Risk-driven therapy in acute myeloid leukaemia
Session type: Parallel sessions
The median age patients with acute myeloid leukaemia (AML) is 68 years, so the choice of treatment is conditioned by the estimated risk-benefit for each patient with the probability that intensive therapy with curative intent will not shorten life. The cytogenetic and molecular heterogeneity of the disease associates with treatment response. Many prognostic factors are emerging which negatively or positively influence outcome, so there is great interest that these factors are not only prognostic but are also predictive of response to a particular therapy. Many factors may not been validated on a separate dataset. The therapeutic implications are limited with respect to what treatment is or is not, predictive. At the moment the use of risk directed therapy in patients who are suitable for curative chemotherapy is who should be transplanted and who should not.
Cytogenetics divides patients into favourable/ intermediate or poor risk categories, with a consensus view that myeloablative transplants should be given to poor risk, but avoided in favourable risk. There remains controversy about its use in the 60% of patients defined as intermediate risk. The issues arise: I) how should transplant be assessed? ii) optimal definition beyond cytogenetics? and iii) what happens to a patient if transplant is delayed until a second remission? Mutations of FLT3 and NPM1 for example are frequent and of negative and favourable prognostic value respectively. This is complicated by the combination of both in some patients where NPM1 gives a protective effect from the negative impact of FLT3. There are several examples of molecular factors where the prognosis is dictated by whether they occur alone or in combination.
Defining patient subsets who derive most benefit from a particular drug based on risk assessment using simple factors like the score which predicts a 10% improvement in survival if given the immunocojugate, gemtuzumab ozogamicin, or the molecular cluster with is most responsive to higher doses of daunoribicin in induction.