ESMO 2010

Professor Rob Coleman - University of Texas M.D. Anderson Cancer Center, Houston, USA

GOG-218: Bevacizumab plus standard chemo for ovarian cancer

Tell us a bit about the GOG study for people who don’t know so much about it, here in Europe.

Sure. GOG-218 was a very large randomised trial that was addressing a question of progression free survival with the addition of Bevacizumab to Carboplatinum and Paclitaxel. So it was a placebo-controlled trial, there were three arms in the trial so it was looking at standard treatment and then a maintenance question. So it was addressing Paclitaxel, Carboplatinum and Bevacizumab followed by a placebo as one of the arms of the trial. And then the third arm of the trial was the combination of the chemotherapy plus Bevacizumab followed by Bevacizumab.

So the three arms of the trial were essentially addressing some very important questions, as you can imagine, one was to assess whether or not there was an increase in the progression free survival, and then ultimately for overall survival. So we were fortunate enough to be able to present the data at the last ASCO meeting and at that time we had met the minimum number of progression free events to make an assessment about that. Of course the overall survival data still is maturing.

And have things changed since ASCO?

No, right now the reanalysis of the data essentially is confirming what we saw before. The curves are getting a little bit less hairy, so to speak, the number of sensor advance are reducing.

Any other interim results, let’s take the induction question? Induction chemotherapy plus or minus Bevacizumab, can you say anything about that yet?

Well it’s a really interesting question, I wish we had better interpretation of that data. It appears that the maintenance, continuation of the therapy in a maintenance setting, is really of primary importance. But when you analyse those curves it suggests that there may be some benefit which is extended beyond the completion of the Bevacizumab into the placebo stage and before those curves come back together. So it’s really testing a hypothesis that we’d love to be able to address further, and that’s the duration.

The data will be mature when? Another couple of years?

Yes, the survival data will be probably a couple of years. The difficulty is that the trial was initiated in a sub-optimal group of patients which we would have seen enough events on a shorter period of time. But as the trial was amended to allow for patients with optimal disease it now will be extended a little bit further because those patients will be living longer.

Now Tim Perren is presenting some of the ICON data here and there has been some discussion about the two trials. What’s your view on the ICON stuff?

The ICON trial, I think, is very important and there are some key differences, I’ll just mention them briefly and he’ll do a very good job at the presentation. So one of the key differences of the trial was the use of the placebo; in GOG-218 there’s a placebo which is protecting or preserving the fidelity of the primary endpoint which was progression free survival. The dosage of the Bevacizumab in the two trials is different, so in the ICON7 trial they’re addressing a question of essentially half the dose. And although the duration of time of exposure to Bevacizumab is a little bit shorter in ICON7, the total amount of dose of that treatment is going to be substantially less so I think that the data that we’ll see soon will help to put this into some perspective as to whether or not dose, duration and then what the effects are on progression free and overall survival.

Another key difference, and I don’t think this has actually been realised by many people is that the treatment population in ICON7 is substantially different than it was in 218. There were patients with early stages of disease and less advanced disease allowed in ICON7 and the way that the analysis was done in ICON7 was all the patients completed their maintenance phase before the analysis was done, and that’s fundamentally different than what was done in 218. In 218 once a sufficient number of progression free events were reached, then we reported the results, and then when it was listed as a positive trial, as investigators we were allowed to un-blind our patients. And so therefore there’s crossover which maybe makes the overall survival question a little more difficult.

That could blur it. And can you tell us how many people have chosen to switch arms or is that not clear yet?

Right. No, I think a lot of patients were given the option, I don’t remember exactly the frequency and actually Dr Burger may actually have gotten that information by now to be able to present later this week. But we do know that a substantial proportion, probably more than 15% of the patients, were ultimately treated with Bevacizumab as the next line of therapy, whether it was for progression or not. And so that’s enough to potentially blur that.

There are, however, some mathematical models which have been applied in the large breast cancer trials to try and take care of that complication. And it is a concern when you are doing in-term analysis. So the ICON people will not have a crossover option?

It’s going to be infrequent, just because of the availability of the compound.

And when will the oncologist in the street be changing practice?

It’s a good question. I tell people to ask this question all the time. I think that the science questions have been addressed appropriately in the studies so the question that you ask is actually one that brings in politics and funding and the cost-benefit ratio. I think that’s really key. We do know that the toxicity for these compounds is slightly more, it’s less than we were fearful of, looking at our recurrent patient population. So in the front line setting there are certain things that are different but a lot of things are very similar across those arms of trial. So then it becomes is the gain in progression free survival enough to justify the cost and knowing that there may be an ability to catch up by giving it at some point later, will that make some people wait to get their therapy?

And use it as a salvage?

Exactly.

Professor Coleman, thank you very much indeed.