with postoperative abdominal complications (most of which were ileus) on multivariate analysis.

Conclusions: In this large single-surgeon series where both EP and TP approaches to MIRP are utilized, prior abdominopelvic or inguinal surgery was not associated with an increased risk of perioperative complications.

Introduction

Minimally-invasive radical prostatectomy (MIRP) continues to increase in popularity in the United States. A recent analysis of American urologists who underwent board certification or recertification reported that 67% of radical prostatectomies were performed through a robot-assisted, rather than an open, approach.¹ However, there is concern regarding the feasibility and safety of MIRP in patients with a history of prior abdominopelvic or inguinal surgery. As most MIRP procedures are performed transperitoneally, extensive intraabdominal adhesions from prior abdominal surgery could increase the complexity of MIRP and potentially result in increased complication rates.² Furthermore, prior inguinal hernia surgery has been shown to increase the complexity of radical prostatectomy, and in some cases necessitate abortion of the procedure.^3–6

In this study, we sought to assess the safety and feasibility of MIRP in patients with a history of prior abdominopelvic or inguinal surgery by describing perioperative outcomes, pathological results, and complication rates in this patient group.

Materials and Methods

Patient cohort

A prospectively maintained database of laparoscopic and robot-assisted radical prostatectomies performed by a single surgeon (C.P.P.) between August 2001 and November 2013 was analyzed. Data regarding prior surgical history was collected from patients at the time of the initial clinic visit. Patients with a history of surgery that entered the peritoneal cavity were considered in the abdominopelvic prior surgery group; similarly, patients with a history of inguinal surgery (e.g., inguinal hernia or inguinal orchiectomy) were placed in the inguinal prior surgery group. For the purpose of this analysis, patients with a history of other surgeries, including transurethral urological procedures, and nonabdominopelvic or inguinal surgeries were not considered in the prior surgery groups.

Operative technique

From August 2001 until July 2005, MIRP was performed via a transperitoneal (TP) pure laparoscopic technique using an Aesop (Computer Motion, Santa Barbara, CA) robotic arm strictly for camera control. Beginning in July 2005, MIRP was preferentially performed extraperitoneally (EP), either through a laparoscopic or Da Vinci robot-assisted (Intuitive Surgical, Sunnyvale, CA) approach. Nevertheless, from 2005 onward MIRP was occasionally performed transperitoneally, particularly in the face of previous bilateral mesh EP inguinal hernia repairs, and occasionally when the EP space could not be developed. The TP technique used was a modified Montsouris approach⁷ with the posterior dissection done either antegrade or retrograde depending on the appearance of the rectovesical pouch. Access was established with a Veress needle inserted away from any prior incisions, and preferably periumbilically, followed by the use of a Visiport Plus (Covidien, Mansfield, MA) visual obturator trocar. The EP approach used the Visiport Plus to directly access the EP space infraumbilically without use of a Veress needle; the space of Retzius was then expanded using balloon inflation, with the posterior dissection performed in antegrade fashion. Beginning in 2010, the robotic approach was adopted and used preferentially in all cases.

Statistical analysis

Patient demographic, preoperative disease characteristics, intraoperative outcomes, pathological results following MIRP, and perioperative complications were then collected. Complications were graded using the Clavien-Dindo classification system.⁸ Subset analysis of major complications (Clavien-Dindo III or IV) and abdominal complications, defined as bowel injury, ileus requiring NG tube, or epigastric vessel injury was also performed. Prior surgery and no prior surgery cohorts were compared. Fisher's exact test was used to compare categorical variables and the Wilcoxon rank-sum test was used to compare continuous variables. Logistic regression analysis for overall, high-grade, and abdominal complications was performed controlling for covariates of approach, operative time, blood loss, prostate size, primary Gleason pattern, and case number. All statistical analysis was performed using Stata 13 (Stata Corp, College Station, TX). Two-sided p-values<0.05 were considered significant.

