Clavien Classification of Complications After the Initial Series of Robot-Assisted Radical Prostatectomy: The Cancer Institute of New Jersey/Robert Wood Johnson Medical School Experience

Abstract

Purpose:

To study the safety and feasibility of robot-assisted radical prostatectomy (RARP) for the surgical management of localized prostate cancer, we analyzed perioperative parameters and the pattern of complications in our patients who underwent RARP.

Patients and Methods:

After the performance of more than 600 RARP over a 4-year period by a single surgeon using the daVinci® robot system at the Cancer Institute of New Jersey/Robert Wood Johnson Medical School, we reviewed the medical records of the first 200 patients retrospectively. All patients were divided into four groups according to the order of case numbers to compare intergroup differences in preoperative characteristics and perioperative parameters. Perioperative complications were determined in all patients, and complications were classified according to the Clavien classification system.

Results:

The mean operative time was 212 minutes, and the mean blood loss was 189 mL. The mean length of hospital stay was 1.13 days. Overall, 12% (24 men) experienced various perioperative complications among the 200 patients. Of the total 24 patients, 5 (20.8%) men experienced intraoperative complications, and 19 (79.2%) men showed postoperative complications. Rectal injury occurred in two (8.3%) men, and the injury was repaired primarily using two-layer suture techniques without any sequelae. Three (12.5%) patients had femoral neuropathy, and urinary retention developed in 7 (25.0%) patients. Among our 200 patients, no transfusion was needed intraoperatively and postoperatively. There were nine (4.5%) patients in the Clavien grade I complications category, and another 9 (4.5%) men were classified as grade II complications. Six (3.0%) men had grade IIIb complications, and there were no grade IV or V complications.

Conclusions:

In our initial series of RARP procedures, we experienced low morbidity, with the overall complication rate of 12%. After implementing minor modifications, most of the early complications were prevented. Rectal injuries, if recognized intraoperatively, can be repaired primarily.

Introduction

T he optimal surgical procedure for localized prostate cancer still remains controversial. As the gold standard operation, radical retropubic prostatectomy (RRP) has been modified over two decades to provide excellent cancer control with preservation of continence and erectile function.¹ According to a recent evaluation of the parameters that are pertinent to open RRP, skillful modern surgeons can experience excellent overall outcomes composed of low complication rates, positive surgical margin rates, and biochemical recurrence rates.

Despite the proven results of open RRP, robot-assisted radical prostatectomy (RARP) has become the most commonly performed surgical procedure in patients with localized prostate cancer. Many surgeons eagerly adopted RARP, because robotic technology, composed of three-dimensional vision and seven degrees of freedom of instrument movement, significantly shortened the learning curve of laparoscopic radical prostatectomy (LRP). More importantly, patients and their families enthusiastically accepted RARP because of the perception of decreased morbidity after RARP.

The benefit of RARP, however, has been questioned, because several early reports of LRP and RARP revealed no compelling advantage over open RRP.² More recently, comparison of outcomes between laparoscopic and open RRP using the 2002 to 2007 Surveillance, Epidemiology, and End Results Medicare database revealed decreased blood transfusion rates and shorter length of hospital stay in the laparoscopic group, but this improvement came at the cost of higher incidence of incontinence and impotence.³

The little or no advantage after RARP in these early reports, though, may be because of a steep learning curve. Indeed, since the initial reports on LRP and RARP, technical modifications have been made and improved outcomes have begun to emerge in high volume centers.^4
–6

At the Cancer Institute of New Jersey/Robert Wood Johnson Medical School, more than 600 RARPs have been performed over the last 4 years by a single surgeon. In this study, we have investigated the impact of the learning curve on the rate of complications after RARP by retrospectively analyzing the results of the initial 200 RARPs.

Patients and Methods

To date, more than 600 RARPs have been performed at the Cancer Institute of New Jersey/Robert Wood Johnson Medical School using the daVinci® robot system (Intuitive Surgical, Sunnyvale, CA) by a single surgeon (I K). The charts of the first 200 patients who underwent RARP between January 2006 and December 2007 were reviewed retrospectively after acquiring Institutional Review Board approval at our institution.

All surgical procedures were performed transperitoneally. The National Comprehensive Cancer Network guideline was used to determine the necessity of pelvic lymph node dissections. In appropriate patients, bilateral nerve sparing was performed using the interfascial approach.

