Essential Thrombocythemia

The new england journal of medicine

Clinical Practice

Caren G. Solomon, M.D., M.P.H., Editor Essential Thrombocythemia

Ayalew Tefferi, M.D., and Animesh Pardanani, M.B., B.S., Ph.D.

This Journal feature begins with a case vignette highlighting a common clinical problem. Evidence supporting various strategies is then presented, followed by a review of formal guidelines, when they exist. The article ends with the authors’ clinical recommendations.

A 62-year-old woman was discovered to have a platelet count of 1,004,000 per cubic From the Division of Hematology, Depart- millimeter during evaluation for fatigue. Her medical history was notable only for ment of Medicine, Mayo Clinic, Roches- ter, MN. Address reprint requests to Dr. hypertension and obesity (body-mass index [the weight in kilograms divided by the Tefferi at the Division of Hematology, square of the height in meters], 34). She reported no symptoms other than fatigue, Department of Medicine, Mayo Clinic, and all other blood counts were within normal limits. A review of previous labora- 200 First St. SW, Rochester, MN 55905, or at tefferi.ayalew@mayo.edu. tory reports obtained 8 and 3 years earlier revealed platelet counts of 290,000 and 595,000 per cubic millimeter, respectively. Analysis of peripheral blood was negative N Engl J Med 2019;381:2135-44. DOI: 10.1056/NEJMcp1816082 for JAK2 V617F but revealed an exon 9 mutation in the gene encoding calreticulin Copyright © 2019 Massachusetts Medical Society. (CALR). How would you establish the diagnosis of essential thrombocythemia, and how would you manage this case?

The Clinical Problem hrombocytosis (platelet count ≥450,000 per cubic millimeter) is frequently encountered in clinical practice, with a reported prevalence of 1.9% in a large study involving ambulatory patients who were 65 years of T 1 age or older. In a retrospective study involving 801 adult patients with thrombocytosis in a tertiary care hospital, primary thrombocytosis (myeloproliferative neoplasms) was observed in 5.2% of cases; other (secondary) causes included infection (47.9%), trauma or postsurgical state (24.5%), other cancer (10.7%), and An audio version of this article is 2 iron-deficiency anemia (7.4%). Extreme thrombocytosis (platelet count ≥1 million available at per cubic millimeter) is infrequent, occurring in less than 2% of patients in a large NEJM.org retrospective study involving patients with thrombocytosis.3 In this cohort, approximately one third of patients had hematologic cancers (predominantly myeloproliferative neoplasms); most had more than one identifiable cause of extreme thrombocytosis. The incidence of essential thrombocythemia is estimated at 1.2 to 3.0 per 100,000 population per year.5 In a retrospective study involving 1076 patients with essential thrombocythemia conducted by the Mayo Clinic,4 the median age at diagnosis was 58 years (range, 18 to 96), 67% were women, palpable splenomegaly was present in 17%, and the median platelet count was 876,000 per cubic millimeter (range, 451,000 to 3,460,000). Leukocytosis was documented in 26% of patients, and an abnormal karyotype was detected in 9%; a history of thrombosis at or before diagnosis was present in 21%. Other disease features included headaches, visual disturbances, lightheadedness, dysesthesia, and, much less frequently, erythromelalgia (a burning dysesthesia associated with red discoloration of the

n engl j med 381;22 nejm.org November 28, 2019 2135 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

Key Clinical Points Essential Thrombocythemia • Essential thrombocythemia is associated with an increased risk of thrombosis and bleeding. • Most patients with essential thrombocythemia harbor a mutation in one of three genes: JAK2 V617F (in 60%), CALR (in 20%), or MPL in (3%). • Confirmation of the diagnosis requires a bone marrow biopsy. • The risk of thrombosis is estimated on the basis of a history of thrombosis, the presence of a JAK2 V617F mutation, an age older than 60 years, and cardiovascular risk factors. Patients are classified into four risk categories: very low, low, intermediate, and high. • Low-dose aspirin is a cornerstone of the therapy used to reduce the risk of thrombosis, except in patients with very-low-risk disease. • Cytoreductive therapy is indicated in patients with high-risk disease but is of uncertain benefit in those with intermediate-risk disease. Hydroxyurea or interferon alfa-2a are considered first-line options when therapy is indicated, with the former used more frequently.

