publication date: Nov. 2, 2018
Conversation with The Cancer Letter
Ramirez: We no longer offer minimally invasive radical hysterectomy at MD Anderson
Director, Minimally Invasive Surgical Research and Education,
Department of Gynecologic Oncology and Reproductive Medicine,
Division of Surgery, The University of Texas MD Anderson Cancer Center
When cervical cancer patients were referred to MD Anderson Cancer Center for a prospective, phase III trial testing for noninferiority of minimally invasive vs. open abdominal radical hysterectomy, many requested the minimally invasive approach, because their referring physicians said it was better, said Pedro Ramirez, a professor of gynecologic oncology at MD Anderson.
That turned out not to be the case. Ramirez is the lead author of the authoritative phase III study, which found that women who underwent minimally invasive surgery for early-stage cervical cancers were four times more likely to die from the disease within three years.
“Basically, it was a bias by patients, bias by physicians that was driving this movement towards a growing embracing of the minimally invasive approach, without actually having cancer-related outcome data,” Ramirez said.
The results of the study, “Minimally Invasive vs. Abdominal Radical Hysterectomy for Cervical Cancer,” were published Oct. 31 in The New England Journal of Medicine.
“When surgeons, and even patients, actually look at the results of these studies and say, ‘How do we move forward?’ often, the question that comes up is, ‘Well, are we going to have to go back to the days when you did open surgery and the patients were in the hospital, debilitated for three to four days? Are we really going to go back to that?’
“These are not the patients that were back five, 10 years ago, staying in the hospital four or five days. They’re going home the next day. They’re getting back to functional recovery much sooner.”
Ramirez spoke with Matthew Ong, a reporter with The Cancer Letter.
The trial began in 2008—the impression I got was that you and your fellow researchers went in expecting to find equivalent outcomes. What’s the genesis story of the study back then?
First of all, thank you, Matt, for your interest in our study. The impetus of the study was based on the fact that, around the time when we initiated the principle of establishing the study, there was increasing evidence that minimally invasive surgery was safe in patients undergoing the type of surgery that was required for uterine cancer.
There was evidence that it was safe oncologically as it pertains to cancer-related outcomes, and therefore, we wanted to ask the same question, particularly in patients with cervical cancer. We wanted to make sure that before considering minimally invasive surgery, a standard of care in patients with cervical cancer, that we should really put that question to the test.
So, as you rightly mentioned, we designed the study to determine whether minimally invasive surgery was equivalent to open surgery, and we began the study in 2008. The total aim for the study was 740 patients, but in June 2017, the data safety monitoring committee alerted us that there was a safety signal in one of the arms, and they couldn’t tell us what the safety signal was, nor in which arm, because they were officially not recommending to close the study, but rather to halt the accrual, so that we would gather additional follow-up information, and that they would do a reanalysis.
If in the reanalysis the issue of concern was balanced, then we would be approved to continue to the full 740 patients. Obviously, because the study was not officially closed at that time to new patient accrual beyond the point of reanalysis, they didn’t want to bias the investigators by telling us what was the issue, and in which arm.
So, we stopped any new patient accrual at that time. We gathered information, we encouraged all the centers to send in as much information as possible pertaining to follow-up of the patients, and in October 2017, the statistical team did a reanalysis of the data and they said, “Well, not only is the safety signal persistent, but actually accentuated and now, we’re definitely recommending for the study to stop accrual. We will unblind the results to the investigators, and what we found was that there a higher risk of recurrences—in fact, four times higher risk of recurrence and a high risk of death from cervical cancer—in the minimally invasive arm.”
Of course, this is a very unexpected finding. We then went on to put this information together for presentation, and in March of this year, we presented it at the Society of Gynecologic Oncology meeting. Obviously, since that time, this sent a shockwave through the field of gynecologic oncology, because of the unexpected findings of the study.
Why were you shocked or surprised that minimally invasive procedures had worse outcomes for cervical cancer?
