IMPAKT Breast Cancer Conference 2013

How do we design trials that improve outcome?

Dr Lisa Carey - University of North Carolina, Chapel Hill, USA

Great to get the chance to talk with you because you’re giving one of the really important talks here, although a bit of a difficult one to give because you’re talking about how to design trials. What do you mean by this and how do we need to redesign trials in the genomic era?

Well, as we have more understanding of the molecular biology and the heterogeneity of breast cancer we can’t design trials that are broadly non-specific; we have to design trials that are predicated on that understanding, that takes advantage of the understanding of biology.

It’s a bit confusing at the moment, though, isn’t it?

It’s confusing and it’s complicated. So these are increasingly going to be complex biology driven trials and I think that is a challenge, although it’s a very good challenge. As they say, it’s a nice problem to have.

What’s the balance at the moment between looking at clinical factors and treatment patterns as compared with genomic factors in determining what’s happening to a patient?

One of the things that I think gets lost sometimes in all the excitement about the science of cancer is the fact that the clinical variables remain really important. So I think we have to remember that a stage 1 breast cancer is still a stage 1 breast cancer, no matter what biology is underpinning it and that’s going to be a low risk cancer. So we have to take into account both things.

Although there’s less interest in node status these days, isn’t there?

Well I think a little bit. We had a lot of writing on things like nodes when that’s all you had. Now I think it’s pretty clear that nodal status is just part of a clinical staging and so it’s now just a piece of the puzzle as opposed to being determinative.

What are the key genomic factors, then, that need to be borne in mind when you’re planning new studies?

It actually depends on the nature of the study. So the fact is that we don’t take into account genomic things in adjuvant studies, for example, right now. The denominator remains hormone receptors and HER2 because those are the things we know we can target. On the other hand, if you are moving into the newer molecularly targeted era then the genomic issues you have to take into account will depend on what you’re trying to target. For example, at the meeting this week we’re going to hear a lot about the PI3 kinase pathway and targeting that pathway so analyses that are related to the signalling in that pathway are going to be key to any trial you’re designing there.

You’ve been concerned with triple negative disease, it could be obvious what that is but how do you define it?

Triple negative is defined really by the clinical assays; it’s defined by what it’s not. It doesn’t have hormone receptor and it doesn’t have HER2 signalling defined on just standard assays. Which, of course, leads us to the question of what is it. I think one of the things we’ve learned increasingly from all of these molecular biology initiatives is that it’s a very heterogeneous disease and I think our difficulty with finding improvements in therapy is really understandable when you think about how heterogeneous it is.

Heterogeneous in many ways and perhaps there could be genomic targets.

Absolutely. I think there are quite a few studies now that are finding that within what we call triple negative disease there are different subsets that may be themselves targetable but, of course, they get mixed together whenever we’re doing a clinical trial.

And when you’re planning clinical trials what’s the sort of decision making process between whether to use therapy neoadjuvantly or to use as it as adjuvant therapy or in any other way?

The neoadjuvant approach, or using the medical therapy pre-operatively, is a terrific paradigm for clinical trials because the trials can be much smaller, can be tissue intensive so you can actually collect information about the biology. We talk about bench to bedside, this allows you to go from the bedside back to the bench and learn from what you’re doing. The challenges are that the drugs have to be ready to be given to patients who are likely to be cured so it’s not a good place to put really novel drugs that you don’t know the toxicity of. So you have to put it into context. And there are implications for surgical management, radiation, the neoadjuvant approach is a far more multidisciplinary approach and requires that kind of structure than a more traditional surgeon and medical and radiation approach.

How big is the choice of agents that you can use in the neoadjuvant setting then, right now?

There’s quite a lot. You really need to have safety data, you have to know the appropriate dosing and you have to know that the drugs you’re combining can be combined safely.

But you’re talking more about chemotherapy than some of the targeted therapies?

No, I think actually the targeted therapies also but most of them, most of the trials, are trying to add the targeted therapies to chemotherapy and you have to know that that’s safe.

Your mission here in Brussels, how to improve outcomes or how to design trials to improve outcomes, what’s your ideal trial then?

