Jerry L. Spivak
POLYCYTHEMIA VERA
INTRODUCTION
Polycythemia vera (PV) is a clonal disorder of a multipotent hematopoietic stem cell in which overproduction of morphologically normal red cells, white cells, and platelets occurs in the absence of an apparent cause. The commonest of the chronic myeloproliferative disorders, PV, is uncommon, occurring at an average frequency of 2/100,000, but with increasing age, rates as high as 18/100,000 have been observed. Females predominate, particularly below age 40 years.
PATHOGENESIS
The etiology of PV is unknown. Abnormalities of chromosomes 1, 8, 9, 13, and 20 have been identified in up to 30% of PV patients, but they are neither specific for the disorder nor necessary for its pathogenesis; in many instances they appear to occur as secondary events and their expression can be enhanced by exposure to chemotherapeutic agents (1). Erythropoietin-independent in vitro erythroid colony formation is a characteristic feature of PV, although not specific for it, since this behavior has also been observed in primary myelofibrosis (PMF) and essential thrombocytosis. Constitutive activation of JAK2 (2), which is the cognate tyrosine kinase for type 1 hematopoietic growth factor receptors such as the erythropoietin, thrombopoietin, and granulocyte colony-stimulating factor receptors, has been identified to be the molecular basis for such growth factor independence in PV and its companion myeloproliferative disorders PMF and essential thrombocytosis.
The mechanism for constitutive JAK2 activation in the chronic myeloproliferative disorders is an acquired point mutation in the autoinhibitory JH2 domain of the JAK2 gene, replacing valine with phenylalanine (V617F). JAK2 is located on the short arm of chromosome 9 and loss of heterozygosity for 9p is a common cytogenetic lesion in PV, leading to homozygosity for the JAK2 V617F mutation; in some patients, there is also reduplication of chromosome 9. In PV, approximately 90% of patients express the JAK2 V617F mutation, of which approximately 35% are homozygous for it. Approximately 5% have a JAK2 exon 12 activating mutation. No clinical differences have been identified between heterozygotes and those homozygous for JAK2 V617F, nor are there any clinical differences between PV patients expressing JAK2 V617F and those who do not. Thus, although the JAK2 V617F mutation provides an explanation for the hematopoietic growth factor independence of PV hematopoietic cells in vitro, their apoptosis resistance and their uncontrolled growth in vivo, the absence of the mutation in some patients with classical PV, and its expression in PMF and essential thrombocytosis patients strongly suggest that other as yet unidentified molecular lesions are involved in the pathogenesis of these disorders.
CLINICAL FEATURES
PV is extremely variable in its presenting manifestations as well as its clinical features, which also change over the course of the disorder. Because its onset can be insidious, an abnormal blood count is often the first sign of the disease. In approximately 40% of patients, there will be an increase in red cells, white cells, and platelets. In approximately 15% of patients, erythrocytosis will be the sole presenting manifestation. In approximately 5%–10% of patients, an elevated platelet count may be the first manifestation of the disease, while in the rest, erythrocytosis and thrombocytosis or leukocytosis are the presenting blood abnormalities. Extramedullary hematopoiesis as manifested by palpable splenomegaly occurs in approximately 40% of patients at the time of diagnosis; rarely, myelofibrosis can be the initial manifestation of PV with erythrocytosis becoming evident later on. Since PV is a hypercoagulable state, arterial or venous thrombosis may also be the first manifestation of the disease. Classically, in young women, the thrombosis most commonly involves the hepatic veins, often as the presenting manifestation and often with an apparently normal hematocrit due to concomitant plasma volume expansion. Pruritus, usually aquagenic, is also not uncommon as a presenting manifestation, but PV is often not initially recognized as its cause. Erythromelalgia, in which the extremities become warm, red, and painful; migraine headaches; or other neurologic disturbances such as vertigo or visual disturbances are also characteristic symptoms that indicate the presence of an elevated red cell mass or thrombocytosis.
LABORATORY ABNORMALITIES
In addition to increases in the red cell, granulocyte, and platelet counts, the MCV can be low if red cell mass expansion or gastrointestinal blood loss depletes body iron stores. An elevated leukocyte alkaline phosphatase and serum vitamin B12 and vitamin B12 binding capacity due to increased release of granulocyte transcobalamin III reflect neutrophil activation, presumably due to JAK2 V617F, which is also responsible for the increased expression of granulocyte PRV-1 mRNA (CD177). When the platelet or leukocyte counts are elevated, spurious hyperkalemia may be observed, as can hypoglycemia and a low pO2 if blood samples are not collected on ice and in the presence of sodium azide. Elevation of the serum alkaline phosphatase occurs with extramedullary hematopoiesis and becomes more marked after splenectomy.
Abnormalities of coagulation in PV are largely limited to platelet function. These include defective platelet aggregation to ADP, epinephrine, or collagen alone or in combination and loss of alpha granules and dense bodies. When the platelet count exceeds 1,000,000/ml, higher molecular weight von Willebrand multimers will be absorbed by the platelets and degraded, leading to a reduction in ristocetin cofactor activity and an acquired form of von Willebrand’s disease, although spontaneous bleeding due to this is uncommon.
DIAGNOSIS
Elevation of the red cell mass is the sine qua non of PV, the only feature that distinguishes it from its companion myeloproliferative disorders, PMF and essential thrombocytosis, and the feature of the disease that is responsible for its most frequent serious consequences, thrombosis and hemorrhage. Unfortunately, erythrocytosis is not unique to PV, and in recent years the very means for identifying the presence of erythrocytosis, direct determination of the red cell mass by isotope dilution, has become unavailable in many medical centers. Attempts to resolve this issue by the use of surrogate markers for direct red cell mass determination have not proved to be useful (3). For example, specific hematocrit or hemoglobin levels are woefully inadequate as indicators of the red cell mass unless the hematocrit is >60% (hemoglobin >20 g/dl) in a man, or >52% in a woman (hemoglobin >17 g/dl). The reasons for this are a consequence of blood rheology and the unique pathophysiology of PV with respect to blood volume regulation.
For example, when erythrocytosis occurs as a consequence of hypoxia, there is a simultaneous reduction in the plasma volume as the body attempts to maintain a normal total blood volume. This contributes to the observed increase in hematocrit. In PV, however, particularly in women, as the red cell mass rises, the plasma volume either rises or fails to decrease. Furthermore, with splenomegaly, there is a compensatory increase in plasma volume. Both of these situations lead to hematocrit values that are spuriously low with respect to the actual red cell mass (4). As a corollary, a decrease in the plasma volume alone can lead to a falsely elevated hematocrit, when in fact the red cell mass is normal.
The recent discovery of the JAK2 V617F mutation has greatly simplified the evaluation of a high hematocrit and the diagnosis of PV. This is because first, benign disorders causing erythrocytosis are more common than PV (Table 26-1) and second, because surrogate markers for the latter lack sensitivity and specificity. For example, while the serum erythropoietin level is lower in PV than in other disorders causing erythrocytosis, the serum erythropoietin level can also be normal in PV as well as in secondary forms of erythrocytosis. Similarly, the bone marrow examination can be normal in PV or even mimic that of PMF or essential thrombocytosis. Cytogenetic abnormalities are present in only 30% of PV patients and are not pathognomonic for the disease, while other markers such as elevation of the leukocyte alkaline phosphatase and endogenous erythroid colony formation are merely consequences of the constitutively active JAK2.
