Prevalence of 46/1 JAK2 Haplotype in Patients with Budd-Chiari Syndrome with and without JAK2V617F and TET2 Mutations.

Abstract 434

Budd-Chiari Syndrome (BCS) is a group of disorders resulting from obstruction to hepatic venous outflow; myeloproliferative disorder (MPD) accounts for 10-40% of cases. We previously described latent MPD in 58.5% of patients with idiopathic BCS, detected with allele-specific PCR for the JAK2V617F mutation and proposed its use as a screening tool for occult MPD. A predisposing germline JAK2 haplotype (designated 46/1) has since been described as a strong genetic risk factor for MPD and may further characterise latent MPD in BCS. We studied 28 patients with BCS (23 from our original cohort; female n=16, mean age 30.3 years, SD 10) presenting between 1985 and 2008; 14 with the JAK2V617F mutation. Genomic DNA was obtained from archived bone marrow films, fractionated and unfractionated peripheral blood or bone marrow leucocytes. Skin biopsy or CD3+ cells were used as a source of constitutional DNA. DNA was analysed by pyrosequencing for 2 SNPs (rs12340895, rs12343867) which tag the 46/1 JAK2 haplotype. The 46/1 haplotype was detected in 16/28 (57.1%) subjects; 50% of those with the JAK2V617F mutation and 64.3% of those without it. The prevalence in those lacking the JAK2V617F mutation is significantly higher than the frequency in the Wellcome Trust Case Control Consortium cohort of 24% (P=0.0023). 3/28 subjects had previously diagnosed JAK2V617F positive Polycythemia Vera (PV) and all had the 46/1 haplotype, resulting in a prevalence of 36.4% in those with JAK2V617F positive latent MPD. Age at presentation of BCS was significantly lower in those with the 46/1 haplotype (26.4 years compared to 34.8 years, P=0.03). This difference remained significant in those lacking the JAK2V617F mutation (24.0 years compared to 37.6 years, P=0.024) but was not seen in those with the JAK2V617F mutation (P=0.547). There was no difference in presenting clinical features, haematological parameters or treatment between those with and without the 46/1 haplotype. Overall survival in 26/28 patients was 76.9% (median 90 months, range 2 days to 266 months). 17/28 subjects underwent OLT of which 14/17 (11/12 with 46/1 haplotype) are alive at a median of 90 months post transplant (range 9-266 months). 3/17 patients developed post-OLT veno-occlusive disease, all with the JAK2V617F mutation and 2/3 with the 46/1 haplotype. Overt MPD has not developed in any patient without the JAK2V617F mutation; repeat JAK2 mutational analysis was undertaken in 3/14 (2/3 with 46/1 haplotype) and none have acquired the mutation at a mean of 54 months. 19/28 cases were genotyped using SNP markers (Affymetrix SNP6); 3/19 have acquired uniparental disomy (aUPD) on 9p overlapping the JAK2 gene. As TET2 has been postulated as a ‘pre-JAK2' aberration, we sequenced the complete TET2 gene using massively parallel high throughput sequencing (Roche 454); 2/15 patients samples were positive for TET2 mutations. One of our cases had a familial history of PV; the patient, his father and uncle all have JAK2V617F positive PV and were heterozygous for the 46/1 haplotype in DNA extracted from a skin biopsy. 2/3 were homozygous for both the 46/1 haplotype and JAK2V617F mutation in bone marrow granulocytes with SNP6 array data confirming aUPD on 9p. JAK2V617F was detected in cultured in vitro colonies from all 3 family members. All 3 affected family members had normal cytogenetics and normal TET2 gene. 3 unaffected siblings were heterozygous for the 46/1 haplotype both in peripheral blood, CD3+ cells and granulocytes but negative for JAK2V617F mutation and lacked aUPD on 9p. We have found a highly significant prevalence of the 46/1 haplotype in our cohort of BCS, as well as in family members of a patient with JAK2V617F positive BCS and PV. The 46/1 haplotype was detected in patients with idiopathic BCS with and without the JAK2V617F mutation, suggesting a predisposition to idiopathic BCS independent of JAK2V617F mutation acquisition and latent MPD. The prevalence and lower age of presentation in those with the 46/1 haplotype lacking the JAK2V617F mutation supports an alternate, as yet unknown, mechanism predisposing to BCS. The presence of the 46/1 haplotype in unaffected relatives of our JAK2V617F BCS patient suggests that additional germline variation may predispose to or protect from acquisition of JAK2V617F positive disease. Alternatively the 46/1 haplotype may directly confer a cellular growth advantage via increased responsiveness of JAK2 to cytokine stimulation.