What I really talked about is a series of work being done under funding from Stand Up to Cancer and the AACR that has involved 7 institutions, 65 investigators across multiple institutions and our focus has been on cancer in women. Women’s cancers in particular have a target, an enzyme called PI3K which stands for phosphoinositide 3-kinase. That enzyme was discovered in my laboratory more than 25 years ago and it turns out to be one of the most mutated genes in cancer. Certainly in women’s cancers it’s very frequently mutated, breast cancer in particular, endometrial cancers and it’s also amplified in ovarian cancers, cervical cancers and other cancers. So we decided as a team to try to figure out how best to use PI3K inhibitors in those cancers. And in particular although we’re seeing responses with single agents, there are some patients who have remarkable responses to single agent drug inhibitors of PI3K, most women do not show such dramatic responses. So we’re trying to figure out how to use drug combinations that will work better than single agents and also try to identify who are the patients who will respond to single agents. If we can understand that then we could get those women the drug very quickly and then figure out drug combinations that will work for other patients.

How do you find these women?

We’re doing clinical trials that are based on preclinical models we have in which we either take pieces of the patient’s tumour out of the patient and transplant it into a mouse or we genetically engineer a mouse to have the same mutations as we see in human disease. Then we test the drug as a single agent or in combination with other drugs that are already in the clinic to see which drug or drug combinations works best in which mutational background.

Now, as we go forward with the clinical trials, with the drug combinations that make most sense based on what we find in the preclinical studies, then from those patients we do a full exome sequencing and something called RNA-seq on the tumours from every one of the patients. The goal there is to retrospectively determine why this patient responded and why this patient didn’t. So we go in and deeply interrogate all the molecular events going on in each of those tumours, one patient responds, the other doesn’t, and try to figure out why did this person respond.

Now, in other cases we have patients who have dramatic responses but then eventually become resistant to the drug. There we try to get a second biopsy at the time the tumour becomes resistant and then compare the initial tumour that was responsive to the secondary tumour that became resistant and ask what were the molecular events there? What mutations occurred during that interval that caused the resistance? If we can learn that then we can be smarter up front in anticipating the mechanism of escape and give a drug combination that will prevent that escape. So that’s the science behind what we’re trying to do in these trials.

Any examples of successes so far?

One of the combinations we’re very excited about is combining a PARP inhibitor, in fact a drug recently recommended for approval here in Europe, in England, called olaparib made by AstraZeneca, and combining that drug with a PI3K inhibitor from Novartis called BKM120. So we found in our preclinical models that for patients that have p53 mutations, which are a very common event in triple negative breast cancer, and also had either BRCA mutations or PTEN mutations, that for those particular subsets of patients that drug combination, certainly in our preclinical models, those combinations indicate a very strong response to that drug combination. Now olaparib alone works well in some of these same patients but we find that combining the two together has a much more dramatic effect than either drug individually. So we’ve now started a trial in the United States, multiple institutions, that has now enrolled 70 patients and we’ve seen a number of responses. We’re hoping now to go forward into a phase II trial where this will be open in more hospitals and potentially in Europe as well. So we’re very excited about the preliminary studies with that combination.

What response have you seen so far?

We have one patient with a complete response, a triple negative breast cancer patient with a metastasis to the lung who had failed all other therapies. We saw really a complete response, the lung metastasis completely disappeared which is extremely rare at that late stage in the disease. We’re doing this not only in triple negative breast cancer but also in ovarian cancer because we often see that triple negative breast cancer and ovarian cancers have very similar mutational events. In fact, in the case of germline BRCA mutations, those are the two diseases that most frequently occur in women. So our models that test the triple negative breast cancer in a BRCA background we think will be predictive of responses in ovarian. That seems to be playing out in the clinic in that we’re seeing responses in both types of cancers.

We’re entering an era that’s really unprecedented in cancer research where we have the tools to, for a relatively small amount of money compared to the total expense of treating cancer, we can do full exome sequencing, even full genome sequencing to determine really what’s going on at the molecular level in the cancer. I think in going forward in clinical trials this will become standard, that we will do this for every single patient in the future. Right now only a few trials do this but the ones that do have the chance of learning, even from their failure if the patient doesn’t respond you learn something as much as for the patient who does respond. That’s the way we should be doing clinical trials where everyone who goes onto the trial we learn something from their experience in the trial; even if we can’t benefit them we’re still going to learn something from their experience.