TABLE 26-1 CAUSES OF ERYTHROCYTOSIS
Figure 26-1 illustrates an algorithm for the evaluation of the patient with a high hematocrit. When red cell mass and plasma volume determinations are not available, it is reasonable to start with an assay for JAK2 V617F with the knowledge that a positive assay only indicates the presence of a myeloproliferative disorder, while a negative assay does not exclude such a disorder. In the absence of a red cell mass determination, a positive JAK2 V617F assay in a patient with a high hematocrit obligates the physician to phlebotomize the patient to the normal hematocrit for gender as discussed below.
FIGURE 26-1 Algorithm for the diagnosis of polycythemia vera. The first requirement is to establish the basis for an elevated hemoglobin or hematocrit. If it is determined that there is an elevated red cell mass, an assay for JAK2 V617F will establish the diagnosis in over 90% of patients with polycythemia vera. A negative JAK2 V617F assay does not, however, exclude a myeloproliferative etiology and in the absence of splenomegaly, leukocytosis, or thrombocytosis; further studies will be required.
NATURAL HISTORY
Most classical hematology textbooks suggest that the natural history of PV follows an inevitable course from erythrocytosis through myelofibrosis and myeloid metaplasia to acute leukemia if the patient does not die first from some other complication or comorbidity. This depiction ignores the clinical heterogeneity of the disease, its modification by improved therapies, and the earlier stages at which PV is now usually recognized. In this regard, prognosis does not appear to be influenced by the presence or the absence of JAK2 V617F or whether this mutation is expressed homozygously or heterozygously.
The complications of PV are listed in Table 26-2. Erythrocytosis, not thrombocytosis, is responsible for the major thrombotic complications of PV; minor transient thrombotic or ischemic complications such as erythromelalgia, ocular migraine, or digital infarction do involve the platelets but are exacerbated by erythrocytosis, which promotes platelet activation, platelet- leukocyte interactions, as well as endothelial cell activation and damage, all of which enhance thrombogenesis. Erythrocytosis can also cause hypertension, splenomegaly, and exacerbate aquagenic pruritus. Acid-peptic disease leading to gastrointestinal hemorrhage and iron deficiency occur at a higher frequency in PV patients than in the general population; the roles of vascular stasis, excess histamine, or other cytokine production are unknown, but the frequency of Helicobacter infection is increased in PV.
TABLE 26-2 THE COMPLICATIONS OF POLYCYTHEMIA VERA
Over time, there will be a gradual increase in the leukocyte and platelet counts, but the leukocytosis is not usually progressive unless there is disease acceleration, while asymptomatic thrombocytosis requires no therapy. The development of excessive extramedullary hematopoiesis with massive splenomegaly and hepatomegaly is a serious complication of PV, occurring in about 10%–15% of patients. These patients are at a higher risk of subsequent leukemic transformation (5).
Splenomegaly can lead to mechanical discomfort, easy satiety, portal hypertension, and cachexia. Marrow fibrosis is another expected event in the natural history of PV. It is essential to distinguish between the development of increased marrow reticulin as a consequence of marrow cell hyperplasia and the hematopoietic stem cell disorder, PMF. There is no evidence that myelofibrosis in PV represents a bad prognostic sign or that it impairs marrow function in the absence of exposure to agents that damage the bone marrow; it is stem cell failure that is the problem. Rarely, pulmonary hypertension has developed with long-standing disease; in some patients this may be due to extramedullary hematopoiesis, while in others there may be pulmonary fibrosis.
Spontaneous acute leukemia develops in PV at an incidence of approximately 1.5%–2.5%; this usually occurs within the first 8 years of the disease and most commonly in patients older than 60 years. Chemotherapy or radiation-induced acute leukemia occurs at rates as high as 10% when these patients are exposed to 32P or alkylating agents. The role of hydroxyurea as a leukemogen has been a matter of debate, but in one randomized prospective clinical trial (6, 7), hydroxyurea was associated with a 10% incidence of acute leukemia after 10 years; hydroxyurea is also a proven tumor promoter when used in conjunction with 32P or an alkylating agent or with UV light exposure.
TREATMENT
PV is generally an indolent disease in which survival is measured in decades in the majority of patients. Most estimates of disease survival have failed to take into account the toxic forms of therapy that have been generally employed, the inadequate use of phlebotomy, and the later stages at which the disease was previously recognized clinically. Furthermore, it is now apparent that PV is a heterogenous disorder with both indolent and aggressive forms and that aggressive chemotherapy has not improved survival (8). There is currently no curative therapy for PV with the possible exception of allogeneic bone marrow transplantation, a therapy not suitable for the older patients who most commonly develop this disorder (9). Thus, treatment should be tailored to disease manifestations. Unfortunately, in contrast to PMF, prognostic risk stratification according to laboratory features has not yet been possible with the exception that a prior history of thrombosis is an adverse risk factor for recurrent thrombotic events.
Erythrocytosis is the greatest initial threat to health because of the adverse effects of hyperviscosity (thrombosis, hemorrhage, hypertension, headache, and impaired cognitive function). Therefore, the red cell mass should be lowered by phlebotomy to achieve a hematocrit of ≤42% (hemoglobin ≤12 g%) in women and ≤45% (hemoglobin ≤14 g%) in men (8). This can be done quickly in all but the frailest because phlebotomy stimulates rapid plasma volume expansion. Repeated phlebotomies will be necessary to maintain the hematocrit at a normal level and to induce iron deficiency, but once this is achieved, the need for phlebotomy will diminish. Phlebotomy therapy actually improves platelet function, does not contribute significantly to thrombocytosis, and does not lead to myelofibrosis, and it must be remembered that the higher the hematocrit, the greater the extent of tissue damage with thrombosis (8). Pruritus, usually aquagenic, is a distressing symptom in approximately 30% of patients. There is no single effective remedy. Phlebotomy, antihistamines, PUVA light therapy, interferon alpha, and hydroxyurea have all been effective but none uniformly. Hyperuricemia (uric acid >10 mg/dL) responds well to allopurinol.
Platelet-related microvascular complications include migraine, visual auras, transient ischemic attacks, erythromelalgia, and digital infarction. Aspirin is a specific remedy for erythromelalgia but with migraine, it may be necessary to lower the platelet count as well to achieve relief using conventional remedies. Symptomatic thrombocytosis causing acquired von Willebrand’s disease will also require platelet count reduction. Asymptomatic thrombocytosis without a significant reduction in ristocetin cofactor activity (<30%) requires no treatment in the absence of a thrombotic risk factor. In this regard, it is important to emphasize that there is no correlation between the platelet count and thrombosis, and no study to date has demonstrated that in the absence of hematocrit control, platelet count reduction prevents arterial or venous thrombosis. Hydroxyurea does appear to be more effective than anagrelide in the prevention of transient ischemic attacks but not venous or arterial thrombosis. The use of prophylactic low dose aspirin therapy is no substitute for adequate control of the red cell mass and has not been demonstrated to have clinical efficacy in asymptomatic PV patients who are adequately phlebotomized.
