WIN 2011, 6-8 July, 2011, Paris

Early stage molecular profiling

Professor Razelle Kurzrock – MD Anderson Cancer Centre, Houston, USA

The approach that we’ve taken is that as patients are referred to us, one of the first things that we do, along with the history and physical examination and blood work and X-rays is to try and determine their molecular profile, that essentially their molecular profile should be part of their diagnosis. Once we determine what abnormalities are present in their tumour, we try to match them with a drug that targets that particular abnormality. If a patient is sent to us we might request analysis for several different mutations and we will find out that the patient, for instance, might have melanoma but that they also have a Raf kinase mutation. So we will match them up with a drug based on the presence of that mutation and if they have a Raf kinase mutation, as an example, we will put them on a Raf kinase inhibitor. If a patient was to come to us and we found that they have a PI3 kinase mutation, we would try to match them up with a drug that inhibits PI3 kinase.

We try to identify the biomarker in everybody, we are successful in identifying a biomarker in about 40% of patients. And probably even in those patients there might be more than one, we may not be identifying all the biomarkers but that leaves about 60% of patients where we cannot identify a biomarker with current technology. We put them on a clinical trial even if we haven’t identified a biomarker and we do that the same way that everybody else does. If they have, for instance, breast cancer we will try to put them on a trial that we think might be best for breast cancer, based on laboratory data or animal data. Or if they have colon cancer, the same thing – we will base our choice on what we know about the drug and what its effect might be on colon cancer.

Can patients from outside the US take part in clinical trials at the MD Anderson Cancer Centre?

Absolutely they could contact us to come into a clinical trial but they would have to come to the United States to be treated on a clinical trial.

Future targeted therapies are going to be combinations of drugs rather than just one?

That’s a very important question. The basis for believing that combinations will be needed is because cancer is complicated and there is a strong belief that there is more than one abnormality for many cancer cells. I think that that’s true but we have to also modify what we’re saying. A lot of the concern about how many drugs will be needed for an individual patient is based on studies with patients that have very advanced disease and in those patients the disease is probably very complicated and I wouldn’t be surprised if there were five or ten, or even more, abnormalities in those patients. To me, the real question is how complicated is newly diagnosed cancer? It’s probably still complicated but not nearly as much so as the type of patients that I treat. So, to me, the question is, if you have newly diagnosed disease, when you come in with a small tumour that may be 1cm or 2cm, is there one abnormality in some of those? I’m sure that there will be one, sometimes there might be two or three but I don’t think there will be ten, usually, as there are in my patients. So for combination strategies, we will need combinations but I really believe we need to get away from looking at very advanced patients as our models and go earlier in the disease and see what is essential earlier in the disease.

What impact will this have on trial design?

It’s going to take a lot of thought. The leukaemia groups have done it very well. So the example that I have used is chronic myelogenous leukaemia and, for that disease Gleevac or imatinib, which has really revolutionised the disease, it was a targeted therapy, a targeted BCR-ABL, which is the fundamental abnormality in chronic myelogenous leukaemia, and the first patients we treated had blast crisis which is end-stage disease and is very similar to the patients that I treat. Only 15% of those patients responded and people say, well why didn’t the others respond? The real question is why did even that 15% respond with such end-stage disease? But they responded and that gave a clue that this was a very interesting drug.

So two approaches could then have been taken: one was to figure out what are the ten abnormalities in blast crisis and how can we give a mixture of drugs to blast crisis to treat those ten abnormalities. The other way to proceed was let’s look at newly diagnosed disease – how complicated is that? It turned out that if you treat newly diagnosed disease, newly diagnosed chronic myelogenous leukaemia with Gleevac, almost 100% of patients respond and with some second and third generation drugs that are similar but more potent the average survival has gone from 3½ years to 25 years. So clearly that was the right way to go; that transformed the disease.

If we were still working on blast crisis we probably would have improved the survival maybe by one or two years but we wouldn’t have transformed the disease. So, again, my feeling is we need to think about the same thing in solid tumours – do we want to continue pursuing end-stage disease which is very complicated or do we want to go the other route and start to look at very early disease and how should we really be treating that? Can we refine our treatments? Chemotherapy, radiation, surgery – they’re important but a lot of them are debilitating, disfiguring and I think it’s time to start looking at this in a new way. If we’re going to transform solid tumours, I think it’s going to be early stage disease.