ASH 2015

Variants for increased risk of osteonecrosis in paediatric acute lymphocytic leukaemia identified

Dr Seth Karol - St Jude Children's Research Hospital, Tampa, USA

You’ve been looking at osteonecrosis, now this is in kids with acute lymphoblastic leukaemia. Why were you looking at this?

Correct. So ALL is the most common malignancy in childhood and over the last several series of clinical trials osteonecrosis, or avascular necrosis, has emerged as a really therapy limiting toxicity. The reason that we think that might be is twofold; the first is that we’ve tried to intensify steroids as well as asparaginase II, therapies with probably less long-term side effects, in our most recent series of clinical trials. The other is as we’ve become aware of osteonecrosis we look for it more, so we’ve probably found it more as well.

So to what extent is this a problem in treatment - curing kids from their ALL but causing bone toxicity?

Yes, so in older children, the children between 10 and 20 years of age we’re seeing osteonecrosis in anywhere from 10% to even 30% of children who get ALL therapy, depending on therapeutic details. In these younger children that we’ve looked at in the most recent study the incidence is about three- or fourfold lower, so between 2-4% of children.

Now you’ve been doing a genome wide search for the factors that might be causing this or might be associated with it. What did you do?

We identified patients from the Children’s Oncology Group protocol 0331 and we took their DNA, their germline DNA, things that they’ve inherited and looked at it using a series of SNP arrays to see single nucleotide polymorphisms, these single DNA base changes that seem to change proteins, either changing their levels or changing the actual structure of the protein, and looked at how those changes may impact the risk of osteonecrosis.

And where did glutamate receptor signalling come into this?

Our top variant, the top nonsynonymous variant, so a variant that changes the structure of the protein was in a glutamate receptor called GRID2. What we also did is we looked at all of the polymorphisms, all of the variants in the DNA, and matched them up to genes and then we categorised those genes in different pathways. So the top pathway that was also associated with osteonecrosis was also a glutamate receptor signalling pathway.

Another issue you raised is BMP7, this is a bone morphogenetic protein. What’s that and what does it do?

Bone morphogenetic protein 7 is a protein that seems to affect the differentiation of mesenchymal stem cells or mesenchymal precursor cells within the bone marrow. It seems to affect the pattern of differentiation, causing them to go either more to osteoblasts, these bone-forming cells, or to adipocytes, fat cells.

So, you’ve got these genetic influences on the bone and, oddly, also on whether you get fat or not. How have you been able to put this together in terms of something that could help doctors who are treating patients with ALL?

The variant that we identified in our population was located in BMP7, it was the top association with the development of osteonecrosis in these younger children. What we would like to do in the future is to try to confirm this finding. If so, you could potentially use that finding as a biomarker for the risk of osteonecrosis but before you would want to change therapy based on that you would want to first make sure that patients that had that variant weren’t at greater risk of relapse if you change therapy. So we’re doing some work in the lab to look at our leukaemia models and our animal models of osteonecrosis to make sure that those variants do in fact incur additional risk of osteonecrosis. Then second-fold that mice that have those variants are at lower risk of relapse. So if both of those things are true then you might conceptually think about decreasing therapy for those patients. But we don’t want to compromise the excellent cure rates we have for acute lymphoblastic leukaemia by trying to make changes to effect osteonecrosis risk.

So it’s a bit of a knife edge then, isn’t it, because you’ve got good treatment. What should doctors be making of this at this stage?

Right now I think what it tells us is that we know drugs that are seeming to be impacting osteonecrosis, it’s these steroids and asparaginase. I think the clinicians in the larger co-operative groups are already making changes to their regimens based on those thoughts and those assumptions. I think our findings reinforce those decisions that are being made and hopefully give us opportunities for future work to try to separate the two, separate cure from the side effect osteonecrosis.

I realise this is pushing it a bit into the future but are you just learning about the disease or do you think you’ll get some real progress in making kids well?

I hope it’s both. I hope that by learning more about the disease we’ll be able to find what parts of therapy are causing osteonecrosis and only then will we be able to truly separate the parts of therapy that are curing the leukaemia from the parts that are also giving the toxicity.

So the take-home message for doctors is what?

The take-home message for doctors, I would say, is that osteonecrosis remains a major toxicity for therapy and that we’re making progress towards understanding what’s causing it. Hopefully that progress will really enable us to separate those two aspects, the cure from the toxicity.