European Urology

Bladder Cancer

This issue of European Urology presents interesting results from the International Robotic Cystectomy Consortium regarding the learning curve for robot-assisted radical cystectomy (RARC) [1]. Pooled data of 496 patients who underwent RARC at the hands of 21 surgeons from 14 institutions from 2003 to 2009 are analyzed. The authors are to be commended for addressing this important issue. Several points in their report deserve our attention.

All of the surgeons had prior experience with robot-assisted radical prostatectomy, and two-thirds of them had performed >50 of these procedures before starting with cystectomies. This experience, together with previous experience with open radical cystectomies and the level of their individual skills, are nonquantifiable factors that may have favorably influenced the results.

Patient selection could also have affected their results. The patients included from most of the centers in this study are not from a consecutive series of cystectomy patients. Indeed, 34% of the RARC patients had non–muscle-invasive urothelial cancer and only 36% had pT3/4 disease. Although it was reasonable, if not imperative, to operate initially on patients with less advanced disease, this fact must be taken into account.

Easily measurable data such as duration of surgery, estimated blood loss, and number of lymph nodes found by the pathologist are reported but are difficult to interpret because only the interquartile range (IQR; ie, the range including 50% of the patients closest to the median value) is given rather than the full range of data. No information is provided on the 25% of patients with the longest operative times, the highest blood loss, or the longest hospital stays, and no information is given for the 25% of patients with the lowest values (eg, the minimal number of retrieved lymph nodes). Although the IQR may be of interest to statisticians, the full range of data that includes the outliers is more meaningful for patients and clinicians. Despite the lack of such important data, it is noteworthy that the mean blood loss still was 408 ml (standard deviation ±392 ml) and that the likelihood of blood transfusions rose in direct proportion to the increase in the number of consecutive patients. This may be attributable to a less restrictive patient selection and underscores how carefully the reported data must be interpreted.

Much more important for bladder cancer patients than the duration of surgery is the assurance that they will have the best possible chance of cure and that the functional results, which may heavily affect their future quality of life, are optimized. Although the 7% positive margin rate appears at first glance to be at the upper limit of the acceptable range, it must be assessed in respect to the relatively favorable patient selection. Only 36% of the patients in their series had pT3/4 disease, but 16% of these had a positive margin—an unacceptably high rate. Regarding the learning curve, the authors state in the second-to-last sentence of the results section that “for patients having a pathologic stage of pT3 or higher the number of patients required for a surgeon to reach a 15% positive margin rate was estimated at 24 patients.” To expose this number of patients to an undue risk at the beginning of the surgeon's learning curve as well as the suggested goal of a 15% positive margin rate for pT3/4 bladder cancer is unacceptable. It is futile to make comparisons to old cystectomy series in which no patient selection was possible, poorer preoperative imaging technologies were available, and no neoadjuvant chemotherapy was routinely used for advanced cases. A positive margin is the worst single prognostic factor, representing a death sentence for most patients [2]. Consequently, it is imperative that the patient selection criteria and/or the surgical technique be refined to obtain at least the same oncologic safety as with open surgery.

Another point of concern is that only 25% of all patients undergoing RARC received an orthotopic bladder substitute. In a selected group of patients such as this, one would expect only about half of patients to receive an ileal conduit. Furthermore, no information is given about the postoperative potency or continence rate in the patients receiving orthotopic bladder substitutes. If nerve sparing cannot be attempted on at least one side, the nocturnal continence rate would be rather poor [3], and [4]. Furthermore, because the urinary diversion was usually performed extracorporeally, the devascularized ureters had to be left long, and an increased stricture rate is to be expected. Such outcome data, affecting patients for the rest of their lives, are more important than the length of the scar or the time to first bowel movement.

The pooled data of 495 patients who underwent RARC are extremely valuable but also put the urologic community in a very unpleasant situation. If we are no longer sure that the oncologic results and perhaps even the functional results with RARC are at least as good as those with open cystectomy, we must ask whether we are still allowed, especially from an ethical point of view, to further expose patients who undergo RARC to possibly increased risks both in terms of oncologic safety and quality of life. It is legitimate for patients to expect that results obtained with novel techniques be at least as good as those obtained with established techniques from the very beginning. In the present instance, this means that the results with RARC should match those from open radical cystectomy combined with extended pelvic lymph node dissection and, if possible, a well-functioning continent urinary diversion.

The members of the International Robotic Cystectomy Consortium must be commended for the frank reporting of their results, although some are worrisome. In doing so candidly, they remind us that new surgical techniques or technologies, such as those for cystoprostatectomy, should only be accepted (and paid for by the health care providers) as alternatives to standard treatment if efficacy has been proven by at least one prospective randomized trial that is sufficiently powered to show at least equivalence to the current standard of treatment.