hands or feet). After a median follow-up of 10 cation system defines major and minor criteria years (with some patients followed for up to 47 for the formal diagnosis of the specific myelo- years), death had occurred in 43%, leukemic proliferative neoplasm.6,7 Table 1 outlines the transformations in 4%, myelofibrosis in 13%, WHO criteria used to distinguish essential throm- and new thrombosis in 21%. The median overall bocythemia from prefibrotic myelofibrosis, which survival was 18 years (26.7 years among patients include morphologic assessment of bone mar- with a low risk of thrombosis), a significantly row.8 Figure 1 shows features of bone marrow shorter survival rate than that of age- and sex- that are less frequent causes of primary throm- matched controls from the general population.5 bocytosis and that can mimic essential thrombocythemia.9,10 Testing can start with peripheral- Strategies and Evidence blood screening to detect driver mutations; the majority of patients with essential thrombocy- Diagnosis and Evaluation themia (approximately 60%) express the JAK2 In patients with persistent acquired thrombocy- V617F mutation, whereas the remainder harbor tosis, the presence of any of the following symp- the mutations in CALR (approximately 20%) or toms or conditions suggests underlying myeloid MPL (approximately 3%) or none of the three cancer: vasomotor symptoms (e.g., migraine or driver mutations (10 to 20% are triple-negative).11 pruritus), constitutional symptoms (e.g., night Although a bone marrow biopsy is required for sweats and weight loss), splenomegaly, throm- conclusive diagnosis, the presence of a driver bosis at multiple sites or at an unusual site (e.g., mutation, combined with appropriate clinical the hepatic, portal, or splanchnic veins), or an features (e.g., normal levels of hemoglobin and abnormal blood smear (e.g., detection of leuko- serum lactate dehydrogenase and the absence of erythroblastosis or dacryocytes [teardrop-shaped leukoerythroblastosis or dacryocytes), strongly cells]). The exclusion of secondary or reactive suggests a diagnosis of essential thrombocy- causes of thrombocytosis is based on clinical themia.6 assessment, supplemented by findings on peripheral-blood smear. Additional laboratory tests (e.g., Risk Stratification serum ferritin level and erythrocyte sedimenta- Survival and Leukemic and Fibrotic Transformation tion rate) and clinically directed imaging studies Despite the overlap in some of the clinical, for the diagnosis of underlying infection or can- pathologic, and molecular features of essential cer may be indicated in select cases. thrombocythemia and prefibrotic myelofibrosis, Common causes of primary (clonal) throm- a key step in prognosis is distinguishing the two bocytosis, aside from essential thrombocythe- conditions, given their disparate clinical out- mia, include primary myelofibrosis (either the comes (Table 1). In a large international study early or prefibrotic variant) and polycythemia vera. involving patients with essential thrombocythe- The World Health Organization (WHO) classifi- mia or prefibrotic myelofibrosis, the respective

2136 n engl j med 381;22 nejm.org November 28, 2019 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. Clinical Practice

Table 1. WHO Diagnostic Criteria for Essential Thrombocythemia and Prefibrotic or Early-Stage Myelofibrosis.*

Essential Thrombocythemia Prefibrotic or Early-Stage Myelofibrosis Diagnosis requires all major criteria or the first three major Diagnosis requires all major criteria and at least one mi- criteria plus a minor criterion. nor criterion. Major criteria Platelet count ≥450,000 per cubic millimeter Megakaryocytic proliferation and atypia, without reticulin Bone marrow biopsy showing proliferation mainly of the fibrosis >grade 1, accompanied by increased, age- megakaryocytic lineage, with increased numbers of adjusted bone marrow cellularity, granulocytic prolif- enlarged, mature megakaryocytes with hyperlobulated eration, and in many cases decreased erythropoiesis nuclei; no substantial increase or left shift in neutrophil Criteria for BCR-ABL1-positive chronic myeloid leukemia, granulopoiesis or erythropoiesis; in rare instances, minor polycythemia vera, essential thrombocythemia, my- (grade 1) increase in reticulin fibers elodysplastic syndrome, or other myeloid neoplasm Criteria for BCR-ABL1–positive chronic myeloid leukemia, not met polycythemia vera, primary myelofibrosis, or other JAK2 V617F, CALR, or MPL mutation or presence of an- myeloid neoplasm not met other clonal marker or of minor reactive reticulin fi- JAK2 V617F, CALR, or MPL mutation brosis in bone marrow† Minor criteria Presence of clonal marker or of evidence of reactive throm- Anemia not attributed to a coexisting condition bocytosis Leukocytosis (≥11,000 cells per cubic millimeter) Palpable splenomegaly Lactate dehydrogenase level above upper limit of normal of institutional reference range

* Data are from Arber et al.7 † In the absence of any of the three major clonal mutations, a search for other mutations associated with myeloid neoplasms (e.g., ASXL1, EZH2, TET2, IDH1, IDH12, SRSF2, and SF3B1 mutations) may be helpful in determining the clonal nature of the disease. Minor (grade 1) reticulin fibrosis caused by infection is noteworthy, as are autoimmune disorders or other chronic inflammatory conditions, hairy-cell leukemia or other lymphoid neoplasms, metastatic cancer, or toxic (i.e., chronic) myelopathies. rates of 10-year survival were 89% and 76%, neoplasms that may improve prognostication.13,15 leukemic transformation 0.7% and 5.8%, and In a recent collaborative study involving 502 pa- progression of fibrosis 0.8% and 12.3%.8 The tients with essential thrombocythemia who under- most important determinant of survival among went targeted next-generation sequencing, approx- patients with essential thrombocythemia is age, imately 15% were found to harbor one or more with median survival ranging from 8.1 years nondriver mutations in SRSF2, U2AF1, SF3B1, or among patients older than 70 years of age to TP5312,13; patients who had mutations in SRSF2 34.7 years among patients younger than 40 years and SF3B1, in SF3B1 and U2AF1, or in TP53 had of age (Fig. S1 in the Supplementary Appendix, significantly lower rates of overall, myelofibrosis- available with the full text of this article at free, and leukemia-free survival, respectively, than NEJM.org). The marked effect of age on survival patients who did not have these patterns of mu- is often overlooked, and patients are often quoted tation. These molecular predictors have been com- median survival figures that do not necessarily bined with clinical variables (age >60 years and age categories. According to the International male sex) in a newer prognostic model for sur- Prognostic Score for Essential Thrombocythe- vival among patients with essential thrombocy- mia (IPSET), an age of 60 years or more, leuko- themia: Mutation-Enhanced International Prog- cytosis (leukocyte count ≥11,000 per cubic milli- nostic Systems for Essential Thrombocythemia meter), and a history of thrombosis are key (MIPSS-ET)12; validation of the predictive role of variables for survival.14 However, further refine- these mutations and of the model is needed. ments have since been published that incorpo- rate molecular markers of disease.12 Thrombosis and Bleeding More recent studies have identified patho- Thrombosis in essential thrombocythemia may genic genetic aberrations beyond the traditional be arterial or venous. In the Mayo Clinic study driver mutations relevant to myeloproliferative noted above,5 arterial thrombosis was present at

n engl j med 381;22 nejm.org November 28, 2019 2137 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

