ESMO 2010 Dr Amit Oza - Princess Margaret Hospital, Toronto, Canada; Dr Robert Burger - Fox Chase Cancer Center, Philadelphia, USA; Professor Nicoletta Colombo - European Institute of Oncology, Milan, Italy The role of Bevacizumab in the treatment of ovarian cancer Interviewed by Gordon McVie GM: Let’s just take the issue of how we manage ovarian cancer now, before all the trials are solid, which we’re going to talk about in a minute or two. In Canada?
Dr Amit Oza - Princess Margaret Hospital, Toronto, Canada; Dr Robert Burger - Fox Chase Cancer Center, Philadelphia, USA; Professor Nicoletta Colombo - European Institute of Oncology, Milan, Italy
The role of Bevacizumab in the treatment of ovarian cancer
Interviewed by Gordon McVie
GM: Let’s just take the issue of how we manage ovarian cancer now, before all the trials are solid, which we’re going to talk about in a minute or two. In Canada?
AO: So in terms of ovarian cancer, the approach is very much a multi-disciplinary approach – surgery is very much of the standard treatment and for systemic therapy, chemotherapy, our general standard is Taxol and Carboplatin.
GM: The same in Fox Chase?
RB: Yes, the standard combination has been around since 1996.
GM: That’s a long time, if you don’t mind me saying so.
RB: Right, so then the carat standard hasn’t really changed except for maybe the mode of delivery. So the options would include intraperitoneal therapy, systemic therapy alone, neoadjuvant chemotherapy before surgery and, most recently, the dose then taxane concept merged with platinum. So I think that those are used on an individual basis.
GM: So these are variations on a theme really.
RB: They’re variations on the same theme.
GM: Nicoletta – Milan, Europe?
NC: Yeah, we’re still with the Carboplatin and Paclitaxel, usually the three-weekly schedule but of course, as he said, there is some possibility that the weekly will improve the data, but we do have only one trial which is a Japanese trial, so we are waiting for the results of other on-going trials to confirm this information.
GM: And you and I met each other many years ago on the intraperitoneal platinum circus, but we’re not talking about intraperitoneal drugs today. Now we’ve got two interesting trials, haven’t we? We’ve got an ICON trial and we’ve got the GOG trial and they’re not entirely the same, there are some interesting differences amongst them and I’d like to explore them but they really look like they’re important trials and we are maybe going to move forward after sixteen years. Bob, tell us about the other one.
RB: Well the genesis of this is, of course, the disappointing results we have with standard therapy and the outcomes being less desirable, and then fundamental properties of ovarian cancer progression in terms of angiogenesis being one of the fundamental processes involved in tumour progression. So these trials were developed with the concept that vascular endothelial growth factor was a fundamental driver of angiogenesis and we had an agent that would specifically neutralise this that was found in earlier studies to generate single agent activity in recurrent ovarian cancer with Bevacizumab, both in terms of response rates and progression free survival. So these trials were developed to look at the possibility of melding Bevacizumab into front line therapy. So the 218 trial…
GM: That’s the GOG trial.
RB: Which was the GOG study, really tested the hypothesis that extending Bevacizumab beyond the combination with cytotoxic therapy would be more beneficial, perhaps, than concomitant therapy alone.
GM: Now this is a three arm trial?
RB: Right, and so the prior studies in non-gynaecologic tumours that led to regulatory approval of Bevacizumab, they utilised a two arm design with concomitant followed by maintenance therapy compared to standard systemic therapy alone.
GM: With no Bevacizumab?
RB: Right. So we threw in a third arm to really address this issue of whether maintenance was important by having a control which was systemic chemotherapy plus Bevacizumab alone confined to that initial period. So it was three arm design, placebo controlled one to one to one randomisation of patients with stage 3 and 4 ovarian cancer, so all advanced. They had to be patients that were not surgically receptible, so a fairly high risk population.
GM: And not debulked?
RB: They had surgery up front but they were not patients that could be completely resected. So their prognosis was relatively poor from the outset.
GM: And anything to add to that?