Results

A total of 1165 men underwent MIRP during the study period. Of these, 300 (25.8%) had a history of prior abdominopelvic or inguinal surgery, with 154 (51.3%) having prior abdominopelvic surgery alone, 112 (37.3%) with prior inguinal surgery alone, and 34 (11.3%) having had both. Of the 188 abdominopelvic surgeries, 108 (57.5%) were appendectomies, 43 (22.9%) cholecystectomies, 10 (5.3%) colectomies, 7 (3.7%) resection of abdominal viscera (partial gastrectomy and splenectomy), and 7 (3.7%) nephrectomies. Of the 146 inguinal surgeries, 139 (94.6%) were uni- or bilateral inguinal hernia repairs, while the other 7 (5.4%) were inguinal orchiectomies. Table 1 lists the clinicopathologic characteristics of the cohort. Patients with a history of prior surgery were slightly older (60 vs. 58, p=0.0009) and more likely to have undergone robotic MIRP (31.7 vs. 22.1%, p=0.0001). An EP approach was used in the majority of cases regardless of prior surgery. The percentage of cases performed on patients with a history of prior surgery was similar in the first half of the study period as compared to the second half [159 (53%) vs. 141 (47%), p=0.4]. Men with a history of surgery had similar rates of lymph node dissection (70.7 vs. 65.9, p=0.135) and bilateral nerve sparing (83 vs. 84.1%, p=0.84) as men who did not have prior surgery.

Table 1.

Clinicopathologic Characteristics of Patients With and Without a History of Prior Surgery Undergoing Minimally Invasive Radical Prostatectomy

	No prior surgery (n=865)	Prior surgery (n=300)	p-Value
Age	58 (53–63)	60 (55–64)	0.0009
Approach, n (%)
Laparoscopic	674 (77.9)	205 (68.3)	0.0001
Robotic	191 (22.1)	95 (31.7)
Access, n (%)			0.139
Extraperitoneal	602 (69.6)	198 (66.0)
Transperitoneal	263 (30.4)	102 (34.0)
Operative time (hours)	3 (2.5–3.5)	3 (2.5–3.5)	0.738
EBL (mL)	200 (150–250)	200 (150–250)	0.787
Prostate weight (grams)	45 (40–57)	54.7 (40–59)	0.576
Primary Gleason pattern, n (%)			0.485
3	728 (84.5)	242 (80.9)
4	122 (14.2)	54 (18.1)
5	6 (0.7)	1 (0.3)
Unavailable	6 (0.7)	2 (0.7)
Lymph node dissection performed	569 (65.9)	212 (70.7)	0.135
Number of nerves spared			0.836
0	17 (2.0)	6 (2.0)
1	116 (13.3)	45 (15.0)
2	727 (84.1)	249 (83.0)
Unavailable	5 (0.6)	0
Complications, n (%)
All complications	87 (10.1)	24 (8.0)	0.361
Major complications	25 (2.89)	6 (2.0)	0.533
Abdominal complications	19 (2.2)	0 (0)	0.006

Values are median (inter-quartile range) unless otherwise specified.

EBL=estimated blood loss.

The overall positive margin rates were comparable when comparing patients with no surgical history to those with a history of prior surgery (18.3% vs. 19.0%, p=0.80). Likewise, there was no significant difference in median length of stay, operative time, estimated blood loss, or transfusion rate.

Postoperative complications occurred in 111 (9.5%) cases, with the majority of complications 80 (72.1%) being minor (Clavien grade I or II). Major complications (Clavien grade III) accounted for 27.9% of all complications and occurred in 31 patients (2.7% of total cohort). There were no Clavien grade IV or V complications. The most common major complications were deep vein thrombosis or pulmonary embolism in 12 patients, symptomatic lymphocele requiring drainage in 3, and postoperative hemorrhage in 2, and notably intraoperative rectal injury in 1 and obturator nerve injury in 1. Abdominal complications occurred in 19 (1.63%) cases–13 ileus requiring nasogastric tube decompression, five cases of epigastric vessel injury, and one postoperative inguinal hernia that occurred shortly after MIRP. When stratified by history of prior surgery, both overall and high-grade complication rates were similar between groups, but interestingly abdominal complications were more common in patients with no history of prior surgery (2.2 vs. 0%, p=0.0006). Among patients with prior surgery, overall and major complications did not differ for laparoscopic versus robotic techniques (overall 8.8% vs. 6.3%, p=0.465; major 2.4% vs. 1.1%, p=0.426) or for TP versus EP approaches (overall 9.8% vs. 7.1%, p=0.410; major 2.0% vs. 2.0%, p=0.972). Furthermore, there was no difference in overall complications rate for patients with abdominal or inguinal surgery (8.4% vs. 7.5%, p=0.7) or major complications (2.6% vs. 1.7%, p=0.5).

