The JAK2 46/1 haplotype in splanchnic vein thrombosis

To the editor:

We read with interest the recently published report by Smalberg et al, which describes the frequency of the JAK2 46/1 haplotype among patients with splanchnic vein thrombosis (SVT).¹  Indeed, recent studies revealed that the 46/1 haplotype located in the JAK2 gene is strongly associated with the development of a JAK2V617F (JAK2VF)–positive myeloproliferative neoplasm (MPN)^2-4  but also with JAK2 exon 12– or MPL gene mutation–positive MPNs.^5,6 

We studied this haplotype in a cohort of 170 patients with SVT (sex ratio M/F = 1.12). Approval was obtained from the PV-Nord Institutional Review Board. Informed consent was obtained in accordance with the Declaration of Helsinki. As controls we studied 58 non-MPN patients with peripheral vein thrombosis (PVT) and 31 patients with JAK2VF-positive polycythemia vera (PV). We studied 2 tag single-nucleotide polymorphisms (rs10974944 and rs12343867) on peripheral blood DNA using Taqman allelic discrimination assays (Applied Biosystems). In 170 patients with SVT, the frequency of the 46/1 haplotype was 0.28 (Table 1), which is not significantly different from the published population controls from the Wellcome Trust Case Control Consortium (WTCCC) (0.24; P = .11 by the Fisher exact test). Thus, in accordance with Smalberg et al, our results suggest that the 46/1 haplotype is not a susceptibility locus for the development of SVT.

In our control populations, the frequency of the 46/1 haplotype was similar to the WTCCC controls in 58 non-MPN patients with PVT (P = .24) and was significantly higher in the group of 31 JAK2VF-positive PV patients (0.52; P = .0005), in agreement with previous studies.^2-4  However, we found no difference in the frequency of the 46/1 haplotype between JAK2VF-positive (0.27; n = 75) and JAK2VF-negative patients (0.28; n = 95) (P = .90). This finding is in contrast with the study by Smalberg et al, which reported a significantly higher frequency of the 46/1 haplotype in 54 patients with JAK2VF-positive SVT compared with controls (0.43 vs 0.27, respectively; P < .01).

The discrepancy between the 2 studies may depend on the populations studied. The proportion of JAK2VF-positive versus -negative patients is higher in our study (75/95 vs 54/145), making it possible that the high 46/1 haplotype frequency found by Smalberg et al would not be confirmed in a larger cohort with greater statistical power. Next, it recently appeared that the 46/1 haplotype frequency is correlated with the JAK2VF allele burden, VF-positive patients with low mutation burden displaying 46/1 haplotype frequency similar to VF-negative patients.^7-9  In our study, concordantly with the literature,¹⁰  mutant allele burden was < 50% in 89% of patients (median: 19%) whereas in the study by Smalberg et al, JAK2VF allele burden was available in 45 patients only. In summary, in a larger cohort of JAK2VF-positive patients with comprehensive mutant allele burden data, we did not confirm the higher frequency of the 46/1 haplotype in SVT patients with MPNs found by Smalberg et al.