The shock was because when we looked at the retrospective data—granted, retrospective data is not as high quality as prospective randomized, and not as good level of evidence—but that’s all we had, and the retrospective data has shown, at least in the studies that mention oncologic outcomes, there seem to be no difference.
But one thing that needs to be highlighted is that, when the minimally invasive surgical approach was introduced, the primary focus of most manuscripts was outcomes around the time of surgery. In other words, blood loss, length of stay, readmissions, and getting back to functional daily activity, but not really focusing on cancer-related outcomes.
In fact, there were very few that had this question in mind as a primary objective. So, there was limited data related to the cancer. No one really focused on is there a difference in cancer-related outcomes from the open approach to the minimally invasive approach, and everyone was just focusing on the immediate recovery of the patient.
Basically, no one was looking at whether these procedures would impact survival outcomes via dissemination of malignant tissue or increase the interval or rate of recurrence.
You’re absolutely right. The other thing also is, for papers that actually mention it—because one could also ask, “Well, for the ones that did mention it and show that it was equivalent, why is it different from this prospective study?”—and I think it’s important to look at the fact that when most hospitals took on minimally invasive surgery, they stopped doing, for the most part, open surgery.
You’re looking not at a concurrent comparison. You’re looking more at a sequential comparison, and what I mean by that is, if you look at a timeframe of the last 20 years, for most academic centers, the first 10 years were open surgery; the last 10 years were minimally invasive surgery.
The first 10 years of the open surgery group, generally, these were bigger tumors, not as great imaging to select the ideal patients for surgery. When patients needed adjuvant treatment to the surgery, they only received for the most part radiation therapy.
Today, we have much better imaging; tumors we’re operating on are much smaller. Today, if somebody needs additional treatment, they get chemotherapy and radiation, and no one really questions, “Why is it that this first unfavorable-looking group and this second favorable-looking group have the same cancer-related outcomes?”
What’s happening with this second group, meaning the minimally invasive group that should be doing much better, why are they having the same outcome as the group that shouldn’t be doing so well? So, I think that the value of the prospective study is that you’re looking at the population that you’re studying at the same time frame. The same treatment, if needed, the same imaging quality, the same patient selection, and the only thing you’re testing is the question of the surgery.
And I think that this is what came to light, that with the prospective design of the study in the same timeframe that we now see that there was significant disadvantage in minimally invasive surgery.
And I think, also just to finish on that comment, one of the highlights of this study—and this has been very vibrant in the themes of the messages that I’ve seen in the last 24 hours since the study was published—is that, we as a community of surgeons and as academic centers are realizing that surgical approach should be put to the test more frequently with higher-scrutiny, evaluation, because we do have that for many of the drug-related trials. Many of the drugs that we use today are put through very rigorous testing.
A lot of the surgical practices often are based on retrospective data. We got to actually put them in for prospective randomized trials. That, certainly, has been an overwhelming response from everyone saying, “Good thing you did this, because now we know that we should be testing other things that we’re doing.”
That’s really great to hear. Just to make sure I understand this thoroughly—I’ve written a lot about malignancies of the body of the uterus, but not so much cervical cancer—how do you decide when to use minimally invasive procedures for cervical cancer and at what stage of the disease? Your study focuses on early-stage, but in the standard of care, when do you use it and when do you not use it?
That’s a very good question and I’m glad you asked to clarify. Patients with early-stage disease, stage I are the ones that are generally recommended to undergo surgery. Anyone above stage I typically is treated with chemotherapy and radiation.
Now, I think there’s a second side to your question as well: how do surgeons choose? Well, for many years now, most surgeons would choose minimally invasive surgery based on retrospective data. Interestingly, we often would see patients who would tell us, “Look, I think it’s great that you’re doing this study, but I don’t want to be randomized to the open approach, because my doctor that referred me to you said that the minimally invasive surgery was better. I want the minimally invasive surgery.”
So, basically, it was a bias by patients, bias by physicians that was driving this movement towards a growing embracing of the minimally invasive approach, without actually having cancer-related outcome data.
So, few women were undergoing open surgery for cervical cancer?