Control of extramedullary hematopoiesis involving the spleen and liver is the most challenging therapeutic problem in PV but fortunately not one that involves every patient. Interferon alpha, and its pegylated congener in particular, is the drug of choice for this because it lacks the potential for bone marrow damage (10). A recent study demonstrated that durable molecular remissions could be achieved with pegylated interferon (11). Since interferon’s side effects can be significant with chronic use, in the absence of a complete molecular remission, intermittent use is a prudent strategy. In some patients, splenomegaly may be refractory to interferon and chemotherapy, and mechanical discomfort, cachexia, and portal hypertension will demand treatment. If bone marrow transplantation is not an option, the newly approved nonspecific JAK2 inhibitor, ruxolitinib, is the drug of choice in this situation (see the primary myelofibrosis chapter). Low-dose thalidomide is another option worth considering with surgery as the choice of last resort because of the high complication rate associated with it. The postoperative complications of splenectomy include wound dehiscence, hernias, bleeding, portal or mesenteric vein thrombosis, exuberant hepatic extramedullary hematopoiesis, and extreme leukocytosis and thrombocytosis, all of which can be very difficult to control. Splenic irradiation is only a temporary solution and not advisable unless surgery is not an option (12).
PREGNANCY
The opportunity for pregnancy should not be denied to women with PV who have no medical contraindications and prior thrombosis is not one of these. The major threat to a successful outcome is failure to maintain the red cell mass at a safe level. Since there is an expansion of the plasma volume with pregnancy normally, there will be masking of the expanded red cell mass. A normal hematocrit in a pregnant woman is never normal and this is doubly true in PV. It is essential to phlebotomize these patients to a hematocrit of <33% and avoid iron supplements; folic acid supplementation is mandatory. Thrombocytosis and splenomegaly may mandate the use of interferon alpha. Given the elevation of von Willebrand factor that occurs during pregnancy, aspirin therapy may be prudent but this is unproved.
PRIMARY MYELOFIBROSIS
INTRODUCTION
Primary myelofibrosis (PMF) is the least common and most enigmatic of the chronic myeloproliferative disorders. Most frequent after age 60 years, PMF has an incidence of approximately 1/100,000 with male predominance. Previously known as agnogenic myeloid metaplasia, idiopathic myelofibrosis, primary osteomyelofibrosis, or myelofibrosis with myeloid metaplasia, it is important to note that both the first and last appellations actually describe a pathologic process that is not restricted to the disease PMF but can be caused by a variety of benign and malignant processes (Table 26-3). Like its companion myeloproliferative disorders, PV and essential thrombocytosis, PMF is a clonal hematopoietic stem cell disorder, but in contrast to them, it is associated not only with overproduction of blood cells without an obvious cause but also, in many patients, with anemia, leucopenia, or thrombocytopenia.
PATHOGENESIS
The etiology of PMF is unknown. Although irradiation and exposure to organic chemicals such as toluene and benzene can cause marrow fibrosis, no other consistent environmental risk factors have been identified for PMF and familial transmission is rare. Cytogenetic abnormalities occur in more than 50% of patients but generally involve the same chromosomes as in PV and essential thrombocytosis, and none appear to be involved in its pathogenesis. Myelofibrosis is the hallmark of the disorder, but there is good retrospective histologic evidence that a premyelofibrotic phase of the disease exists (13), supporting other evidence that the fibrosis is a consequence of the disease, not its cause.
TABLE 26-3 DISORDERS CAUSING MYELOFIBROSIS
Malignant
Acute leukemia (lymphocytic, myelogenous, megakaryocytic)
Chronic myelogenous leukemia
Hairy cell leukemia
Hodgkin disease
Idiopathic myelofibrosis
Lymphoma
Multiple myeloma
Myelodysplasia
Metastatic carcinoma
Polycythemia vera
Systemic mastocytosis
Nonmalignant
HIV infection
Hyperparathyroidism
Renal osteodystrophy
Systemic lupus erythematosus
Tuberculosis
Vitamin D deficiency
Thorium dioxide exposure
Gray platelet syndrome
Normally, hematopoiesis is extravascular, and hematopoietic progenitor cell proliferation and differentiation in the marrow are supported by accessory cells such as macrophages, adipocytes, reticulum cells, fibroblasts, and endothelial cells, all of which are embedded in an extracellular matrix of collagens. PMF represents the deposition of additional collagen fibrils that are thicker and contiguous. The earliest phase of marrow fibrosis is the deposition of reticulin, which represents collagen fibrils coated with matrix substances such as hyaluronic acid that are argyrophilic and stain with silver. As the quantity of collagen increases relative to matrix substances, the fibrils eventually become reactive with the classical histological collagen stains.
Osteosclerosis in PMF represents only the deposition of minerals on the marrow trabeculae as opposed to the combined osteoblastic and osteoclastic activity that characterizes metabolic bone disease, and is thought to be due to overproduction of osteoprotegerin, which is an osteoclast inhibitor. With advancing myelofibrosis, there is also a reduction in the number of hematopoietic cells in the marrow with the exception of the megakaryocytes. Disease duration, spleen size, and prognosis do not correlate with the degree of myelofibrosis or the type of collagen staining pattern.
The stimulus for the marrow fibrosis and osteosclerosis that are central features of PMF are not entirely understood, but both megakaryocytes and monocytes appear to be involved through the elaboration of the fibrogenic cytokines, particularly TGF-b and thrombopoietin. Animal models of myelofibrosis and osteosclerosis, which have been created by overexpression of thrombopoietin, impaired expression of the hematopoietic transcription factor GATA-1, or transplantation of hematopoietic cells overexpressing JAK2 V617F, further implicate megakaryocytes and other hematopoietic progenitor cells in these processes.
Importantly, a variety of techniques have been employed to definitively demonstrate that the fibroblastic component of PMF is not monoclonal but rather reactive, in contrast to the hematopoietic cells in this disorder, which are clonal and primary. The latter exhibit the hematopoietic growth factor hypersensitivity and growth factor-independent in vitro colony-forming activity that is characteristic of all three chronic myeloproliferative disorders. Bone marrow neoangiogenesis is another characteristic feature of PMF that is thought to be due to increased VEGF production.
CLINICAL FEATURES
As with the other chronic myeloproliferative disorders, PMF may first be recognized during a routine health maintenance evaluation due to abnormal blood counts or a palpable spleen. However, in contrast to the other chronic myeloproliferative disorders, PMF can present with significant constitutional symptoms such as fever, night sweats, anorexia, pruritus, weakness, fatigue, and weight loss. In some patients, particularly men, thrombocytosis alone may be the first manifestation of the disease and, less commonly, isolated leukocytosis. In general, the cases of so-called essential thrombocytosis take approximately 4–7 years to develop the complete myelofibrosis phenotype, but in these patients, bone marrow examination may initially reveal changes inconsistent with the diagnosis of essential thrombocytosis. This situation has been designated as the cellular or premyelofibrotic phase of PMF (13).
