We did a study looking at more than 100,000 patients, 148,000 patients to be precise, in collaboration with Foundation Medicine. What we did was to look for the presence of BRCA1 and BRCA2 alterations in these genes in these patients. As we know, taking a step back, what is the role of BRCA1 and BRCA2 gene alterations? We know, based on studies done in ovarian cancer and breast cancer, that patients harbouring these mutations, either in germline or in the tumour, they may respond to treatment with PARP inhibitors because they have the deficiency in the double-strand DNA break. So if you use PARP1 inhibitors in these patients which lead to single strand DNA breaks and they do not have the capability, or tumours do not have the capability, of repairing the double strand breaks this leads to something known as synthetic lethality and these cancer cells die. Based on this principle multiple PARP inhibitors are now approved for treatment of ovarian cancer and also in breast cancer patients.

What is the application of these drugs, or potential application of these drugs, in other cancers? That was the main question we asked – what is the prevalence of BRCA1 and 2 alterations in other cancers? Surprisingly, or maybe not so surprisingly, we found that multiple other cancer types, many of the common cancer types such as prostate cancer, stomach cancer, pancreatic cancer, bladder cancer or urothelial cancer, they also were harbouring these mutations. So should we extrapolate these findings to efficacy of PARP inhibitors in these patient populations? Do we assume that these PARP inhibitors will be efficacious just because they are harbouring BRCA1 and 2 alterations? The answer is maybe not.
So, should we assume that PARP inhibitors are going to be effective in other cancers also? The answer is maybe not. Why? Because we also know that PARP inhibitors are a lot more likely to work if there is a certain level of genome-wide loss of heterozygosity, or LOH. That is predicted by the homozygosity or biallelic nature of these BRCA1 and 2 alterations. For example, in our study patients who were harbouring BRCA1 or 2 alterations on only one allele, and the other allele was completely normal, their genome wide loss of heterozygosity score was as much as those without these alterations. On the other hand, patients who are harbouring BRCA1 and 2 alterations in both alleles, whether similar mutations or different mutations, which is homozygosity or biallelic loss, their genome-wide loss of heterozygosity was much higher and they were more likely to respond. Based on the ovarian cancer trials we know they are more likely to respond to PARP inhibitors.

So the question arises, what should we do with patients in other cancer types who may not have genome-wide loss of heterozygosity as prevalent as we see with ovarian and breast cancer? The answer is very simple – we need to look for zygosity status. So the simple presence, or mere presence, of BRCA1 and 2 alteration may not be sufficient in these patients. Patients with pancreatic, bladder cancers, prostate cancer, in those patients it’s very important to also look for the homozygosity or biallelic loss of BRCA1 and 2 alteration which will likely extrapolate to increase the genome-wide LOH and increase the likelihood of their response to PARP inhibitors.

For these others who do have the response markers, is there room for N of 1 trials for getting real personalised medicine, for saying, ‘We didn’t expect to see this mutation in this tumour but you’ve got it and we’ve got a treatment,’ seeing if anyone is willing and ready to participate and embrace this possible personalised medicine approach?

This is personalised medicine. We are first identifying the presence of these alterations, either in germline or in the tumour. Then we are determining the zygosity status of these alterations. Then we are offering them treatment with PARP inhibitors. And that’s a perfect example of personalised medicine. The good news for our patients is that these testings are available on a commercial basis – we can order these testings in the context of genetic counselling. So I do want to mention the presence of a genetic counsellor in the team, multidisciplinary team, which is taking care of these patients is extremely important.

Then for these other patients, the genetic counsellor must also play the role of trying to communicate that it’s not that they’ve got a worse mutation or a better mutation but that it’s just approaching what can be done next, what is the best case scenario, what is the best care for these patients.

Genetic counsellors play an important role in my view, a very important role, because many of these mutations may seem like mutations but they may not be pathogenic. So when you do this panel testing and you see results in these genes, that’s not only BRCA1 and BRCA2, there are a lot of non-canonical pathways or genes which may be affected and may or may not be associated with what we call homologous recombination defects. Who is going to tell those patients? Genetic counsellors play a big role here.

Going beyond the patients, if these patients are harbouring these mutations in their germline it is the genetic counsellor who is going to counsel the patient and the patient’s family about the implications of those mutations and may want to recommend further testing like screening of family members.

Are there any plans for using a similar technique to try and assess patients for any other mutations which are similarly linked to disease progression – BRAF, MEK mutations, that may not have the same awareness in the general populous as BRCA mutations but are still linked to disease onset in different organs?

This is the million dollar question. When we do comprehensive genomic profiling of tumours in our patients with different kinds of cancers we often see many different types of genomic aberrations. What is the implication of that in the treatment of these patients? We know in many cancers, such as in lung cancer, mutations in EGFR or ALK predict response to certain therapies which are approved based on randomised trials utilising these drugs in these patients with specific mutations. But in my view, based on the current data, most of the patients who are found to have these genomic aberrations in comprehensive genomic profiling, we do not have targeted therapies available. It doesn’t mean we are not going to have targeted therapies available for them but at this point in time for most of the patients we do not.
But this is just the start and as the field moves forward newer targets are being identified, I’m sure newer drugs will be approved in the near future and will improve survival outcomes in our patients.