Reactive thrombocytosis Essential thrombocythemia Prefibrotic myelofibrosis

Myelodysplastic and myeloproliferative overlap with ring sideroblasts Usually no anemia, and red-cell Anemia and possibility of indexes, including MCV and RDW, abnormal red-cell indexes are often normal (↑RDW, ↓MCV)

Peripheral-blood smear, Peripheral-blood smear, leukoerythroblastosis leukoerythroblastosis and and dacrocytosis absent dacrocytosis may be present

Serum LDH level usually Serum LDH Chronic myeloid leukemia within normal reference range level often elevated

With JAK2 mutation, With JAK2 mutation, mutant allele burden usually mutant allele burden may <20% exceed 20%

Figure 1. Morphologic Features of Megakaryocytes in Essential Thrombocythemia and Related Disorders. In reactive thrombocytosis, megakaryocytes have normal morphologic features. In essential thrombocythemia, megakaryocytes are large, appear to be mature, and form loose clusters. In contrast, in prefibrotic myelofibrosis, megakaryocytes have hyperchromatic and irregularly folded nuclei and form tight clusters. In cases of myelodysplastic–myeloproliferative overlap, the characteristic morphologic feature is increased erythropoiesis with dyserythropoiesis. Ring sideroblasts constitute 15% or more of erythroid precursors, and iron- laden mitochondria form a perinuclear ring (arrow). Megakaryocyte numbers are increased and have pleomorphic morphologic features that are similar to those seen in myeloproliferative neoplasms. In chronic myeloid leukemia, the megakaryocytes are smaller than normal (so-called dwarf megakaryocytes). LDH denotes lactate dehydrogenase, MCV mean corpuscular volume, and RDW red-cell distribution width.

or before diagnosis in 14% of the patients, and essential thrombocythemia as defined by the venous thrombosis was present in 8%; during a WHO criteria who were recruited from multiple median follow-up of 9.9 years, another 15% of international centers of excellence.16 the patients had arterial thrombosis and another Conventional risk factors for thrombosis in 8% had venous thrombosis. These findings were essential thrombocythemia are an age greater similar to those described in 891 patients with than 60 years and a history of thrombosis.17,18

2138 n engl j med 381;22 nejm.org November 28, 2019 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. Clinical Practice

Recent studies have identified additional risk essential thrombocythemia is prevention of vas- factors; the revised IPSET-thrombosis model cular events, and treatment decisions are guided takes into account conventional risk factors as by risk stratification in accordance with the re- well as the presence of the JAK2 V617F mutation vised IPSET-thrombosis model (Table 2).21 and cardiovascular risk factors to categorize In the absence of cardiovascular risk factors, patients into four groups according to level of very-low-risk disease may be managed with ob- risk.16,19-21 In patients with conventionally low risk, servation alone.19,20,24 For patients with higher the absence of the JAK2 V617F mutation and levels of risk, low-dose aspirin (81 mg per day) cardiovascular risk factors defines “very-low-risk is the cornerstone of antithrombotic therapy, disease.” The presence of JAK2 V617F is associ- especially in those with JAK2 V617F mutations. ated with a significantly increased risk of throm- Although data from randomized trials are lack- bosis (categorized as “low risk”).19-21 Patients ing, benefit is inferred from data from a con- older than 60 years of age who have neither a trolled trial involving patients with polycythemia history of thrombosis nor the JAK2 V617F muta- vera26 and from retrospective studies.25,27,28 Some tion are considered to be at “intermediate risk”; experts routinely test for acquired von Wille- their risk of thrombosis is significantly lower than brand’s disease in patients with a platelet count that of patients at “high risk,” who have a his- of 1 million or more per cubic millimeter and tory of thrombosis or the JAK2 V617F mutation. withhold aspirin if the activity of von Wille- In patients whose high-risk status was defined brand’s factor is less than 30% or if there is a by prior thrombosis, no incremental risk of re- history of bleeding. Emerging data suggest that current thrombosis was conferred by the JAK2 once-daily low-dose aspirin is insufficient to V617F mutation or cardiovascular risk factors.19-21 produce adequate (24-hour) suppression of plate-