AO: So ICON7 took a similar approach but there were a number of key differences to the GOG-218 and both of these trials were designed at more or less the same time. The key difference from the GOG-218 study was, first of all, the dose of Bevacizumab used in ICON7. The dose of Bevacizumab used in ICON7 is 7.5mg/kg every three weeks as opposed to 15mg and the reason for that is that was the previously approved dose related to colorectal cancer. The second key difference in the trial was that it allowed patients who were optimally debulked, so women who had been optimally debulked, as well as sub-optimally debulked women. So it allowed a slightly different patient population to participate in the study. And I guess the third difference, which is not such a major difference, is the duration of treatment that was given in terms of maintenance. So maintenance was given for an additional six months following completion of standard chemotherapy plus Bevacizumab. So the duration of the total therapy was slightly shorter than GOG-218.
I guess one other difference that’s important is the difference between placebo control or not. So ICON7 is an open label, two arm study, it did not have a placebo control arm. And this was discussed at the outset and there are a number of reasons for that, one of them being that the endpoints that were going to be looked at were overall survival and PFS. And the second is that there was a practical issue related to delivery of the placebo in the control arm. This is a pretty challenging thing to do and I think kudos to GOG-218 in terms of doing that.
GM: Nicoletta, comments on the designs of these two?
NC: I think that they are quite complementary because ICON7 addresses the more general population, so what we are facing now with our daily practice, which is tumour completely debulked or not completely debulked. So it is more generalised, the type of results. And this GOG-218 was much higher risk because no-one was completely resected. So it’s a kind of different population and then again the issue of dose is interesting because then you can see what you can achieve with one dose and the other dose and the duration, which was likely longer in the GOG-218. But overall, again these are quite complementary studies.
GM: So we got some preliminary data at ASCO and that generated a lot of enthusiasm and a lot of excitement, and we’re going to get some updated data here in ESMO and then the European Gynae Oncology Group meet in Prague in a couple of weeks’ time. So do you want to run us through the Gynaecological Oncology Group results up till now and where it’s going?
RB: Very well. So there were almost 1900 patients enrolled to the three arms, really no differences in terms of key prognostic factors across the arms. Of note, actually two-thirds of patients that were involved were either sub-optimally debulked, had greater than 1cm residual disease after surgery, or stage 4 disease. The median progression free survival, progression free survival being the primary endpoint of the trial, was 10.3 months in the control group and just over 11 months in the concomitant therapy only, without maintenance.
GM: So Bevacizumab plus chemotherapy?
RB: Right. So there was a slight difference in progression free survival.
GM: But not a lot.
RB: But not statistically significant. Where in the concomitant plus maintenance Bevacizumab arm, which was arm 3 of the trial, the median progression free survival was 14.1 months and the hazard ratio for the difference in progression free survival was 0.717, which met the bar for statistically significant improvement in progression free survival, favouring arm 3 versus arm 1. The adverse events profile was very, very similar to what had been seen in prior phase III trials of non-gynaecologic cancers. One of the things that we were worried about was gastro-intestinal perforation going into this trial and there were some safety signals that came out of the phase II studies and some of the historical studies of Bevacizumab used in the recurrent disease setting, which is fairly widespread in the US off-label. But in this trial the overall event rate for GI perforations was less than 3% across the arms and it was between 2.6 and 2.8% for the experimental arms and about 1.5% for the control arm.
GM: So that’s probably in the same ballpark as the colorectal?
RB: Right. And then the only statistically significant adverse event, in terms of differences across the arms, was for hypertension which we feel expected but not for some of the less common events like arterial thrombosis and other critical events.
GM: And the adverse event pattern between the Bevacizumab maintenance and the non-maintenance? So adverse events, in other words, out.
RB: Seemed to be very similar, other than hypertension. There are some effects that are very interesting to look at so when adverse events are reported on trials like this, they are reported within thirty days of last infusion. So certain events, such as pain at any body site, might be increased in a population that is receiving a drug for a longer period of time but that goes with prolonged treatment where you find a lower incidence in an arm where patients are coming off therapy earlier because it is no longer being reported as an adverse event. So we saw that for pain but we thought it was more confounding.
GM: This is pain in sites of metastasis?
