IGCS 2010, 23-26 October, Prague

Advances in targeted therapies for ovarian cancer

Professor Martin Gore – Royal Marsden Hospital and ICR, London, UK

Martin Gore, welcome here to ecancer.tv, great to have you. You’re taking a huge amount of interest in therapy for ovarian cancer, here at the meeting in Prague and there seems to be a lot of discussion, perhaps progress too. Could you update us on what you regard as the state of the art with therapy for ovarian cancer?

Ovarian cancer has really entered a new era and that’s the era of targeted therapies, particularly the two randomised trials of antiangiogenic therapy as maintenance using bevacizumab which were presented at ASCO and ESMO and have been re-presented here that show a progression free survival benefit for bevacizumab with chemotherapy followed by maintenance. Now the progression free survival benefit is a median of 3.1 months improvement and 1.7 months improvement retrospectively in the two trials, so there is lots of discussion about whether or not that is clinically worthwhile because there’s no overall survival benefit yet. But I think the importance of these two trials is that there is a clear clinical signal for maintenance antiangiogenic therapy after induction therapy in ovarian cancer. That’s really a very exciting step forward because it’s a proof of concept.

And the two trials are?

The two trials are chemotherapy alone versus either chemotherapy with bevacizumab or chemotherapy with bevacizumab followed by maintenance, that’s the GOG study, and the ICON7 study is just the two arm study of standard chemotherapy against chemotherapy with bevacizumab and bevacizumab maintenance afterwards.

What do you think about using progression free survival as an endpoint?

Let me start with the, perhaps, non-controversial statement that in terms of getting a signal of biological effect and importance of strategy, it is a very important endpoint because if one uses CT scanning, for instance, as the hard endpoint, the data, the CT scan, that is a very hard useful signal. If one gets perhaps more sophisticated and uses functional imaging, then that is also a pretty hard endpoint and can be a very useful proof of principle. The easy end point, as a proof of principle, is to alter the trajectory of the CA125 rise or to wait for the CA125 to rise from normality. The only problem with that as an endpoint is that some targeted agents may alter the biology of the CA125 molecule itself. But, generally speaking, as a proof of principle for a particular new strategy or new target, PFS is very good, I think it’s accurate, provided it’s done properly, and of course you get the answer quickly. There is absolutely no point at all in waiting for overall survival when one is really on a path, on a journey, of a new strategy.

That’s antiangiogenesis, what about other targeted therapies?

I think the same will apply to other targeted agents; the principle is the same, not just in ovarian cancer but in solid tumour oncology in general, that we need to get this sort of output quickly because you need to do the trials quickly. PFS as a proof of principle is completely reasonable.

Moving this along to refractory disease, though, and resistant ovarian cancer, what there?

I think the same is true there. Response rate in refractory disease is pretty important as well and it depends a little bit on what type of treatment you’re putting in but the old endpoints like that are important. I’m not really talking about changing standard of care, when it comes to changing standard of care, there is a slightly different argument. There the argument has to be around a certain magnitude of improvement in PFS is important, not just any magnitude that is statistically significant. The absolute is a value there because there is still, of course, the question of whether PFS is a useful surrogate endpoint for overall survival in ovarian cancer and I think the jury is still out on that. I have my own view but I think the answer to the question is more complicated than simply is it a good surrogate endpoint or not.

What are the options being tried and what’s the evidence so far beyond paclitaxel, carboplatin.

In terms of standard of care? What people are discussing are two major therapeutic interventions. One is that one I’ve already mentioned about bevacizumab maintenance but the other is IP therapy up front in patients who are optimally debulked. The jury is still out on that, mainly because of the design of the trials that we’ve had to date.

And there’s a study with trabectedin, doxorubicin, in the second line therapy too.

Yes, and that obviously has an importance and there, like with all second line trials, the important thing there is the clinical magnitude of the benefit in terms of what you’re seeing, together with toxicity, convenience of treatment and how do patients perceive that and is it of real benefit to them or not. There’s clearly a benefit to the combination, there is the question of whether or not patients find that acceptable in terms of toxicity and whether they find the magnitude of benefit worthwhile. There is an additional interesting side question to trabectedin and Caelyx given together and that is if you give it at second line and you delay the necessity of putting in platinum at second line in sensitive patients, do you gain more by pushing out the platinum free interval or do you simply gain more by giving people that extra option. So it may well have an important place.

So you could possibly use platinum third line?

