Diagnostic Performance of Erythropoietin Levels in Polycythemia Vera: Experience at a Comprehensive Cancer Center

Introduction: Considering the evolving diagnostic criteria of polycythemia vera (PV), we analyzed the utility of serum erythropoietin (EPO) as a predictive marker for differentiating polycythemia vera (PV) from other etiologies of erythrocytosis. Patients and Methods: We conducted a retrospective study after a review of electronical medical records from January 2005 to December 2016 with diagnosis of erythrocytosis using International Classification of Disease–specific codes. To evaluate the diagnostic performance of EPO levels and JAK2-V617F mutation, we constructed a receiver-operated characteristic curve of sensitivity versus 1-specificity for serum EPO levels and JAK2-V617F mutation as predictive markers for differentiating PV from other causes of erythrocytosis. Results: We surveyed 577 patients with erythrocytosis. Median patient age was 59.2 years, 57.72% (n = 329) were male, 86.3% (n = 491) were white, and only 3.3% (n = 19) were African American. A total of 80.88% (n = 351) of those diagnosed with PV had a JAK2-V617F mutation compared to only 1.47% (n = 2) whose primary diagnosis was secondary polycythemia. When comparing JAK2-V617 mutation to the EPO level, the area under the curve of JAK2-V617 (0.8970) was statistically larger than that of EPO test (0.6765). Therefore, the PV diagnostic methodology using JAK2-V617 is better than the EPO test. An EPO level of < 2 mIU/mL was > 99% specific to predict PV but was only 12% sensitive. Conclusion: In the appropriate clinical setting, cytogenetic and molecular studies such as JAK2 mutation status prevail as the most useful tools for PV case identification. The use of isolated EPO to screen patients with erythrocytosis is not a good diagnostic approach.