In a retrospective study involving 891 patients let thromboxane A2 synthesis owing to acceler- with essential thrombocythemia for whom the ated platelet turnover in essential thrombocythe- median follow-up from the time of diagnosis mia but that twice-daily aspirin does result in was 6.2 years (range, 0 to 27.0), major bleeding effective suppression.29,30 Thus, twice-daily aspirin (most commonly gastrointestinal) was observed has been recommended for some patients (e.g., in 4% at or before diagnosis and in 6% at follow- those with high-risk disease and prior arterial up22; predictors of bleeding included previous thrombosis),19,31 but data from controlled trials hemorrhage and aspirin use. Bleeding in patients to inform efficacy and safety are currently lack- with essential thrombocythemia is in some ing.32 For patients older than 60 years of age who cases associated with acquired von Willebrand’s are not carriers of JAK2 V617F or who have no disease, which results from a selective loss of history of thrombosis (i.e., those with intermedi- large von Willebrand’s factor multimers. ate-risk disease), once-daily aspirin therapy is advised; additional therapy is generally not indi- Treatment cated in asymptomatic patients. In those with Despite having a higher risk of death than the high-risk disease (a previous vascular event or general population, patients with essential throm- the presence of JAK2 V617F in an older patient), bocythemia, especially those with morphologi- the combination of hydroxyurea therapy and cally confirmed disease, have life expectancies once-daily aspirin therapy is the standard of care that are only slightly lower than those of per- and is recommended on the basis of data from sons without the disease and generally better controlled studies that have shown a significant than those of patients with other myeloprolifera- reduction in the risk of thrombosis with this tive neoplasms.8,23 Drug therapy proffered in ac- treatment. cordance with the criteria for essential thrombo- Limited data from controlled trials are avail- cythemia as defined by the WHO has not been able to guide selection of therapy in patients with shown to increase survival or prevent progres- high-risk essential thrombocythemia. Although sion of leukemia or fibrosis in controlled trials not approved by the Food and Drug Administra- (Table S1), although these trials were not pow- tion (FDA) for this indication, hydroxyurea is the ered to detect differences in these end points. most frequently prescribed first-line drug for The major treatment objective in patients with patients with essential thrombocythemia given

n engl j med 381;22 nejm.org November 28, 2019 2139 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

Table 2. NCCN and ELN Guidelines for Risk Stratification and Treatment in Patients with Essential Thrombocythemia.*

Guideline Very Low Risk† Low Risk† Intermediate Risk†‡ High Risk NCCN33 Patient characteristics Age ≤60 yr, no prior throm- Age ≤60 yr, no prior throm- Age >60 yr, no prior throm- Age >60 yr, no prior thrombosis, JAK2 V617F mu- bosis, JAK2 V617F mu- bosis, JAK2 V617F mu- bosis, JAK2 V617F mu tation absent tation present tation absent tation present Rate of thrombosis 0.44%/yr, with no cardio- 1.59%/yr with no cardio 1.44%/yr with no cardio- 2.36%/yr with no cardio vascular risk factors; vascular risk factors; vascular risk factors; vascular risk factors; 1.05%/yr with risk 2.57%/yr with risk 1.64%/yr with risk 4.17%/yr with risk factors factors factors factors Management of cardiovas- Aspirin, 81–100 mg/day for Aspirin, 81–100 mg/day for Aspirin, 81–100 mg/day for Aspirin, 81–100 mg/day for cular risk factors vascular symptoms§ vascular symptoms§ vascular symptoms§ vascular symptoms§ Treatment Cytoreductive therapy Cytoreductive therapy Cytoreductive therapy First-line therapy with hy- not recommended not recommended not recommended droxyurea or interferon as initial treatment¶ as initial treatment¶ as initial treatment¶ alfa-2a or anagrelide, second-line therapy with hydroxyurea, interferon alfa-2a,‖ or anagrelide, or referral to clinical trial ELN18 Patient characteristics** Rate of thrombosis** — Score of 0–1, 1.03%/yr Score of 2, 2.35%/yr Score ≥3, 3.56%/yr Management of cardiovas- — Low-dose aspirin for micro- Low-dose aspirin for micro- Low-dose aspirin for micro- cular risk factors vascular symptoms§ vascular symptoms§ vascular symptoms§ Treatment First line — Cytoreductive therapy Cytoreductive therapy First-line therapy, hydroxy- not recommended not recommended urea or interferon alfa-2a for initial treatment¶ for initial treatment¶ Second line — — — Cytoreductive therapy with interferon alfa-2a or anagrelide

* The guidelines from the National Comprehensive Cancer Network (NCCN) and European LeukemiaNet (ELN)are based on the International Prognostic Score for Essential Thrombocythemia (IPSET). The NCCN guidelines are in accordance with the revised IPSET guidelines (i.e., IPSET-thrombosis), published in 2015. The ELN guidelines are in accordance with the original IPSET guidelines, published in 2012. † The risk of thrombosis is estimated on the basis of a history of thrombosis, the presence of a JAK2 V617F mutation, an age older than 60 years, and cardiovascular risk factors. ‡ There was a significant difference in the risk of thrombosis between patients at low risk and those at very low risk (P<0.001 for thrombosis-free survival) and between those at intermediate risk and those at high risk (P = 0.008 for thrombosis-free survival). The difference in the risk of thrombosis between those at low risk and those at intermediate risk was not significant.21 § Aspirin should be used with caution in patients with a platelet count of 1 million or more per cubic millimeter, acquired von Willebrand’s disease, or bleeding. One retrospective study indicated that aspirin therapy may not reduce the risk of thrombosis and may instead increase the risk of bleeding among patients with low-risk, CALR-mutated essential thrombocythemia.24 Data from another retrospective study suggest that outcomes are not improved with aspirin use in patients with essential thrombocythemia who have “very-low-risk” disease (i.e., those who are ≤60 years of age, have no history of thrombosis, and do not have the JAK2 V617F mutation) and who do not have cardiovascular risk factors.25 Expert recommendations for consideration of twice-daily low-dose aspirin in patients with essential thrombocythemia include patients at low risk for thrombosis with cardiovascular risk factors, patients at intermediate risk for thrombosis, and patients with arterial thrombosis who are at high risk for repeat thrombosis. ¶ Potential indications for starting therapy include development of new thrombosis, presence of acquired von Willebrand’s disease or disease- related major bleeding, extreme thrombocytosis (platelet count ≥1 million per cubic millimeter) or extreme leukocytosis (e.g., white-cell count >15,000 per cubic millimeter), splenomegaly, vasomotor symptoms not responsive to aspirin therapy, or other uncontrolled, disease- related symptoms (e.g., pruritus and night sweats). ‖ Interferon α-2b, peginterferon α-2a, or peginterferon α-2b may be considered in younger patients (≤60 years of age) or in pregnant patients in need of therapy or in those in need of therapy who prefer to not use hydroxyurea. ** Risk stratification is calculated according to patient characteristics. A score of 0 to 1 point indicates low risk, 2 points intermediate risk, and 3 points or more high risk. Risk factors include the JAK2 V617F mutation (2 points), a history of thrombosis (2 points), older age (>60 years, 1 point), and cardiovascular risk factors (1 point).16