RB: No, not really, it’s just all; it could be abdominal pain, joint pain. So you saw there was an increased incidence of this in the maintenance arm but we think it may be confounded by being on treatment longer and the events not being reported in the arms of the study where patients came off treatment earlier, where events were no longer being reported. This is why quality of life is so important to study because it’s independent of treatment status, disease status, just essentially follows a timeline.
GM: So half the dose of the Bevacizumab, the test agent in the ICON study. Tell us about that.
AO: So the overall findings of the results are that the progression free survival for the whole group of women that participated in the experimental arm and received Bevacizumab did benefit significantly more than the control group. If you take the entire group of patients together then the average benefit is in the region of about two months, in terms of progression free survival. Now it’s a little bit complicated in terms of describing the benefit because there are a number of differences and one of the things is that the benefit that was seen is proportional to the time over which the treatment was perhaps given, so the benefit varies at different time points through the patient’s treatment. And the peak benefit, I guess, was seen at about twelve months when there was a 15% difference between the patients that were treated with Bevacizumab compared to the patients that did not receive treatment with Bevacizumab.
GM: And remind me how long the maintenance Bevacizumab was given for?
AO: Six months, so it was that the total treatment finished at twelve months, so at the time that the total treatment finished the benefit was 15%.
GM: Was best.
AO: That was the best. So it raises the intriguing possibility whether the benefit starts tapering off when the Bevacizumab is actually discontinued. So that was one thing. The second difference is looking at the patients who were optimally debulked as opposed to sub-optimally debulked. So if you actually separate out those two populations…
GM: Just let me interrupt you. Did you stratify for that?
AO: Yes it was.
GM: Stratified to optimal debulking versus non-optimal?
AO: That’s right. And if you look at the patients that were sub-optimally debulked there is a difference in outcome that’s probably more reminiscent of what was seen in GOG-218, the actual absolute difference between the two groups is greater in sub-optimally debulked patients compared to optimally debulked patients. And that’s quite interesting because I think a priori we had thought that perhaps the patients who would benefit most from an adjuvant type of strategy would be the patients who have the least amount of disease and, if anything, the results seemed to be pointing in the opposite direction, that the patients who perhaps have a worst prognostic category and sub-optimally debulked patients may seem to be benefitting more.
GM: So is that maybe because angiogenesis is a bigger factor in those cases, and this is the target of the drug?
AO: Perhaps, and again one of the things that we’ve seen is that there is a difference in terms of the response rates in the patients who had residual disease and sub-optimally debulked when Bevacizumab is given together with the chemotherapy as well as subsequently. So there is a better response.
NC: I think maybe also because these are patients with a higher risk of relapse so you can more easily see the effect because they are going to relapse more. So if the treatment is effective you are going to see the benefit in this population. But this does not mean that you don’t have the benefit in the other population, it’s just the quality of the results are similar but the quantitative is a little bit less.
RB: The ability to detect is less.
NC: Yes, the ability because of course the number of events is lower so it may be just a matter of being a risk population also.
GM: So is there a view amongst the three of you that this is actually now on the point of changing practice in terms of the maintenance Bevacizumab? Let’s get back to the up front in a minute. Do you think we’re there? Obviously the data is going to mature, we haven’t got any overall survival results yet.
AO: That’s right, there’s no overall. There’s a little trend at the moment in ICON7 that favours the experimental arm but we’re still about eighteen months short of the maturity of the survival data.
GM: In ICON, and in GOG?
RB: We were required to analyse overall survival at the time of the primary progression free survival analysis, but it’s clearly immature data because 76% of the patients enrolled were still alive at the time of that analysis and there were numerically fewer total deaths in arm 3 versus the other two arms, but there was no statistically significant difference at the time of that analysis.
GM: 76% is a pretty good number, isn’t it, for these very bad prognostic patients?
RB: Well yes, it’s still early. I think that the data are not yet mature enough. We basically chose an analysis time point that was based on a predetermined number of progression events in our population whereas in ICON7 all patients had to complete therapy in order for that analysis to be done.
GM: This dose issue in the maintenance, was that reflected in the toxicity pattern?