Few complications were directly related to prior surgery or to obtaining laparoscopic access to the pelvis. Epigastric vessel injuries occurred in five patients who had no surgical history during trocar placement. All of these were repaired intraoperatively with suture ligation. There were no small or large bowel enterotomies in our series, but one patient, with a history of multiple prior TURPs and several prostate biopsies, experienced a rectal injury.

To determine whether the higher percentage of robotic cases in men with a surgical history accounted for the comparable outcomes between the prior surgery and no prior surgery groups, a subset analysis was performed including only men with surgical history. This analysis compared the laparoscopic to the robot-assisted cases in the following outcomes: operative time, estimated blood loss (EBL), mean number of nodes removed, number of nerves spared, positive margin rates, length of stay, discharge hematocrit, and transfusion rates. All outcomes were comparable between the two groups, with the exception of length of stay, which was shorter in patients who underwent robot-assisted MIRP (1.61 vs. 1.72 days, p<0.0001).

Logistic regression models were created to predict the risk of overall, major, and abdominal complications. In univariate analysis (Table 2), neither prior abdominopelvic surgery alone nor abdominopelvic or inguinal surgery predicted overall, major, or abdominal complications. Significant predictors of overall complications were operative time (odds ratio [OR] 1.56 95% confidence interval [CI] 1.26–1.93, p<0.0001), EBL (OR 1.002, 95% CI 1.000–1.0003, p=0.017), and case number (OR 0.99, 95% CI 0.998–0.999, p=0.010). While increasing operative time and EBL were associated with more complications, the association with case number was an inverse one, as might be expected. The only predictor of major complications was operative time (OR 1.71, 95% CI 1.22–2.41, p=0.0002). In terms of abdominal complications, increasing operative time was associated with greater complications (OR 1.9, 95% CI 1.27–2.87, p=0.0002), and case number was inversely correlated (OR 0.99, 95% CI 0.998–0.999, p=0.10). To assess for independent risk factors for complications, prior abdominal surgery along with all variables with p-values<0.25 were used to construct multivariable models for overall, major, and abdominal complications (Table 3). In multivariable analysis, operative time was the only significant risk factor for both overall complications (OR 1.4, 95% CI 1.04–2.07, p=0.029) and abdominal complications (OR 2.0, 95% CI 1.07–3.72, p=0.030), and there was a trend toward significance for major complications (OR 1.61, 95% CI 1.00–2.61, p=0.051).

Table 2.

Predictors of Overall, Major, and Abdominal Complications: Univariate Analysis

	Overall			Major			Abdominal
	OR	95% CI	p-Value	OR	95% CI	p-Value	OR	95% CI	p-Value
Age	0.98	0.957–1.01	0.308	1.0	0.95–1.05	1.00	0.94	0.88–1.00	0.068
Prior abdominopelvic surgery only	1.05	0.63–1.76	0.849	1.0	0.38–2.63	1.00	0.95	0.273–3.29	0.935
Prior inguinal surgery only	0.75	0.39–1.43	0.382	0.47	0.11–2.01	0.311	a
Prior abdominopelvic or inguinal surgery	0.89	0.569–1.41	0.640	0.69	0.28–1.69	0.41	0.52	0.151–1.81	0.307
Robotic (vs. laparoscopic)	0.74	0.454–1.20	0.225	0.89	0.38–2.10	0.80	0.30	0.081–1.55	0.17
EP (vs. TP)	0.75	0.54–1.13	0.182	0.72	0.34–1.49	0.37	0.77	0.3–1.99	0.6
Operative time (hours)	1.56	1.26–1.93	0.000	1.71	1.22–2.41	0.0002	1.9	61.27–2.84	0.002
BMI	1.04	0.994–0.109	0.090	1.05	0.98–1.13	0.19	1.1	0.97–1.16	0.20
PLND	1.25	0.814–1.93	0.306	2.08	0.85–5.11	0.11	1.35	0.49–3.86	0.54
EBL	1.002	1.000–1.003	0.017	1.00	1.0–1.00	0.38	1.00	0.99–1.00	0.395
Prostate weight	1.00	0.99–1.016	0.453	1.01	0.99–1.03	0.24	0.986	0.95–1.01	0.379
Case number	0.99	0.998–0.999	0.010	1.00	1.00–1.00	0.21	0.99	0.998–0.999	0.010
Primary Gleason pattern
3 (reference)	—						—
4	0.66	0.35–1.22	0.188	0.39	0.09–1.63	0.197	0.64	0.147–2.81	0.556
5	3.6	0.689–18.0	0.129	5.61	0.65–48.14	0.116	—	No complications

^aAll abdominal complications occurred in patients without history of prior inguinal surgery.