2140 n engl j med 381;22 nejm.org November 28, 2019 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. Clinical Practice its ease of administration, good profile regard- feron alfa-2a or busulfan. The advantages of ing efficacy and side effects, and low incidence pegylated interferon alfa-2a include the absence of treatment resistance.18,31,33 A controlled trial of genotoxic and leukemogenic effects, high rates with a median follow-up of 27 months showed of complete hematologic response, and anticlonal significantly fewer thrombotic events in patients activity, with a minority of patients having a with high-risk disease who were treated with complete molecular response.44,45 However, longer- hydroxyurea than in those who received no myelo- term follow up has shown a yearly discontinua- suppressive therapy (3.6% vs. 24%; P = 0.003).34 tion rate of 5 to 16% owing to drug-related Possible exceptions to the use of hydroxyurea toxicity, leukemic and fibrotic transformation, include younger patients (<40 years of age), par- or loss of response.44 Researchers who conduct- ticularly women of childbearing age, owing to ed a phase 3 trial in which pegylated interferon concerns regarding gonadal toxicity and muta- alfa-2a was compared with hydroxyurea in pa- genicity. Hydroxyurea also has teratogenic and tients with high-risk essential thrombocythemia embryotoxic effects and is consequently not con- or polycythemia vera who had not previously sidered safe for use during pregnancy. Observa- received treatment did not report a significant tional studies have not indicated an increased difference in complete hematologic response rates risk of leukemic transformation with hydroxy- in the two treatment groups.46 Subcutaneous urea use among patients with myeloproliferative pegylated interferon alfa-2a is usually initiated at neoplasms; in contrast, treatment with radio a dose of 45 μg once a week and gradually in- active phosphorus (P32), alkylators (e.g., chloram- creased to 180 μg, depending on side effects. bucil), or pipobroman has been associated Busulfan therapy is also associated with high with significantly higher leukemic risk, particular hematologic response rates; a concern regarding ly among in patients receiving sequential ther an increased risk of leukemia has not been sup- apies.35-41 Hydroxyurea is typically started at ported by retrospective studies.31,47 Oral busulfan 500 mg twice daily, and the dose is adjusted to is started at 4 mg per day, and treatment is with- keep the platelet count in the normal range. held when the platelet count falls to levels be- Anagrelide, which selectively reduces the tween 400,000 and 600,000 per cubic millimeter; platelet count, has been approved by the FDA for when the platelet count rises, treatment is re- essential thrombocythemia but is infrequently sumed at a lower dose (e.g., 2 mg per day). used owing to conflicting results from con- In a recent trial, ruxolitinib (a JAK1 and JAK2 trolled trials involving patients with essential inhibitor) was compared with the best available thrombocythemia who are at high risk for vas- therapy in patients with high-risk essential cular events and to the potential for cardiovascu- thrombocythemia who did not have a response lar toxicity, particularly in older patients. Where- to hydroxyurea or who had intolerable side ef- as the ANAHYDRET (Anagrelide vs. Hydroxyurea fects. The primary end point was a complete — Efficacy and Tolerability Study in Patients response rate. Patients in both groups had simi- with Essential Thrombocythemia) trial showed lar rates of response.48 that anagrelide was noninferior to hydroxyurea 42 in preventing vascular events, a trial in which Guidelines hydroxyurea plus low-dose aspirin was compared with anagrelide plus low-dose aspirin in Guidelines for the management of essential patients with essential thrombocythemia who thrombocythemia have been published by the Na- were at high risk for vascular events showed that tional Comprehensive Cancer Network (NCCN) hydroxyurea plus aspirin was superior to anagrelide and European LeukemiaNet. The guidelines are plus aspirin.43 Differences in study populations similar but differ slightly in risk stratification and in the use of aspirin in the latter trial may and in recommendations for first-line cytoreduc- explain the discrepant results. tive therapy (Table 2). In patients for whom the side effects of Our approach is generally concordant with hydroxyurea therapy are unacceptable, options these guidelines, with some caveats. We prefer (other than anagrelide) include pegylated inter- the four-tier risk-stratification model outlined in

n engl j med 381;22 nejm.org November 28, 2019 2141 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. The new england journal of medicine