For the purpose of comparing intergroup differences in terms of preoperative patient characteristics and perioperative parameters, all patients were divided into four groups according to the order of case numbers (1–50, 51–100, 101–150, and 151–200). Preoperative demographics, including age, body weight, serum prostate-specific antigen (PSA) value, biopsy Gleason score, and clinical stage, were analyzed. Also, perioperative parameters (operative time, intraoperative blood loss, hospital stay, and complication rate) were presented.

All patients were included in the analysis of perioperative complication pattern, and complications were classified according to the Clavien classification system updated in 2004.⁷ This system determines severity of complication using five grade scales. Grade I complications are designated as any deviation from the normal postoperative course and without the need for therapy (with exceptions of antiemetic, antipyretic, analgesic, and antidiarrheal drugs). Also, grade I includes wound infections opened at the bedside. Grade II complications necessitate pharmacologic treatment with drugs other than the drugs mentioned in grade I complications. This grade includes blood transfusion and hyperalimentation. Grade III complications are defined as complications necessitating surgical, endoscopic or radiologic intervention. This grade is subdivided into grades IIIa and IIIb based on the need for general anesthesia. Grade IV includes life-threatening complications necessitating intensive care with residual disability, and grade V represents death of a patient from complications. If the patient has a complication at discharge, the suffix “d” (for “disability”) is highlighted to the respective grade.

Finally, we compared our outcomes with representative contemporary series of RRP, LRP, and RARP in terms of perioperative complication rates. The results of data analysis were expressed in descriptive statistics consisting of mean value and interval. To compare the perioperative characteristics, the Student t test was used. If the P value was less than 0.05, the results were considered statistically significant.

Results

Preoperative patient demographics are shown in Table 1. The mean age of the men who were undergoing RARP was 58.8 years, and mean body mass index (BMI) was 28.4 kg/m². The mean serum PSA value was 6.31 ng/mL, and the mean biopsy Gleason score was 6.41. There were no significant differences among the four groups with respect to these four characteristics. Overall, 9.5% of patients had biopsy Gleason score ≥8, and 80% had clinical T_1c cancers.

Table 1.

Preoperative Patient Demographics

Case number	Mean age (y)	Mean BMI (kg/m ² )	Mean PSA	Mean Gleason score	% Gleason ≥8	%T_1c
1–50	59.4 (45–70)	28.0 (21.0–39.9)	6.89 (0.9–52.4)	6.6 (6–9)	12%	78%
51–100	57.7 (39–67)	28.4 (21.7–35.9)	6.19 (1.4–20.4)	6.47 (6–9)	10%	82%
101–150	58.9 (48–73)	28.2 (20.2–38.6)	6.11 (0.6–17.7)	6.24 (6–8)	4%	88%
151–200	58.9 (45–74)	28.7 (21.3–42.0)	6.16 (1.1–31.3)	6.38 (4–9)	12%	76%
Total	58.8 (39–74)	28.4 (20.2–42.0)	6.31 (0.6–52.4)	6.41 (4–9)	9.5%	80%

Numbers in parentheses are the range.

BMI = body mass index; PSA = prostate-specific antigen.

The mean operative time was 212 minutes, and the mean estimated blood loss (EBL) was 189 mL (Table 2). In patients 1 to 50, the mean operative time (250 min) was relatively longer than that of the other three groups. The mean length of hospital stay was 1.13 days, and there was no significant difference among the four groups. Overall, 24 of the 200 men had perioperative complications (12%); relatively high complication rates were seen in the first (16%) and fourth groups (18%).

Table 2.

Perioperative Parameters

Case number	Mean OR time (min)	Mean EBL (mL)	Mean hospital stay (days)	Perioperative complications (%)
1–50	250 (165–540)	205 (50–800)	1.28 (1–3)	8 (16%)
51–100	215 (140–355)	200 (100–800)	1.14 (1–4)	4 (8%)
101–150	213 (135–315)	202 (50–700)	1.00 (1–1)	3 (6%)
151–200	169 (110–280)	148 (50–400)	1.12 (1–4)	9 (18%)
Total	212 (110–540)	189 (50–800)	1.13 (14)	24 (12%)

Numbers in parentheses are the range.

OR = operating room; EBL = estimated blood loss.

Table 3 shows the comparison of the perioperative patient characteristics between the group with and without complications. There were no significant differences between the two groups in age, BMI, biopsy Gleason score, and operative time. However, the patients with complications, however, had a higher PSA level, larger EBL, and longer length of hospital stay than the patients without complications.