BMI=body mass index; EP=extraperitoneal; PLND=pelvic lymph node dissection; TP=transperitoneal.

Table 3.

Predictors of Overall, Major, and Abdominal Complications: Multivariable Analysis

	Overall			Major			Abdominal
	OR	95% CI	p-Value	OR	95% CI	p-Value	OR	95% CI	p-Value
Age	Not included in model			Not included in model			0.96	0.89–1.03	0.25
Prior abdominopelvic surgery only	0.95	0.52–1.74	0.88	1.28	0.47–3.51	0.63	a
Robotic (vs. lap)	1.45	0.57–3.70	0.43	Not included in model			0.46	0.046–4.2	0.52
EP (vs. TP)	1.17	0.62–2.19	0.62	Not included in model			1.22	0.27–5.56	0.76
Operative time (hours)	1.4	1.04–2.07	0.029	1.61	1.00–2.61	0.051	2.00	1.07–3.72	0.030
BMI	Not included in model			1.06	0.97–1.16	0.20	1.00	1.00–1.00	0.35
PLND	Not included in model			2.50	0.96–6.51	0.061	Not included in model
EBL	1.00	0.999–1.00	0.41	N			Not included in model
Prostate weight	Not included in model			1.01	0.99–1.03	0.173	Not included in model
Case number	0.99	0.997–1.00	0.43	1.00	1.00–1.00	0.743	1.00	1.00–1.00	0.73
Primary Gleason pattern (reference 3)							Not included in model
4	0.80	0.41–1.56	0.51	0.38	0.086–1.67	0.160
5	4.20	0.78–22.5	0.095	4.96	0.53–46.47	0.16

All variables with p<0.25 from the univariate analysis were included in each respective model.

^aAll abdominal complications occurred in patients without history of prior surgery.

Discussion

Despite the increasing popularity of MIRP, some practitioners may be reluctant to pursue a minimally invasive approach to radical prostatectomy in patients with a history of prior abdominopelvic or inguinal surgery due to concerns for technical difficulties or higher rates of adverse patient outcomes. In the current series, at least 25% of patients had previously undergone abdominopelvic or inguinal surgery. As such, concerns regarding the safety of MIRP in these patients could potentially influence surgical decision making in a large percentage of patients with prostate cancer. This study, however, demonstrates the feasibility of MIRP in patients with a history of prior surgery. In addition, it demonstrates the safety of laparoscopic access using a Veress needle and optical trocar to establish access for TP cases, and the safety of direct entry into the space of Retzius with an optical trocar for EP cases, in experienced hands. No enterotomies (other than one rectal injury unrelated to access) were encountered in this study.

When comparing men with no surgical history to those with a history of prior abdominopelvic or inguinal surgery, we found no significant differences in operative time, estimated blood loss, performance of pelvic lymph node dissection, performance of a nerve-sparing procedure, positive margin rates, length of stay, or transfusion rate. Furthermore, we found that complication rates were comparable regardless of a patient's prior surgical history. These results argue that patients should not be discouraged from undergoing MIRP on the basis of prior surgical history, particularly if both TP and EP approaches can be offered and selected depending on the prior surgical history and location of the previous incision lines.

Our findings corroborate those of prior studies reporting favorable outcomes following TP robotic prostatectomy in patients with a prior surgical history.^9,10 Ginzburg et al. compared patients with prior surgery to those with no such history and found similar operative times, surgical margin positivity, and complication rates between groups.⁹ Siddiqui et al. performed a similar comparison, and reported no differences in EBL, operative time, or complication rates between groups.¹⁰ These studies, however, exclusively investigated patients treated with robotic radical prostatectomy, as opposed to the current analysis in which a majority of patients underwent laparoscopic radical prostatectomy.

In this study, we noted that a larger percentage of patients with a surgical history underwent robot-assisted, as opposed to pure laparoscopic, MIRP. To assess whether this confounder accounted for the equivalent outcomes between groups, we performed a subset analysis comparing laparoscopic to robotic outcomes in men with prior surgery. We found no difference in outcomes between the laparoscopic and robotic cases, with the exception of length of stay, which was shorter in the robot-assisted cases. We feel that this difference in length of stay is not due to the robotic technology, but rather that the robotic cases were the most recent cases, when our postoperative pathway had been refined to allow patients to be discharged earlier. Thus, we do not feel that robotic technology specifically accounts for the safety and efficacy of MIRP in patients with prior surgery. In addition, the robotic cases were done later in the surgeon's experience, so we accounted for case number in our analyses; increasing case number was associated with decreasing overall and major complications on univariate analysis, whereas an association between robotics and complication rate was not found.

