Polycythemia Vera Mayo Clin Proc, February 2003, Vol 78174
Mayo Clin Proc. 2003;78:174-194 174 2003 Mayo Foundation for Medical Education and Research
Polycythemia Vera: A Comprehensive Review and Clinical Recommendations
From the Division of Hematology and Internal Medicine, Mayo Clinic,Rochester, Minn.
Address reprint requests and correspondence to Ayalew Tefferi, MD,Division of Hematology, Mayo Clinic, 200 First St SW, Rochester,MN 55905 (e-mail: email@example.com).
AYALEW TEFFERI, MD
More than a century has elapsed since the appearance ofthe modern descriptions of polycythemia vera (PV). Dur-ing this time, much has been learned regarding diseasepathogenesis and PV-associated molecular aberrations.New information has allowed amendments to traditionaldiagnostic criteria. Phlebotomy remains the cornerstonetreatment of PV, whereas myelosuppressive agents mayaugment the benefit of using phlebotomy for thrombosisprevention in high-risk patients. Excessive aspirin use iscontraindicated in PV, although the use of lower-dose aspi-rin has been shown to be safe and effective in alleviatingmicrovascular symptoms including erythromelalgia andheadaches. Recent studies have shown the utility of selec-tive serotonin receptor antagonists for treating PV-associ-ated pruritus. Nevertheless, many questions remain unan-swered. What is the specific genetic mutation or alteredmolecular pathway that is causally related to the disease?In the absence of a specific molecular marker, how is aworking diagnosis of PV made? What evidence supportscurrent practice in the management of PV? This articlesummarizes both old and new information on PV; pro-
The clinical phenotype of polycythemia vera (PV)(plethora, engorged veins) was appreciated long be-fore the disease was formally described by Vaquez1 in 1892and subsequently by Osler2 in 1903.3-7 By 1910, it wasevident that erythrocytosis in PV was often associated withleukocytosis, thrombocytosis, and panmyeloid hyperplasiaof the bone marrow.8-10 The development of myelofibrosisand acute leukemia as part of the natural history of PV wasfirst reported in 1935 and 1938, respectively.11,12
In 1951, Dameshek13 classified PV as a chronic myelo-proliferative disorder (CMPD) along with other relatedmyeloid disorders, including chronic myeloid leukemia(CML) and agnogenic myeloid metaplasia (AMM), be-cause of similarities in both clinical and laboratory fea-tures. Between 1967 and 1981, Fialkow et al14-21 showedthat CMPDs are biologically interrelated on the basis ofbeing clonal stem cell disorders with involvement of bothmyeloid and lymphoid lineage.
Phlebotomy used as treatment of PV was recommendedby Osler22 in the first decade of the 20th century. During
ACE = angiotensin-converting enzyme; ADP = adenosine di-phosphate; AMM = agnogenic myeloid metaplasia; CML =chronic myeloid leukemia; CMPD = chronic myeloprolifera-tive disorder; COPD = chronic obstructive pulmonary dis-ease; EORTC = European Organisation for Research andTreatment of Cancer; EPO = erythropoietin; EPOR = EPOreceptor; ET = essential thrombocythemia; GP = glycopro-tein; IFN- = interferon ; IGF-1 = insulin-like growth factor1; MMM = myelofibrosis with myeloid metaplasia; 32P = radio-active phosphorus 32; PRTE = postrenal transplant erythro-cytosis; PRV-1 = polycythemia rubra vera-1; PTP = proteintyrosine phosphatase; PV = polycythemia vera; PVSG = Poly-cythemia Vera Study Group; RCM = red blood cell mass;SH2 = Src (family of tyrosine kinases) homology 2; SP =secondary polycythemia; STAT = signal transducer and acti-vator of transcription; TPO = thrombopoietin
poses a modern diagnostic algorithm to formulate a work-ing diagnosis; and provides recommendations for patientmanagement, relying whenever possible on an evidence-based approach.
Mayo Clin Proc. 2003;78:174-194
this early period, external beam irradiation of the spleen,long bones, and vertebrae also was being used in the treat-ment of PV23 until the introduction of intravenous therapywith radioactive phosphorus 32 (32P) by Lawrence24 in1938. By the early 1950s, alkylating agents and antime-tabolites were introduced to the therapeutic armamen-tarium of PV, and the additional benefit of these agents inreducing thrombotic complications was touted.25 However,questions were raised and needed to be resolved about thevalidity of such claims as well as the possible leukemoge-nicity of both 32P and chemotherapeutic agents.26 This andother factors led to the formation of an international Poly-cythemia Vera Study Group (PVSG) in 1967, under theauspices of the National Cancer Institute with Dr Louis R.Wasserman as principal investigator.27 The objectives ofthe PVSG were to describe the natural history of PV andto define optimal therapy through the institution of long-term therapeutic trials. It is reasonable to consider theperiod that started with the PVSG studies as the modern erain PV.
