Robotic Cystectomy Versus Open Cystectomy: Do We Really Need to Discuss?

We are going to focus on a radical cystectomy procedure that we’re familiar with, but understand several key points associated with it; the morbidity, the learning curve required, the cost associated with the procedure, and the importance of local control, and oncologic outcomes. And unfortunately, today the average resident actually completes four cystectomies and diversions during their entire residency. We’re taking a complicated procedure and in essence we’re putting our patients in this situation. And so in the attempt is to move to minimally invasive surgery. I mean you all are all familiar with the robotic approaches and the attributes that people have claimed to be helpful in terms of doing robotics as opposed to an open approach.

In the U.S., this is a slide that goes out to 2012, and right now in 2017 about 30 to 35% of radical cystectomies are done via robotic-assisted method, and it’s been a slow increase, very different from the rapid rise of radical prostatectomy, and there are different reasons for that.

Different I think four main comparison points looking at robotic versus open procedures and open radical cystectomy, and I’m going to go over each of these and have a variety of data, and that’s part of the issue when it comes to looking at robotic versus open. You can consider the prostatectomy data, which is basically nonexistent in terms of quality when looking at randomized trials or anything like that, and you can look at a number of single institution reviews, meta-analyses, you can look at retrospective reviews, and you can do comparative effectiveness research in terms of comparison studies. And there’s a wide variety, and a couple of studies that I’m going to focus on; one is this meta-analysis that looked at a variety of trials, both randomized control trials as well as prospective, as well as retrospective. But in an attempt that actually had direct comparison points, and this meta-analysis showed some data that favored the robots, some data that showed actually favored the open approach, and we’ll focus on different things specifically on those four areas of comparison.

First, if we look at morbidity and complications I think key points would be complications themselves, the mortality rate associated with it, and re-admission, and I’m going to start off with radical cystectomy itself, the standard open approach that we know is actually high risk. I don’t know if the complication rate is now at 60%, but it’s been published that high, with a mortality rate of about 3% historically. If you look at our review of more than 700 patients over a seven-year time period this is what we tell our patients. We have a readmission rate of about 20%, and overall within that 90-day time period of almost more than actually 25% of our patients come back to a hospital, either our location or some other facility, and we quote a mortality rate of 7% within the first 90 days. And that’s a significant mortality rate that needs to be explained and discussed with every patient.

If you look at the reasons for this, the number of complications I think are very common and familiar to those that do the procedure. The most common are bowel related as well as infection related, and our overall re-operation rate is 2%. Of those that are re-admitted almost 10% actually undergo a procedure.

Looking at the data that’s been published at other series this is not too far off from other large institutional series looking at open radical cystectomy rates, and the concern is not only with re-admission but overall mortality rate.

You look at that and then you look at early series looking at the robot. Now, importantly again none of these were direct. These are basically here’s our institution, this is what we’re able to achieve, a complication rate of 20/25, 30/35%. If you look at, and then the meta-analysis that I just told you about the overall complication rate, if you took all complications, the actual complication rate in this meta-analysis actually favored the robotic approach. But if you break down and look at every single grade, grade 1, grade 2, grade 3, complications, the only difference really, if you go back, the only difference that you can see was actually in grade 4. The most serious complication is grade 5. There was no difference between the open and robotic in this meta-analysis. And the lower grade complications were not any more frequent with a robot versus the open.

We looked at a specific issue with urethral stricture rates, and we had a feeling, or we thought at our own institution that we had a higher stricture rate associated with a robotic approach. And in talking to those people that do quite a few robotic approaches I think early on there clearly was a tendency to over dissect our ureters, and there was a higher stricture rate. But in essence in following them out for a period of time there did not seem to be a specific indication or incidence rate associated with a higher stricture rate.

What about the learning curve? We talked about the morbidity. The learning curve I think is essential. This is actually a group of so-called experts that got together, looking at open radical cystectomy, and what those individuals thought were actually important in terms of what’s the minimum required to do a cystectomy in a way that would be oncologically effective? That you’d have to do at least ten cases per year, overall positive margin status rate should be less than 10%, your standard pelvic lymph node dissection should be done at least in 80% of cases, and the lymph node yield 10 to 14, and this was published more than ten years ago.