Palpable splenomegaly, which can be modest or extreme, is the most common clinical finding, but occasionally patients are encountered before palpable splenomegaly has developed, putting them in a diagnostic limbo. Hepatomegaly is less common and not seen in the absence of splenomegaly. Lymphadenopathy is very uncommon and, when localized, should suggest another diagnosis.
LABORATORY ABNORMALITIES
Any combination of blood count abnormalities can be encountered in PMF. Anemia is the most common abnormality, and a normal hematocrit in a patient with splenomegaly should suggest the presence of PV; indeed, retrospectively approximately 10% of patients in most published series of PMF actually had PV. The anemia is usually normochromic and normocytic and a hemolytic component is rare. Folic acid deficiency, however, can complicate the disorder due to an increased turnover of marrow cells. Leukocytosis is common but not usually to the degree found in chronic myelogenous leukemia. The platelet count is usually normal or elevated, but modest thrombocytopenia can be seen in approximately 25% of patients. Nucleated red blood cells, myelocytes, promyelocytes, and even blast cells may be present in the blood, creating the classical leukoerythroblastic blood picture. Tear drop-shaped red cells reflect the presence of splenomegaly. The leukocyte alkaline phosphatase can be low, normal, or high. With hepatic extramedullary hematopoiesis, the serum alkaline phosphatase will be increased. The JAK2 V617F mutation occurs in approximately 50% of PMF patients and is often homozygous in its expression, but this abnormality does not correlate with disease activity. The more important abnormality, which does correlate with disease phenotype, is clonal dominance by the malignant clone. Platelet function abnormalities in PMF parallel those in the other chronic myeloproliferative disorders, but PMF patients are particularly prone to bleeding (14).
Bone marrow is inaspirable when there is marrow fibrosis, necessitating a biopsy. The bone marrow biopsy in PMF may reveal a hypercellular marrow with myeloid hyperplasia, an increase in large dysplastic megakaryocytes occurring in clusters and, in some patients, widening of the bony trabecula. An increase in collagen and osteoid deposition, sinusoidal dilatation, neoangiogenesis with extramedullary hematopoiesis, and a reduction in cellularity with erythroid islands and megakaryocyte sparing may also be encountered. The reticulin stain will show a dense pattern of argyrophilic fibers in a contiguous pattern with sinusoidal accentuation. If collagen deposition is extensive, the trichrome stain will be positive. It is important to remember, however, that marrow histology is not uniform with respect to biopsy sampling, and marrow histology cannot be relied on to stage this disorder.
RADIOLOGIC ABNORMALITIES
Osteosclerosis but not myelofibrosis can be detected radiologically, most commonly as an increase in medullary bone density in the proximal long bones. Rib and vertebral involvement are also common and even the skull can be affected. The presence of radiologically evident osteosclerosis suggests involvement of at least 40% of the marrow cavity and is related to the extent of the myelofibrosis and splenomegaly but not disease duration or prognosis. Hypertrophic osteoarthropathy with painful periostitis and onion skinning of the tibiae is an uncommon complication of PMF and could be another consequence of the neoangiogenesis that characterizes this disorder.
CYTOGENETIC ABNORMALITIES
Cytogenetic abnormalities are more common in PMF than in its companion myeloproliferative disorders, but they are mostly nonspecific. They include 20q-, 13q-, trisomy 8, trisomy 9, partial trisomy 1q, 5q-, -5, 7q-, -7, 12p-, i(17q), and 9p reduplication. In contrast to its companion myeloproliferative disorders, however, trisomy 8 and 12p- as well as certain complex chromosomal abnormalities appear to confer a poor prognosis in PMF (15).
DIAGNOSIS
Given the inability to aspirate bone marrow when myelofibrosis is present, distinguishing PMF from the many other disorders that cause myelofibrosis (Table 26-3) is a difficult task. Clinical criteria have been formulated to facilitate this (Table 26-4), but they are either inadequate or uncritical. For example, the presence of extramedullary hematopoiesis as defined by circulating nucleated red cells and myelocytes is a feature of many disorders other than PMF (Table 26-5). Although splenomegaly has been considered an optional diagnostic criterion in one classification, splenomegaly is present in over 90% of patients at the time of diagnosis. Thus, at a minimum, in the absence of splenomegaly, it is probably not possible to distinguish PMF clinically from the many disorders that mimic it.
TABLE 26-4 DIAGNOSTIC CRITERIA FOR MYELOFIBROSIS WITH MYELOID METAPLASIA
Necessary criteria
1) Diffuse bone marrow fibrosis
2) Absence of the Philadelphia chromosome or BCR-ABL rearrangement in peripheral blood cells
Optional criteria
1) Splenomegaly of any grade
2) Anisopoikilocytosis with teardrop erythrocytes
3) Presence of circulating immature myeloid cells
4) Presence of circulating erythroblasts
5) Presence of clusters of megakaryocytes and anomalous megakaryocytes in bone marrow biopsy sections
6) Myeloid metaplasia
A diagnosis of MMM is acceptable if the following combinations are present: the two necessary criteria plus any other two optional criteria when splenomegaly is present; the two necessary criteria plus any four optional criteria when splenomegaly is absent
(From Italian consensus conference on diagnostic criteria for myelofibrosis with myeloid metaplasia. Br J Haematol. 1999; 104: 730–737.)
TABLE 26-5 CAUSES OF EXTRAMEDULLARY HEMATOPOIESIS AND A LEUKOERYTHROBLASTIC REACTION
Carcinoma metastatic to the bone marrow
Lymphoma involving the bone marrow
Idiopathic myelofibrosis
Polycythemia vera
Chronic myelogenous leukemia
Myelodysplasia
Acute hepatic injury
Hemolytic anemia
The most important disorders with respect to differential diagnosis are chronic myelogenous leukemia, PV, acute myelofibrosis, myelodysplasia, hairy cell leukemia, primary bone marrow lymphomas, multiple myeloma, metastatic carcinoma, and systemic mastocytosis. The most difficult of this group to identify are acute myelofibrosis and myelodysplasia with myelofibrosis. The former is a rapidly progressive form of acute leukemia, which can have extramedullary hematopoiesis without palpable splenomegaly; the latter is somewhat more indolent but carries an equally poor prognosis. In either instance, an increase in marrow blast cells, micromegakaryocytes, specific chromosome abnormalities, and an increase in marrow CD34+ cells favor an acute myeloid malignancy or myelodysplasia rather than PMF, in which there is an increase in circulating CD34+ cells (16).
The JAK2 V617F mutation can be used to distinguish PMF from chronic myelogenous leukemia and nonmyeloid hematopoietic malignancies in approximately 50% of cases; absence of the mutation, however, is not helpful. From a diagnostic perspective, bone marrow aspiration and biopsy with cytogenetics, flow cytometry using peripheral blood, marrow immunohistochemistry with respect to CD34+ cells, and a JAK2 V617F assay should suffice to differentiate PMF from the other disorders that mimic it.
