ESMO 2012

Phase I study examining an API2 kinase inhibitor for various tumours

Dr Pamela Munster – University of California, San Francisco, USA




Can you tell us about the study you presented at ESMO?

We presented a phase I study on PI3 kinase inhibitors. The PI3 kinase is a pathway that’s often activated in cancer, it’s also often mutated in cancer. So we were hoping that using a PI3 kinase inhibitor we could test a drug in a pre-defined population that is enriched for PI3 kinase mutations. We used this drug sponsored by GSK and did a standard design phase I dose escalation trial using a 3+3 design. We started with a very low dose of 0.1 and escalated all the way up to 3mg which was found to be the highest and non-tolerable dose. We learned that giving it once a day is probably not optimal so we went back and gave the drug twice a day. We also then wanted that the dose that we found to be the maximally tolerated dose, we wanted to expand in a select cohort that was either selected for having paired tumour biopsies or having PET-CTs or having a closer look at what insulin and glucose levels do before patients take this drug and after, after a meal. Then, of course, we had a select group of patients where we were hoping to enrich for PI3 kinase mutations.

So, as the study went we enrolled 170 patients of which they were men and women who were equal in numbers. A large group of patients had tumours that were kidney cancers, endometrial cancer, breast cancer and bladder cancer, for which we were hoping to select for PI3 kinase mutations. About 30% of the patients had different varieties of tumours. We looked for breast cancer with the PI3 kinase mutation, however, in our dose escalation the patient didn’t have a prospective analysis of the mutation, only in the cohort expansion did we have a prospective analysis of PI3 kinase or KRAS status.

We found that there are responses in select groups but in summary we are not quite clear whether the patients who had a PI3 kinase mutation are the ones who benefit the most or whether this benefits in patients who don’t have mutations. In other words, we found efficacy in kidney cancer and bladder cancer and these patients did not have a PI3 kinase mutation. The endometrial cancers had wild type KRAS but the responses were somewhat low. And in breast cancer we have seen one response out of eight patients who had a PI3 kinase mutation. So we are, in the end, not quite clear whether the PI3 kinase mutation really needs to be there, whether this predicts response to this drug or, as our discussant aptly put it, the negative predictive value of the mutation may not be there.

Where do you go next from here?

So, where are we going to go next? I think the total response rate of 5-10% is in keeping with what other sponsors making these drugs have seen, is that the response rates to PI3 kinase inhibitor across the board is probably 5-10%. I think we need to really go back to the drawing board and find out what is our biomarker, the PI3 kinase - is that not the right biomarker because we have seen responses lasting over two years in patients who don’t have the mutation. So did we miss it? Did we not look for the right mutation?

Could there be something we haven’t found yet?

Either we need to have a better means to look at the mutation via better testing or we need to look for different markers, maybe another mutation needs to be mutated or up-regulated in conjunction with the PI3 kinase pathway activation in order to see the responses.

What about side effects?

The side effects were very much in keeping with drugs in its class. Typically we see rash, diarrhoea, hypoglycaemia; we found that the insulin levels go up as the drug concentrations go up. I think this is a very important pathway in cancers and we haven’t quite figured out how to use these drugs effectively.