When you have this learning curve I think transitioning you need to understand that you have a difficult procedure that then you take into account a learning curve. You’ve only increased I think the difficulty being able to do it successfully, and you take into account the number of cystectomies normally done by those finishing training it becomes a very daunting task.

These are actually slides that are courtesy of Raj Pruthi, who has looked at perhaps where is the inflection point in terms of if I’ve gotten enough experience to do this in a safe and a successful manner. So, if you look at blood loss it’s about probably 20 to 25 cases where beyond that there wasn’t a significant difference in terms of the blood loss. The first 20 to 25 actually contributed to the learning curve. If you look at operative time the case point is about probably between 35 and 40 where you’ve significantly improved your operating room time, but beyond that you don’t get much better. And if you look at lymph node count really early on within 15 to 20 cases your lymph node count does not increase in any number. So, those numbers seem reasonable, but understand this is a learning curve that continues to show significant difference early on in your experience.

So, when looking at then, associated with the learning curve, certain things I think would take into account okay how much have you been able to learn? How much have you been able to do? And certain things I think do show a benefit in the robotic approach, and the meta-analysis would agree with that. I think in every retrospective institutional series and every meta-analysis, and actually in the prospective trials I’ll talk about there is a decreased blood loss associated with a robotic approach, much like the data associated with robotic prostatectomy. In line with that, in verifying that is a decrease or a lower transfusion rate associated with a robotic approach, and perhaps a decrease in length of stay.

Third comparison point; cost. This is actually an evaluation looking at if we looked at certain fixed costs associated with a radical cystectomy, and when we looked at variable costs, and then you look at overall hospitalization rates associated with a robotic versus an open cystectomy. And in this series you see that the difference is almost 2,000 dollars in favor of an open approach. If you look at the OR fixed costs the robotic is actually much higher, and that’s associated with the equipment that’s required. The OR variable rates really were not that much different. And if you look at the hospital stay there was a slight benefit in terms of the robotic approach, but overall a higher cost for the robotic approach. And I think this is true in this series, and in the memorial series, and I’ll talk about this study specifically later on in the talk, there also was a similar difference in cost favoring the open approach. If you look at an orthotopic diversion that difference was almost 4,000 dollars, and if you look at the open approach the difference is very similar to that previous study, about a 2,000 dollar advantage for the open robotic approach, I’m sorry, the open approach versus the robotic approach.

Study data is important. It really is the foundation for much of what we do. But if you talk with the health outcomes folks as well they’ll tell you that it’s important to also take into account real-world experiences, real-world accounts. So, Jim Hu looked at actually SEER Medicare data and looked at actually the overall cost of a robotic versus an open approach, and interestingly within the hospitalization with actual patients that undergo a cystectomy, either robotic or versus open, not in a trial situation or in a comparison standpoint, there was actually not a difference in cost statistically. A slightly higher number of robotic versus open, but you can actually see the difference in the 30 day and 90 day mortality rate, or not mortality but cost, not really sure why. There’s some variables I’ll talk about, but more costly over time with the robotic approach as compared to the open approach, verifying what we found in the trial data.

All right, most importantly, and I think that what we’re still gaining information on is what about the oncologic issues associated with this? An ineffective surgery here in these patients is truly a life-altering and a life-threatening ongoing process. And we know that the impact of an effective surgery is essential to these patients’ overall survival and success rates. This is actually data based upon the SWOG trial looking at certain surgical indicators that predicted how patients actually did. If you looked at this improved post cystectomy survival was associated with negative margins, and increased number of lymph nodes removed. If you had a positive margin your chance of dying was three times greater than if you had a negative margin. Now, some of that may be associated obviously with disease biology, but clearly that’s an independent risk factor for dying from disease when a positive surgical margin is left behind. In addition, the greater the lymph nodes removed seems to also correlate with survival. Predictors of local recurrence are also associated with that positive margin status and the number of lymph nodes removed, and clearly the quality of the surgery impacts patients’ outcomes.