There’s certainly less than there were by minimally invasive surgery. I think that when you look at most academic centers and training centers, most of them were doing minimally invasive surgery.
Actually, it’s been quite dramatic, that since we presented the abstract in March of this year, there has been a trend towards moving away from minimally invasive surgery to open surgery. There were many centers that were saying, “Well, you know, this is very compelling, but I want to wait to see that this is actually published in a peer reviewed journal, and then I’m going to make my decision.”
And certainly, the fact that this is published in what we consider the highest-ranking medical journal, I think this will carry tremendous weight for further transitioning and acceptance of these results to the open surgical approach.
One thing I want to add, Matt, is that when surgeons, and even patients, actually look at the results of these studies and say, “How do we move forward?” often, the question that comes up is, “Well, are we going to have to go back to the days when you did open surgery and the patients were in the hospital, debilitated for three to four days? Are we really going to go back to that?”
In fact, one thing that has to be highlighted is that, today, the approach to the care of patients around the time of surgery is very different than it was five years ago, 10 years ago. An example of that is we have now what is called Enhanced Recovery After Surgery program, the acronym is ERAS, and through the implementation of these programs, patients undergoing this same surgery, a radical hysterectomy, they’re generally going home after an open radical hysterectomy one day later.
Therefore, these are not the patients that were back five, 10 years ago, staying in the hospital four or five days. They’re going home the next day. They’re getting back to functional recovery much sooner.
I want to make sure we have the numbers: about 600,000 women undergo hysterectomies overall every year in the U.S., but how many are undergoing open and minimally invasive radical hysterectomies for cervical cancer?
I actually recently looked that up and on average, there is about anywhere from 1,500 to 1,700 women undergoing radical hysterectomies each year in the U.S. Drawing from the numbers of the national registry database is also published in the NEJM by Dr. [Jose Alejandrio] Rauh-Hain, about 56 to 60 percent undergo the minimally invasive surgery, and the other 40 percent go the open surgery route.
I would venture to say patients at most academic center undergo minimally invasive surgery. But I think that this has an even bigger impact worldwide, because, as you know, in the U.S., cervical cancer is not so prevalent, but around the world, it is a leading of cause of death of women.
I think it also needs to be highlighted that this was a study that encompassed 33 centers around the world, and I think that these results will also have an effect on the patterns of practice, not only in the U.S., but also around the world.
Together with your colleague Dr. Rauh-Hain’s epidemiological survey, could you describe in greater detail the implications of your findings for gynecologists and gynecologic oncologists everywhere? Many are saying that these are immediately practice-changing results.
I agree. When we at MD Anderson learned of these results, we didn’t hesitate. We felt that this was very evidence, and that it would have impact on patients’ lives. We no longer offer minimally invasive radical hysterectomy at MD Anderson.
I think that there are a number of institutions that are actually making that change right away, and I think also the fact that the publication is now in a very reputable medical journal and has gone through the highest level of scrutiny—that is supported by a second study by Dr. Rauh-Hain—and the fact that immediately, even after just an abstract, the National Comprehensive Cancer Network guidelines for 2019 have made a change in their statement. That’s very impactful and most likely practice-changing.
Correct me if I’m wrong, but I understand that there was no fragmentation of known disease in your patient cohort. What do you think contributed to a higher rate of recurrence and shorter overall survival post-operation?
Obviously a question we all have. Very important to highlight before going on to the specific details of my answer is that, when we look at both of these groups—meaning, in the open group and in the minimally invasive group—one of the things that we saw was that, in terms of baseline characteristics, meaning, stage of disease, histology, in terms of risk factors, lymph node status, in terms of margins of disease, tumor size, residual disease—all of those were balanced.
So, in one group, presumably here, the minimally invasive group was at a disadvantage, because it had higher risk factors. Both groups were balanced in both elements of risk, be it baseline and be it post-operative pathology, the only thing that was different is your surgical intervention.
When that is the case, naturally, one has to say, “Well, then, what is it about that surgical intervention that’s different from the other?”
It makes the results more solid, because there was nothing different in these two groups except your intervention. So, in that setting, then what could be the possibility, to your question?
