An increased incidence of vascular complications and a progression to myelofibrosis or acute leukemia/myelodysplasia are the major causes of morbidity and mortality. In a large randomized PV trial, previous thrombosis, age, phlebotomy, and rate of phlebotomies contributed to the increased risk of thrombosis.49 The age of the patient (>60 years) and previous thrombotic events are now universally acknowledged major risk factors for new major vascular complications in PV.133 Thus, PV patients are staged to low risk and high risk, with previous thrombotic events, including transient ischemic attacks and age greater than 60 years, defining the high-risk category. The assigned risk factor has a major impact on therapeutic decisions; high-risk patients are treated with cytoreductive drugs. Other stratifications of risk include a formulation of an intermediate-risk category that includes PV patients with other cardiovascular factors, such as hypertension.134 Other risk factors also play a role in the pathogenesis of thrombosis, including the presence of leukocytosis135,136 and the JAK2 V617F mutational allele burden,137,138 but have not yet been incorporated into therapeutic decision making. Bleeding is more frequent in patients with platelet counts in excess of 1500 x 109/L and this is thought to be a result of an acquired von Willebrand disease. An increased platelet count does not increase the risk of thrombosis.102
There is a need for prospective clinical studies with stratification of patients according to their baseline leukocyte counts and mutant JAK2 allele burden. A number of criteria of response essential for prospective studies have been formulated (Table 86–2).139 However, until such evidence is available, the decision on how to manage patients with high leukocyte levels and/or high JAK2 V617F mutational burden should continue to follow conventional criteria.
The mainstay of therapy of PV remains nonspecific myelosuppression, which many practitioners supplement by phlebotomies. Additional measures are medications to prevent thrombotic events (aspirin) and to relieve symptoms. Promising therapies are pegylated interferon preparations, which are better tolerated, and JAK2 inhibitors, which are being evaluated mainly in the post-PV–myelofibrotic stage.
It is useful to consider treatments in the plethoric and the spent phases separately.
The Plethoric Phase
The treatment of patients in the plethoric phase of the disease is aimed at ameliorating symptoms and decreasing the risk of thrombosis or bleeding by reducing the blood counts. This is best accomplished by the use of myelosuppressive drugs and, in some patients, by combination therapy consisting of myelosuppression, phlebotomies, and platelet-reducing agents, or, alternatively, by using interferon- therapy. Table 86–3 summarizes the advantages and disadvantages of various forms of therapy.
Myelosuppression decreases blood counts, decreases the risk of vascular events, and ameliorates symptoms, thus increasing an overall sense of well-being. Although there is also a clinical impression that it increases patients’ long-term survival, there are no long-term clinical studies to document this.
Hydroxyurea is the most common myelosuppressive agent used in the treatment of PV. Hydroxyurea is effective therapy for controlling the erythrocyte, leukocyte, and platelet count and it decreases the risk of thrombosis during the first few years of therapy when compared to an historical cohort treated with phlebotomy alone.50 Its suppressive effect is of short duration; consequently, continuous rather than intermittent therapy is required. Because it is short acting, it is relatively safe to use, as the blood counts rise within a few days of decreasing the dose or of stopping the drug. Moreover, because it is not an alkylating agent, it has less potential for causing acute leukemic transformation than other myelosuppressive agents. Although some studies suggest that hydroxyurea presents a higher risk of acute leukemic transformation, the observations did not reach statistical significance. An analysis of 1638 PV patients enrolled in a prospective observational study did not find an increased incidence of leukemic or myelodysplastic transformation (hazard ratio: 0.86; 95% confidence interval 0.26-2.88; P = 0.8021).143 In one study, the incidence of acute leukemia was slightly, but not significantly, higher than that in patients treated with phlebotomy alone,140 a trend that remains controversial because many, if not all, patients treated with hydroxyurea who developed leukemia had also received an alkylating agent at some time in their disease course. Experience in the use of hydroxyurea in the treatment of ET has indicated a marked reduction of thrombosis to approximately 30 percent of the untreated patient.141 Limited evidence indicates that hydroxyurea treatment in adults with sickle cell disease is not associated with leukemia.142 Unfortunately, despite its safety and effectiveness, a number of patients administered hydroxyurea discontinue the drug because of adverse effects (skin ulcers or gastrointestinal intolerance).