In the United States, the vast majority of minimally invasive robotic prostatectomy procedures are performed transperitoneally.¹¹ This approach may be problematic in patients with a prior history of abdominal surgery.¹² Extensive intra-abdominal adhesions are known to increase the risk of inadvertent bowel injury when establishing pneumoperitoneum and placing laparoscopic ports.¹³ Furthermore, many patients with prior abdominal surgery may require extensive lysis of adhesions to gain access to the pelvis, further subjecting patients to potential bowel injury and increasing operative time. In fact, a prior series reporting robotic prostatectomy outcomes in patients with a history of prior abdominal or inguinal surgery found that 23.2% of such patients required adhesiolysis, and in 3.8% adhesiolysis increased the operative time by >30 minutes.¹⁰ Relevant to these data, our series demonstrates a correlation between increased operative time and overall and abdominal complications.

In the current series, the majority of procedures were performed via an EP approach. This approach is particularly valuable in patients with prior abdominal surgery, as it avoids potential complications related to access in a hostile abdomen, and obviates the need to perform a lysis of adhesions except as related to prior inguinal or pelvic surgery such as inguinal hernia repair (to access the pelvic lymph nodes). The ability to perform EP MIRP in a majority of patients, particularly in patients with multiple prior intraabdominal surgeries, likely reduced operative time and may have contributed to the low rate of surgical complications observed in the current series.

In men with prior abdominal surgery, the EP approach allows for entry into a virgin space. Even prior colonic surgery usually spares this space and allows for EP access if the pelvis is approached infraumbilically and in avoidance of prior incision lines. As an alternative to a pure EP approach, TP access can be obtained, and if extensive adhesions are encountered in the rectovesical pouch, the surgeon can perform the posterior dissection (seminal vesicles and vasa deferentia) in an antegrade fashion, obviating pelvic adhesiolysis. Conversely, in patients with prior laparoscopic mesh inguinal hernia repairs, particularly if bilateral, it is prudent to obtain TP access and carefully take the bladder down rather than to try to redevelop the previously operated-on EP space. In sum, caution and a familiarity with a variety of approaches to the prostate and surrounding structures allow the surgeon to avoid adhesions, scars, and concomitant bowel injury.

There are several limitations to this study. The surgical technique used for MIRP evolved over time and was thus heterogeneous; both TP and EP approaches were used, and pure laparoscopic technique was eventually replaced by robot-assisted laparoscopic technology. Our data are not sufficient to establish the superiority of one MIRP approach (TP or EP) or technique (laparoscopic or robotic) over another when approaching a patient with prior surgical history. However, we also see this as a strength of the study, as it shows that the ability to vary surgical approach and technique from one patient to another allows for a safe and oncologically sound operation in almost every patient. Additionally, the MIRP analyzed in the current study were performed at a tertiary referral center. Therefore, it is possible that some postoperative complications were managed locally and not captured in our data set, potentially biasing our results. This is unlikely, however, as nearly all our patients had follow-up at approximately 3 months postoperatively. Furthermore, this study encompasses results from a single high-volume surgeon. Results may not be generalizable to surgeons earlier in their learning curve. Finally, although all data were collected prospectively, the analysis is retrospective and thus subject to the inherent limitations of retrospective nonrandomized analyses. There is a likely a degree of selection bias in both referral patterns and choice of treatment modality that cannot be accounted for in this analysis.

Conclusions

In a large series of patients treated with MIRP for prostate cancer, over 25% had a history of prior abdominal, pelvic, or inguinal surgery. We found that perioperative outcomes in these men, as compared to those in patients with no surgical history, were similar. Furthermore, we found similar complication rates between the two groups, suggesting that MIRP can safely be performed in men with a history of prior abdominopelvic and/or inguinal surgery, particularly when the options of operating EP or TP, and dissecting the posterior structures antegrade or retrograde, can be offered.

Abbreviations Used

BMI

body mass index

EBL

estimated blood loss

EP

extraperitoneal

MIRP

minimally invasive radical prostatectomy

TP

transperitoneal

Disclosure Statement

No competing financial interests exist.