CURRENT DISEASE CLASSIFICATIONDiscordant with the traditional scheme proposed byDameshek,13 current classification systems consider CMLseparate from the other CMPDs because of its specific
Mayo Clin Proc, February 2003, Vol 78 Polycythemia Vera 175
association with the Philadelphia chromosome transloca-tion (bcr-abl tyrosine kinase)28 as well as its unique treat-ment response to both interferon (IFN-)29 and imatinibmesylate.30 Therefore, the term chronic myeloproliferativedisorder is currently reserved for only PV, essential throm-bocythemia (ET), and myelofibrosis with myeloid meta-plasia (MMM). However, the CMPDs are joined by CMLand myelodysplastic syndrome as members of a broadercategory of chronic myeloid disorders that also includes afourth category that groups unclassified myeloprolifera-tive/myelodysplastic disorders, chronic myelomonocyticleukemia, juvenile myelomonocytic leukemia, hyper-eosinophilic syndrome, systemic mast cell disease, andchronic neutrophilic leukemia under an operational desig-nation of atypical chronic myeloid disorders (Figure 1).31
All 3 CMPDs, including PV, are characterized by vari-able degrees of bone marrow hypercellularity and atypicalmegakaryocytic hyperplasia and clustering.32 In addition,all 3 disorders may or may not display splenomegaly,leukocytosis, thrombocytosis, and clonal cytogenetic ab-normalities. The clinical distinction among the CMPDs is
Figure 1. Operational classification of the chronic myeloid disorders. *Atypical chronicmyeloid disorders include chronic neutrophilic leukemia, chronic eosinophilic leukemia/hypereosinophilic syndrome, systemic mast cell disease, juvenile myelomonocytic leuke-mia, chronic myelomonocytic leukemia, and a chronic myeloid process that displaysoverlapping features of both myelodysplastic syndrome and chronic myeloproliferativedisorder. Reprinted with permission from Tefferi.31 Copyright 2000 Massachusetts Medi-cal Society. All rights reserved.
made by the demonstration of clonal erythrocytosis inPV,26 substantial bone marrow fibrosis in AMM,31 andclonal thrombocytosis that is not associated with eithererythrocytosis or high-grade myelofibrosis in ET.33,34
Note that currently there is no specific disease marker,molecular or otherwise, for PV or the other CMPDs andthat a working diagnosis is made on the basis of theconstellation of clinical and bone marrow histologicalfindings.
EPIDEMIOLOGYPopulation-based epidemiological studies done in Roches-ter, Minn, have suggested a stable incidence trend for PV ofapproximately 2.3/100,000.35,36 Other studies have reportedeither similar37,38 or lower39-43 incidence figures. A higherdisease incidence has been suggested in persons of Jewishancestry44-48 and among parent-offspring pairs.49,50 Medianage at diagnosis of PV is approximately 60 years with aslight (1.2:1) male preponderance.51 Approximately 7% ofpatients are diagnosed before age 40 years,51 and childrenare rarely diagnosed with PV.52-54
Polycythemia Vera Mayo Clin Proc, February 2003, Vol 78176
No strong evidence supports disease association withenvironmental exposure,55 although an excess risk has beensuggested in embalmers and funeral directors,56 as well asin persons exposed to benzene,57 petroleum refineries,58
and low doses of radiation.59 In contrast to patients withacute leukemias and CML,60 a high risk of developing PVin atomic-bomb survivors of Hiroshima and Nagasaki hasnot been shown.
PATHOGENETIC MECHANISMSOrigin of the PV Clone
It is now well established that PV is a clonal stem celldisease with trilineage myeloid involvement.16,61 However,some studies have suggested clonal heterogeneity includ-ing clonal involvement of B lymphocytes62 and polyclonalgranulopoiesis in certain patients.63-66 Also, normal he-matopoietic stem cells are known to coexist with clonalstem cells in PV and may gain growth advantage underexperimental conditions.67,68 Regardless, unlike with CML,the clonogenic molecular lesion in PV remains elusive.Recall that cytogenetic studies have played a major role indeciphering the causal genetic mutation in CML.28,69-73
However, karyotypic abnormalities in PV are infrequent(13%-18% in chemotherapy-nave patients) and nonspe-cific, and they include trisomies of chromosomes 9 and 8and deletions of the long arms of chromosomes 13 and20.74-76 Not surprisingly, these chromosomes were targetedfor further studies that showed cryptic interstitial deletionson chromosome 20q,66 9p chromosomal gains that are notapparent by conventional techniques,77,78 and a high inci-dence of heterozygosity loss involving chromosome 9p.79
Whether further inquiry into these preliminary observa-tions leads to specific pathogenetic information remains tobe seen.
Growth Factor Independence and Hypersensitivityof Erythroid Progenitor Cells
A frequent area of pathogenetic investigation in PVinvolves erythropoi