Busulfan is a useful second-line agent in patients whose disease is difficult to control or who have adverse reactions to hydroxyurea. The administration of busulfan is a convenient and effective means for the treatment of PV. Marrow suppression produced by this drug is long-lasting and, as a consequence, it can be given intermittently at a dose of 2 to 6 mg daily for a period not exceeding several weeks; the counts continue to fall for several weeks after drug administration is discontinued and may then remain within normal range for many months or even years. In one large study, the median first remission duration of busulfan-treated patients was 4 years.144 This prolonged depression of marrow activity that is brought about by busulfan is its major advantage in the treatment of PV, but it also poses a hazard of long-term pancytopenia. The incidence of transformation to acute leukemia in patients treated intermittently with busulfan is relatively low. Of 145 patients followed from 2 to 11 years, 3 developed acute leukemia.144 Exposure to busulfan had an independent role in producing an excess risk for progression to acute leukemia in the multicenter study of more than 1600 PV patients.143
32P therapy was one of the first effective modes of treatment used. Extensive investigations of the long-term outcome of treatment with 32P have been documented.48,145 Good control of the disease usually can be achieved with initial doses of 2 to 4 mCi. It is rarely used at present, but it may be the treatment of choice for older patients and patients who may be difficult to follow on chemotherapy.146,147
Often, the initial treatment for patients with uncomplicated PV is phlebotomy.28,148 The rationale for phlebotomy of patients with PV is based on a widely quoted study that suggested that the risk of thrombosis in PV was proportional to the elevation in hematocrit.149 The underlying mechanisms causing thrombosis in PV are not fully understood, but the hematocrit is unlikely to be the only risk factor. The lack of a significant risk of thrombosis in large studies of patients with polycythemia caused by chronic exposure to high altitude or by Eisenmenger syndrome150 and other cyanotic heart diseases151 argue against hematocrit as the only factor causing thrombosis. For example, when 100 cyanotic patients with congenital heart disease were observed for a total of 748 patient-years, no patient with polycythemia developed cerebral arterial thrombosis.152 Studies of transgenic mouse with extreme polycythemia caused by constitutive overexpression of erythropoietin did not have an increase in thrombotic complications.153 Additionally, the increased risk of strokes in Chuvash polycythemia is not statistically different in those affected patients whose hematocrit is controlled by phlebotomies. Furthermore, the European Collaboration on Low-Dose Aspirin in the Polycythemia Vera study, which included 1638 patients from 12 participating countries and 94 centers, has not found differences in thrombotic complications in the range of hematocrits between 40 and 50 percent.
When phlebotomy is instituted, the hemoglobin may be reduced to normal or near-normal values by the removal of 450 mL of blood at intervals of 2 to 4 days for average-size patients, with removal of smaller amounts being from patients who weigh less than 50 kg. Patients with impaired cardiovascular function are better treated with smaller phlebotomies at more frequent intervals.
Phlebotomy is an effective way in which to lower or normalize the elevated blood viscosity of patients with PV. Phlebotomy may result in improvement of symptoms such as headaches or a feeling of “increased pressure” in some patients. It neither reduces the leukocyte nor platelet count, nor does it affect symptoms such as pruritus or gout. Iron deficiency and a resulting microcytosis are the usual consequences of repeated phlebotomies. The iron-deficient state may help to control the hemoglobin concentration in the long run but it increases the sense of fatigue. It may increase the platelet count in some patients. Viscosity of the blood is a function of the hematocrit, and is independent of the number of red cells154; deformability of iron-deficient erythrocytes is virtually normal.155
A randomized study156 comparing phlebotomy alone with treatment with 32P or with chlorambucil indicated that the life span of patients treated only with phlebotomy was better than that of patients treated with chlorambucil and no worse than of those given 32P. However, patients undergoing phlebotomy suffered more thrombotic episodes than patients treated with myelosuppressive therapy, although the risk seemed limited to the first 3 years of therapy. This documented increased risk of phlebotomy is balanced by a lower incidence of acute leukemia late in the patient’s course. Surprisingly, there was no correlation between the level of the platelet count and the development of thrombotic complications. Many patients can be well controlled by phlebotomy alone during much or all of their disease course; the role of myelosuppressive therapy in the treatment of PV has sometimes been questioned.157 Patients younger than age 50 years who have no prior history of thrombosis might be treated with phlebotomy alone,158 but no rigorous data are available to support this recommendation.
