ASH 2015

Incidence of and predicators for early death in paediatric patients with acute promyelocytic leukaemia

Dr Oussama Abla - The Hospital for Sick Children, Toronto, Canada

You’ve been talking here at the ASH meeting on acute promyelocytic leukaemia and in particular you’ve done an international retrospective study of the predictors and the incidence of early death. Now, why were you particularly interested in that aspect of APL?

APL, as we all know, is a rare subtype of acute myeloid leukaemia and the outcome of these children with APL has actually improved a lot in the last 15-20 years.

Thanks to things like ATRA.

Exactly. With ATRA and chemotherapy now the ten year event free survival of these patients is around 80%. However, we still see a percentage of patients, between 4 and 7.5%, that suffer from early death in children. We don’t seem to know exactly what are the exact predictors of early death in children with APL whereas we know that in adults.

So you conducted a study, what did you do?

We conducted an international retrospective study; we tried to look at incidence and predictors of early death in childhood APL. Basically we asked several international groups from Europe, Canada and the US to collect data from their centres and we collected data on the incidence and also on some clinical treatment features to see if we can determine predictors of early death.

What aspects in particular did you focus in on?

We tried to look at age, gender, initial white blood cell count, the body mass index or the weight and height of the patient, but also on biological features like the platelet count, the initial white blood cell count, some cytogenetic markers like FLT3 ITD and whether they received ATRA in induction, chemotherapy or not, what other drugs they received in induction. So those are the main specific things that we focussed on.

Quite a lot of variables there.

Exactly.

What did you find?

We enrolled 707 children overall and the incidence of early death was around 6.3%. So among 707 children 45 had an early death and early death, by the way, was defined as death within 60 days from presentation. Among the patients who died we found out that ATRA induction was extremely important, which is not surprising. In fact, those who had not had ATRA in induction, the early death rate was 41% and those who had ATRA during induction the early death rate was only 5%, so that’s a huge difference. So this actually confirms the importance of starting ATRA and starting it as soon as possible.

How might you do that? It’s very fascinating that ATRA, which itself was an amazing drug when it was found to work in this situation, but how do you get that earlier diagnosis, earlier treatment?

The most recent routine recommendations for APL is to start ATRA as soon as possible. The minute you look at the microscope and you suspect APL by morphology, whether you’re in children or in adults, then you should start ATRA. You should not wait for the molecular test or the cytogenetic tests which can sometimes take two to three days. There have been adult studies that have looked at the timing of ATRA starting from presentation and it was found that if you start ATRA the first day the early death is close to zero percent. If you start it the next day the early death goes to 35%. If you start it the third day the early death rate goes to 75% but that’s in adults. We tried to look at that in children, unfortunately we had so many missing data about the timing so we could not really determine that. But it was obvious that ATRA in induction certainly is extremely important.

The other finding that we found that we had 33 patients among the total number had central nervous system bleeding and among those only 15% of them survived. In other words, if you have a CNS bleed in children with APL there’s an 85% chance of dying early on whereas if there was no CNS bleed the early death rate was only 2%.

Is that something potentially you could control?

This is an event that is one of the most challenging things in APL is to control early bleeding and thrombosis. What we need is really to try to optimise management of bleeding and thrombotic complications, especially in high risk patients. In APL high risk is defined as anyone who comes through the door with a white cell count of more than 10,000, that’s a high risk patient. So those patients are those who need to be more aggressively treated with transfusions and to optimise the coagulopathy to prevent bleeding.

Could you run me through ideally what the doctor should be doing the moment she or he suspects APL?

The minute that they suspect APL by the morphology they immediately should start ATRA orally.

That’s without waiting for confirmation?

No, just morphology is enough, just give ATRA immediately. That’s the first thing you have to do so, in fact, we always suggest to have ATRA tablets in emergency departments or even on the in-patient ward. The second thing, they need to keep platelets above 50,000, they need to keep the fibrinogen above 1.5, so transfusing fresh frozen plasma or cryoprecipitate very aggressively. The anaemia, the haemoglobin they don’t need to keep it above a certain level because actually if the white cell count is very high you don’t want the haemoglobin to be high because that will increase the risk of viscosity and that will increase the risk of stroke. So you need to transfuse platelets immediately, FFP or fresh frozen plasma as soon as possible, at the same time give ATRA as soon as possible.

Any ideas about reducing the risk of CNS bleeding?

As I said, this is a major problem in both adults and children. The way to reduce it is by doing these three things immediately, that hopefully will help. But unfortunately there’s a subset of patients that no matter what you do they die from CNS bleeding. Why? Because they may get to the hospital at the plate and they may actually die before we start treatment with ATRA. In fact we had, among the cause of early death, 60% of them were caused by CNS bleeding and it was about 28 patients. 3 of them actually died before we started any treatment so that’s the big problem. The other problem is cerebral thrombosis, so not only bleeding but thrombosis. That’s even more difficult to treat than bleeding, extremely more difficult to treat, because one of the treatments for thrombosis is to give heparin but at the same time heparin can cause bleeding. So it’s even more difficult to treat than bleeding, but both of them need to be tackled immediately, especially in high risk patients.

So what’s the bottom line message, then, coming out of this for cancer doctors?

The bottom line message is that bleeding and thrombosis are still very frequent in children with APL and we need to tackle them early on by starting ATRA and aggressive transfusion immediately, as soon as possible, without waiting for molecular and cytogenetic tests to come back. But we really need new drugs to counteract the coagulopathy. We need prospective studies to study predictors of early death in childhood APL.