Table 3.

Perioperative Characteristics of Patients With and Without Complications

	Complication (−) group (n = 176)	Complication (+) group (n = 24)	Total (n = 200)	P value
Age (y)	58.5 (39–74)	60.4 (50–68)	58.8 (39–74)	NS
BMI (kg/m²)	28.5 (20.2–42.0)	27.1 (21.0–33.9)	28.4 (20.2–42.0)	NS
Preoperative PSA	6.04 (0.2–31.3)	8.40 (2.63–52.4)	6.31 (0.2–52.4)	0.043
Biopsy Gleason score	6.38 (4–9)	6.68 (6–9)	6.41 (4–9)	NS
OR time (minutes)	210 (110–540)	224 (130–330)	212 (110–540)	NS
EBL (mL)	180 (50–800)	260 (50–600)	189 (50–800)	0.014

Numbers in parentheses are the range.

NS = not significant; BMI = body mass index; PSA = prostate-specific antigen; OR = operating room; EBL = estimated blood loss.

Table 4 displays the pattern of perioperative complications. Of the 24 patients with complications, 5 (20.8%) men experienced intraoperative complications, and 19 (79.2%) men had postoperative complications. Three (12.5%, 3/24) patients, all from the first group (case numbers 1–50), were converted to open RRP because of lack of progress. Rectal injury occurred in two (8.3%) men during a wide resection of the prostate in patients with high-risk localized prostate cancer. Both injuries were recognized intraoperatively and repaired primarily using the two-layer suture technique in which the mucosa was closed followed by the reinforcement of the serosa without tension. Subsequently, a perivesical fat flap was harvested and interposed to cover the repaired rectum. At the 2-year follow-up, both patients remain free of any sequelae.

Table 4.

Pattern of Perioperative Complications

Case number	1–50	51–100	101–150	151–200	Total (24)	Complication grade
Conversion to open	3				3	IIIb
Rectal injury^a				2	2	IIIb
Wound infection	1	1	2		4	I
DVT				1	1	I
Femoral neuropathy^b	3				3	I
Urinary retention^c			1	6	7	II
Ileus	1				1	II
BNC		1			1	IIIb
Transfusion					0	II
Others		2 (SVT, clot retention)			2	I/II
Total (24)	8	4	3	9	24

Both rectal injuries occurred during the wide resection of high–risk prostate cancers. Both were repaired primarily.

No femoral neuropathy occurred after switching from Allen stirrup to spreader bar.

At case 151, Rocco posterior urethral plate repair was implemented.

DVT = deep vein thrombosis; BNC = bladder neck contracture; SVT = supraventricular tachycardia.

Postoperative complications were distributed evenly in the four groups except femoral neuropathy and urinary retention. Femoral neuropathy was seen in only three patients in the first group (case numbers 1–50) and did not occur again after switching from Allen stirrup to leg spreader bar. Urinary retention eveloped in six (25.0%) patients in the fourth group (case numbers 151–200) after we implemented the Rocco posterior urethral plate repair at case 151. No transfusion was needed intraoperatively and postoperatively.

According to the 2004 Clavien classification system, there were nine (4.5%) patients with grade I complications consisting of wound infection, deep vein thrombosis, supraventricular tachycardia, and femoral neuropathy. In addition, nine (4.5%) men had grade II complications that included urinary retention, ileus, and clot retention. The one case of ileus resolved with bowel rest and nasogastric tube suction for 2 days. Finally, six (3.0%) men had grade IIIb complications—open conversion, rectal injury, and bladder neck contracture (BNC). The lone patient with BNC presented at the 3-month follow-up visit and underwent transurethral incision of bladder neck contracture. This patient remains pad free 3 years later. Clavien grade IV or V complications were not identified in our patients (Table 5).

Table 5.

Perioperative Complications of Prostatectomy in the Published Literature

	RRP	LRP	RARP
Grade	Constantinides et al ¹⁷	Hu et al ⁵	Badani et al ¹⁰	Present study
I	34 (3.4%)	68 (19.1%)	221 (8.0%)	9 (4.5%)
Id	39 (3.9%)
II	128 (12.8%)	45 (12.6%)	102 (3.7%)	9 (4.5%)
IId
III			13 (0.5%)
IIIa	26 (2.6%)	3 (0.8%)
IIIb	38 (3.8%)	30 (8.4%)		6 (3.0%)
IV			2 (0.01%)
V	3 (0.3%)		1 (<0.01%)
Total	268/1,000 (26.8%)	99/358 (27.7%)	339/2276 (12.3%)	24/200 (12.0%)

RRP = radical retropubic prostatectomy; LRP = laparoscopic radical prostatectomy; RARP = robot-assisted radical prostatectomy.

