Can novel genetic polymorphisms predict response to therapy in acute myeloid leukemia?

Acute myeloid leukemia (AML) arises on the background of genetic alterations in the hematopoietic progenitor cells that influence their growth, differentiation, and proliferation. Besides some genetic disorders, such as Down syndrome and Fanconi anemia, little is known about the genetic predisposition to AML. A genome-wide association study (GWAS) on therapy-related AML showed an association with certain single nucleotide polymorphisms (SNPs) [1], but larger studies in this area are needed. Therapy often includes Ara-C (cytarabine) and an anthracycline, with remission rates ranging from 35 to 80% depending on patient characteristics and acquired cytogenetic abnormalities. However, little is known about the impact of germline mutations in treatment response.

In this issue of Leukemia and Lymphoma, Zhong and colleagues [2] present data on the role of a Janus kinase 2 (JAK2) SNP, A830G, in treatment outcomes in 152 patients with AML. More specifically they show that JAK2 830*AA is associated with higher complete response compared with the JAK2 830*GG genotype. Furthermore, in a previous study, the same investigators found that JAK2 A830G is associated with susceptibility to acute leukemia. This is not the first report on the role of SNPs in outcomes and response to therapy in patients with AML. A SNP in the WH1 (rs16754) was found to be a favorable marker in cytogenetically normal AML [3], and a BCL2 polymorphism (rs2279115) was found to predict outcomes in patients with AML [4]. Others have found that polymorphisms in DNA repair genes can predict response to therapy [5]. Some of the SNPs in question are associated with an amino acid change, whereas others, like the JAK2 A830G, are synonymous, meaning that they do not cause an amino acid substitution. So can a synonymous SNP predict risk of disease or response to therapy? The answer is yes. Synonymous mutations can result in conformational changes in proteins [6], aberrant splicing [7], premature ending of transcription [8], and decreased levels of mRNA expression [9]. Furthermore, these analyses do not necessarily reveal a significant SNP, but more often than not a significant genomic region.

The significance of SNPs in identifying high-risk individuals is still questionable. The past few years have seen several GWASs being published for diseases ranging from cancer to Alzheimer's. Several SNPs have been identified as being potentially causative. Some have been found in regions corresponding to known genes, whereas others have been found in the extragenic part of DNA. Furthermore, this process has produced new genes and increased our understanding of the pathogenesis of diseases. However to date, at least in cancer, these studies have not helped identify a high-risk population. Known epidemiologic risk factors have overshadowed the importance of the SNPs found in GWASs. The era where an individualized genetic profile will accurately assess the risk for disease is far away. And this may be due to the limitations from studies looking at individual SNPs. SNPs are not the only genetic changes in the DNA. Deletions, insertions, translocations, and epigenetic phenomena are examples of other DNA related changes that may predispose to a specific disease. But identifying potential causative SNPs is still an important task.

Can we tailor patient therapy by identifying germline or tumor related mutations? The prototype drug that took advantage of our genetic knowledge was imatinib. Since then, others have found their way into clinical practice in hematologic malignancies. Even in solid tumors, the introduction of Poly-(ADP-ribose) polymerase (PARP) inhibitors shows great promise in carriers of deleterious mutations in BRCA1 and BRCA2 [10]. Taking the information from GWAS as well as other SNP analyses can help us identify pathways that are important in the pathogenesis of the disease and lead to treatments targeting that area. Therefore, the study by Zhong and colleagues can help identify individuals with AML who can benefit from therapy targeting the JAK/STAT pathway.

The potential significance of a region in JAK2 being associated with treatment response is high. JAK2 plays an important role in hematopoiesis. A recent report that an acquired activating mutation of JAK2, V617F, is present in the majority of patients with Philadelphia chromosome negative (Ph−) myeloproliferative disorders, has led many drug companies to develop and test JAK2 inhibitors [11]. Currently, several JAK2 inhibitors are in clinical trials, mostly in patients with myelofibrosis (MF), with early results showing an improvement in splenomegaly and constitutional symptoms, although treatment of other hematologic malignancies such as AML also shows promise [12].

Studies such as the one from Zhong and colleagues have the potential to point us toward causative genes with direct implications for prevention, early detection, and treatment. The current study is a small step forward in that direction, and larger studies evaluating many more SNPs are needed for real progress in the field.