Personalizing Busulfan Therapy for Children Undergoing Hematopoietic Stem Cell Transplantation

Anuj K Tyagi; Patricia Huezo-Diaz; Marc Ansari


Personalized Medicine. 2014;11(5):463-466. 

In This Article

Pharmacogenetics as a Tool to Predict PK of BU & Clinical Outcomes

Studies have demonstrated that a single variant or haplotypes in genes encoding GST enzymes alters BU clearance (CL).[7,11] GSTA1*B haplotype and GSTM1-null genotype may contribute to diminished metabolism of BU and higher plasma levels owing to the diminished activity of the enzyme irrespective of the route used for administration of BU.[7,11–13] However few other studies did not find this association.[14,15] In a recent report by our group we also showed existence of subhaplotype groups within GSTA1*A and *B haplotypes and their association with PK of BU, explaining the importance of haplotype-based subgroup analysis, which may partly explain why previous studies did not find an association of GSTA1 with BU PK.[7] Another recent study conducted in adult patients also found reduced intravenous BU CL in individuals with combined GSTM1 and GSTT1-null genotypes, but not in individuals with GSTM1-null genotype alone indicating the importance of gene–gene interactions,[16] whereas others reported absence of such association.[11–14] Thus, presence of haplotype subgroups, gene–gene interactions, nonhomogeneous patient population, sample size and treatment protocols or drug administration routes may account for discrepancies seen across studies performed till date to understand association of BU PK with GST gene polymorphisms.[7]

Bartelink et al. demonstrated a link between the PK and BU toxicity, as well as event-free survival for patients receiving BU before HSCT and thus it was possible that a correlation between the GST polymorphisms and clinical outcomes of BU treatment may also exist.[1] Many factors are important for determining the clinical outcomes of HSCT in addition to BU exposures, which includes underlying disease, immunological factors, transplant characteristics, pre-existing risk and particularly other drugs administered concomitantly with BU.[1,11] However, our recent study[7,10] rather documented a higher incidence of sinusoidal obstruction syndrome in individuals carrying nonfunctional haplotypes of GSTA1*B and BU Css above 600 ng/ml. Even in multivariate analysis, GSTA1*B was independently associated with higher incidence of sinusoidal obstruction syndrome, suggesting its influence on BU CL and thus the outcomes. We also observed higher incidence of acute graft-versus-host disease in GSTM1-null allele carriers, thus highlighting the reduced CL of BU in these individuals in relation to acute graft-versus-host disease incidence via enhanced tissue damage also reported previously.[7,12] We also observed improved event-free survival in carriers of normal GSTA1 haplotypes (*A and *A2) and individuals with BU Css below 600 ng/ml with lower incidence of treatment related mortality in these individuals.[7,10] These observations suggest that BU CL and BU Css may be predicted from GST genotypes and that the therapeutic window of BU in children might be different to that of adults and could be targeted below or near 600 ng/ml for better outcomes. An ongoing European Blood and Marrow Transplantation study at our center is validating these results to establish the evidence for these observations to be implemented in clinics.