Plasma cytokines in polycythemia vera: Phenotypic correlates, prognostic relevance, and comparison with myelofibrosis


  • Conflict of interest: Nothing to report.


Plasma cytokine milieu is abnormal in primary myelofibrosis (PMF) and correlates with disease phenotype and prognosis. In this study, we show that several plasma cytokines are also abnormally expressed in polycythemia vera (PV; n = 65), compared to normal controls (n = 35), but with a significantly different pattern than that of PMF (n = 127). Direct phenotypic correlation in PV included levels of IL-12 with hematocrit; IL-1b, IL-2, IL-7, FGF-b, and HGF with leukocytosis; and IFN-α and IFN-γ with thrombocytosis. In univariate analysis, levels of 13 cytokines (out of 30 analyzed) correlated with survival but only MIP-1β remained significant on multivariable analysis that included the other cytokines as covariates. Increased level of MIP-1β (P < 0.01), older age (P < 0.01), and leukocytosis (P = 0.03) maintained their association with shortened survival, on multivariable analysis. This study provides preliminary observations that warrant a larger scale study and suggests the value of plasma cytokines as prognostic biomarkers in PV. Am. J. Hematol., 2012. © 2012 Wiley Periodicals, Inc.


Aberrant cytokine expression is hypothesized to contribute to the disease phenotype in myeloproliferative neoplasms (MPN) and the associated bone marrow and spleen stromal changes [1]. We have recently described altered plasma cytokine levels in primary myelofibrosis (PMF) [2] and myelodysplastic syndromes (MDS) [3] and demonstrated their phenotypic correlates and prognostic relevance. In PMF, increased levels of interleukin (IL)-8 or IL-2R were associated with presence of constitutional symptoms, transfusion need, and leucocytosis, as well as inferior overall and leukemia-free survival. The latter effect was independent of risk stratification according to the dynamic international prognostic scoring system (DIPSS)-plus [4]. In MDS, increased levels of C-X-C motif chemokine (CXCL)10, IL-7, and IL-6 correlated with inferior overall survival [3], independent of other prognostic factors, including risk stratification according to international prognostic scoring system (IPSS) [5]. Prognostic relevance of plasma cytokines has also been demonstrated in acute myeloid leukemia (AML) [6].

Furthermore, cytokine levels in myelofibrosis have been shown to predict response to treatment with pomalidomide [7], and JAK inhibitor therapy-induced down regulation of pro-inflammatory cytokines has been correlated with response in constitutional symptoms [8]. Taken together, these observations suggest an important role for cytokines in the pathogenesis, disease phenotype, prognostication, and monitoring of treatment response in MPN. In this study, we examined the plasma cytokine profile in polycythemia vera (PV), when compared with both normal controls and patients with PMF, and its correlation with clinical features and survival.


This study was approved by the Mayo Clinic institutional review board. All patients provided written informed consent for study sample collection as well as permission for use in research. Inclusion in this study required availability of archived plasma at the time of first referral to the Mayo Clinic, Rochester, MN, and at a time point that is within 1 year of initial diagnosis. The diagnosis of PV was strictly according to the 2008 WHO diagnostic criteria [9]. Conventional criteria were used for diagnosis of post-PV leukemic transformation (AML) and post-PV fibrotic transformation [10]. Survival data was obtained via review of patient charts, social security death index or a telephone call to the patient in July 2011.

Peripheral blood was collected under a Mayo Clinic protocol for patients with myeloid malignancies, and standard procedures were followed to centrifuge samples at 4°C and store aliquots at −80°C. Concentrations of 30 plasma cytokines/chemokines were analyzed in duplicates by using multiplex bead based Luminex technology (Invitrogen, Carlsbad, CA): interleukin-1β (IL- 1β), IL-1RA, IL-2, IL-2R, IL-4, IL-5, IL-6, IL-7, IL-8, IL-10, IL-12, IL-13, IL-15, IL-17, epidermal growth factor (EGF), eotaxin, fibroblast growth factor-basic (FGF-b), granulocyte-macrophage colony-stimulating factor (GM-CSF), granulocyte colony-stimulating factor (G-CSF), hepatocyte growth factor (HGF), interferon alfa (IFN-α), IFN-γ, IFN-γ-inducible protein 10 (IP-10), monocyte chemotactic protein-1 (MCP-1), monokine induced by IFN-γ (MIG), macrophage inflammatory protein 1α (MIP-1α), MIP-1β, regulated on activation normally T-cell expressed and secreted (RANTES), tumor necrosis factor α (TNF-α), and vascular endothelial growth factor (VEGF). Measurements were performed on a Luminex 200 analyzer (Luminex, Austin, TX), and the resulting data were evaluated by using Star Station Software Version 2.3 (Luminex).