the NCCN guidelines. Whereas these guidelines versus anagrelide is being evaluated in another recommend a low dose of aspirin (81 mg) once ongoing trial involving patients with essential daily, we consider twice-daily aspirin use in se- thrombocythemia that is resistant to hydroxyurea lected patients who have vasomotor symptoms or for whom its side effects are not tolerable that are not responsive to once-daily therapy. We (ClinicalTrials.gov number, NCT03123588). In pa- often use busulfan as a second-line agent in older tients who also have venous thromboembolism, patients. Finally, we do not use a platelet count the appropriate duration of anticoagulation ther- of 1,500,000 per cubic millimeter or higher as an apy remains uncertain. Observational data sug- automatic trigger to start cytoreductive therapy, gest a higher rate of recurrent thrombosis among unless disease-related hemorrhage or other symp- these patients than among patients with venous toms warrant such treatment. thromboembolism who do not have a myeloproliferative neoplasm, despite long-term treatment Areas of Uncertainty with vitamin K antagonists, and a doubling of the risk of recurrence after anticoagulation ther The most appropriate target platelet count in apy is discontinued.50 patients with essential thrombocythemia who are receiving therapy is uncertain. One observa- Conclusions and tional study suggested that time in remission Recommendations (defined as a platelet count of <600,000 per cubic millimeter) rather than platelet count per se has In the patient in the vignette, the presence of a significant (inverse) correlation with the throm- extreme thrombocytosis that has been progres- bosis rate in patients with essential thrombocy- sive over 3 years and an exon 9 mutation in CALR themia.17 In controlled trials, the target platelet are highly suggestive of essential thrombocythe- count for therapy in patients with high-risk dis- mia. A bone marrow biopsy should be per- ease has been less than 600,000 per cubic milli- formed to distinguish essential thrombocythe- meter or less than 400,000 per cubic millimeter. mia from its mimics, particularly prefibrotic In practice, it is not always possible to achieve a myelofibrosis.6 Once the diagnosis of essential normal platelet count because of intolerable side thrombocythemia has been confirmed, the pa- effects, and patients receive maintenance ther- tient’s risk of thrombosis should be stratified in apy on a maximum tolerated dose, regardless of accordance with the revised IPSET-thrombosis platelet count. It is also unclear whether a reduc- model (Table 2). tion in leukocyte count is needed for therapy to Since this patient is at intermediate risk for most effectively reduce the risk of thrombosis thrombosis, we would recommend low-dose as- among patients with essential thrombocythemia; pirin therapy, noting the presence of a cardiovas- in a recent meta-analysis, the presence of leuko- cular risk factor (hypertension) and after exclud- cytosis was shown to be significantly associated ing von Willebrand’s disease, given the extreme with an increased risk of arterial thrombosis.49 thrombocytosis at presentation. Although “ex- Unlike hydroxyurea, which provides a general treme thrombocytosis” is included in interna- effect, anagrelide does not correct leukocyto- tional society guidelines as one of the reasons to sis.42,43 Data are needed to better determine the consider cytoreductive therapy in patients at inter- benefits and risks of strategies for decreasing mediate risk,18,33 we would defer the therapy pend- the residual risk of thrombosis among patients ing development of thrombohemorrhagic com- with high-risk essential thrombocythemia (3.5% plications or other disease-related symptoms. If per year among those receiving current state-of- cytoreductive therapy is indicated, we would the-art treatment) (Table 2). select hydroxyurea as first-line therapy. If the In an ongoing trial (EudraCT number, 2016- patient is of childbearing age, we would use in- 002885-30), investigators compared the effects terferon alfa-2a; in patients for whom the side of twice- and thrice-daily aspirin regimens with effects of hydroxyurea are unacceptable or who once-daily use on thromboxane A2 production have resistance to its use, we would use pegylated in patients with intermediate- or high-risk essen- interferon alfa-2 in younger patients and busulfan tial thrombocythemia.32 The use of ruxolitinib in older patients (i.e., those ≥65 years of age).

2142 n engl j med 381;22 nejm.org November 28, 2019 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. Clinical Practice

No potential conflict of interest relevant to this article was Disclosure forms provided by the authors are available with reported. the full text of this article at NEJM.org.

