I will be talking about precision systems medicine which essentially means that we try to tailor the treatment of cancer patients with all the possible information available – DNA, RNA, protein functional analysis, follow-up of patients – and try to tailor treatment and help drug development.
Are you focusing on genomics, transcriptomics and proteomics?
Yes and no. Absolutely, everybody is talking about systems biology in cancer. What we are doing maybe differently is real time application in patient cancer treatment and the other unique aspect that we are working on is patient derived cancer models that we try to test with a panel of drugs. So typically for each patient we scan the genome, we scan the transcriptome, we scan the proteome of the cancer cells, we take cells, put them in culture and expose them to 500 drugs and read the response and integrate everything and try to make sense of it.
Are you also working with organoids?
Many people are working on the organoids and we do that as well. But in some diseases, like we work a lot on leukaemias so you don’t really need any organoid formation, you just take the cells from the patient and you can directly put them in microtiter plates and do high throughput functional screens on them. So that actually is a low-hanging fruit, in that sense, to be working on the leukaemias as opposed to solid tumours where you have a much more complicated scenario – you have stroma and other things to take into account. That’s why we started this thing from leukaemias but are now working on solid tumours as well.
What is the storage and interrogation of the data?
This is by far the biggest challenge. Not only are we dealing with lots of data being produced on patient samples but we are talking about the ability and the need to integrate and understand that data at the level of n of 1, which means a single patient. So you can’t use the regular statistics that people have made use of to make use of high throughput data, you have to analyse it at the level of n of 1 and you have to do it in a clinically meaningful timeframe. In order for this information to be ever useful in the clinic for the treatment of that patient you need to return the data in a matter of days to weeks and not months to years, as our typical lifespans of biomedical science are.
This ties into personalised medicine and tumour heterogeneity then?
This is why it’s then incredibly informative to follow the patients and take multiple samples. What I didn’t say in the beginning to keep things simple was the other aspect of systems medicine which is to follow up the patients and see what happens. What we have seen in many occasions is that we can actually clearly kill cancer at the level of sub-clones. So there are parts of the cancer that the drugs can wipe off but, as we know, the cancers are heterogeneous and then there’s another clone that takes over. So the challenge in the future is really the combinatorics of it and we are not able to do it today but if we can follow the patients, try to make sense of it, get the rules of the play in terms of what drugs work with what sub-clones. So essentially we’ve coined the term sub-clone based precision medicine. So you have a stratified medicine where you stratify patients in the sub-population, you could talk about individualised medicine where you look at the drug that matches a given patient but at the end we actually need sub-clone specific precision medicine, drugs that work on the particular sub-clones that the patients have. So things are getting more difficult, more refined. Maybe in the future it’s single cell precision medicine that we need to target cancer.
If you move to cellular treatment do you run the risk of distancing yourself from the patients’ needs?
Obviously if you look at it from the research point of view we can go deeper and deeper into the cellular features but in this type of translational research where we try to help the cancer patients at the same time, try to work with pharmaceutical industry to help them to promote their drug development, we have ways and means that actually keep us more focussed. The focus has to be at the patient, has to be improving the patients’ lives, has to be in helping the system, be it drug development or healthcare, to work with the problem at the end. But it is an incredibly complicated process.
Can you report on any successes?
We certainly have had some sort of early successes of these things and these typically relate to drug repositioning. So drugs that have been indicated to a particular disease have been found through this process to be active in another disease. The most remarkable example is the story of axitinib which is a renal cancer drug that works on VGFR, it’s an angiogenesis inhibitor. In these functional screens that we did on leukaemia cells, particularly drug resistant leukaemia that had been treated with ABL inhibitors, once you have a drug resistance mutation of the ABL protein you actually assume sensitivity or the cells become sensitive to a totally different drug, axitinib. So in a sense we’ve been able to suggest, and this requires clinical trials to prove it, we’ve been able to suggest that axitinib could work in drug resistant leukaemia, essentially a fast-track to repositioning an existing drug to a new indication and therefore helping patients much more quickly than starting a new drug development process. And similar other stories are cooking in other leukaemias and other cases as well that we think that there’s a lot of potential in drug repositioning through this systems medicine approach that I mentioned.
Does it save a lot in terms of regulatory approval?
There’s a lot of that that at the end will be interesting to see how the system could adapt to this type of approach where we individually tailor treatment and how should drugs be approved in such a context when sometimes the people who benefit will be individual patients with fifty different types of