I think that there are two things. And first of all, I say “I think,” because the study was not designed to answer that question, we don’t have definitive answers, because we were evaluating equivalency, not assuming that minimally invasive surgery was going to be inferior.
One of them is that in minimally invasive surgery, we inflate the abdomen with carbon dioxide gas, and there’s been data from animal studies that suggest that the combination of carbon dioxide gas in the setting of cervical cancer, there may be a propensity for these cells to implant along the surface of the lining of the abdomen and pelvis—and therefore, by implanting, increasing tumor growth.
The second element is that in order to perform minimally invasive surgery, we need to have a tool that moves the uterus around while we’re operating, and therefore we use something called a manipulator. What that is, is basically an instrument that is placed through the cervix into the uterus to mobilize that tissue during the surgery.
So, it’s been proposed that in the setting of tumor exposed to the cavity with carbon dioxide, with this manipulator, that perhaps one is causing an increase in the spread of cancer cells.
Anatomically, how is treating or operating on cervical cancers different from malignancies of the body of the uterus? Since cervical cancer is an epithelial disease, how does en bloc removal—especially via minimally invasive methods—increase the risk of dissemination?
Great question, and I’m actually glad you asked that, because that’s important to have and understand, “Well, how come in cervix you get this, but in uterine, you don’t?”
I think that there’s a great distinction: in uterine cancer, typically, the tumor is small, frequently microscopic within the confines of the uterus. It is never exposed to the abdominal and pelvic cavity. It’s removed intact, and the tumor surface never actually sees any of the abdominal and pelvic cavity.
Cervical cancers are very different. You actually have gross tumor on the cervix, when you’re doing the surgery by a minimally invasive approach, that tumor is exposed to the abdominal and pelvic cavity for quite some time during the surgery.
Also, you add to that the component of a manipulator that may be potentially shedding some of those cells throughout the abdomen and pelvis, and then therefore, that may be the instigating factor.
Of course, we don’t know for sure, there haven’t been studies designed now to answer this question, but I’m assuming this will generate a lot of interest in proving or disproving that theory.
Since this is standard of care, has there been any effort that you know of to prevent the tumor on the cervix from touching any other tissue surface, or is that just kind of unavoidable?
Very good question. Generally, through the minimally invasive approach, no.
There are no active efforts to prevent that tumor from having exposure to the pelvic cavity. However, there is a subgroup of surgeons—particularly, this has been advocated for more in Europe—where they have proposed that this might be a problem with minimally invasive surgery.
And actually, they do, as part of their standard practice, not use a manipulator and they actually go vaginally for the end of the procedure to try to prevent this exposure.
And of course, up until the results of this study, everyone in our community of gynecologic oncology would hear some of these proposals and say, “Well, I’m not really sure there is any evidence to suggest that that may be the case,” but obviously, surgeons who have had that approach as part of their standard practice are now saying, “We’re going to look at our data and see whether we see the results that you’re seeing.”
At this point, it almost sounds like a no-brainer that exposed surface tumors are at higher risk of being disseminated in a minimally invasive setting, but do you have any case studies or examples of patients in that arm of your trial, where the cancer had metastasized at the points of contact?
We did not have patients who had implantation of the port sites. You’re getting exactly to the point of, if you do something different to protect the tumor from exposure, are you getting the same issue of recurrence?
There’s a very interesting study out of Korea. It was a retrospective study, but what they did was they had a group of patients where the entire procedure was done laparoscopically, even until separating the entire specimen from its vaginal attachment and proceeding with the removal of the specimen vaginally, versus the entire procedure laparoscopically except the point where you actually separate the cervix and uterus from the vagina.
In other words, in that second group, they took the precaution of just removing the cervix and uterus while minimizing exposure of that tumor to the cavity. Interestingly, in the first group the recurrence rate was 16 percent. And in the second group, the recurrence rate was 5 percent.
Basically, if you take these extra precautions to remove the specimen vaginally, perhaps you will reduce that likelihood of recurrence.