Among 113 patients with PV who had thrombocytosis, the administration of anagrelide produced a platelet response in 85 cases (75%).159 The starting dose was 0.5 or 1.0 mg given four times daily, and a response was noted in most patients within a week. The average dose required to control the platelet count was 2.4 mg per day. Adverse events included headache, palpitations, diarrhea, and fluid retention and were occasionally sufficiently severe to require discontinuation of the treatment.160 The United Kingdom randomized trial indicated superior results for hydroxyurea compared to anagrelide for the control of elevated platelet count, myelofibrosis, and hemorrhagic complications in essential thrombocythemia.24
Symptomatic Therapy for Pruritus
Many of the symptoms of polycythemia are controlled either with myelosuppressive therapy or by phlebotomy. Pruritus is sometimes an exception. It tends to be more severe when the disease is active and becomes milder or disappears when control is achieved by myelosuppression. Evidence indicates that the mutant JAK2 can stimulate an agonist of pruritus in basophils161 and the JAK2V617F allelic burden correlates with thromboses on multivariate analysis.162 The same analysis suggested a negative correlation between arterial thrombosis and pruritus.162 Nonetheless, in some patients pruritus becomes a nearly intolerable annoyance. Because bathing or showering usually intensifies the itching, often the best advice that can be offered is to bathe less frequently. Photochemotherapy with psoralens and ultraviolet light is helpful.163 Antihistamines are often given but are usually not very effective. Aspirin164 and cyproheptadine90 may be useful. Interferon- has been helpful in some patients.165–167 The use of JAK2 inhibitors can decrease pruritus in patients with post-PV PMF.168,169
An aspirin and dipyridamole have been used to prevent thromboembolic events in PV. The early trials using 300 mg of aspirin daily showed an increase in the incidence of bleeding without a measurable impact on the incidence of thrombotic episodes.170 The administration of low-dose aspirin has been suggested in patients who have a vascular occlusion.171 A pilot-controlled trial showed that low-dose aspirin was well tolerated by PV patients and is sufficient to fully inhibit synthesis of the platelet aggregating compound thromboxane, but not the endothelial cell protectant prostacyclin.172 The European Collaboration on Low-Dose Aspirin in Polycythemia Vera study showed that daily low-dose aspirin decreases arterial and venous thromboses, albeit to a small extent.47 Because most of the thrombotic complications were not prevented, this study suggests that only a minor fraction of thromboses are attributable to platelets and raises a question regarding the mechanism of thrombosis in PV. The studies showed a correlation between the neutrophil count and the incidence of thrombosis. Multivariate analysis of a study of hydroxyurea in sickle cell disease revealed a benefit to decreasing blood neutrophil levels in prevention of sickle cell vascular events.173 The possibility that cytoreductive agents affect platelets and/or endothelial cells is being explored.
Since the pioneering work of interferon- in PV,174 many other studies documenting the efficacy of interferons in PV have been reported.175 Although these studies have many design similarities, they do not lend themselves to accurate meta-analysis as various formulations of interferon (2a, 2b, human leukocyte interferon, peg-2a, peg-2b) were used and because heterogeneous criteria were employed to measure response. For example, in studies of PV, complete response was defined in some cases as control of hematocrit with freedom from phlebotomy, whereas in other studies it was defined as normalization of platelet count, disappearance of splenomegaly, and resolution of disease-related symptoms such as pruritus. Furthermore, in the small number of patients studied thus far, administration of interferon- has led to the JAK2 V617F mutation becoming undetectable175,176 and clonal hematopoiesis converting to a polyclonal state.13 In addition to its capacity to induce hematologic responses and to ameliorate disease-related symptoms, studies show molecular responses in a subset of patients treated with pegylated interferon- (Pegasys). Using JAK2 V617F as a molecular marker, a French group175 studied 40 patients with PV treated with Pegasys and reported that of 37 evaluable patients, 94 percent had a complete hematologic response. Approximately 25 percent of patients with PV and ET treated with interferon discontinue treatment, half within the first year. The hematologic toxicities include anemia, thrombocytopenia, and neutropenia. Other potential untoward effects of interferon include depression, mood changes, skin toxicity, hair loss, nausea, diarrhea, weight loss, liver function abnormalities, and cardiac and neurologic toxicity. Immunologic abnormalities in the form of autoimmune processes (e.g., hypothyroidism, autoimmune hemolytic anemia, polyarthritis, glomerulonephritis, connective tissue diseases, and asymptomatic antinuclear antibodies) may occur as a consequence of interferon therapy.177 Conceivably, the development of interferon-induced autoimmune processes reflects the immunomodulatory activity of the drug through which at least part of its antitumor activity is mediated. Using serologic analysis of tumor antigens through screening an expression complementary DNA library, tumor antigens that elicit immune responses in chronic myelogenous leukemia178 and PV patients who achieved remission after treatment with interferon were identified.179–181 Interferon is the drug of choice in pregnant patients with PV (see Chap. 7).
