IMPAKT Breast Cancer Conference 2013

Gene expression test distinguishes between breast cancer patients at high or low risk of late recurrence

Prof Michael Gnant - Medical University of Vienna, Austria

It’s great to see you here and you’re addressing a very important topic because you’re looking at late recurrence because although a lot of recurrences, if they’re going to happen, happen within the first five years, they’re happening at ten years, beyond five years, ten years, even fifteen years. What can you do about that as a doctor and what are you addressing right here at the meeting in Brussels?

You’re absolutely right, late metastasis is, particularly for hormone receptor positive breast cancer, one of the most burning clinical research questions. Because, in fact, and this is especially true for the overall good prognosis hormone receptor positive breast cancers, we lose more patients after five years than we lose in the first five. There are several studies out there which have shown that extending the duration of treatment can help in reducing these late metastases.

And there’s a lot of talk about having longer courses of hormonal therapy, isn’t there?

Most recently the ATLAS trial has demonstrated that ten years of Tamoxifen may be beneficial for some patients, as compared to five years. There are a number of other trials of, how we call it, extended adjuvant therapy, mainly aromatase inhibitors after five years of endocrine treatment. The issue with this is that in order to obtain benefit we have to treat thousands, ten thousands of patients and most of them won’t derive any benefit because they are actually at very low risk. So this is what sparked our interest in researching modern diagnostic tools in order to be able to more accurately predict the risk for late metastasis.

Of course we’re here at a conference where molecular markers, genes, DNA, RNA, gene expression is the order of the day. What do you have to offer us in this respect?

What we did was that we have done a study on a large clinical trial, ABCSG-8, using the multi-genetic test called PAM50, basically investigating fifty relevant genes, and we validated that test in our large clinical data set.

You used PAM50 but of course there are quite a number of different possible gene tests that you could have chosen. Could you explain to me why you went for PAM50, although it is called the PAM50 ROR, risk of recurrence, isn’t it?

The read-out of this is basically the risk of recurrence and this again is grouped in low and intermediate and high risk groups of patients as determined by the overall risk for metastases at five years, ten years, fifteen years. There are other tests out there, we chose PAM50 because it’s RNA based and it can be done in a decentralised way so you don’t have to send all the material to one single place in the world but it can basically be done in any qualified pathology laboratory around the globe which I think in the long run will help to make this technology available to patients.

So what did you find?

What we found, and we particularly focussed on late metastases, metastatic events after year five, is that the PAM50 ROR defined risk groups pretty well differentiate between patients at extremely low risk for late metastases, those with a somewhat intermediate risk and a high risk of metastasis. So what is new about this is that all the risk assessment tools usually describe the first five years, maybe the first seven years, we basically excluded the first five years because what we do against that risk is we apply adjuvant chemotherapy and we focussed on the follow-up period after five years up to fifteen years. We have reliable data for, let’s say, eleven or twelve years because this addresses the most important question of late metastasis.

Now just how good is this test, though because you might well be tempted to prescribe hormonal therapy just in case?

Well the issue is that prescribing something just in case is what we currently do and it usually leads to what I call the over-treatment disaster. So if we can safely identify a sub-group of patients at particularly low risk we would be very relieved, patients and physicians, in not having to prescribe these patients. This is exactly what we demonstrate using the PAM50. To give you some numbers, for the low risk group of patients, low risk for late metastasis, the overall rate of patients experiencing metastases between years 6 and 10 is 1.3%, so that’s really tiny, less than 0.3% per year. Most physicians around the world would agree that with such a tiny remaining risk there is no justification to recommend to these patients the extension of endocrine therapy with all its related side effects.

OK, that’s still quite a small number of patients who are being spared the therapy although a significant number globally, of course.

No, that’s not a small number because it’s 40% of all breast cancer patients. It’s just that these patients have an excellent prognosis but within that excellent prognosis subgroup, using these tests, we have been able to identify a subgroup of excellent prognosis patients with particularly excellent prognosis and with very limited risk of late metastasis. So we’re not suggesting that these patients should not have any treatment, we’re just saying that we know that the standard duration of treatment is pretty much as good as it gets for these patients.

Now it’s one thing sparing treatment but what about the other patients? Is there any evidence that your risk scoring can help you to treat better the patients who really do need therapy?

Definitely. On the other hand, in high risk patients all these data come from a trial that did not use adjuvant chemotherapy so these patients had endocrine treatment only. But if I now see that after ten years or fifteen years in the high risk group of patients we can identify individuals with a 20-25% even 30% risk of metastasis I would, at least in hindsight, say we should have treated these patients with additional chemotherapy for example. This is more difficult to prove because clearly now you would have to do a prospective trial to demonstrate that this is something that can be reliably done but still I think individualising treatment is based on individualising risk assessment and if we can differentiate late risk between 0.3% and almost 10% in the high risk group, that’s going to make an impact on the clinical decision making process.

What about intervening at that earlier stage where you’re deciding between hormonal and chemotherapy?

Well, this is about the addition of chemotherapy. I had demonstrated those data at the San Antonio meeting last year so we are not focussing on it this time but it’s already established now in two large clinical trials that using the PAM50 score you can identify a so-called higher risk group of patients who should receive adjuvant chemotherapy in addition and on top of their endocrine intervention.

OK, and you’ve been doing this on large numbers of patients, you’re confident of your data, what do you advise practising doctors to do right now?

There is always a gap between research and clinical routine. The PAM50 test has now a CE mark in Europe; I understand that in the United States the FDA is scrutinising the data and defining whether this will be approved in a way for a diagnostic tool. But I think that for me as a practising physician if I have that individualised information about late risk available at the time when the patient and myself discuss the extension of the duration of adjuvant therapy, this is going to be very helpful. It’s going to indicate whom to suggest that a prolonged treatment would be most likely beneficial and, maybe even more important, who could be spared the side effects of unnecessary treatment.

And bearing in mind the fact that you have to get your head around the fact that many patients are already at low risk, could you interpret these findings in terms of how many cases of unnecessary treatment could you save globally and how much morbidity might you prevent by this?

It’s a little bit difficult to give you an accurate answer to this because this hugely depends on the healthcare environment we live in. Obviously in an African country this may not be the most important issue. One can calculate that, let’s say, in the US and in Europe this would be a little over 100,000 breast cancer patients who could safely be spared unnecessary treatment if we, with additional data, validate these results and get them to be available for the general public.

So what would you like doctors to take home from all of this?

I think the take home message is that risk assessment is a very important part of clinical decision-making. Just prescribing to be on the safe side does not cover all the dimensions of treatment because some of these treatments with the side effects exceed the potential benefit and this heavily correlates with the general background risk of relapse. We are approaching an era where we really have modern diagnostic tools with these multi-genetic assays to be better able to fine tune and individualise treatment for our patients and I think that’s very encouraging.

Michael, thank you for joining us on ecancer.tv.

My pleasure.