All statistical analyses considered clinical and laboratory parameters obtained at time of first referral to the Mayo Clinic, which coincided in all instances with time of plasma collection for cytokine analysis and a period of less than 1 year from initial diagnosis. Differences in the distribution of continuous variables between categories were analyzed by the Mann-Whitney test (for comparison of two groups). Overall survival analysis was considered from the date of first referral to the Mayo Clinic (i.e., date of plasma collection) to date of death (uncensored) or last contact (censored). Overall survival curves were prepared by the Kaplan-Meier method and compared by the log-rank test. A Cox proportional hazards regression model was used for multivariable analysis. P values <0.05 were considered significant. The Stat View statistical package (SAS Institute, Cary, NC) was used for all calculations.

Results and Discussion

Clinical characteristics of the study cohort (n = 65) at time of referral are outlined in Table I. Median age was 61 years and 36 (55%) patients were males. After a median follow-up of 6 years, 15 deaths (23%), 3 (5%) leukemic transformations, and 10 (15%) fibrotic progressions were documented.

Table I. Clinical Characteristics of 65 Patients with Polycythemia Vera with Cytokine Profiling Available Within One Year of Diagnosis
Clinical characteristicsN = 65Cytokines with significant correlation on univariate analysis (P < 0.05)
  • a

    Associated with decreased cytokine levels as opposed to increased levels.

  • FGF-b, fibroblast growth factor-basic; GM-CSF, granulocyte macrophage colony-stimulating factor; HGF, hepatocyte growth factor; IFN, interferon; IL, interleukin; IP-10, IFN-γ–inducible protein 10; MCP-1, monocyte chemotactic protein-1; RANTES, regulated on activation normally T-cell expressed and secreted; TNF-α, tumor necrosis factor α.

Age in years; median (range)61 (25–90)Eotaxin
Males; n (%)36 (55)
Hemoglobin, g/dL; median (range)18.5 (15.7–23)IL-6a, FGF-ba
Hematocrit; median (range)56 (46–71)IL-6a, IL-12
Leucocyte count, × 109/L; median (range)11.4 (4.3–26.5)IL-2R
Leukocytosis (>10.5 × 109/L); n (%)38 (58)IL-1β, IL-2, IL-7, FGF-b, HGF
Platelet count, × 109/L; median (range)408 (44–1703)
Thrombocytosis (≥450 × 109/L); n (%)28 (43)IFN-α, IFN-γ, GM-CSFa
Palpable spleen; n (%) (N = 63)24 (38)Eotaxina, MCP-1a
Pruritus; n (%) (N = 57)23 (40)
Vasomotor symptoms; n (%) (N = 57)22 (39)IL-4, FGF-ba, HGFa, RANTESa
Arterial thrombosis before/at diagnosis; n (%)12 (18.5)IL-1RAa, FGF-ba
Venous thrombosis before/at diagnosis; n (%)1 (1.5)
↑ Lactate dehydrogenase; n (%) (N = 36)18 (50)IL-12
Leukoerythroblastic smear; n (%) (N = 58)9 (16)
JAK2 mutation; n (%) (N = 53)52 (98)
Abnormal karyotype; n (%) (N = 47)5 (11)IL-2, IL-4, 1L-7, 1L-13, 1L-17, GM-CSF, MIP-1α
Overall survival; median (months)77.8 (1.3–367)IL-1β, IL-4, EGF, IL-5, IL-7, IL-10, IL-17, TNF-α, IFN-α, GM-CSF, MIP-1α, MIP-1β, MCP-1
Fibrotic transformation; n (%)10 (15)IL-1β, IL-5, IL-6, IL-10, IL-12, IL-15, IL-17, IP-10, eotaxina

Correlation with phenotype

In univariate analysis, cytokine excess was significantly associated with the following: eotaxin with older age; IL-12 with elevated hematocrit and LDH; IL-1β/IL-2/IL-7/FGF-b/HGF with leucocytosis; IFN-α/IFN-γ with thrombocytosis; and IL-4 with microvascular symptoms (Table I). Lower levels of certain cytokines were associated with following: IL-6/FGF-b with hemoglobin level; GM-CSF with thrombocytosis; eotaxin/MCP-1 with splenomegaly; FGF-b/HGF/RANTES with microvascular symptoms; and IL-1RA/FGF-b with arterial thrombosis.