Discussion

Radical prostatectomy has been established as the most effective treatment modality for patients with localized prostate cancer.⁸ Recently, RARP has been recognized as a new therapeutic option since its first introduction by Abbou and associates⁹ in 2001. Indeed, RARP has been accepted as the surgical treatment of choice by many urologists and patients.¹⁰ It was 60% of total volume of radical prostatectomies performed by American urologists in 2008.¹¹

Although the long-term effect of RARP on outcome still remains controversial,³ there has been a growing number of reports showing better outcomes in the context of preservation of continence and erectile function as well as favorable cancer control with refinements in technique during RARP procedures.¹²

Among the recent reports, Badani and colleagues¹⁰ presented the largest series of patients undergoing RARP with the longest follow-up to date. Over a 6-year period, 2766 consecutive men underwent RARP. The mean surgical time and robot console time were 154 minutes and 116 minutes, respectively. Mean EBL was 142 mL, and 96.7% of patients were discharged within 24 hours of surgery. The authors reported a 7.3% PSA recurrence rate at a median follow-up of 22 months and an 84% 5-year biochemical recurrence-free survival rate. Overall, the authors demonstrated that RARP can be performed with favorable outcomes while they can minimize perioperative complications (339/2276, 12.3%).

In our initial series of 200 RARP procedures, we experienced very low morbidity with overall 12.0% complication rate (24/200). The mean operative time, the mean EBL, and the mean length of hospital stay of our series were 212 minutes, 189 mL, and 1.13 days, respectively.

We applied the Clavien classification system in our 200 patients, demonstrating complication rates of 4.5%, 4.5%, and 3.0% for grades I, II, and IIIb (Table 5). Of the total 24 complications in our series, there were 2 rectal injuries (grade IIIb) in patients who were undergoing wide resection because of high-risk localized prostate cancer. Both rectal injuries were recognized intraoperatively and repaired primarily using two-layer suture techniques without any sequelae. Rectal injury is an uncommon complication of radical prostatectomy, with an incidence of 0.5% to 9% in general, less than 0.5% in contemporary open series, and 0.2% to 0.8% in the recent RARP series.¹³ Although the incidence of rectal injury during RARP is low, the sequelae can be devastating if the injury is not recognized intraoperatively and adequately repaired. We managed the two rectal injuries uneventfully with a meticulous two-layer closure reinforced by a perivesical fat flap.

In our current cohort of 200 patients who underwent RARP, no blood transfusion was necessary intraoperatively and postoperatively. Currently, we have reached 600 patients without any blood transfusion at our institution. Our perfect record of blood transfusion rate is identical to that of Patel and coworkers¹⁴ (0/200 RARP) and El-Hakim and associates¹⁵ (0/373 RARP) while Murphy and colleagues⁶ and Badani and coworkers¹⁰ reported very low blood transfusion rates of 2.5% (10/400 RARP) and 1.5% (41/2776 RALRP), respectively. Although one institution has reported that RARP and open RRP have similar rates of transfusion,¹⁶ our results suggest that the transfusion rate is significantly lower when the results of expert surgeons at high-volume centers are analyzed.

Analysis of our data by 50-patient blocks revealed three cases of transient femoral neuropathy in the first 50-patient group. After switching from the Allen stirrup to leg spreader bars, no further incidence of femoral neuropathy occurred. Based on this observation, the use of the leg spreader bar is mandated at our institution during RARP.

More interestingly, a sudden increase in urinary retention necessitating prolonged Foley catheter drainage was observed in patients 151 to 200. In this group of patients, the Rocco posterior urethral plate repair was implemented with the goal of achieving early continence. With this increased urinary retention rate, however, the Rocco repair has been abandoned at our institution.

