AML17

Between 2009 and 2014 AML17 recruited over 3,500 patients aged generally under 60 to what is the largest ever trial in Acute Myeloid Leukaemia run in the UK.

Patients were recruited from the UK, Denmark and New Zealand. AML is a condition with just over 2,000 new cases every year and a median age of diagnosis of 67. This means that there are only about 800 cases of AML in this age group in the UK annually. In order to test new treatments rapidly, the UK NCRI AML trials use factorial designs to answer a number of questions at the same time as efficiently as possible. Patients are randomly allocated to different treatments at different parts of their treatment journey, and AML17 has also tested the use of targeted therapies where treatment is based upon different genetic features of the disease. First results are appearing and these are summarised as follows:

• We tested the best dose of a drug called Mylotarg. We have already shown that the drug improves survival; here we asked if increasing the dose would lead to greater improvements. We found no improvement but more toxicity with higher doses
• We tested whether increasing the dose of a standard drug called daunorubicin improved outcomes. Other groups have found some benefit, but their outcomes are generally worse than ours. We found that increasing the dose did not seem to help except possibly in a group of patients with a genetic mutation known as FLT3-ITD.
• We tested whether a drug called everolimus would improve outcomes. This did not improve outcomes and was closed early.
• We found that in patients with a subtype of disease called APL, we could do away with chemotherapy altogether, and give a combination of two drugs, ATRA and arsenic trioxide. We found this significantly reduced the risk of the disease coming back.

We are in the process of analysing data on a drug called lestaurtinib for patients with FLT3-ITD mutations, and whether a drug called clofarabine can improve outcomes in patients with poor risk disease. We are also collecting follow-up data to determine the best number of courses of treatment.

Finally, we have recruited over 400 patients to a randomisation designed to find out if regularly monitoring patients using sophisticated laboratory techniques can improve outcomes. This will be carried forward into our new AML19 trial, but we have already identified a test which picks out a group of patients previously thought to be good risk who have a poor prognosis. The challenge is now to find an effective treatment for these patients, a challenge which is taken up in AML19.