Comparison between PV (n = 65), normal controls (n = 35), and PMF (n = 127)

Compared to normal controls, patients with PV demonstrated significantly higher levels of IL-1RA, IL-5, IL-6, IL-7, IL-8, IL-12, IL-13, IFN-γ, GM-CSF, MIP-1α, MIP-1β, HGF, IP-10, MIG, MCP-1, and VEGF. Levels of EGF and RANTES were significantly lower in PV compared to normal controls (Table II). Levels of the following cytokines were significantly higher in PMF compared to PV: IL-1β, IL-1RA, IL-2R, EGF, IL-10, FGF-b, IL-12, IFN-α, and RANTES (Table II). Levels of the following cytokines were significantly higher in PV compared to PMF (P < 0.05): IL-7, IFN-γ, GM-CSF, MIP-1α, IP-10, MIG, eotaxin, and VEGF.

Table II. Cytokine Levels in 65 Patients with Polycythemia Vera Within One Year of Diagnosis and Their Comparison with 35 Normal Controls and 127 Patients with Primary Myelofibrosis
CytokinesNormal N = 35PV N = 65P-valueaPMF N = 127P-valueb
Median (pg/mL)Range (pg/mL)Median (pg/mL)Range (pg/mL)% with >3SDMedian (pg/mL)Range (pg/mL)% with >3SD
  • a

    Mann–Whitney comparison between polycythemia vera and normal controls.

  • b

    Mann–Whitney comparison between polycythemia vera and primary myelofibrosis.


Prognostic relevance of cytokine levels in PV

In univariate analysis, elevated levels of 13 cytokines showed significant association with inferior overall survival: IL-1β (P = 0.003), IL-4 (P = 0.01), EGF (P = 0.006), IL-5 (P = 0.003), IL-7 (P = 0.004), IL-10 (P = 0.002), IL-17 (P = 0.002), TNF-α (P = 0.004), IFN-α (P = 0.0007), GM-CSF (P = 0.004), MIP-1α (P = 0.0008), MIP-1β (P = 0.0002), and MCP-1 (P = 0.005). On multivariable analysis of cytokines, only MIP-1β remained significant. When age and leucocytosis were added as covariates, MIP-1β remained significant (P = 0.0005), as did both age (P = 0.002) and leucocytosis (P = 0.03). In univariate analysis, elevated levels of the following cytokines were significantly associated with fibrotic transformation: IL-1β, IL-5, IL-6, IL-10, IL-12, IL-15, IL-17, and IP-10. Cytokine association with leukemic transformation could not be reliably assessed since only three patients developed leukemic transformation in our cohort.

This study suggests that the cytokine milieu in PV is significantly altered, in a pattern that is different from PMF, and identifies MIP-1β as a potentially important and prognostically independent biomarker. MIP-1β, also known as CCL4, belongs to a class of chemokine subfamily that is secreted by mononuclear cells and serves as a chemoattractant for migration of specific lymphocyte subsets to areas of inflammation [11]. Its effects are closely related to MIP-1α (CCL3) and both are known to play a key role in clonal B-cell expansion via B-cell receptor signaling [12]. Prognostic relevance of both MIP-1α and MIP-1β has been previously demonstrated in chronic lymphocytic leukemia, where elevated levels both were independently associated with inferior survival [13, 14]. Little is known about the role of these chemokines in myeloid malignancies; in a study that included 57 patients with MDS, higher levels of CCL4 were found in the sub-group with high-risk MDS compared to low-risk MDS. Additional studies with larger sample sizes will be necessary to confirm our findings and also decipher the biological role of these cytokines in MPN and their utility as drug targets.

Author Contributions

Rakhee Vaidya collected data, participated in data analysis and the writing of the paper; Naseema Gangat participated in data analysis; Thitina Jimma, Christy M. Finke, and Terra L. Lasho participated in data collection; Animesh Pardanani contributed patients; Ayalew Tefferi designed the study, contributed patients, collected data, performed the statistical analysis, and wrote the paper. All authors approved the final draft of the paper.