To compare our results of complication analysis with that of other centers, we used the Clavien classification system (Table 5). Comparison of outcome data between different centers is limited mainly by the lack of consensus on the definition of complications and stratification of severity. The Clavien system was originally devised in 1992 to standardize and compare complications using a reliable and valid method, and it has been modified several times to improve some drawbacks. The most recent revision was reported in 2004 with evaluation in a cohort of 6336 patients, resulting in the validation of its usefulness.⁷

Many authors have reported their outcomes using this system to present their negative surgical events. Constantinides and colleagues¹⁷ reported their complication rates of RRP procedure for grades I, Id, II, IIIa, IIIb, and V of 3.4%, 3.9%, 12.8%, 2.6%, 3.8%, and 0.3%, respectively. Also, Hu and associates⁵ reported 19.1%, 12.6%, 0.8%, and 8.4% for grades I, II, IIIa, and IIIb complications of their LRP procedure. As for RARP, Badani and coworkers¹⁰ demonstrated 8.0%, 3.7%, 0.5%, 0.01%, and <0.01% for grades I, II, III, IV, and V, whereas our present study showed 4.5%, 4.5%, and 3.0% for grades I, II, and IIIb, respectively, without grade IV or V complications. Overall complication rates were represented as 26.8%, 27.7%, and 12.3% for these contemporary series of RRP, LRP,, and RARP, respectively, whereas our study demonstrated the lowest rate as 12.0%.

Conclusions

Our initial experience of RARP series demonstrated promising short-term outcomes with rare perioperative complications. We experienced low morbidity with an overall complication rate of 12% and most of the early complications were prevented with minor modifications. By carefully analyzing the perioperative outcomes after every 50 patients, any adverse trends can be recognized and changes can be implemented early to improve the safety and surgical outcomes after RARP.

Footnotes

Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Walsh

. Anatomic radical prostatectomy: Evolution of the surgical technique. J Urol, 1998; 160:2418–2424.

Lepor

, Kaci

. Contemporary evaluation of operative parameters and complications related to open radical retropubic prostatectomy. Urology, 2003; 62:702–706.

, Gu

, Lipsitz

et al. Comparative effectiveness of minimally invasive vs open radical prostatectomy. JAMA, 2009; 302:1557–1564.

Kaul

, Savera

, Badani

et al. Functional outcomes and oncological efficacy of Vattikuti Institute prostatectomy with Veil of Aphrodite nerve-sparing: An analysis of 154 consecutive patients. BJU Int, 2006; 97:467–472.

, Nelson

, Wilson

et al. Perioperative complications of laparoscopic and robotic assisted laparoscopic radical prostatectomy. J Urol, 2006; 175:541–546.

Murphy

, Kerger

, Crowe

et al. Operative details and oncological and functional outcome of robotic-assisted laparoscopic radical prostatectomy: 400 cases with a minimum of 12 months follow-up. Eur Urol, 2009; 55:1358–1366.

Dindo

, Demartines

, Clavien

. Classification of surgical complications: A new proposal with evaluation in a cohort of 6336 patients and results of a survey. Ann Surg, 2004; 240:205–213.

Roehl

, Han

, Ramos

et al. Cancer progression and survival rates following anatomical radical retropubic prostatectomy in 3,478 consecutive patients: Long-term results. J Urol, 2004; 172:910–914.

Abbou

, Hoznek

, Salomon

et al. Laparoscopic radical prostatectomy with a remote controlled robot. J Urol, 2001; 165:1964–1966.

10.

Badani

, Kaul

, Menon

. Evolution of robotic radical prostatectomy: Assessment after 2766 procedures. Cancer, 2007; 110:1951–1958.

11.

Dasgupta

, Patil

, Anderson

, Kirby

. Transition from open to robotic-assisted radical prostatectomy. BJU Int, 2008; 101:667–668.

12.

McCullough

, Barret

, Cathelineau

et al. Role of robotics for prostate cancer. Curr Opin Urol, 2009; 19:65–68.

13.

Yee

, Ornstein

. Repair of rectal injury during robotic-assisted laparoscopic prostatectomy. Urology, 2008; 72:428–431.

14.

Patel

, Tully

, Holmes

, Lindsay

. Robotic radical prostatectomy in the community setting—the learning curve and beyond: Initial 200 cases. J Urol, 2005; 174:269–272.

15.

El-Hakim

, Leung

, Tewari

. Robotic prostatectomy: A pooled analysis of published literature. Expert Rev Anticancer Ther, 2006; 6:11–20.

16.

Smith

Jr , Herrell

. Robotic-assisted laparoscopic prostatectomy: Do minimally invasive approaches offer significant advantages? J Clin Oncol, 2005; 23:8170–8175.

17.

Constantinides

, Tyritzis

, Skolarikos

et al. Short- and long-term complications of open radical prostatectomy according to the Clavien classification system. BJU Int, 2009; 103:336–340.