Introduction

Radical cystectomy is the gold standard for the treatment of muscle invasive bladder cancer as well as recurrent high-grade non-invasive bladder cancer. The traditional open radical cystectomy (ORC) is fraught with postoperative complications, reported to be around 30–50 %, due in part to the age and the often severe comorbidities that are inherent in this patient population. In recent years, the use of minimally invasive surgery for the treatment of bladder cancer has been employed in attempts to reduce the morbidity of what has traditionally been the most invasive of urologic procedures. The use of the robotic platform in performing cystectomies was first described in 2003. Since then, robotic-assisted radical cystectomy (RARC) has gained in popularity but has for the most part been limited to high-volume referral centers.

The transition from open to robotic-assisted cystectomy has come in large part due to the urologists’ familiarity with the robotic set-up and pelvic anatomy, as it parallels that seen during robotic-assisted radical prostatectomy (RARP). The vast majority of prostatectomies in the United States, over 80 %, are now being performed robotically [1]. This is true in our health care setting as well. We have previously reported on the transition from open radical prostatectomies to robotic-assisted radical prostatectomies and have shown similar perioperative and pathologic outcomes [2]. Similar to the evolution of RARP, robotic technology has attracted the attention of urologists and patients alike who prefer a minimally invasive surgery for muscle-invasive bladder cancer. The da Vinci robotic technology offers 3-D vision, seven degrees of freedom of movement, and lack of tremor while providing results similar to the benchmark open procedure.

After a journal literature search using PubMed in March 2012, we did not find any reports evaluating initial robotic-assisted radical cystectomy experience compared with open radical cystectomy in a community-based, non-tertiary health care setting. In the present study, we compared intraoperative and perioperative outcomes to assess the safety and feasibility of the transition from ORC to RARC.

Materials and methods

Patient population

From 2003 to 2010, 14 consecutive patients underwent an ORC followed by 14 unselected consecutive patients who underwent a RARC by a single urologic surgeon in a community urologic practice. The transition was made after the initiation of the robotic technique by the surgeon and was not based on patient characteristics. This transition was after 147 robotic-assisted laparoscopic prostatectomies had been performed by the urologist. After institutional review board approval, a retrospective review and data collection of the first 14 RARC patients and the last 14 ORC patients was performed. These were consecutive patients and previous pelvic surgery or radiotherapy did not preclude a robotic approach to cystectomy.

Operative technique and postoperative care

ORC was performed in the traditional fashion [3]. RARC was performed as previously described [4]. Figures 1 and 2 show port placement. All urinary diversions were performed after extension of the midline camera port to an open surgical incision after the RARC and lymph node dissection was performed. In this series, all patients had an ileal conduit urinary diversion. The postoperative care for all patients was directed by the standardized departmental care pathway that included a nasogastric tube until evidence of bowel function by passing of flatus.

Figs. 1–2
figure 1

Set-up for robotic-assisted cystectomy

Full size image

Data collection

The data collected included age, gender, body mass index, operative time, estimated blood loss, pathologic stage, number of lymph nodes, complications, days to return of bowel function, and length of hospital stay. Complications were graded using the modified Clavien score [5]. Minor complications were defined as Clavien grade 1–2 and major complications as Clavien grade 3 or greater.

Statistical analysis

Analyses were performed using SAS Proprietary Software version 9.3. All tests employ a significance level of 0.05 and are two-tailed. Complication rates were assessed using Kruskal–Wallis tests. Unpaired t-tests, χ2 tests of independence and Fisher’s exact tests were performed on length of hospital stay (in days), length of surgical procedure, time to return of bowel function, and estimated blood loss.

Results

In total, 28 patients underwent radical cystectomy with extended lymph node dissection and were retrospectively reviewed (Table 1). From 2003–2010, 14 consecutive patients underwent an ORC and from 2010–2012 another 14 consecutive patients underwent a RARC. The primary variable analyzed was complications. The secondary outcome variables included oncologic outcomes (number of lymph nodes and surgical margin status), length of hospital stay, operative time, time to return of bowel function (postoperative ileus was defined as greater than 5 days), and intraoperative blood loss. The mean patient age was 71 years for RARC and 67 years for ORC. There were three females in the RARC group and no females in the ORC group. Of the patients in each group, all elected to receive an ileal conduit. Regarding body mass index (BMI), 35 % of patients in the RARC group were considered obese (BMI ≥30) compared to 28 % in the ORC group.

Table 1 Baseline patient demographics and perioperative outcomes
Full size table

There were two significant differences found between the ORC and RARC groups. The operative time was significantly longer in the RARC group (6 h 23 min vs. 4 h 28 min; p < 0.05), and RARC was associated with significantly lower intraoperative blood loss compared with ORC (470 ml vs. 942 ml; p < 0.05). The median estimated blood loss (EBL) for the RARC group was 350 ml (range 100–1,300 ml) and for ORC was 800 ml (range 150–2,200 ml). The RARC group had one patient who received a blood transfusion, whereas the ORC group had four patients who received blood transfusions. Figure 3 displays the trend in EBL and operative time during the study period, showing a trend towards improved operative time while maintaining a stable EBL.

Fig. 3
figure 2

a Estimated blood loss and b surgery time comparing ORC with RARC

Full size image

Regarding complications, 21 % of RARC patients experienced major complications versus 14 % of ORC patients. Overall, there was no statistically significant difference in complication rates between the ORC and RARC groups. Postoperative ileus was the most common complication, occurring in 57 and 58 % of the RARC and ORC patients, respectively. There was no significant difference between lengths of hospital stay, with patients staying on average 11.4 days following ORC versus 12 days following RARC.