Well that is the question. We attempted, in the UK, to do that study of randomising sensitive patients to receive Caelyx second line with delayed platinum against platinum and then Caelyx but unfortunately we couldn’t get funding to do the trial, that was the UK group tried to do that. That question is still an important one to answer and actually trabectedin with Caelyx second line followed by platinum based chemotherapy third line, randomised against the drugs given the other way round would be still a very important trial to do.

And where do we stand with the regulatory authorities with which drugs to use when?

A lot of drugs are licensed for second line treatment and I don’t think there is a right drug to give second line. Most people are giving Caelyx as a default position, if you like, in patients who have disease that is either refractory or has relapsed under six months. Then over six months it’s really carboplatin together with Caelyx or gemcitabine or giving carboplatin and paclitaxel again. So that is the broad brushstroke framework of it but, of course, single agent topotecan or gemcitabine in patients with disease that has relapsed under six months, or indeed under twelve months, is perfectly acceptable. Then perhaps the last thing to say about that is we have this very interesting emerging data of using paclitaxel weekly with or without carboplatin and response rates seem to be high, even in resistant disease and it is a very, very well tolerated treatment given that way. So we’ve got lots of options at the moment. They are much of a muchness, I don’t think there are huge differences between the approaches; I do think we need to do more trials to tie down exactly what the benefits and risks are to the different approaches though.

Given that, with optimal debulking, first line therapy with platinum combinations has already achieved quite a lot, how much is it reasonable to extend that still further in second and third line?

If you mean using carboplatin and taxol together at second line, is that what you’re saying?

How much additional extension of survival can we expect, and good quality of life, by adding second and third line therapies?

By adding further drugs to the backbone of carboplatin and paclitaxel? If we leave aside the intraperitoneal debate for a moment, the really exciting things that are going on in ovarian cancer research at the moment from the clinical aspect is the translation of what we are learning about the biology of the disease and that ovarian cancer is not just a series of diseases based on morphology but actually is a series of diseases based on what we know about the molecular events that are happening within ovarian cancer. Of course, the story of BRCA1 and BRCA2 mutations and the PARP inhibitors is a tremendously important paradigm in the journey we have trying to find better treatments for patients with ovarian cancer. Then when you add to that the bevacizumab story that has just come out and our need to find markers for who is going to benefit from antiangiogenic therapy, we’ve already got a tremendous amount to do just on those two advances. Coming up behind we’ve got the PI3 kinase pathways, we’ve got the Src inhibitors, there are all sorts of other targets that are coming up behind which we also need to investigate. So the PARP inhibitors and the bevacizumab maintenance story is really the beginning of a very exciting new therapeutic chapter for patients with ovarian cancer.

So lots of hope in the future but already the intraperitoneal therapy is being discussed here in Prague. What’s your assessment of the scope of that?

For me the problem about the intraperitoneal therapy debate is that we don’t really have data that asks one of two questions, the trials don’t ask one of two questions. The first question is: is exactly the same chemotherapy given IV as IP, different in any shape or form? In other words, you have to do a trial where the dose of the drugs is the same given IV as IP. That really was only done in the ‘90s by Alberts using cisplatinum and cyclophosphamide, which of course isn’t standard of care. So we actually have to do that trial, asking the question if you give this drug IV and you give the same drug at the same dose IP, is there a difference, is there an advantage to giving IP therapy? That’s one trial design. The other trial design is taking any IP regimen you care to name, and it doesn’t matter what it is, and you compare it against IV carboplatin plus paclitaxel given at standard dose. In other words, you ask the traditional randomised trial question – here is standard of care, in the experimental arm it is an experiment. And really for neither of those two designs do we have satisfactory randomised data on yet and I think we need to do that.

So the jury is still out?

For me, the jury is still out because the GOG IP study really doesn’t answer one of those two questions.

Could you wrap it up for the busy cancer doctor, what practical messages that could be applied now are coming out of this meeting here in Prague?

The messages that are coming out of Prague are that there is quite a biological step forward in ovarian cancer in terms of targeted agents and targeted agents particularly used for molecularly selected patients and also in the maintenance setting. This is a really important message in terms of ovarian cancer that has come out of Prague and the busy clinician needs to really keep his ear to the ground about the new targeted agents and new targeted era for ovarian cancer because it will change practice quite radically and, my prediction is, quite quickly within the next five years.

Martin, great to have you here on ecancer.tv, thank you for your wise words and we’ll keep an eye open for all of your publications in the future.

Thank you very much.