The JAK2V617F and exon 12 mutations are critical for the diagnosis of polycythemia vera (PV) and JAK2V617F is the molecular marker of 50-60% of patients (pts) with essential thrombocythemia (ET) [Arber, DA et al. Blood. 2016;127:2391-2405]. Distinguishing these illnesses in their early stage remains a clinical problem because of this shared mutation and phenotypic characteristics [Silver, RT et al. Blood. 2013;122:1881-1886][Ancochea, A et al. Br J Haematol. 2014;167:411-417]. When the hematocrit (HCT), red blood cell (RBC) count, or hemoglobin (HGB) are increased, we distinguish PV from JAK2V617F -mutated( JAK2V617F +) ET using chromium-51 (Cr-51) labeled RBC to measure red cell mass (RCM) and iodine-135 (I135) to determine plasma volume. To define polycythemia, the World Health Organization (WHO) 2016 criteria used HCT and HGB values derived from a retrospective study which have never been confirmed [Arber].

Materials and Methods

We diagnosed PV by establishing JAK2V617F positivity, a Cr-51 measured RCM>125% of expected volume, and a marrow biopsy consistent with PV [Silver]. Utilizing a receiver operating characteristic (ROC) analysis [IBM SPSS, 23.0], we determined threshold values for HCT, HGB, and RBC to distinguish 69 PV pts from 31 JAK2V617F + ET pts who presented with "high-normal" HCT or HGB values.


Relevant demographic and hematologic data are shown in Table 1. The Cr-51 determined RCM of all 69 PV pts was >125%, establishing polycythemia [Silver], and was measured in 28/31 JAK2V617F + ET pts; it was normal in all (mean=95.6 ±13.1%).

ROC analyses were performed to determine threshold HCT, HGB, and RBC values to distinguish PV from JAK2 V617F + ET pts. For men and women, respectively, the HCT threshold values were 49.3 and 45.8%, for HGB 16.8 and 15.2 g/dL, and for RBC 5.79 and 5.125 x106/uL . These values and those proposed by WHO 2016 are shown in Table 2. The area under the curve (AUC) for the Cr-51 RCM was 1, indicating 100.0% diagnostic specificity (spec) and sensitivity (sens) for distinguishing PV from JAK2 V617F + ET (AUC measures how well groups are distinguished and can range from 0 to 1, with 1 representing a perfect test). Figure 1 shows ROC curves representing the spec and sens of using HCT to distinguish PV from JAK2V617F+ ET. For men, the AUC for HCT was 0.846 (spec=100.0%, sens=64.9%), for HGB 0.753 (spec=100.0%, sens=55.6%), and for RBC 0.761 (spec=100.0%, sens=77.4%). For women, the AUC for HCT was 0.974 (spec=95.7%, sens=78.1%), for HGB 0.875 (spec=95.5%, sens=77.4%), and for RBC 0.924 (spec=90.9%, sens=88.0%). Blood values were less sensitive in men than in women, whereas the specificities were consistent (Table 2). Distribution curves plotted for HCT, HGB, and RBC values are shown in Figure 2. For pts with PV or JAK2V617F+ ET, the HCT values overlap in 40.8% of men and 18.3% of women. The HGB values overlap in 49.0% of men and 53.6% of women, and the RBC values overlap in 53.3% of men and 50.0% of women.


The Cr-51 RCM was 100% specific and sensitive for distinguishing PV from JAK2V617F+ ET. Our HCT and HGB values derived at the time of the diagnostic Cr-51 RCM from PV pts are concordant with those of WHO 2016 for diagnosing PV [Arber]. However, there is substantial overlap in HCT values occurring in 40.8% of men and 18.3% of women, and in HGB values occurring in 49.0% of men and 53.6% of women, which indicates low sensitivity for distinguishing PV from JAK2V617F+ ET. In the absence of performing a Cr-51 RCM, a marrow biopsy is mandatory for distinguishing them. It has been suggested that JAK2V617F + ET pts are at higher risk for thrombosis than calreticulin-mutated ET [Rumi, E et al. Blood. 123;10:1544-1551]. Since median HCT values were not reported and since neither isotope studies nor systematic marrow biopsies were performed, some alleged JAK2V617F+ ET pts who had a higher risk of thrombotic events may actually have had PV.


Of importance, this is the first confirmation of WHO 2016 values for the diagnosis of PV. However, these values do not provide unequivocal separation of PV from JAK2V617F+ ET. In these cases, in the absence of a Cr-51 RCM, a marrow biopsy is mandatory because its distinction has serious prognostic and therapeutic implications.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.

Sign in via your Institution