Therapeutic targeting of childhood medulloblastoma

Share :
Published: 23 Nov 2016
Views: 2408
Rating:
Save
Dr Sheila Singh - McMaster University, Hamilton, Canada

Dr Singh speaks with ecancertv at NCRI 2016 about the pathways through which cancer stem cells drive tumourigenesis and subvert treatments, in the specific setting of medulloblastoma.

She describes research with patient-derived xenograft models to modify treatment regimens, giving insight into opportunities to avoid the systemic toxicities associated with current therapies.

I just spoke in the stem cells and development session and in that session we’re examining the role of multiple ways that stem cells have to potentially drive cancer and to evade or escape current therapies. Stem cells have mechanisms by virtue of their ability to self-renew or regenerate themselves; they have the ability to try to evade and escape common therapies that are delivered for cancers. So I study childhood brain cancer, medulloblastoma, which is the most aggressive and common paediatric brain cancer. In this cancer we believe that there are primitive cell populations that may be driving the cancer and these are tricky cells, they use all kinds of tricks to evade chemotherapy and radiation therapy. These are standard therapies for this type of childhood cancer and these therapies tend to be very toxic. So my whole programme is looking at trying to develop novel therapies that are more targeted and more specific to avoid this horrible systemic toxicity of chemo and radiotherapy that may target the mechanisms by which these stem cells self-renew.

What is the benefit of these xenograft models?

My lab uses patient derived xenograft models and in fact my institute, the McMaster Stem Cell and Cancer Research Institute, is focussed purely on using human stem cell models of disease. The reason that we like to use human models is because we believe it’s one step closer to translating any therapies. As you may know, mouse models are excellent for learning about mechanisms of disease and small organisms and animal models that we use are fantastic but they have their drawbacks. The drawbacks are very often discoveries that we make in these model organisms don’t translate into human beings and, in fact, very often we find those things don’t recapitulate in the human system. So in a patient derived xenograft model you’re already starting with the material from the patient directly and simply using the mouse as a carrier, as a host, to explore the mechanisms of what that human disease will do. So in my particular patient derived xenograft model we take the same chemotherapy and radiation therapy that we use to treat children with brain cancer and we’ve tinkered with it, the whole protocol, in order to deliver it to immuno-compromised mice. So we engraft the mice with the human patient derived tumour, we treat the mouse with the same regimen and protocol of chemo and radiation therapy and then we’re able to read out what happens when the tumour inevitably relapses or recurs. We’re able to read out all the molecular mechanisms at every step of the way which often isn’t possible in human beings because in patients you’re not often able to sample the cancer at all those different time points but in our model we can.

What are some of the challenges that you have faced?

That’s an excellent question to ask about how patient to patient variability confounds a lot of the findings that we make in basic science research simply because we can’t extrapolate one beautiful mechanism we discover to every single patient. Quite simply, there’s no one drug that works for every patient. So that’s why developing these model systems that we use in my lab is trying to advance towards two goals – one goal is to try to model the dynamic nature of cancer, so cancer is not a static disease and very often when we simply biopsy a patient’s tumour and then study that tiny specimen to the greatest depth possible it’s excellent because it yields information but the problem is the tumour is evolving and changing in the patient. So very often the founding basis for our discoveries may not actually predict what’s happening in the patient. So one goal of our model systems is to try to model the dynamic nature of cancer and to look at something that will be sampled multiply over time. The second virtue of the system is that we model individual patient tumours almost in a personalised medicine approach. This will be the way of the future no matter what model or assay you use, it really should be adapted to each patient in order to understand the dynamics of that patient’s particular tumour.

What are some of the challenges with childhood cancer?

A very large problem in terms of funding for childhood cancer is that childhood cancer is very rare and this is a boon, of course, we’re very glad that so few children get cancer but it does make something difficult to study. For example companies that make drugs or therapies for diabetes have very large programmes that they can disseminate around the world because diabetes is such a common and such an invasive problem in our society so it’s something that we have large resources to ramp up for that type of problem. But very often I find it’s not engaging charities that is the problem it’s engaging pharma or industry because industry, very necessarily being financially driven, is interested in large problems that generate large revenues but rare and small problems it’s often hard to find that niche, to find industry partners who are interested in problems in paediatric cancer. That’s been by far the greatest challenge. On the other hand, finding public funding or charity funding for brain tumour research has not be difficult because children are very good advocates and people are very willing to form charities around children who are afflicted unfairly with cancer. However, I’d like to be able to capture industry partners more for childhood cancer.

What’s next for your work?

The future for my lab programme is definitely moving towards translation of basic science discoveries into cures for patients. Through all the mechanisms that we’ve discussed, simply finding model organisms and model systems that best capture the dynamic and evolving nature of cancer; by looking at more of a personalised medicine approach and modelling each patient’s tumour in such a way that we can predict in some ways what will happen shortly in their future and to be able to react quickly such that we can design empirical therapeutic regimens that are individualised for each patient and based very rationally on the patient’s own tumour’s molecular biology.