Abstract
Background: Patients with transfusion-dependent thalassemia (TDT) require blood transfusions generally every 2-5 weeks, depending on disease severity. The high iron burden that results from frequent blood transfusions can lead to organ dysfunction, therefore iron chelation is a critical part of therapy for all patients with TDT. Deferasirox (DFX) is a tridentate oral iron chelator indicated for patients ≥ 2 years old with TDT. However, ~30% of TDT patients do not respond to DFX, for reasons that are not understood. To evaluate if pharmacokinetic differences in patients might impact DFX efficacy and/or toxicity, a single center, prospective pharmacokinetic study was undertaken at UCSF Benioff Children's Hospital over 2 years. This study is focused on evaluating the impact of selected pharmacogenetic variants in key genes in DFX pharmacology.
Patients and Methods: Blood samples were obtained for pharmacogenomic testing from patients (age ≥ 2 years) enrolled in this study during their regular clinic visits, prior to blood transfusion. Institutional Review Boards at the University of Florida and UCSF approved the study.DNAfrom these samples were genotyped for rs887829 and rs4124874 in UGT1A1 and rs2273697 in ABCC2 SNP using TaqMan-based genotyping assay (QuantStudio3). Association between SNP genotypes and DFX efficacy markers 1) serum ferritin levels (above or within target range of 500-1000 mg/mL) and 2) LIC (above or within target range of <7 mg/g) was tested using logistic regression analyses (RStudio v4.5.1). The Genotype-Tissue Expression (GTEx) portal was utilized to explore the relationship between respective SNP and gene expression in liver.
Results: Overall, 49 patients were included, the median age of the participants was 34.5 years (range 2-63), with 47% identifying as female and 53% as male. SNP rs2273697 (G>A) in ABCC2/MRP2 (multidrug resistance-associated protein 2), which is involved in the excretion of DFX from hepatocytes into bile, was associated with increased serum ferritin levels above the target range (p <0.001). Gene expression data from GTEx database shows that the rs2273697 (G>A) change is associated with higher intron expression ratio of ABCC2/MRP2 in the liver. As part of the UDP-glucuronosyltransferase (UGT) family, UGT1A1 is involved primarily in DFX inactivation, while UGT1A3 plays a minor role. We tested 2 SNPs in UGT1A1/UGT1A3, rs4124874 (T>G) and rs887829 (C>T) for association with DFX efficacy in this cohort. For patients with the rs887829 T allele (TT and CT genotypes), 47% had LIC exceeding the target range compared to only 21% with CC genotype. Similarly, for rs4124874, 42% patients with GG and GT genotypes had LIC exceeding the target range compared to only 13% in those with TT genotype (both p <0.001). For rs887829 presence of T allele (CT and TT genotype) was also associated with high serum ferritin levels (p=0.002). Both the UGT SNPs were also significantly associated with higher hepatic gene expression levels of UGT1A1 and UGT1A3 in data from GTEx. These results imply decreased effectiveness of DFX and thus elevated LIC or serum ferritin levels above the target range.
Conclusions: Overall, our data show that 3 SNPs in ABCC2/MRP2 and UGT1A1/UGT1A3 are significantly associated with reduced DFX efficacy in our study population. Despite a small patient cohort, our finding highlights that genetic testing has the potential to be incorporated in clinical management when initiating DFX monotherapy or concomitant therapy to optimize clinical outcomes. Our ongoing work is focused on expansion of the list of SNPs to include other key variants in DFX pharmacology and further investigate the association of candidate SNPs with DFX-induced toxicities (grade ≥3) such as renal failure, hepatic failure, and gastrointestinal hemorrhage based on selected toxicity markers. Overall, our goal is to improve therapy and quality of life for patients with TDT, and reduce toxicity-related mortality.
