AACR 2016

Mutational clonality and intratumour heterogeneity as biomarkers for cancer immunotherapy

Dr Nicholas McGranahan - Crick Institute, London, UK

In the presentation I was exploring how exactly we go about looking at heterogeneity in tumours and first of all how heterogeneous are lung tumours and whether that has clinical relevance, first of all in the immunotherapy naïve setting, so before patients have received any immunotherapy and patients that won’t necessarily receive immunotherapy whether the heterogeneity of the tumours has a clinical impact and, moreover, whether we can predict the number of neoantigens. So for each mutation whether it might give rise to a peptide which can be recognised by the patient’s own HLA type and then presented. Then we can assess whether that has a clinical relevance as well, combine those two aspects.

The idea is that rather than targeting specific… whereas a cancer has lots of mutations and some of those will be drivers and some of those will be passengers, with the neoantigens we can pretty much use the spectrum of mutations that will make that cancer be recognised as non-self by the immune system.

Have there been any clinical studies?

So what we were doing is we were taking the studies that have already been performed and then looking, if we look at the heterogeneity, whether that adds an extra layer. So so far it has been shown that the number of mutations can lead to improved response rate to immune checkpoint blockades. So, for instance, in lung cancer and melanoma there’s a clear relationship there but then still there’s  some of them can’t be explained so there’s a few outliers where they’ve got a large number of mutations and therefore a large number of neoantigens but they still don’t respond to the immune checkpoint blockade. So what we wanted to say is what if we add the layer of heterogeneity, so if we look at whether the neoantigens are not just present but whether they’re present in every tumour cell within the tumour, whether that also has an impact, and our results suggested that yes, it appears that not just presence but whether they’re in all cancer cells or only a subset with the homogenous tumours appearing to do better than the heterogeneous tumours.

What’s next?

First of all we need to validate it in a larger set. So far we’ve used all the data we could get hold of but it’s rather disparate studies and there are also patients that have been treated with various different things before receiving the immune checkpoint blockade. So I think we need to make sure that we’re absolutely correct. Then in terms of moving this forward to actually benefitting patients the idea is how can we use this, can we use this to help in combination to try to improve so we know which patients are more likely to respond and then also whether it could be used in some sort of vaccine type strategy so you can then isolate the clonal neoantigens and then you could perhaps inject them back in.

So it remains patient minded?

The whole idea of the neoantigens and everything is in some ways the ultimate personalised therapy because it’s not only just selecting the cohort of patients that are going to do well, you want to identify the exact neoantigens in that specific patient and that’s what we’ve been finding in these analyses is that in general every patient’s tumour is very unique so therefore the neoantigens they have are very unique as well.

What are the limitations?

Currently we’re at the stage where to fully understand a tumour, for instance one project we’re doing is the TRACERx study where we take multiple regions of tumours and then analyse these to work out how the tumours have evolved and how they’ve changed over time and which neoantigens are present in all tumour cells and which are present in a subset. But this isn’t necessarily feasible for every single patient at this current stage so we need to move the technology forward as well so we can see how can we do this in as minimally invasive way as possible and get as much information as possible. This is where potentially using things like circulating biomarkers or where you’re essentially just taking a blood biopsy or something like that, whether that could then replace these very extensive sampling techniques that we’re currently doing as well.

What is your take home message?

In reference to the future of genomics I think it’s definitely an incredibly exciting time because there’s all this data that’s out there. Especially from my perspective, I’m a bioinformatician so my job is to analyse the data, we’re learning new ways to look at the same data we’ve had before. So initially the initial study might just have been to sequence the genome to work out which are driver mutations and which are passenger mutations but now we’re going beyond that to look at the passenger mutations and identify the neoantigens there as well. So it really is a very exciting time for cancer researchers in terms of understanding the genome and hopefully also for patients as well where we’re seeing treatment strategies that will revolutionise the way we see cancer.