NATURAL HISTORY
The natural history of PMF is highly variable. In its most aggressive form, there is progressive bone marrow failure with expanding extramedullary hematopoiesis. The consequences of this include splenic and hepatic enlargement, portal hypertension, anemia, thrombocytopenia, leukocytosis or leukopenia, hyperuricemia, cachexia, and in some patients, pulmonary hypertension or transformation to acute leukemia. The actuarial frequency of transformation in one series was 21% at 8 years, a frequency much higher than in PV or essential thrombocytosis. No organ or body cavity is immune to the development of extramedullary hematopoiesis, which can invade the lymph nodes, kidneys, adrenals, ovaries, dura, spinal canal, skin, mediastinum, mesentery, pulmonary parenchyma, and the pleural, peritoneal, and retroperitoneal spaces (17). Progressive, disseminated extramedullary hematopoiesis can also be a harbinger of leukemic transformation. Fortunately, however, this scenario is not the fate of every patient.
Although previous estimates of survival indicated that life expectancy in this disorder was distinctly inferior to its companion myeloproliferative disorders, subsequent studies have indicated that PMF is not a monolithic illness and, with risk stratification, it has been possible to identify patients whose disease is indolent rather than progressive. For this purpose, two useful risk stratification schemes have recently been proposed: the first for risk stratification at diagnosis (the International Prognostic Scoring System [IPSS]) (18) (Table 26-6), and the second during the course of the disease (the Dynamic IPSS plus [DIPPS plus]) (15). Age >65 years, anemia (hemoglobin <10 g/dL), leukocyte count (>25,000/ml), circulating blast cell count (≥1%), and constitutional symptoms are the criteria for IPSS risk stratification, a scoring system similar to that employed for risk stratification in myelodysplasia. The DIPSS plus adds thrombocytopenia (platelets <100,000/ml), transfusion dependence, and unfavorable cytogenetics. JAK2 V617F expression correlated only with older age at diagnosis and a history of thrombosis or pruritus but not prognosis, while the impact on prognosis of newly described but low frequency mutations in PMF has not yet been assessed.
TABLE 26-6 INTERNATIONAL PROGNOSTIC SCORING SYSTEM FOR PMF
TREATMENT
There is no specific therapy for PMF and allogeneic bone marrow transplantation is the only potentially curative therapy. Unfortunately, this approach has been most effective in patients under age 45 years with good prognosis disease, and when viewed with the context of the DIPSS plus, survival correlated with the scoring stage with the low-risk patients faring better than all other groups. Transplant-related mortality was high at 28%; relapse-free survival was 100% at 5 years in low-risk patients, but was 51%, 54%, and 30% for int-1, int-2, and high-risk patients, respectively, which was not better or even worse than their intrinsic natural history (19). Recently, reduced intensity conditioning was found to decrease transplant-related mortality and achieve remission rates of greater than 70% (20). However, prospective studies will be required to establish the most effective conditioning regimen (21). Splenomegaly and myelofibrosis per se did not appear to impact negatively on engraftment.
Bone marrow failure and progressive splenomegaly are the two most pressing problems in the management of PMF. Anemia is the most common problem and can be multifarious with respect to etiology, which can include hemodilution, blood loss, hemolysis, and folic acid or vitamin B6 deficiency. In patients with constitutional symptoms, prednisone therapy may be effective in alleviating anemia as well as the constitutional symptoms since a cytokine storm of varying extent is a feature of PMF. If the serum erythropoietin level is less than 125 mU/ml, a trial of recombinant erythropoietin is worthwhile with the caveat that this could increase spleen or liver size. Impeded androgens such as danazol have been tried in this situation with modest success, but these agents have side effects that make their long-term use unattractive and much of their benefit may actually be virtual since these agents also contract the plasma volume.
Progressive splenomegaly with or without hepatomegaly is a difficult therapeutic problem in PMF since it gives rise to mechanical problems such as early satiety, diarrhea, abdominal discomfort, sequestration of leukocytes and platelets, splenic infarction, portal hypertension, and esophageal varices. Cachexia is an inevitable complication. A number of therapies have been tried in this situation, including low-dose alkylating agents, hydroxyurea, interferon alpha, imatinib mesylate, and thalidomide. Alkylating agents such as busulfan and melphalan at doses of 2–4 mg/day have proved effective but have the potential for substantial hematologic and nonhematologic toxicity and are also leukemogenic; their use should be reserved for specific situations where other remedies have not been effective. Hydroxyurea is effective in controlling leukocytosis and thrombocytosis but can exacerbate anemia. Neither interferon nor imatinib has proved effective in advanced PMF, but the former may be effective in the early stages of the disorder. However, both appear to have substantial toxicity in this group of patients. Thalidomide at low doses in combination with prednisone has proved to be effective in ameliorating anemia as well as thrombocytopenia in PMF patients and also reducing spleen size in approximately 20% (22). Lenolidamide has also been used in PMF with minimal success but has the disadvantage of being myelotoxic in contrast to thalidomide. Low-dose alkylating agents have also been used but are myelotoxic and genotoxic, particularly if used with hydroxyurea.
The most important therapeutic advance for PMF patients is the recent development of nonspecific JAK2 inhibitors, the first of which, ruxolitinib (Jakafi), has been FDA approved for patients with advanced disease, regardless of the presence of a JAK2 mutation (23, 24). Given orally with a recommended starting dose of 15 mg bid, ruxolitinib has proved effective in reducing splenomegaly by approximately 35% in at least 50% of patients and alleviating constitutional symptoms by suppressing inflammatory cytokine production with improvement in exercise tolerance and weight gain in up to 50% of patients. Symptomatic improvement and reduction in spleen size can be seen within 12 weeks with a durable effect as long as the drug is continued. Ruxolitinib’s major toxicity is marrow suppression with the exacerbation or induction of anemia or thrombocytopenia. Discontinuation of the drug will also lead to the reappearance of constitutional symptoms within 7 days, and thus, the drug should be tapered slowly or the use of glucocorticoids considered should the cytokine rebound be severe. To date, ruxolitinib has had no significant impact on the JAK2 V617F allelic burden or the size of the malignant stem cell pool. Nevertheless, it is an important new therapy, which should reduce the need for splenectomy and improve the lives of PMF patients without the risk of increasing genetic instability in the involved hematopoietic stem cell clone, as may occur with hydroxyurea.
In some patients, splenectomy may be necessary for massive splenomegaly because of the failure of other treatment options. This is a major undertaking with significant postoperative complications including hemorrhage, splenic vein thrombosis, infection, hepatomegaly, exuberant leukocytosis or thrombocytosis, and abdominal hernias. Neither anemia nor thrombocytopenia is significantly improved in most patients. In one series, the incidence of acute leukemia increased postsplenectomy. Splenic irradiation has been employed in patients thought to be unfit for surgery. This is often effective in reducing spleen size and ameliorating symptoms temporarily and can be repeated, although not always with the same result. However, myelosuppression is frequent and the mortality rate as a consequence can be as high as 50%. By contrast, irradiation can be useful in controlling localized soft tissue sites of extramedullary hematopoiesis or periostitis. In summary, lacking specific therapy for PMF, treatment in this disorder must be tailored to the individual patient.