I think there are a lot of ideal trials, it depends on what question you’re trying to answer. I do think that the neoadjuvant trials already are giving us a lot of information and I think that will be, as we go forward, an approach that’s far more nimble than adjuvant trials for answering questions in early breast cancer. I think we get a lot of information from window trials which are trials where you give a brief exposure to a novel drug and then look at essentially how it impacts on the tumour which gives you a lot of information about resistance pathways and whether the drug is actually doing what you think it’s doing. So I think you’ll be seeing a lot more of those kinds of studies. And in the metastatic setting the challenge for us is you can design lots of trials with metastatic disease and you can design them in multiple lines of therapy going forward but how do you understand the tumour? One of the things that we ignore is the fact that many times when we’re doing biomarkers we’re looking at a tumour from five years ago and trying to see if that can explain behaviour today. So I think many of the newer biologic initiatives have to do with liquid biopsies and trying to get information from the blood about the biology of the tumour today.

Do you think that you can get early information in one particular patient in order to be able to individualise therapy later on, perhaps when the disease recurs?

I hope so, I hope so. I think that’s what people are trying to do and even more to the point, information serially during therapy to see how the tumour evolves. Otherwise we’re biopsying patients over and over again which, of course, is very difficult.

However gene expression can change, can’t it?

It can and it does. We already know that even our traditional markers like ER, PR and HER2, if you look at the metastatic disease, change about 15% of the time.

And at this meeting here in Brussels there’s quite a lot of talk about preventing death five, ten, fifteen years down the line. Do you think some of your early probes with neoadjuvant therapy and the different agents could help in that?

I do, I do. The truth is that changing survival in breast cancer is in the purview of definitive trials. I think the neoadjuvant trials can be designed that way but that’s not traditionally been the mechanism for them. I think you still need to have definitive trials that do survival as endpoints. We’re doing much better in early breast cancer than we were because of advances in chemotherapy and endocrine therapy and HER2 targeting and I think we’re going to continue to see advances there. We haven’t made as much of a dent in metastatic disease and so I think those trials actually are part of our challenge going forward.

How much does all of this affect the job of the busy cancer doctor?

The busy cancer doctor is affected primarily because you really need to engage in the trials in order to help us move forward and the trials are complicated so you really need to become a bit of a molecular biology voyeur in order to understand the trials well. I have to say I think meetings like this are very helpful in that regard.

There are some biomarkers that are floating to the top, aren’t there?

There are, there are but, as one of my colleagues likes to say, a bad assay is as bad as a bad drug. So I think we have to be as rigorous about our biomarkers as we are about our drugs.

PI3K CA marker is quite interesting, isn’t it?

It is, it is and I think the question to which we don’t have an answer yet is how important are mutations? Are those determinative in terms of response to treatments in that pathway or are there other markers that are just as good?

And p53 that you can’t do a lot about.

p53 is frequently mutated in certain triple negative breast cancers. We don’t know how to target that yet.

You’re actually planning some studies at the moment, you’re doing some dual therapy in the neoadjuvant setting, tell me about that study.

So at ASCO will be the first presentation of CALGB-40601 which is a reasonably large phase III neoadjuvant study looking at dual HER2 targeting versus single HER2 targeting in early HER2 positive breast cancer. There has been a group of these neoadjuvant studies that I think help inform each other. In this study one of the most interesting things you can do with neoadjuvant treatment is we required research biopsies of every patient so we actually are able to look at the predictors of responsiveness. In truth these are very expensive drugs so who needs dual targeting and who can get by with just one drug, for example, is one of the questions I think all of these trials will try to answer.

So another message for doctors is to be quite meticulous about your biopsies and sampling?

Indeed, indeed. That’s going to be increasingly important.

What about surgical aspects, I know you’re researching this as well?

Interestingly also in an oral session at ASCO will be a surgical viewpoint; in that CALGB study we designed a companion study that looked at surgical practice patterns. As we change the way we approach medical therapy for breast cancer the surgeons and the radiation doctors similarly need to pivot and design interesting trials about what’s the best way to administer those modalities. It’s a multidisciplinary treatment paradigm and so we have to be mindful of our colleagues in the local therapy realm.

So if you were to sum up the sort of practical message coming out of your recent research, it’s quite diverse but tantalisingly promising too, for the busy doctor, what would that message be?

I think the message is that the world of breast cancer therapy is increasingly complicated but in a way that is itself transformative. Our understanding of the disease as it is, because it isn’t really a disease it’s multiple diseases, puts us really at the cusp of real breakthroughs; within a heterogeneous disease I think we’re going to have multiple approaches and I think that’s going to be good for our patients.

Thank you very much for joining us.

Thank you very much.