Therapy of acute myeloid leukemia: towards a patient-oriented, risk-adapted approach

The successful use of differentiating treatment for patients with acute promyelocytic leukemia (APL) suggests that other acute myeloid leukemias (AML) may benefit from tailored and subtype-specific therapy. Despite the fact that new drugs specifically targeting AML genetic lesions have not yet been developed, distinct karyotypic categories have been identified which may deserve differentiated treatment. In addition, molecular assays to assess response to therapy more sensitively are now available for several AML subsets. In this review, we discuss the role of genetic characterization in the therapy of AML, and the investigative efforts which we believe are still needed for the design of tailored treatment for each and every patient with this disease.

The authors have been working in this field for many years and have contributed original papers, the data of which are incorporated in this article. In addition, the material analyzed in this overview includes articles and reviews covered by the Science Citation Index and Medline as well as some more recent unpublished personal observations.

Modern therapeutic approaches to AML tend to differentiate post-induction treatment intensity according to cytogenetically defined risk categories. Such prognostic categorization is largely unsatisfactory. In fact, following the advent of newly developed molecular assays (e.g. RT-PCR and FISH), specific and prognostically relevant lesions are frequently found in patients with an apparently normal karyotype, and these patients are, therefore, re-assigned to more appropriate prognostic categories. In addition, recent studies suggest that some patients may benefit from an increase in induction intensity; rapid genetic characterization will be needed for future differentiation of initial therapy. However, preliminary investigation of AML by integrated karyotypic/molecular analyses show that no specific abnormalities are detectable in at least half of the cases. Therefore, use of genetic criteria for prognostic stratification is currently feasible in only a proportion of patients.

The prognostic role of genetic lesions, currently identified by karyotypic studies, needs to be validated in large series of AML patients prospectively characterized by advanced molecular/cytogenetic analyses and treated uniformly. In addition, searches for new clinically relevant genetic abnormalities, and diagnostic tools for their rapid identification are urgently needed to identify prognostic categories better. Elucidation of AML gene alterations should foster basic investigation aimed at developing new drugs targeted to the specific lesion in the individual patient. Before these more specific therapeutic agents are developed, diagnostic genetic characterization should add to other well-established prognostic factors to optimize the use of the presently available therapies.