References 1. Msaouel P, Lam AP, Gundabolu K, et al. Vera (MIPSS-PV). Blood 2018;132:Suppl: observation in low-risk essential throm- Abnormal platelet count is an indepen- 578. abstract. bocythemia with a CALR mutation. Haema- dent predictor of mortality in the elderly 13. Tefferi A, Lasho TL, Guglielmelli P, tologica 2016;101: 926-31. and is influenced by ethnicity. Haemato- et al. Targeted deep sequencing in polycy- 25. Alvarez-Larrán A, Cervantes F, Pereira logica 2014;99: 930-6. themia vera and essential thrombocythe- A, et al. Observation versus antiplatelet 2. Rose SR, Petersen NJ, Gardner TJ, mia. Blood Adv 2016;1:21-30. therapy as primary prophylaxis for throm- Hamill RJ, Trautner BW. Etiology of 14. Passamonti F, Thiele J, Girodon F, et al. bosis in low-risk essential thrombocythe- thrombocytosis in a general medicine A prognostic model to predict survival in mia. Blood 2010;116:1205-10. population: analysis of 801 cases with 867 World Health Organization-defined 26. Landolfi R, Marchioli R, Kutti J, et al. emphasis on infectious causes. J Clin Med essential thrombocythemia at diagnosis: Efficacy and safety of low-dose aspirin in Res 2012;4:415-23. a study by the International Working polycythemia vera. N Engl J Med 2004; 3. Hsieh RW, Ravindran A, Hook CC, Group on Myelofibrosis Research and 350:114-24. et al. Etiologies of extreme thrombocyto- Treatment. Blood 2012;120:1197-201. 27. De Stefano V, Rossi E, Carobbio A, sis: a contemporary series. Mayo Clin Proc 15. Grinfeld J, Nangalia J, Baxter EJ, et al. et al. Hydroxyurea prevents arterial and 2019;8: 1542-50. Classification and personalized progno- late venous thrombotic recurrences in pa- 4. Girodon F, Bonicelli G, Schaeffer C, sis in myeloproliferative neoplasms. N Engl tients with myeloproliferative neoplasms et al. Significant increase in the apparent J Med 2018;379: 1416-30. but fails in the splanchnic venous district: incidence of essential thrombocythemia 16. Barbui T, Finazzi G, Carobbio A, et al. pooled analysis of 1500 cases. Blood Can- related to new diagnostic criteria: a popu- Development and validation of an Inter- cer J 2018;8:112. lation-based study. Haematologica 2009; national Prognostic Score of Thrombosis 28. De Stefano V, Za T, Rossi E, et al. Re- 94:865-9. in World Health Organization-essential current thrombosis in patients with poly- 5. Szuber N, Mudireddy M, Nicolosi M, thrombocythemia (IPSET-thrombosis). cythemia vera and essential thrombocy- et al. 3023 Mayo Clinic patients with mye Blood 2012;120:5128-33. themia: incidence, risk factors, and effect loproliferative neoplasms: risk-stratified 17. Cortelazzo S, Viero P, Finazzi G, of treatments. Haematologica 2008;93:372- comparison of survival and outcomes data D’Emilio A, Rodeghiero F, Barbui T. Inci- 80. among disease subgroups. Mayo Clin Proc dence and risk factors for thrombotic 29. Pascale S, Petrucci G, Dragani A, et al. 2019;94: 599-610. complications in a historical cohort of Aspirin-insensitive thromboxane biosyn- 6. Barbui T, Thiele J, Gisslinger H, et al. 100 patients with essential thrombocy- thesis in essential thrombocythemia is The 2016 classification and diagnostic themia. J Clin Oncol 1990;8:556-62. explained by accelerated renewal of the criteria for myeloproliferative neoplasms: 18. Barbui T, Barosi G, Birgegard G, et al. drug target. Blood 2012;119:3595-603. document summary and in-depth discus- Philadelphia-negative classical myelopro- 30. Dillinger JG, Sideris G, Henry P, Bal sion. Blood Cancer J 2018;8:15. liferative neoplasms: critical concepts and dit Sollier C, Ronez E, Drouet L. Twice 7. Arber DA, Orazi A, Hasserjian R, et al. management recommendations from Eu- daily aspirin to improve biological aspirin The 2016 revision to the World Health Or- ropean LeukemiaNet. J Clin Oncol 2011; efficacy in patients with essential throm- ganization classification of myeloid neo- 29:761-70. bocythemia. Thromb Res 2012;129:91-4. plasms and acute leukemia. Blood 2016; 19. Barbui T. Refining prognostication of 31. Tefferi A, Barbui T. Polycythemia vera 127:2391-405. thrombosis in ET. Am J Hematol 2016;91: and essential thrombocythemia: 2019 up- 8. Barbui T, Thiele J, Passamonti F, et al. 361-3. date on diagnosis, risk-stratification and Survival and disease progression in essen- 20. Haider M, Gangat N, Lasho T, et al. management. Am J Hematol 2019;94:133- tial thrombocythemia are significantly Validation of the revised International 43. influenced by accurate morphologic diag- Prognostic Score of Thrombosis for Es- 32. De Stefano V, Rocca B, Tosetto A, et al. nosis: an international study. J Clin Oncol sential Thrombocythemia (IPSET-throm- The Aspirin Regimens in Essential Throm- 2011;29: 3179-84. bosis) in 585 Mayo Clinic patients. Am J bocythemia (ARES) phase II randomized 9. Breccia M, Cannella L, Diverio D, et al. Hematol 2016;91:390-4. trial design: Implementation of the serum

Isolated thrombocytosis as first sign of 21. Barbui T, Vannucchi AM, Buxhofer- thromboxane B2 assay as an evaluation chronic myeloid leukemia with e6a2 BCR/ Ausch V, et al. Practice-relevant revision tool of different aspirin dosing regimens ABL fusion transcript, JAK2 negativity of IPSET-thrombosis based on 1019 pa- in the clinical setting. Blood Cancer J 2018; and complete response to imatinib. Leuk tients with WHO-defined essential throm- 8:49. Res 2008;32:177-80. bocythemia. Blood Cancer J 2015;5:e369. 33. National Comprehensive Cancer Net- 10. Wang SA, Hasserjian RP, Loew JM, et al. 22. Finazzi G, Carobbio A, Thiele J, et al. work. NCCN guidelines: myeloprolifera- Refractory anemia with ringed sidero- Incidence and risk factors for bleeding in tive neoplasms, version 2. 2019 (https:// blasts associated with marked thrombo- 1104 patients with essential thrombocy- www.nccn.org/professionals/physician_gls/ cytosis harbors JAK2 mutation and shows themia or prefibrotic myelofibrosis diag- pdf/mpn_blocks .pdf). overlapping myeloproliferative and myelo- nosed according to the 2008 WHO crite- 34. Cortelazzo S, Finazzi G, Ruggeri M, dysplastic features. Leukemia 2006;20: ria. Leukemia 2012;26:716-9. et al. Hydroxyurea for patients with essen- 1641-4. 23. Hultcrantz M, Kristinsson SY, Anders- tial thrombocythemia and a high risk of 11. Tefferi A, Pardanani A. Myeloprolif- son TM, et al. Patterns of survival among thrombosis. N Engl J Med 1995;332:1132-6. erative neoplasms: a contemporary review. patients with myeloproliferative neo- 35. Fruchtman SM, Mack K, Kaplan ME, JAMA Oncol 2015;1: 97-105. plasms diagnosed in Sweden from 1973 Peterson P, Berk PD, Wasserman LR. 12. Tefferi A, Guglielmelli P, Lasho TL, to 2008: a population-based study. J Clin From efficacy to safety: a Polycythemia et al. Mutation-Enhanced International Oncol 2012;30: 2995-3001. Vera Study Group report on hydroxyurea Prognostic Systems for Essential Throm- 24. Alvarez-Larrán A, Pereira A, Gugliel- in patients with polycythemia vera. Semin bocythemia (MIPSS-ET) and Polycythemia melli P, et al. Antiplatelet therapy versus Hematol 1997;34:17-23.