Currently available JAK2 inhibitors target the catalytic site of the enzyme and are therefore active against wild-type JAK2 and JAK2 V617F; some also inhibit other JAKs, including JAK3.182 In addition to technical obstacles, certain pathobiologic characteristics of the JAK2V617F-positive myeloproliferative disorders suggest that targeting JAK2 V617F is a suboptimal therapeutic strategy; data from several laboratories indicate that the mutant enzyme is not the disease-initiating step in Philadelphia-chromosome-negative MPDs34,183 and that these disease processes are characterized by clonal heterogeneity that is a consequence of genetic instability.184 Indeed, initial studies in humans suggest that currently available inhibitors of JAK2 have modest clinical activity and little, if any, effect on the mutant allele burden.169,185 The clinical studies of the inhibitors have been largely confined to JAK2 V617F-positive PMF. Five JAK2 inhibitors are already in clinical testing (INCB018424, TG101348, CEP-701, AZD1480, and XL019). Overall, these JAK2 inhibitors have reduced myelofibrosis symptoms. The Incyte compound INCB018424 has completed phase II studies and has reduced spleen size and splenomegaly-associated symptoms, as well as general symptomatology such as night sweats, itching, and fatigue; however, the clinical, hematologic, and molecular responses have not matched that of imatinib mesylate use in chronic myelogenous leukemia and do not induce a complete or substantial remission of the disease.168,186
There is growing evidence of abnormal epigenetic gene regulation as a mechanism potentially contributing to the pathogenesis and the phenotypic diversity of myeloproliferative disorders, especially PMF or post-PV myelofibrosis.187 The clinical trials with epigenetic drug modifiers are in progress. The use of these agents holds considerable promise.187
Summary of Therapeutic Approach
The current general approach to treatment of patients not participating in clinical trials is:
The Spent Phase
Ultimately, sometimes after only a few years and usually after 15 or more years, erythrocytosis of PV patients gradually abates, phlebotomy requirements decrease and cease, and anemia develops. During this “spent” phase of the disease, marrow fibrosis becomes more marked and the spleen becomes greatly enlarged (Fig. 86–3A). Instead of phlebotomies, transfusions or erythropoietin may be required.188 The platelet count may remain high or may decline, even to thrombocytopenic levels. Marked leukocytosis may occur with the appearance of immature granulocytes in the blood. At this point, the disease mimics closely primary myelofibrosis (Chap. 91) and the condition is termed post-PV myelofibrosis. Treatment of this phase of the disease is difficult and requires the judicious use of a combination of therapeutic approaches, depending on the circumstances of each individual patient. Hydroxyurea given together with erythropoiesis-stimulating agents is often helpful and splenectomy should be considered in selected patients. Allogeneic stem cell transplantation should be considered in younger patients; nonmyeloablative allogenic marrow transplantation has been successful in clinical trials of patients up to 65 years of age.188
Splenectomy may be warranted (see Fig. 86–3B), particularly in patients with severe fatigue, with cytopenias, and in those where a greatly enlarged spleen produces physical discomfort.189 However, a large Mayo Clinic series reported significant mortality and morbidity of splenectomy at this stage of the disease.190
Hematopoietic Stem Cell Transplantation for Treatment of polycythemia vera.
A few younger patients have undergone successful allogeneic stem cell transplantation.191 Nonmyeloablative stem cell transplantation, a procedure that may be performed in otherwise healthy people even in the sixth decade of life (see Chap. 21), shows promise192 and is, at present, the only curative approach to this disease stage.188
Course and Prognosis
Thrombotic complications discussed in the preceding sections are the dominant cause of morbidity and mortality in patients with PV. In addition and in contrast to other polycythemic disorders, PV has an increased risk of evolution to acute leukemia. Although several clinical stages of PV are recognized (plethoric or proliferative phase, stable phase, spent phase or postpolycythemic myelofibrosis phase, and acute leukemia), it is not clear that these stages represent a sequential progression of the disease.
The Polycythemia Vera Study Group48 found that the median survival from the beginning of treatment was 13.9 years for those treated by phlebotomy alone, 11.8 years for 32P-treated patients, and 8.9 years for chlorambucil-treated patients. Thrombosis was the most common cause of death, accounting for 31 percent of the fatalities. Nineteen percent of the patients died of acute leukemia, 15 percent from other neoplasms, and approximately 5 percent each from hemorrhage or the development of the spent phase. Similarly, a large French study revealed a median survival of 13.5 years of PV patients initially treated with 32P, only slightly less that the 15.2 years of age-matched controls.193 Others suggested that PV is a disease that is compatible with normal or near-normal life for many years.194,195 However, most studies agree that there is excess mortality attributable to thrombotic complications and acute leukemia transformation as a direct consequence of PV.47 Acute leukemia occurs even in patients who have been treated only by phlebotomy, although its incidence is increased by the various forms of cytotoxic therapy that have been employed. While acute myelogenous leukemia is most common, acute lymphoid leukemia196 and chronic neutrophilic leukemia197 have occurred as well.