ESSENTIAL THROMBOCYTOSIS
INTRODUCTION
Essential thrombocytosis (ET) is the most nebulous of the chronic myeloproliferative disorders, since its only identifying marker, thrombocytosis, is not specific for it. Like PMF and PV, ET is a clonal disorder involving a multipotent hematopoietic stem cell. However, unlike its companion myeloproliferative disorders, hematopoiesis is not globally disturbed, women predominate, and overall life span is superior. The frequency of ET is approximately 2/100,000. The frequency of the disorder increases with age with a mean age at diagnosis of 51 years. In women, the incidence appears to be biphasic with a peak at age 50 years and a second at age 70 years.
PATHOGENESIS
The etiology of ET is unknown. Although thrombopoietin is essential for the survival of primitive hematopoietic stem cells, overproduction of thrombopoietin does not recapitulate ET in animal models or in familial thrombocytosis due to mutations in the 5′ UTR of the thrombopoietin gene or the thrombopoietin receptor gene (MPL). In contrast to PV, where the plasma level of erythropoietin is severely reduced due to the expansion of the red cell mass, in ET, the thrombopoietin level is normal or elevated despite expansion of the megakaryocyte mass, preventing its distinction from secondary forms of thrombocytosis on this basis.
A number of epigenetic abnormalities found in PV and PMF, such as increased expression of granulocyte PRV-1 mRNA and reduced expression of the thrombopoietin receptor, Mpl, in megakaryocytes and platelets, are also found in ET. Cytogenetic abnormalities similar to those found in PV and PMF are also present in ET, but at a much lower frequency. The frequency of the JAK2 V617F mutation is also lower in this disorder than in the other chronic myeloproliferative disorders, and homozygosity for the mutation is rarely present. Some investigators have claimed that ET patients expressing JAK2 V617F have a “PV-like” phenotype. However, a consistent failure on the part of these investigators to exclude PV by performing a red cell mass determination renders these claims specious.
CLINICAL FEATURES
First recognized in 1920, ET has been known by a variety of names, including hemorrhagic thrombocythemia, idiopathic thrombocytosis, and primary thrombocytosis. This ambivalence reflects the lack of a specific diagnostic marker for ET and the fact that the thrombocytosis can be associated with either thrombosis or hemorrhage. Furthermore, with the advent of electronic particle counters, thrombocytosis is now being recognized in individuals who are asymptomatic. This was most often true in women and did not vary with age. Microvascular occlusive syndromes such as migraine, transient ischemic attacks, visual disturbances, dizziness, or erythromelalgia are the most common presenting complaints but are, of course, not specific for the disease. Hemorrhage, usually involving the mucous membranes and generally mild, has been more common in some series than thrombotic episodes, which could be arterial or, less frequently, venous. Interestingly, hemorrhage was more common with platelet counts greater than 1,000,000/µl, and thrombosis when the platelet count was lower.
The physical examination in ET is usually normal. Splenomegaly is present in less than 30% of reported patients at diagnosis and even then is minimal in extent. Significant splenomegaly, isolated hepatomegaly, or lymphadenopathy should suggest another cause for the thrombocytosis.
LABORATORY ABNORMALITIES
Thrombocytosis is the major laboratory abnormality in ET with the platelet count averaging 1,000,000/µl or greater in most large studies. It is not possible, however, to distinguish reactive thrombocytosis from ET simply on the basis of platelet number. Anemia is uncommon and usually mild, and an elevated hemoglobin or hematocrit level should suggest PV. A mild neutrophilic leukocytosis is common, but when the leukocyte count is greater than 15,000/µl or there is significant anemia or a leukoerythroblastic reaction, another diagnosis should be considered. Many patients are iron deficient but paradoxically correction of the deficit usually does not influence the platelet count. Pseudohyperkalemia occurs as a consequence of platelet potassium release during blood clotting when the platelet count is elevated. A very high platelet count can also cause pseudohypoglycemia and hypoxemia if blood for glucose and oxygen tension measurements is not collected on ice and in the presence of a metabolic inhibitor. It is of interest that the serum erythropoietin level can be low in ET, making this test not useful for distinguishing between PV and ET.
Coagulation abnormalities in ET are a consequence of intrinsic platelet abnormalities or the platelet count (25). Abnormalities of platelet structure include an increase in mean platelet volume and distribution width, loss of alpha granules and dense bodies, and disorganization of the platelet microtubular and canalicular systems. Surface expression of CD41 and the thrombopoietin receptor, Mpl, are decreased, while the expression of P-selectin and thrombospondin are increased; the intracellular ADP, PF4, and 5-HT content are reduced. The majority of patients have increased platelet aggregation in response to epinephrine, ristocetin, ADP, and collagen. Paradoxically, however, the bleeding time is increased in less than 20% of patients. Thromboxane excretion is frequently increased and suppressible by salicylate therapy, suggesting continuous intravascular platelet activation. However, there is no correlation between the platelet abnormalities and thrombosis in this disorder.
Acquired von Willebrand’s disease is an interesting feature of ET as well as the other chronic myeloproliferative disorders (26). As the platelet count increases, generally above 1,000,000/µl, the platelets adsorb and destroy the highest molecular weight plasma von Willebrand multimers, leading to a reduction in ristocetin cofactor activity. Patients with this abnormality are at risk of bleeding, particularly if exposed to salicylates.
CYTOGENETIC ABNORMALITIES
Cytogenetic abnormalities are uncommon in ET and none is pathognomonic for the disorder. The common cytogenetic abnormalities include trisomy 1, 8, 9, and 21, 1q-, 13q-, and 20q-. Since both chronic myelogenous leukemia and the 5q- syndrome can present with thrombocytosis, cytogenetic analysis constitutes an important part of the diagnostic evaluation.
DIAGNOSIS
Establishing a diagnosis of ET is more difficult than for the other chronic myeloproliferative disorders because ET lacks any unique identifying characteristics or a specific diagnostic marker, and because thrombocytosis can be the initial manifestation of PV or PMF, either of which may not become clinically apparent for many years after the onset of the thrombocytosis (27, 28). Furthermore, there is as yet no agreement as to what platelet count threshold should be used for the diagnosis of ET. A number of diagnostic criteria have been developed, but they rely on the exclusion of other disorders and their complexity emphasizes the difficulties inherent in the diagnosis of this disease. The extent of the problem can be simply visualized from the number of benign and malignant disorders that can cause thrombocytosis (Table 26-7). Furthermore, it is apparent from epidemiologic studies of JAK2 V617F expression, platelet Mpl expression, and clonality that there is substantial heterogeneity among ET patients with respect to these abnormalities and, except for lack of JAK2 V617F homozygosity, no specificity. Attempts to distinguish the cellular phase of PMF from ET solely on the basis of bone marrow morphology have not been convincing, but a JAK2 V617F allele burden greater than 50%, anemia, significant leukocytosis (>15,000/µl), or a leukoerythroblastic reaction should suggest the presence of a different myeloproliferative disorder (29).