Looking at the positive surgical margin rate this is actually early data looking at organ confined versus non-organ confined disease, and again, you can see that right now there’s not a lot of data comparing one versus the other in this review in terms of the ultimate outcomes. But if you look at the meta-analysis, I’ll go back, if you look at the meta-analysis of again more than 19 studies looking at a positive surgical margin rate there actually was no statistical difference between the open versus robotic approach. Again, assuming that that will correlate to oncologic outcomes is a bit of a leap of faith.

What about lymphadenectomy, another quality indicator? The question is should it be here at this iliac artery bifurcation or should it be higher up in the common iliac artery? Should it be at the aortic bifurcation? Should it be higher? Should it be at the IMA? And I think Dr. Lerner, who was kind enough to share his talk with me, is going to talk a little bit about that as a quality indicator again associated with cystectomy. And the issue with robotic is are we able to actually get the number of lymph nodes? None of these trials show the actual geographic or rather anatomic distribution of their lymph node dissection and just went with a simple node count, which is perhaps a surrogate but probably is not a total complete picture of what’s going on with the dissection. But you know, in these “me too” series, look what I’ve done, you have an overall lymph node count of 18, 20, 25, 30, etcetera.

If you look at the meta-analysis actually in the comparison of these trials there was a slightly higher yield within the robot in terms of the number of lymph nodes removed. Now, in essence, and I’ll talk in terms of my conclusion what I personally think, but there clearly is not a decreased number of lymph nodes removed with robotic approach, and that’s been verified I think in other studies as well.

Ultimate outcomes; what about survival? This is a slide that’s been shown many, many times I the USC series in terms of how patients do with organ-confined disease the ultimate success rate is quite high. Those with extravesical disease obviously that decreases, but even in patients with actually lymph node positive disease a proportion of patients still survive. So, this again talks about and emphasizes the importance of an effective cystectomy.

Early on, people were wondering about this, and so this is short-term data, 12 months, 2 years, less than 2 years. you have recurrence-free survival rates that are actually quite good. Way too early to make a conclusion, and these were early on series regarding very well and carefully selected patients. Those patients that were higher risk were not put in these categories but these were early proof of principle that we think were doing a good job. And to these authors’ credits, each one of these that I know personally, they were very careful in terms of selecting their patients.

The early reviews actually showed that early on there seems to be a benefit. Memorial actually reviewed their open series versus the robotic series in a longer follow up, and showed that basically with the open you had follow up between 34 to 120 months, and this short-term follow up with the robotic-assisted series. Early reviews show that it seems to be very good in terms of overall cancer-specific success.

These are some of the more recent kind of my institution, my retrospective review. This is an interesting one because this is from Roswell Park, and the physicians there have basically said that they have not done an open cystectomy for more than five years, not a single patient at their institution has had an open bladder removal, and this is basically all comers they say, all have gotten a robotic cystectomy, and this is what their five-year results are. And if you look at overall the success rate is similar to actually all comers for an open cystectomy series. If you look by stage, as one would expect, for organ confined very similar, 80% five-year survival. It drops as the disease becomes either outside or no longer organ confined. So, similar, not comparative but similar success in terms of five-year recurrence-free survival rates.

And if you look at another series, this is the Cornell series, who also do the majority now of their cystectomies via robotic approach similar, I’m sorry, similar all comers, a success rate five years of 75, 80% in terms of cancer-specific survival at five years.

Again, looking at comparative effectiveness research those were basically single institution, look what we’ve done with experts that have done a big number of robotic cystectomies. This is actually SEER data looking at a match series of patients, robotic assisted versus open, and this was just published actually less than a year ago, and actually for a follow up of actually almost four years there was no difference in cancer-specific survival in those patients that had a robotic versus open approach in Medicare or in SEER data. So, these are all comers, these are claims data for anyone over the age of 65 that had this procedure. This is the success rate so far with cancer specific. Now, again, are these cohorts exactly the same? No. Are they understanding in terms of differences in terms of morbidity and mortality? No, but looking at all comers there was no difference in cancer-specific survival at a follow up at almost four years.

