The authors do not address potential study design issues that may have significantly undermined the final conclusions…. [W]e would caution readers to acknowledge these caveats when interpreting the results of this trial.
—Piyush K. Agarwal, MD (left), Mani Menon, MD (center), and Adam R. Metwalli, MD (right)
We read the letter to the editor in the July 24, 2014, issue of The New England Journal of Medicine entitled, “A Randomized Trial of Robot-Assisted Laparoscopic Radical Cystectomy,” with great interest.1
In the letter, reviewed in this issue of The ASCO Post, Bochner and colleagues report on their randomized trial, NCT01076387, at Memorial Sloan Kettering Cancer Center comparing robot-assisted laparoscopic radical cystectomy to open radical cystectomy. The study enrolled patients with bladder cancer (clinical stages Ta–3, N0–3, M0) with a primary outcome of grade 2 to 5 complications within 90 days of surgery. The trial was halted early after accruing 118 patients due to a predetermined interim futility analysis that stipulated ending the trial if no differences in the primary outcome were seen.
On the basis of an intention-to-treat analysis, 37 patients (62%) who underwent robot-assisted laparoscopic radical cystectomy and 38 patients (66%) who underwent open radical cystectomy had a complication of grades 2 to 5 (P = .66). The authors concluded: “These results highlight the need for randomized trials to inform benefits and risks of new surgical technologies before widespread implementation.”
These results are provocative and force the medical community and the lay public to assess current practice patterns and call into question the purported benefits of minimally invasive surgery for bladder cancer. However, the authors do not address potential study design issues that may have significantly undermined the final conclusions.
First, the primary endpoint evaluating overall complications instead of perioperative outcomes, complications specific to surgical approach, or oncologic outcomes is problematic. Also, a significant expertise differential exists between the open and robotic surgeons in the study. In addition, limited data are presented, preventing an adequate analysis of complications, blood loss, and length-of-stay differences. Therefore, we would caution readers to acknowledge these caveats when interpreting the results of this trial.
Teasing Out Complications
The aim of this study was to detect a difference in complications between open radical cystectomy and robot-assisted laparoscopic radical cystectomy. However, radical cystectomy entails two distinct surgical portions: an extirpative portion, in which the bladder and lymph nodes are removed, and a reconstructive portion, in which a urinary diversion made out of bowel is created and connected to the ureter(s). Teasing out complications resulting from the extirpative portion or the reconstructive portion may be difficult, but we would argue that bowel-related complications are likely a result of the latter while surgical complications (eg, vascular injury) would be a reflection of the former.
Unfortunately, the published letter and its supplementary materials do not list the complications seen, but Memorial Sloan Kettering investigators and others have described gastrointestinal problems (eg, ileus, small bowel obstruction) as the most common early postoperative complications, occurring in up to 29% of patients.2 In the study of open radical cystectomies by Shabsigh and colleagues, American Society of Anesthesiologists score, gender, and type of urinary diversion were significant predictors of any grade complication. Others have also confirmed bowel-related complications as the leading causes of early (within 30 days) readmission to the hospital after radical cystectomy.3
In the Bochner et al trial, while the cystectomy portion was randomized, the urinary diversion was not. Because the majority of complications seen are related to the urinary diversion, we would argue that no difference should be expected to be seen in this trial. This significantly diminishes the benefit of randomization.
Question of Experience
Another point of concern is whether the surgeons in each arm had comparable experience. Indeed, previously published literature suggests that not only does surgical experience and volume impact morbidity, but specific subspecialization in bladder cancer also appears to influence outcomes.4-6 Thus, it is conceivable that the robotic surgeons in this protocol were at somewhat earlier points along their learning curve with this specific procedure compared to their open surgery colleagues.
In this study, the authors used the years after fellowship as a surrogate for surgeon experience. Also, the open surgeons had to have completed a minimum of 100 open cystectomies whereas the robotic surgeons had to have completed at least 100 minimally invasive pelvic procedures, not necessarily 100 robot-assisted cystectomies. A more direct comparison would have been more instructive: How many open cystectomies had the open surgeons performed and how many robot-assisted cystectomies had the robotic surgeons performed at the time the study was conceived?
Perioperative outcomes were not primary endpoints of this trial, but blood loss and length of hospital stay were compared between the groups. A lower blood loss rate was noted in the robotic arm of the trial, with a mean difference of 159 mL. However, no P value is reported.
