I am presenting on the idea of combination therapy and what our experience has been using single agents that are personalised for patients with cancer and what we are finding regarding the need for combinations. Personalised cancer therapy is something that is new and is also rapidly evolving. In the old days we gave therapy to patients rather indiscriminately and what we found was that only a small fraction of patients would respond. The first generation of personalised therapy trials then matched patients with drugs based on the genetic findings that were driving the patients. But in this first generation we were giving single targeted agents to patients and what we found is that we could get great responses but most patients relapsed, the disease came back and they died of the disease. We now understand why – because cancer is complicated and there is more than one abnormality driving a cancer. So it now seems very rational that the next step should involve giving more than one customised therapy to patients, giving a combination that is personalised for that patient.
What have you been doing in your lab specifically?
What we have been doing is doing genomic profiling on, now, several thousand patients. What we see is that virtually all patients, especially patients with metastatic disease, have multiple genomic abnormalities. The average patient has five or six important abnormalities and some patients have fourteen or more abnormalities. So that means that targeting just one of them is not going to reverse the tumours on a permanent basis. In fact, we now feel that we’ve been very fortunate that patients respond when we take out one of those abnormalities, when we target one of those abnormalities but the responses, as I mentioned, are not permanent. So we have been building clinical trials that incorporate combinations of therapy and not just single agents.
What are the most common genetic abnormalities found?
The abnormalities encompass several hundred genes that are related to cancer. I think one of the interesting things that we find is that if you look at a patient with breast cancer they can have almost any one of those several hundred abnormalities so it’s not confined to two or three or ten. The whole spectrum can be abnormal in some subsets of patients. But some of the most common abnormalities are in tumour suppressor genes such as p53 and PTEN and in cell cycle genes such as the CDK4/6 genes and in pathways such as RAS, RAF, MEK and PI3 kinase pathways. There are other pathways that come up frequently as well but the ones I just mentioned are the most frequent. What we are doing with this information is synthesising it, looking at large numbers of patients and putting it together working in partnership with our bioinformatics specialists and our statisticians. What we want to see is whether we can see patterns of abnormalities that emerge or if indeed every patient will need their own customised cocktail of drugs. So we’re still not sure which way that’s going to turn out. We know, or we believe, that patients will need combinations but whether they’ll be a set number of combinations that work for most patients with a certain type of cancer or whether every patient will be like a snowflake and have to have their own custom brew, that we’re still working through.
If you don’t have a drug for an abnormality you want to target, what do you do?
We now know that acquiring the medications is one of the biggest barriers for our patients. Now fortunately there are really several hundred new drugs that have been developed to target these abnormalities and many of the drugs are now either approved, I’m in the United States, by the FDA or they’re in clinical trials. But, even having said that, getting the drugs for the patient may be very difficult. The clinical trial may be taking place in another city so the patient has to travel. If the drug is approved but it’s approved for a different type of cancer, let’s say the drug is approved for melanoma but now we’ve found a similar abnormality in lung cancer and we want to use the drug, sometimes the insurance in the United States will cover that but not always. So we have to, as a nation and really as a global community, develop mechanisms where we can get these drugs to patients because that is probably the most difficult part of the whole thing.
The future research plans involve what I believe are some of our most important protocols. First of all there is WINTHER which involves identifying either genomic or transcriptomic, which is at the RNA, abnormalities and matching patients to drugs. WINTHER was the first clinical trial from WIN and in that trial the matching is mainly for single agents but this is really a very advanced trial because we don’t just use the genetic abnormalities to understand a patient’s tumour, we also look at the RNA, we look deeper in the cell than really just about any other trial that I know. The next trial that we’re developing is a trial that we’ve called SPRING and SPRING will be aimed at lung cancer. It will be the first trial, to my knowledge, whose objective is to look at the genomics, look at the RNA, use the most advanced technology to do the profiling and to use combinations. Our plan is for three drug combinations that target the most important nodes that are driving the tumour. I think this will be the first trial of its sort.
What is your take home message from your talk?
Personalised medicine is making important head roads into the treatment of cancer patients and that we now need to move from the first generation of trials that matched single abnormalities in a patient’s tumour to single targeted agents to the next generation of trials which looks at all the abnormalities of the patient’s tumour and creates a customised cocktail of drugs for that patient that is personalised.