TABLE 26-7 CAUSES OF THROMBOCYTOSIS
Tissue inflammation
Collagen vascular disease, inflammatory bowel disease
Malignancy
Infection
Myeloproliferative disorders
Polycythemia vera, idiopathic myelofibrosis, essential thrombocytosis, chronic
myelogenous leukemia
Myelodysplastic disorders
5q- syndrome, idiopathic refractory sideroblastic anemia
Postsplenectomy, or hyposplenism
Hemorrhage
Iron deficiency anemia
Surgery
Rebound
Correction of vitamin B12 or folate deficiency, post-ethanol abuse
Hemolysis
Familial
Thrombopoietin overproduction, constitutive Mpl activation
From a prognostic prospective, the most serious illnesses associated with thrombocytosis that need to be excluded are chronic myelogenous leukemia, myelodysplasia (5q- syndrome), sideroblastic anemia, PMF, and PV. It also needs to be emphasized that chronic myelogenous leukemia can present with isolated thrombocytosis alone in the absence of leukocytosis or basophilia. From this perspective, a bone marrow aspirate and biopsy for morphology, flow cytometry, cytogenetics, and peripheral blood FISH for bcr-abl, since this can be positive in the absence of the Philadelphia chromosome, are the essential diagnostic tests. A negative assay for JAK2 V617F does not exclude the diagnosis of ET, nor does its presence have any implications with respect to diagnosis or the clinical course.
NATURAL HISTORY
Most but not all studies of ET have found that life span was not significantly different from the general population. A recent study of low-risk ET suggests that the most important risk factor for thrombosis in ET is tobacco use, particularly in the presence of cardiovascular risk factors. Otherwise, the traditional risk factors including age ≥60 years, prior thrombosis, and leukocytosis (>15,000/µl) were not predictive (30). Additionally, rates of venous thrombosis were higher in women than men. Importantly, regardless of the type of therapy employed, the risk of thrombosis appeared to reach a plateau after 9 years. A platelet count of 1,000,000/µl or greater is the major risk factor for hemorrhage. Transformation to myelofibrosis or PV occurs in approximately 20% of patients over the first decade after diagnosis (27, 28). Spontaneous leukemic transformation occurs but is uncommon and most instances are a consequence of myelotoxic drug exposure.
TREATMENT
The first rule of therapy for ET is accuracy in diagnosis, particularly because life span is generally not reduced in this disease and its treatment differs from the other chronic myeloproliferative disorders it mimics. The second rule of therapy is to do no harm. Stated differently, the treatment cannot be worse than the disease. Thrombosis, either macrovascular or microvascular, is the major impediment to health in ET, but there is no correlation between the height of the platelet count and thrombosis, rendering problematic the formulation of a treatment endpoint on that basis. In general, patients with ET who have had a prior major vessel thrombosis should be treated no differently with respect to anticoagulation and risk factor reduction than their counterparts with a normal platelet count. The most difficult decision then becomes how best to manage the platelet count (31).
Patients with ET under age 60 years, who have no cardiovascular risk factors or a prior thrombosis, are not at a greater risk of thrombosis than their age-matched counterparts with a normal platelet count (30, 32). Treatment in these patients should be directed at the alleviation of microvascular symptoms such as ocular migraine or erythromelalgia. Aspirin is a specific remedy for these and can be given daily or on as needed basis. Ibuprofen can be substituted if a shorter acting agent is required. When the platelet count is greater than 1,000,000/µl, ristocetin cofactor activity should be measured before using either agent in a symptomatic patient and, if reduced, platelet count reduction rather than platelet inactivation will be necessary. In some patients, particularly those with migraine, platelet inactivation may not be sufficient to control symptoms. The safest method to lower the platelet count then becomes the major issue.
Current therapy for controlling the platelet count includes hydroxyurea, anagrelide, interferon alpha, alkylating agents, and 32P. All of these agents are usually effective but each has distinct disadvantages. The most serious of these is myelotoxicity leading to acute leukemia, which has been demonstrated unequivocally for the alkylating agents and 32P. Whether hydroxyurea is leukemogenic has been a matter of debate. It also enhances the leukemogenic effect of the alkylating agents and 32P, whether given before or after them. Since the use of chemotherapeutic agents has not been shown to improve longevity in the chronic myeloproliferative disorders, their use should not be routine but restricted to situations where other forms of therapy have been ineffective.
Two randomized clinical trials provide some guidance to this end. In a study of ET patients older than 60 years, hydroxyurea was not more effective than aspirin in preventing arterial thrombosis (33) and failed to prevent venous thrombosis. In a much larger study of high-risk patients with thrombocytosis taking aspirin, in whom the platelet count was normalized, hydroxyurea was not more effective than anagrelide in preventing arterial thrombosis and was actually less effective in preventing venous thrombosis. Hydroxyurea was, however, more effective in preventing transient ischemic attacks (34) because it is a nitric oxide donor. Therefore, in patients over age 60 years who have risk factors for thrombosis and who are experiencing transient ischemic attacks, hydroxyurea is the drug of choice. Otherwise, a safer alternative such as interferon alpha or anagrelide should be used when there is a clinical indication to lower the platelet count. In the case of both, given the side effects associated with long-term use, their use should be intermittent if possible (35). If long-term use is planned, periodic cardiac monitoring is also indicated. Finally, the combination of aspirin and anagrelide has been associated with an increased incidence of gastrointestinal hemorrhage (34).
Acquired von Willebrand syndrome caused by thrombocytosis requires no treatment unless there is a need for surgery or the patient experiences spontaneous bleeding (26). In this instance, platelet count reduction will be required. In an emergent situation, platelet pheresis can be employed but this is not a particularly efficient approach when there is extreme thrombocytosis. Administration of epsilon aminocaproic acid is an effective remedy for bleeding in this situation.
PREGNANCY
Special mention needs to be made about pregnancy since ET is so common in young women. Pregnancy has an ameliorating effect on the thrombocytosis in this disorder and, while first trimester abortions are increased, there is no correlation between platelet count and obstetrical complications. No specific therapeutic intervention has been proved to be uniformly effective, but low-dose aspirin has been recommended as prophylactic therapy and, when there has been prior thrombosis, low-molecular-weight heparin. Interferon alpha can also be given safely during pregnancy if platelet count reduction is necessary. Perhaps the most important recommendation is to be sure that the patient does not actually have PV. Stated differently, a normal hematocrit in a pregnant woman with ET should suggest the presence of PV.
REFERENCES
1. Gangat N, Strand J, Lasho TL, et al. Cytogenetic studies at diagnosis in polycythemia vera: clinical and JAK2V617F allele burden correlates. Eur J Haematol. 2008; 80: 197–200.