Since no data on transfusion rates between the arms was specifically presented, it is hard to determine if any clinical significance is associated with this finding. These data would certainly be informative to put the relatively small difference in estimated blood loss into clinical context. Nevertheless, the difference seen in estimated blood loss has been seen in previous robotic vs open cystectomy series,7 and some data suggest that transfusions alone are associated with poorer survival.8
As for the length of hospital stay, there was no difference between the two groups (mean length of stay of 8 days in each group, P = .53). Although not specified, we would presume that Memorial Sloan Kettering has an enhanced recovery after surgery (ERAS) protocol. This clinical care pathway specifies preoperative, perioperative, and postoperative management, allowing for minimization of hospital stay and optimization of postoperative outcomes.
Implementation of such patient care protocols can result in an earlier time to hospital discharge, among other benefits.9,10 Assuming that these pathways are applied equally to both cohorts of patients, we would not expect a difference in length of hospital stay.
Multicenter Data Awaited
Finally, it is important to note that this is a single-center trial at one of the premier institutions for urologic surgery in the world. Outcomes at such a high-volume, specialized cancer center may not accurately identify real-world differences due to the exceptional, and difficult to reproduce, clinical outcomes they produce.
This trial confirms the results of a previous randomized single-institution study that also demonstrated no significant difference in complications or hospital stay between the open and robotic groups.7 Furthermore, it provides us with thought-provoking data to consider while we eagerly await the results of a multicenter SWOG trial (NCT01157676) that may better address differences between open radical cystectomy and robot-assisted laparoscopic radical cystectomy.
The primary endpoint of the SWOG trial is 2-year progression-free survival difference between randomized arms (open cystectomy and robot-assisted cystectomy). Presumably, differences in this endpoint will reflect true differences in the oncologic or extirpative portion of the procedure. Secondary endpoints will address differences in outcomes such as complications related to technique (eg, injury to surrounding structures) or surgical quality measures such as extent of lymphadenectomy and surgical margin status that may better reflect any true differences between open cystectomy and robot-assisted cystectomy, as well. ■
Disclosure: Drs. Agarwal, Menon, and Metwalli reported no potential conflicts of interest. Drs. Agarwal and Metwalli are employees of the Federal Government and these views represent their own and not those of the Federal Government.
1. Bochner BH, Sjoberg DD, Laudone VP: A randomized trial of robot-assisted laparoscopic radical cystectomy. N Engl J Med 371:389-390, 2014.
2. Shabsigh A, Korets R, Vora KC, et al: Defining early morbidity of radical cystectomy for patients with bladder cancer using a standardized reporting methodology. Eur Urol 55:164-174, 2009.
3. Stimson CJ, Chang SS, Barocas DA, et al: Early and late perioperative outcomes following radical cystectomy: 90-day readmissions, morbidity and mortality in a contemporary series. J Urol 184:1296-1300, 2010.
4. Bhindi B, Yu J, Kuk C, et al: The importance of surgeon characteristics on impacting oncologic outcomes for patients undergoing radical cystectomy. J Urol 192:714-720, 2014.
5. Finks JF, Osborne NH, Birkmeyer JD: Trends in hospital volume and operative mortality for high-risk surgery. N Engl J Med 364:2128-2137, 2011.
6. Leow JJ, Reese S, Trinh QD, et al: The impact of surgeon volume on the morbidity and costs of radical cystectomy in the United States: a contemporary population-based analysis. BJU Int 2014; doi: 10.1111/bju.12749.
7. Nix J, Smith A, Kurpad R, Nielsen ME, Wallen EM, Pruthi RS: Prospective randomized controlled trial of robotic versus open radical cystectomy for bladder cancer: perioperative and pathologic results. Eur Urol 57:196-201, 2010.
8. Linder BJ, Frank I, Cheville JC, et al: The impact of perioperative blood transfusion on cancer recurrence and survival following radical cystectomy. Eur Urol 63:839-845, 2013.
9. Arumainayagam N, McGrath J, Jefferson KP, Gillatt DA: Introduction of an enhanced recovery protocol for radical cystectomy. BJU Int 101:698-701, 2008.
10. Pruthi RS, Nielsen M, Smith A, Nix J, Schultz H, Wallen EM: Fast track program in patients undergoing radical cystectomy: results in 362 consecutive patients. J Am Coll Surg 210:93-99, 2010.
Dr. Agarwal is Head of the Bladder Cancer Section, Urologic Oncology Branch, Center for Cancer Research, National Cancer Institute, Bethesda, Maryland. Dr. Menon is The Raj and Padma Vattikuti Distinguished Chair, Director, Vattikutin Uroloigy Institute, Henry Ford Health System Detroit. Dr. Metwalli is Senior Clinical Staff and Faculty Member, Urologic Oncology Branch, Center for Cancer Research, National Cancer Institute.
Retrospective analyses indicate that robot-assisted laparoscopic surgery for radical cystectomy in patients with bladder cancer is associated with reduced risk of complications and shorter hospital stay compared with open surgery. In a small single-institution randomized trial reported in a letter...