PRV-1 mRNA is overexpressed in most patients with polycythemia vera (PV) and a part of patients with essential thrombocythemia (ET). It is well known that PRV-1 expression levels correlate with JAK2 V617F mutation. While it was reported that JAK2 V617F clones can be significantly reduced or become undetectable by treatment with hydroxyurea (HU)[Girodon et al. Heamatologica 93:1723, 2008].The change of PRV-1 expression after treatment of myeloproliferative neoplasm (MPN) has been examined in only several small-scale studies with inconclusive results. Although PRV-1 overexpression is not included in the diagnostic criteria for MPN, it can be a potentially important biomarker in making diagnosis and predicting outcome. However, its clinical utility has not been fully investigated.
From 1999 to 2011, 309 cases of PV and 451 cases of ET were registered in Chang Gung Memorial Hospital. The diagnosis was reclassified according to the 2008 WHO system. Among them, 203 cases of PV and 228 cases of ET were examined for PRV-1 overexpression and formed the basis of this study. Evolution from ET to PV during clinical follow-up (ET-PV) was observed in 41 patients and 16 of them have PRV-1 data available. Quantification of PRV-1 mRNA was performed by the real-time polymerase chain reaction with TaqMan assay on peripheral blood granulocytes. Data from fifty healthy donors served as normal controls. The results were compared between MPN subgroups (ET vs. PV vs. ET-PV) and correlated with JAK2 V617F allele burden, white blood cell counts, platelet counts and clinical thrombosis or bleeding events. PRV-1 mRNA expression levels of patients before and after HU treatment were compared. In addition, the pre-HU and post-HU data were analyzed in pairs to evaluate the impact of HU therapy on PRV-1 mRNA expression.
Levels of PRV-1 expression were significantly different between controls and ET ( P<0.01), and between ET and PV ( P<0.001). ET-PV patients' pre-HU PRV-1 expression was comparable to other PV (P>0.05) but significantly higher than ET patients (P<0.01). PRV-1 overexpression, defined as pre-treatment PRV-1 /GAPDH ratio greater than 1.13, was found in 127 PV and 46 ET patients, which translated into 90.7% sensitivity and 75.3% specificity in differentiating PV from ET. When the post-HU PRV-1 /GAPDH ratio was used for differential diagnosis between PV and ET, the sensitivity was 90.5% and specificity 71.4%. PRV-1 expression levels were proportional to JAK2 V617F allele burden (r2=0.065, P=0.004). White blood cell counts were proportional to PRV-1 expression for both PV and ET (P<0.001 for PV, P<0.0001 for ET). The PRV-1 expression levels were divided into 4 quartiles, the top quartile had a risk ratio of 6.55 for developing thrombosis compared with the bottom quartile (95% CI 2.01–21.32, P=0.0008). In either PV or ET group, no difference was found between pre-HU and post-HU PRV-1 expression (P=0.579 for PV; P=0.253 for ET and P=0.143 for both subtypes combined). In analysis of the paired matched pre-HU and post-HU samples, there was no significant change of PRV-1 expression after HU treatment in any subtype (N=21, P=0.613 for PV; N=14, P=0.463 for ET and N=35, P=0.794 for both subtypes combined). Pre-treatment levels of PRV-1 expression were highly correlated with levels after HU (r2=0.20, P=0.007). Using 1.13 as the cut-off for definition of overexpression, the pre-HU and post-HU PRV-1 expression status was concordant in 72% of cases. The results of PRV-1 expression status largely unaffected by HU therapy have important clinical relevance in differential diagnosis when dealing with MPN patients who already received partial treatment.
PRV-1 expression level is a useful predictor of risk of thrombosis, highly correlated with JAK2 V617F mutant levels and may serve as a surrogate marker for classification and prognostication of MPN.
No relevant conflicts of interest to declare.
Supported by grants NSC96–2314-B-182-003, CMRPG330303 and DOH100-TD-C-111-006.
Asterisk with author names denotes non-ASH members.