I can’t help but feel a sense of déjà vu here. Not too long ago GYNs everywhere were, or maybe still are, debating the role of power morcellation—or of any kind of morcellation—in uterine tissue. I know we’re not talking about fragmentation in the case of cervical cancers, but are we looking at somewhat similar situations here in terms of dissemination and exposure in what are basically adjacent diseases?
Certainly, one would think that it drives along that same theme, exactly as you say, with regards to tumor spread, tumor implantation.
I would think that there’s an element that directly leads us to the instrument itself like it was with the manipulator, but I think that it does ring a bell along with that same principle of cell implantation and perhaps the gas being a potential etiologic factor in this setting.
There’s an interesting question of who is getting harmed here: I understand that it’s predominantly affluent white women who are being harmed by minimally invasive radical hysterectomies for cervical cancer?
It’s interesting. That’s speaking to Dr. Rauh-Hain’s study. I think that the correlation of the more affluent population getting minimally invasive surgery is the fact that, by nature, it’s been shown that that more affluent population may have a higher likelihood of being at risk through that approach—for whatever reason, be it because of insurance status or location.
I don’t think that patient profile is at higher risk of recurrence, it’s more so if that’s the patient profile that’s getting minimally invasive surgery, then the minimally invasive surgery brings in that risk.
I don’t think your trial was designed to study the risk of spreading unsuspected cervical cancer, but what are the chances that routine minimally invasive hysterectomies for benign indications would also worsen undetected disease?
The general population should understand that there is nothing from this study that indicates that minimally invasive surgery should not be performed in the benign setting, because these were not benign diagnoses.
It’s very important to highlight that patients should not be concerned if they have a diagnosis of endometrial cancer and their surgeon is offering them the minimally invasive surgery—again, in that setting, we feel that there is good evidence that the minimally invasive approach is safe as it relates to oncologic outcomes.
There should not be this overwhelming fear that minimally invasive surgery is not to be used for any cancer. We’re strictly, specifically talking about cervical cancer. I think for women, or men, diagnosed with other disease sites other than gynecological, that they should have a discussion with their doctor regarding the tested efficacy of the surgical approach as it relates to cancer outcomes.
In cervical cancer, can you reliably rule out any sign of malignant disease with existing screening methods?
For the most part. No screening test is absolutely 100 percent effective, but yes, for cervical cancer, patients certainly should be informed that by doing routine screening as indicated by their physicians, it reduces the risk of developing cervical cancer.
What subset of the patient population could minimally invasive procedures continue to be used for the indications in your study, now that we know what we know?
That’s a very important question, because I know that there have been, at least, suggestions in editorials or discussions in conferences, “Is there a subset of the population that can still get minimally invasive surgery?”
Our study was designed to evaluate patients with early-stage disease regardless of tumor size, meaning any patient with tumor less than 4 cm, that’s considered early stage. Therefore, we have stopped performing the radical hysterectomy in patients with early-stage disease.
There have been suggestions, “Well, perhaps there’s a low-risk group that can still undergo minimally invasive surgery.” And our study was not designed to specifically answer the question of one approach vs. the other in that low-risk group, because for that, it would’ve been a much larger study, and therefore we don’t have an answer to that specific question as to whether the same applies only in the low-risk patients.
But because our study included low-risk and high-risk patients, we opted to stop the minimally invasive surgery in any patient that has an indication for a radical hysterectomy.
Are you working on any other studies related to this?
We actually are now in the process of doing an evaluation of low-risk patients, at multiple institutions, again, and this is going to be a data registry study in looking specifically at this low-risk population.
Hopefully, we’ll have this information within the next year or two.
Did I miss anything?
One thing I will stress again, a message to convey to physicians that if your patients are diagnosed with early-stage cervical cancer, and there is a discussion about undergoing a radical hysterectomy, certainly encourage patients to have a very thorough discussion with their surgeons regarding the cancer-related outcomes, and specifically addressing the results of this trial—knowing that by minimally invasive approach, there is a fourfold likelihood of having recurrence of the cervical cancer.