There was some concern, and one that I definitely had, about atypical recurrences, recurrences that were occurring in places that we don’t normally actually see, and this was actually first reported this is from Cornell, and actually this institution, this group actually was the first to bring up the fact that the pneumoperitoneum, that this robotic approach, its spilling of tumor, that location of recurrence may be affected by this technique. They’ve gone back and looked at their series and actually their conclusion now is that those patients with atypical recurrences were actually no different than those that had recurrences at “normal” distant sites and in a common location, so that there seemed to be no difference. Again, not a prospective randomized trial, but in terms of their retrospective review they couldn’t find a difference in terms of more likely having an atypical recurrence associated with the robotic approach.

I told you about the meta-analysis that had 19 different studies, prospective, some retrospective, some two randomized. This is actually the most recent meta-analysis that looked at only the randomized control trials, and these are the conclusions. The robot was better, less blood loss, about 250 cc less, a quicker time to diet, 1.1 days, shorter operative time, failure to the open approach, saved more than an hour doing an open approach versus a robotic approach. No differences in complication rate. No differences in positive surgical margin rate, no differences in lymph node count.

Again, looking at real-world data this is Jim Hu’s study, what were their conclusions? Very similar, shorter length of stay by a day, more likely actually to have more comorbidities, actually sicker patients now were in the robotic arm compared to the open arm, but interestingly, they needed more home health. Now that may have contributed to actually higher costs associated with the follow up, which was the same study that I showed you at 30 days and 90 days that the robot was more expensive. That may be associated with the fact that these are sicker patients to start off with, hard to tell. Can’t tease that out from this data. But similarly, no difference in major complications, respiratory complications, or ICU admissions associated with the robot versus open approach.

In summary, I think, overall I think they’re comparable in terms of lymph node yield and complications, short-term quality of life. I think the benefits less blood loss, maybe half a day, maybe in terms of recovery and length of stay, and I think Dr., I think Peter is going to talk a little bit about optimizing cystectomy and outcomes. Clearly, I think there is a longer time in the operating room. It is more expensive, there is a learning curve, and long-term outcomes still remain uncertain. And so two trials that I’ll mention and focus on; one is this Memorial trial which has gotten a lot of press, was actually first introduced via a letter to the New England Journal, and what they found were very similar findings, increased time, lower blood loss, no difference in complications, no difference in length of stay, margins, etcetera. They found it was more expensive. Caveats to that trial, I applaud them for doing an attempt for a randomized prospective trial. Only 25% of patients actually were randomized. Secondly, you had people who had done thousands of open cystectomies versus a robotic surgeon that was experienced in robotic pelvic surgery, in reality had done very few robotic cystectomies prior to the initiation of this trial. Blood transfusion rate was not included in complications. My understanding was that there was actually a difference in blood transfusions. Quality of life was actually measured at three months, not earlier on where it may have benefited the robot. And there was a significant difference in terms of OR time. Interestingly enough, three-quarters of the patients robotically went to the aortic bifurcation versus less than half in the open, and that may have contributed to the longer operating room time in the robotic approach.

I think this trial is very important, the RAZOR trial, and this is actually one that’s multi-institutional prospective, and the focus here specifically will be a randomized trial that’s actually finished accrual, looking specifically at oncologic efficacy at three-year mark. And hopefully we’ll have that data actually at this year’s AUA.

No clear superior I think. What do I do? I do both. I think ongoing studies are essential to determine the long-term oncologic efficacy, probably three-quarters of cystectomies I do now robotically. Importantly though, and I want to emphasize this, is this data actually really deals with extracorporeal urinary diversion. All right? This is a whole different category when we start talking about intracorporal division where I think the advantages or the disadvantages of robotic approaches may actually lengthen even more with longer OR times, increased cost.

Where do we go from here? You know what? I don’t think we should spend so much time talking about a robotic versus open cystectomy. I think it’s a technique. We need to know the long-term oncologic data, and some of the talks that we’ll get today I think are areas we should focus on more. The importance of clinical pathways and enhanced recovery, the importance of patient education and optimization prior to any of these significant treatments, risk stratification, as well as perhaps focusing cystectomies at centers of excellence. And so, I’m going to stop there and thank you again for your attention.