Poster Presentations

Session Title: Category 13. LATE-BREAKERS
Presentation Date: Apr 15, 2010

GENOME WIDE ANALYSIS OF PATIENTS FROM THE IDEAL STUDY IDENTIFIES A CAUSAL ROLE FOR ITPA GENETIC VARIATION IN RIBAVIRIN-INDUCED HEMOLYTIC ANEMIA

A.J. Thompson¹*, J. Fellay², D. Ge², T. Urban², K. Shianna², M. Sulkowski³, A. Muir¹, N. Afdhal⁴, I. Jacobson⁵, R. Esteban⁶, F. Poordad⁷, E. Lawitz⁸, J. Mc Cone⁹, M. Shiffman¹⁰, G. Galler¹¹, W. Lee¹², R. Reindollar¹³, J. King¹⁴, P. Kwo¹⁵, R. Ghalib¹⁶, B. Freilich¹⁷, L. Nyberg¹⁸, K. Patel¹, H. Tillmann¹, S. Noviello¹⁹, N. Bopari¹⁹, K. Koury¹⁹, L. Pedicone¹⁹, C. Brass¹⁹, J.K. Albrecht¹⁹, D. Goldstein², J.G. Mc Hutchison^1
1Duke Clinical Research Institute, ²Institute for Genome Sciences & Policy, Center for Human Genome Variation, Duke University, Durham, NC, ³Johns Hopkins University School of Medicine,, Baltimore, MD, ⁴Beth Israel Deaconess Medical Centre, Boston, MA, ⁵Weill Cornell Medical College, New York, NY, USA, ⁶Hospital General Universitario Valle de Hebron, Barcelona, Spain, ⁷Cedars-Sinai Medical Center, Los Angeles, CA, ⁸Alamo Medical Research, San Antonio, TX, ⁹Mt. Vernon Endoscopy Center, Alexandria, ¹⁰Virginia Commonwealth University, Richmond, VA, ¹¹Kelsey Research Foundation, Houston, ¹²University of Texas Southwestern Medical Center, Dallas, TX, ¹³Piedmont Healthcare, Statesville, NC, ¹⁴Louisiana State University, Shreveport, LA, ¹⁵Indiana University School of Medicine, Indianapolis, IN, ¹⁶The Liver Institute at Methodist Dallas Medical Center, Dallas, TX, ¹⁷Kansas City Gastroenterology and Hepatology, Kansas City, MO, ¹⁸Kaiser Permanente, San Diego, CA, ¹⁹Schering-Plough Corporation (now Merck & Co., Inc.), Whitehouse Station, NJ, USA. *alexander.thompson@duke.edu

Background/Aims: We performed a genome wide association study on a well characterized genotype 1 HCV treatment cohort to identify genetic determinants of ribavirin (RBV)-induced hemolytic anemia (HA).
Methods: 1604/3070 patients treated with peginterferona (pegIFN) and RBV in the IDEAL study consented to DNA testing. Samples were genotyped using the Illumina Human610-quad BeadChip. After quality control, 97.5% of the single nucleotide polymorphisms (SNPs) included on the chip were used in the analyses. The primary analysis focused on the genetic determinants of quantitative change in hemoglobin (Hb) levels from baseline to week 4 of treatment (to minimize confounding by erythropoietin use), in 3 separate populations (Caucasians, African Americans, Hispanics) by logistic regression, adjusting for: age, gender, weight, liver fibrosis, baseline Hb level, RBV dose, and type/dose of pegIFN. A modified Eigenstrat method controlled for population stratification, and Bonferroni adjustment corrected for multiple testing.
Results: 1,286 patients were included in the final analysis. The SNP rs6051702 on chromosome 20 was strongly associated with Hb reduction at week 4 in the 3 separate populations (overall P< 10^-46). Genotyping of known functional variants demonstrated that the association signal was entirely explained by SNPs in the adjacent ITPA gene (encoding inosine triphosphatase, ITPase): rs1127354 and rs7270101; the minor alleles cause ITPase deficiency and were protective against anemia. A composite ITPase deficiency allele showed an association of P=10^-91. Predicted ITPase deficiency, defined according to functional studies of these variants, strongly protected patients against Hb falling >3g/dL by week 4 (P< 0.0001) (Figure 1). The causal ITPA variants were not associated with SVR.


[Figure 1]

Conclusions: We have identified two functional variants in the ITPA gene that are strongly associated with the risk of RBV-induced HA. ITPA genotyping could help guide clinical decision-making, especially in patients at high risk for anemia or related morbidity.