AO: No, the toxicity, in fact, again the Bevacizumab arm was very well tolerated. The main toxicity difference that emerges significantly is the hypertension effect.
GM: Sorry, hypertension is significantly different from his study?
AO: No, not from his study compared to ours.
GM: In the two arms. So you also got hypertension? That’s what I’m getting at.
AO: We also got hypertension but what I don’t think is emerging at the moment is that there is a clear difference in toxicity between the two studies overall. We need to look at that a little bit more but there wasn’t a major toxicity signal in the arm that received Bevacizumab, other than the fact that some of these toxicities were greater compared to the control arm. But I don’t think that we’re at the stage where we can say that the 7.5mg is different in terms of its safety profile to the 15mg, I think that there are important differences that we still need to look at but the safety was very good in both studies, is how I read it at the moment.
GM: So Nicoletta, where do you think this is going now?
NC: Well I want just to comment on the fact that everybody is focussing on the median progression free survival which is, of course, not surprising, it’s not so striking as a result. But given the fact that the cures, as you pointed out, the benefit varies over time, if you look at the median you just pick up part of the curves where there is not much effect, but if you just look at the curve a little bit earlier then you will see. And also this may mean that you need to prolong the treatment and if you prolong the treatment probably also the median progression free survival would be improved. So we should be very careful just by looking at one point of the curves because this may not be the most meaningful way to interpret the results. So if you look at the curves at twelve months then you see a 15% benefit which is of great importance because it makes these patients going from a platinum resistance situation to a platinum sensitive disease situation which is, from their prognostic point of view, clearly much better.
So I think we should really probably not look too much at the median but just to the overall effect, to the hazard ratio actually, and ICON7 met the primary endpoint and the hazard ratio, which was pre-specified, was reached.
GM: This is good news, this is good news.
NC: Yes, I am positive. I’m an optimist. And also I wanted to say that again it has been more than fourteen years since we had two positive trials looking at the same kind of questions and now we have two trials looking at the same question with differences but yet reaching the endpoint. So that’s very important, I think.
GM: So we have a consensus about the role of maintenance Bevacizumab? Up front Bevacizumab plus platinum taxane, versus just platinum taxane, is there a consensus there? What’s your view on that because you’ve done that, those two arms?
RB: Well my view is that this type of concept of at least maintenance but really we’ve only tested the concept with concomitant antiangiogenic therapy followed by maintenance antiangiogenic therapy, seems to be a strategy that would be within one standard option of treating a patient with disease that would normally be treated with platinum taxane combination therapy. Now we should be able to build on that experience and optimise the conditions for treatment, maybe we can optimise duration, we can factor in the dose, whether there’s a dose that might allow treatment to be delivered in a more cost effective way or identify populations within these trials that may have benefitted versus those that may have derived no benefit.
GM: You’re already well on the way to finding that out. So the up front question – chemotherapy plus Bevacizumab or minus?
AO: I think at the moment, based on the two studies, as Bob said, Bevacizumab has to be given concurrently as well as at maintenance. I don’t think we should start Bevacizumab just at maintenance based on the evidence that we have.
GM: Nicoletta, are you happy with that?
NC: Yes, because we don’t have actually the data.
RB: We didn’t test that.
AO: So the key things are that Bevacizumab certainly seems to be the first new agent in a long time to actually improve outcome in ovarian cancer. Exactly how we should use it in terms of the dose and the duration I think we still need to work out, but clearly this is good news, that this is an improvement in the overall therapy. And the other key thing was that it seems to be fairly safe in terms of actually delivering the Bevacizumab.
GM: I didn’t ask you your GI perforation rate.
AO: GI perforation rate was less than 1%, I think it was about 1%.
GM: So you’re planning your next studies, I suppose, let me second guess you. Are you going to tell me that the next GOG or the next ICON are going to be up front Bevacizumab, a platinum and a taxane plus maintenance of six months of Bevacizumab and then the randomisation into more maintenance Bevacizumab versus nothing, or are you going to go back and look at the dose of Bevacizumab issue? What are you thinking forward?
RB: I would argue that the duration question is the most critical.
GM: And Nicoletta was hinting at this, about the twelve months stop thing.