2. Vainchenker W, Dusa A, Constantinescu SN. JAKs in pathology: role of Janus kinases in hematopoietic malignancies and immunodeficiencies. Semin Cell Dev Biol. 2008; 19: 385–393.
3. Johansson PL, Safai-Kutti S, Kutti J. An elevated venous haemoglobin concentration cannot be used as a surrogate marker for absolute erythrocytosis: a study of patients with polycythaemia vera and apparent polycythaemia. Br J Haematol. 2005; 129: 701–705.
4. Lamy T, Devillers A, Bernard M, et al. Inapparent polycythemia vera: an unrecognized diagnosis. Am J Med. 1997; 102: 14–20.
5. Beer PA, Delhommeau F, Lecouedic JP, et al. Two routes to leukemic transformation following a JAK2 mutation-positive myeloproliferative neoplasm. Blood. 2010; 115: 2891–2900.
6. Najean Y, Rain J. Treatment of polycythemia vera: the use of hydroxyurea and pipobroman in 292 patients under the age of 65 years. Blood. 1997; 90: 3370–3377.
7. Kiladjian JJ, Chevret S, Dosquet C, et al. Treatment of polycythemia vera with hydroxyurea and pipobroman: final results of a randomized trial initiated in 1980. J Clin Oncol. 2011; 29: 3907–3913.
8. Spivak JL. Polycythemia vera myths, mechanisms, and management. Blood. 2002; 100: 4272–4290.
9. Ballen KK, Woolfrey AE, Zhu X, et al. Allogeneic hematopoietic cell transplantation for advanced polycythemia vera and essential thrombocythemia. Biol Blood Marrow Transplant. 2012; 18: 1446–1454.
10. Silver RT. Treatment of polycythemia vera with recombinant interferon alpha (rifnalpha) or imatinib mesylate. Curr Hematol Rep. 2005; 4: 235–237.
11. Kiladjian JJ, Cassinat B, Chevret S, et al. Pegylated interferon-alfa-2a induces complete hematologic and molecular responses with low toxicity in polycythemia vera. Blood. 2008; 112: 3065–3072.
12. Elliott MA, Chen MG, Silverstein MN, Tefferi A. Splenic irradiation for symptomatic splenomegaly associated with myelofibrosis with myeloid metaplasia. Br J Haematol. 1998; 103: 505–511.
13. Thiele J, Kvasnicka HM, Mullauer L, et al. Essential thrombocythemia versus early primary myelofibrosis: a multicenter study to validate the WHO classification. Blood. 2011; 117: 5710–5718.
14. Murphy S, Davis JL, Walsh PN, Gardner FH. Template bleeding time and clinical hemorrhage in myeloproliferative disease. Arch Intern Med. 1978; 138: 1251–1253.
15. Gangat N, Caramazza D, Vaidya R, et al. DIPSS Plus: A refined dynamic international prognostic scoring system for primary myelofibrosis that incorporates prognostic information from karyotype, platelet count, and transfusion status. J Clin Oncol. 2011; 29: 392–397.
16. Barosi G, Hoffman R. Idiopathic myelofibrosis. Semin Hematol. 2005; 42: 248–258.
17. Rumi E, Passamonti F, Boveri E, et al. Dyspnea secondary to pulmonary hematopoiesis as presenting symptom of myelofibrosis with myeloid metaplasia. Am J Hematol. 2006; 81: 124–127.
18. Cervantes F, Dupriez B, Pereira A, et al. New prognostic scoring system for primary myelofibrosis based on a study of the International Working Group for Myelofibrosis Research and Treatment. Blood. 2009; 113: 2895–2901.
19. Ditschkowski M, Elmaagacli AH, Trenschel R, et al. DIPSS scores, pre-transplant therapy and chronic GVHD determine outcome after allogeneic hematopoietic stem cell transplantation for myelofibrosis. Haematologica. 2012; 97:1574–1581.
20. Rondelli D, Barosi G, Bacigalupo A, et al. Allogeneic hematopoietic stem-cell transplantation with reduced-intensity conditioning in intermediate- or high-risk patients with myelofibrosis with myeloid metaplasia. Blood. 2005; 105: 4115–4119.
21. Ballen KK, Shrestha S, Sobocinski KA, et al. Outcome of transplantation for myelofibrosis. Biol Blood Marrow Transplant. 2010; 16: 358–367.
22. Mesa RA, Steensma DP, Pardanani A, et al. A phase 2 trial of combination low-dose thalidomide and prednisone for the treatment of myelofibrosis with myeloid metaplasia. Blood. 2003; 101: 2534–2541.
23. Verstovsek S, Mesa RA, Gotlib J, et al. A double-blind, placebo-controlled trial of ruxolitinib for myelofibrosis. N Engl J Med. 2012; 366: 799–807.
24. Harrison C, Kiladjian JJ, Al-Ali HK, et al. JAK inhibition with ruxolitinib versus best available therapy for myelofibrosis. N Engl J Med. 2012; 366: 787–798.
25. Wehmeier A, Sudhoff T, Meierkord F. Relation of platelet abnormalities to thrombosis and hemorrhage in chronic myeloproliferative disorders. Semin Thromb Hemost. 1997; 23: 391–402.
26. Michiels JJ, Budde U, van der PM, et al. Acquired von Willebrand syndromes: clinical features, aetiology, pathophysiology, classification and management. Best Pract Res Clin Haematol. 2001; 14: 401–436.
27. Jantunen R, Juvonen E, Ikkala E, et al. Development of erythrocytosis in the course of essential thrombocythemia. Ann Hematol. 1999; 78: 219–222.
28. Cervantes F, Alvarez-Larran A, Talarn C, Gomez M, Montserrat E. Myelofibrosis with myeloid metaplasia following essential thrombocythaemia: actuarial probability, presenting characteristics and evolution in a series of 195 patients. Br J Haematol. 2002; 118: 786–790.
29. Barbui T, Thiele J, Passamonti F, et al. Survival and disease progression in essential thrombocythemia are significantly influenced by accurate morphologic diagnosis: an international study. J Clin Oncol. 2011; 29: 3179–3184.
30. Alvarez-Larran A, Cervantes F, Pereira A, et al. Observation versus antiplatelet therapy as primary prophylaxis for thrombosis in low-risk essential thrombocythemia. Blood. 2010; 166: 1205–1210.
31. Schafer AI. Thrombocytosis. N Engl J Med. 2004; 350: 1211–1219.
32. Ruggeri M, Finazzi G, Tosetto A, et al. No treatment for low-risk thrombocythaemia: results from a prospective study. Br J Haematol. 1998; 103: 772–777.
33. Cortelazzo S, Finazzi G, Ruggeri M, et al. Hydroxyurea for patients with essential thrombocythemia and a high risk of thrombosis. N Engl J Med. 1995; 332: 1132–1136.
34. Harrison CN, Campbell PJ, Buck G, et al. Hydroxyurea compared with anagrelide in high-risk essential thrombocythemia. N Engl J Med. 2005; 353: 33–45.
35. Emadi A, Spivak JL. Anagrelide: 20 years later. Expert Rev Anticancer Ther. 2009; 9: 37–50.