n engl j med 381;22 nejm.org November 28, 2019 2143 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved. Clinical Practice

36. Barbui T, Ghirardi A, Masciulli A, et al. Acute leukemia in polycythemia vera: an plasms–Research Consortium (MPN-RC) Second cancer in Philadelphia negative analysis of 1638 patients enrolled in a 112 randomized trial of pegylated inter- myeloproliferative neoplasms (MPN-K): prospective observational study. Blood feron alfa-2a (PEG) versus hydroxyurea a nested case-control study. Leukemia 2005;105: 2664-70. (HU) therapy for the treatment of high 2019;33:1996-2005. 42. Gisslinger H, Gotic M, Holowiecki J, risk polycythemia vera (PV) and high risk 37. Björkholm M, Derolf AR, Hultcrantz et al. Anagrelide compared with hydroxy- essential thrombocythemia (ET). Blood M, et al. Treatment-related risk factors for urea in WHO-classified essential throm- 2018;132:Suppl.577. abstract. transformation to acute myeloid leukemia bocythemia: the ANAHYDRET Study, a 47. Shvidel L, Sigler E, Haran M, et al. Bu- and myelodysplastic syndromes in myelo- randomized controlled trial. Blood 2013; sulphan is safe and efficient treatment in proliferative neoplasms. J Clin Oncol 2011; 121:1720-8. elderly patients with essential thrombo- 29:2410-5. 43. Harrison CN, Campbell PJ, Buck G, et cythemia. Leukemia 2007;21:2071-2. 38. Berk PD, Goldberg JD, Silverstein MN, al. Hydroxyurea compared with anagrelide 48. Harrison CN, Mead AJ, Panchal A, et al. et al. Increased incidence of acute leuke- in high-risk essential thrombocythemia. Ruxolitinib vs best available therapy for mia in polycythemia vera associated with N Engl J Med 2005;353:33-45. ET intolerant or resistant to hydroxy- chlorambucil therapy. N Engl J Med 1981; 44. Masarova L, Patel KP, Newberry KJ, carbamide. Blood 2017;130:1889-97. 304:441-7. et al. Pegylated interferon alfa-2a in pa- 49. Carobbio A, Ferrari A, Masciulli A, 39. Kiladjian JJ, Chevret S, Dosquet C, tients with essential thrombocythaemia Ghirardi A, Barosi G, Barbui T. Leukocy- Chomienne C, Rain JD. Treatment of or polycythaemia vera: a post-hoc, median tosis and thrombosis in essential throm- polycythemia vera with hydroxyurea and 83 month follow-up of an open-label, bocythemia and polycythemia vera: a sys- pipobroman: final results of a random- phase 2 trial. Lancet Haematol 2017;4(4): tematic review and meta-analysis. Blood ized trial initiated in 1980. J Clin Oncol e165-e175. Adv 2019;3:1729-37. 2011;29:3907-13. 45. Quintás-Cardama A, Abdel-Wahab O, 50. De Stefano V, Ruggeri M, Cervantes F, 40. Kaplan ME, Mack K, Goldberg JD, Manshouri T, et al. Molecular analysis of et al. High rate of recurrent venous Donovan PB, Berk PD, Wasserman LR. patients with polycythemia vera or essen- thromboembolism in patients with mye Long-term management of polycythemia tial thrombocythemia receiving pegylated loproliferative neoplasms and effect of vera with hydroxyurea: a progress report. interferon α-2a. Blood 2013;122:893-901. prophylaxis with vitamin K antagonists. Semin Hematol 1986;23: 167-71. 46. Mascarenhas J, Kosiorek HE, Prchal JT, Leukemia 2016;30:2032-8. 41. Finazzi G, Caruso V, Marchioli R, et al. et al. Results of the Myeloproliferative Neo- Copyright © 2019 Massachusetts Medical Society.

specialties and topics at nejm.org Specialty pages at the Journal’s website (NEJM.org) feature articles in cardiology, endocrinology, genetics, infectious disease, nephrology, pediatrics, and many other medical specialties.

2144 n engl j med 381;22 nejm.org November 28, 2019 The New England Journal of Medicine Downloaded from nejm.org by ERIC MCCARTT on May 13, 2020. For personal use only. No other uses without permission. Copyright © 2019 Massachusetts Medical Society. All rights reserved.