RB: Right, and there is actually some historical data and some tumour biology data that would suggest that treatment, even beyond progression, into second line cytotoxic therapy is reasonable to consider. So one strategy might be a trial where patients are randomised to three arms, one being the winner of these trials, for example, and then the second arm treatment until progression, and then the third arm treatment beyond progression and then specifying with the second line regimen. I kind of gave away something that we were planning two weeks from now but if you did that in a way that was well-controlled, you would actually be able to answer the duration question.
GM: And the second intervention through progression could actually be platinum and taxane again. You would not need necessarily to go to a second line combination with a new drug because I think Nicoletta was hinting at the fact you might actually be turning some of these tumours from platinum resistant to sensitive.
RB: Certainly a platinum based regimen would be reasonable.
GM: And what might the ICON be discussing?
AO: I think the different trial designs Bob has already eloquently spoken to. I think that the key questions are what’s the optimal duration of treatment in maintenance and the practicality of giving that. I think that clearly from the ICON7 results the benefit seems to start tapering off as soon as you stop the Bevacizumab so giving a longer duration of Bevacizumab and perhaps even continuing it when there is a recurrence with whatever the second line treatment is, whether it’s platinum again or whether that’s another agent, would be really interesting to see. And that would be a good concept to test out. It challenges us in terms of making that feasible in terms of actually delivering a treatment intravenously for a long time and continuing a treatment beyond progression and adding in another drug. So that has infinite complexities in terms of trial design but I think that that’s a really key thing to test.
GM: And of course we’re not short of antiangiogenesis drugs. If you look at the renal carcinoma, the kidney cancer literature over the last three or four years, that has just been transformed by Sunitinib and Bevacizumab and Berazilomide. They’re really excited now in renal cell carcinoma where they had the same situation as ovarian cancer, absolutely stable, dismal, inevitable bad news and now they’re looking at doubling and maybe more than doubling, trebling survival. And it’s all been down to antiangiogenesis designed drugs. Nicoletta, what are we going to do at the European Institute of Oncology then?
NC: OK, what I wanted to say is that Bevacizumab is very well tolerated but, as he pointed out, the intravenous administration is not ideal. So that’s why there are already on-going trials, and some of them closed trials, looking at the other oral antiangiogenic agents. So we just closed, with a huge collaboration in the Inter Group, a trial with Pazopanib as a maintenance and so we are looking forward to seeing the results. And then we are doing another trial with another antiangiogenic agent which is BIBF 1120, which is an oral drug, so this is a design very similar to actually ICON7 because it’s given concomitantly and then as a maintenance for two years. And so, again, we are going to see the data there. So almost all on-going trials now in ovarian cancer are looking at antiangiogenic therapy as a maintenance and concomitant to chemotherapy. So there is great interest in this field.
GM: So I think you’re just telling me that the experts, you are the experts, are actually in agreement with one another that things have changed and that you need to have an antiangiogenesis active drug and at the moment the only one with hard data looks like Bevacizumab up front with your combination chemotherapy. You need to have some sort of maintenance with, at the moment Bevacizumab, and this has just completely turned ovarian cancer management on its head. Is that a fair summary or am I being over-optimistic?
RB: I’d like to think that we have done that. There are some challenges to this but I think that this is a new chapter.
NC: Yes, I am optimistic so I feel like we’ve achieved a big improvement now but of course there is also the cost issue. We didn’t go into that but of course there is a cost issue and the practicality of a long administration, intravenous administration. And we probably will have to select patients who are going to benefit the most from the treatment. So there are several things that we have to look at, but yes, I am positive.
AO: I agree, I think that antiangiogenic therapy in ovarian cancer is maturing and I think we’ve seen data to show that it improves outcome in women with ovarian cancer. The key challenge now is how do we implement it so that it’s safe and continued for the appropriate length of time and so that we can try and deliver this treatment to women who are most in need of it, and be able to try and see if we can build any selection criteria going forward so that we’re not giving it to women who perhaps may not benefit as much or for whom other therapies may be more appropriate.
GM: Nicoletta, Amit and Bob, thank you very, very much indeed. I compliment you on the rigour of your trials and well done on the outcome.