No pathologic differences were noted between the two groups. The pathology from both series ranged from Tis/T1 to T4. The two groups did not have a statistically significant positive surgical margin rate, with two ORC patients and three RARC patients having positive margins. The lymph node counts were similar between the two groups, with median numbers of 11.9 and 9.5 in RARC and ORC, respectively. Lymph node positivity was found in 7 and 35 % of RARC and ORC patients, respectively.

Discussion

Our study is the first comparative study of ORC and RARC in a community-based, non-tertiary health care setting. Previous studies have all reviewed data from surgeries at high-volume referral centers. Most of these surgeries are performed by either oncology fellowship-trained urologists or urologists who specialize primarily in minimally invasive robotic surgery. However, most practicing urologists do not practice in high-volume centers, yet want to offer a minimally invasive approach to their patients. Even smaller hospitals are instituting the da Vinci robotic system despite the substantial initial capital investment and high maintenance costs associated with it. This is primarily due to patients’ demands for the least invasive surgery possible and the desire of health care systems not to lose the income that can be generated by performing these types of surgeries. In the past, these patients have typically been referred to tertiary centers for their care. In the present study we have demonstrated the safety and applicability of instituting the robotic platform for performing radical cystectomy in a community-based, non-tertiary health care setting.

Our study has several limitations. The clinical and oncologic data were limited to the immediate postoperative period. The examination of oncologic outcomes was limited to pathologic data, which we intended to serve as surrogates for the quality of the operation. This was also a retrospective study at a single institution with a small sample size that was not randomized.

Surgical procedures are associated with a certain number of initial cases that must be performed to obtain proficiency and to assess safety and viability. In addition, cancer control is of the utmost importance when comparing newer procedures to the gold standard and surgical margin status is often reflective of the learning curve associated with adopting a new surgical approach. The learning curve in our series did not have a statistically significant impact on positive surgical margins, with two and three positive margins in the ORC and RARC series, respectively. When adopting a new approach to a surgical procedure, operative times during the initial cases tend to be inherently longer. This was true in our study as well, with operating room time being significantly longer in the RARC group (p < 0.05). Hayn et al. demonstrated a decreased operative time with increasing number of cases in a review of their initial 164 consecutive RARC cases [6]. As both the physician and the operating room staff become more familiar and comfortable with the new procedure, operative times are expected to diminish.

Pelvic lymphadenectomy is an integral part of radical cystectomy and, in particular, the number of lymph nodes removed and percent positive nodes has been important for both overall survival and prognostic purposes. Studies have retrospectively shown a curative benefit in up to 35 % of patients with nodal metastases [7, 8]. To date, there are no long-term (>6 year) oncologic outcome data available for RARC. Because of this, other parameters must be used to help define oncologic equivalency to ORC. Lymph node dissection and lymph node count, and surgical margin status have been used as surrogate markers for oncologic outcomes while survival data continue to mature.

Proponents of ORC have criticized RARC as being an inferior operation, in large part because it is felt that the lymph node yield is less than that obtained during ORC. Robotic arms limit the maneuverability within the abdominal cavity, especially for more proximal dissections. Concern over whether it is possible to perform an adequate lymphadenectomy during RARC seems to have been answered by minimally invasive surgeons from high-volume centers. Studies by Wang et al. compared their RARC series with a contemporary ORC series and did not find any difference in the number of lymph nodes removed [9]. Similar results were also reported by Abraham et al. and Lavery et al. [10, 11]. Our data also demonstrated a similar lymph node yield when comparing ORC and RARC. This reveals that the same oncologic principles can be maintained when transitioning to a robotics platform.

Open radical cystectomy remains a highly morbid procedure. In a recent prospective series from Memorial Sloan–Kettering Cancer Center, 64 % of patients undergoing ORC experienced at least one perioperative complication within 90 days of surgery, and 13 % experienced a high-grade complication (grade 3 or higher) [12]. The majority of complications were gastrointestinal, followed by infectious and wound-related complications. Previous reports from a high-volume referral center comparing ORC with RARC showed a decreased blood loss but found no difference in the rate of 30-day minor or major Clavien complications, length of hospital stay, or 30-day readmissions between the two groups [13]. Our study also shows no significant difference in complications between the two groups. Although the end point lacked the power to clearly define this relationship, there was a trend toward fewer minor complications in the RARC group.

Previous studies analyzing the perioperative outcomes of open to minimally invasive surgery have shown a decrease in EBL and transfusion requirement. In the current study, EBL was significantly less in the RARC group than the ORC group (p < 0.05). Nix et al. previously conducted a randomized, single institution clinical trial comparing 20 ORC with 21 RARC cases and also found significantly less EBL, quicker return of bowel function, and lower use of inpatient narcotics in the RARC group [14].

There was also no difference in length of hospital stay for ORC and RARC. This is largely influenced by return of bowel function, as those with quicker return of bowel function are typically discharged sooner. This was despite slightly fewer of the patients undergoing RARC requiring total parenteral nutrition, which is similar to colorectal and laparoscopic studies demonstrating quicker return of bowel function with minimally invasive approaches. Despite this, it did not appear to affect discharge disposition or length of hospital stay.

Conclusion

Robotic-assisted radical cystectomies can be accomplished in a community-based, non-tertiary health care setting without compromising perioperative or pathologic outcomes during the institution of this minimally invasive procedure. The length of the operation was longer for RARC; however, our RARC patients had a lower EBL, which translated into fewer blood transfusions. These results need to be validated in a larger, multicenter, prospective clinical study.