ESMO 2010

Professor Joan Albanell Mestres - IMIM-Hospital del Mar, Barcelona, Spain

Oncotype DX testing to determine adjuvant therapy for breast cancer patients

So can you tell us a bit more specifically about what you are talking here at ESMO these days?

Yes, I have the pleasure and honour to present a poster for the poster discussion on oncotype. In particular we have done a study at seven hospitals in Spain, seven highly recognised in our country. What we have done is to test prospectively the clinical utility of Oncotype DX testing in our patients.

So what kind of patients? Breast cancer patients, but over-expressing a particular protein? Did you have a subtype of patients, can you tell us a bit more about how these studies were set up?

Yes of course. Oncotype was developed initially in the United States and validated in samples from clinical trials from patients with endocrine-sensitive breast cancer in earlier stage and treated with tamoxifen.

So in our study we chose to select patients with breast cancer with node-negative disease, ER-positive disease, and also HER2-negative, so it’s a very specific sub-population. And then how we did the study is that the oncologist filled in a form with a tumour recommendation, before doing their oncotype; then if the patient agreed to participate in the study, then a specimen of the sample was sent to Genomic Health for oncotype testing. The oncotype result came back and then was taken into consideration for the final clinical recommendation of adjuvant therapy, be it hormone therapy alone or hormone therapy plus chemotherapy. So our goal was to compare whether testing patients in our country for oncotype had an impact on the final decision on the treatment.

So that’s very interesting. From what I understand the molecular profiling of the patient actually had repercussions on the clinical decision-making that you were doing. Is that the first kind of this study that was done in Spain, to your knowledge?

Yes, this study is inspired by a very similar study performed in the United States and published very recently by Shelly Lo and colleagues in The Journal of Clinical Oncology and was the first prospective study to evaluate the clinical impact of testing oncotype in adjuvant therapy recommendations. So it is not only the first study in Spain, but the first study in Europe testing this test to see how it impacts on clinical practice.

We felt it was important to do this study because in Europe, in Spain also of course, doctors are different, patients are different, maybe the treatment decisions are different. So what has been published in the United States was very good, you may be aware that oncotype studies in clinical practice show that treatment recommendation in this patient population changed after oncotype testing in about one out of every three patients, or 30%. So our question was whether a similar percentage of patients was also influenced by oncotype testing in Spain and this was what we did.

And your result was a confirmation of the results of the United States trial?

Yes, we also had very similar results. We had in the study 107 patients and comparing the recommendation of the oncologist before the test and after the test, at the end what it meant was the recommendation was changing for 31.5% of patients, so again almost one out of every three patients has a treatment change because of the oncotype testing.

Yes, that’s a very high percentage. So can I ask you what’s your next move, do you want to export your kind of trial to other countries in Europe, to set up multi-centre trials with other hospitals, not only in Spain, what do you think should be done next?

Well I think that for our country this study will be important. We have to finally analyse everything including cost effective analysis, but if you allow me I can add something. We see in our study that in our population there is a 31.5% change. But what we wanted to do also is to relate the probability of change to specific biological factors such as progesterone-receptor status, proliferation index like Ki-67, or others such as tumour grade.

And we realised that there is within our sub-population of patients with a possibility of a change by doing oncotype, is even more than 30%; it can be maybe 50%, 70% in some specific conditions. So I think that our objective is to analyse this in depth, and really compliment oncotype testing with standard clinical pathological data to make the best decision.

The best results of our study confirm what has been said by ESMO Highlights, for instance, that this kind of genomic testing compliments very well with standard clinical pathological factors and might help to predict better the prognosis of the patients and in particular to predict the magnitude of the benefit of adjuvant chemotherapy. So I think that the study very nicely correlates standard clinical pathological factors, oncotype testing, with the final decision-making. So I think that for our country it is good, as I said, we have to analyse together with all the participant centres and additional colleagues, to see the implications of this. I think that the study will be also interesting for other European colleagues, but for me at present the focus is to go in-depth with the results.

Can you tell us the implication of your research for the patient?

Our study shows that in this patient population that we studied, as I said, there is a change in treatment in one out of every three patients. And that change is mainly to avoid chemotherapy in patients that, when you do the test, are low risk. So I think that avoiding chemotherapy when it’s of course of practical importance to avoid the toxicities of chemotherapy, avoid the cost, and of course if it’s even later when you combine your standard factors with oncotype leads to a final decision of the patient and the doctor that chemotherapy is not to be given because the possibility of benefit is so low then I think it’s a clear benefit to avoid unnecessary chemotherapy.

On the other hand, sometimes you have a patient that the profile may look good but you do the oncotype and there appears to be a high risk, a high recurrence score. And then you know that in this patient population, with high recurrence score, the actual benefit of chemotherapy is in the range of 30% absolute benefit. So you can identify some patients, and in our study it was about 17% of the patients, that we would consider for hormone therapy only; when we did the oncotype they were mostly too risky for chemotherapy also. So this is the numbers and also the feeling of the patients when you do a test and the confidence of the oncologist in the recommendation is improved then, of course, the patient feels better. We did not analyse this in our study, the patient satisfaction, just the feeling. But what we analysed is the degree of confidence with oncologists, in determining their recommendations, and after doing the test, the confidence of the oncologists in clinical recommendation was increased in more than 68% of their patients.

I can see that maybe something you can do in the future, you can think about what would also be an assessment of the patients’ satisfaction.

I think this will be our next step, and my prediction is that the feeling of doctors, oncologists, goes very much in parallel with the confidence of the patient, and we’ll do a study.

How do you see your study within the framework of personalised medicine?

I think that a genomic test like oncotype is of course a step ahead that was waited for for many years in the American oncological community. Tests like oncotype help to define better which patients may benefit more or less from chemotherapy. It was a need, that has started to be covered, and my impression is that it’s the first of a series of tests that will help to improve very much the quality of care of breast cancer and many other cancer types in the near future.

So you don’t see it limited only to breast cancer, but you see the relevance also for other kinds of tumours?

Of course. I think that other tumours like colon cancer, for instance, are already close to having a similar genomic test that can be used whether as a prognostic aid or whether to predict the efficacy of treatment. It’s an area that we are living in now and in years to come we will see a lot of things.

Thank you. Where do you see the difficulties ahead, since you told us about the positive parts or the pros? Do you see any difficulties that could lie in this path towards personalised medicine for other kinds of tumours?

I think that key for us is to have good clinical samples, good biopsies from patients with clinical follow-up and ideally in the context of randomised clinical trials where these samples are already available. So if these kinds of samples are available, then it would be possible to find markers that have clinical utility. So I think this will be a limiting factor.

And the other is the proper integration of these novel technologies in the whole system, how to integrate with standard factors, treatment decisions, and also the reimbursement of these tests.

What about the high cost you mentioned before? Do you think this could be a problem about the cost of this medicine, personalised medicine? Or the reimbursement?

Well I think, as I said, we are starting to live in a new era and we are going to have to work together with scientific points of view, with the medical, and also the administration to find the best way to support these kind of important additions to personalised medicine. I am sure that we will find a way to put everything together and take the ultimate decisions.

Thank you, so teamwork is the future.