Polycythemia vera is a myeloproliferative neoplasm characterized by clonal hematopoiesis and an absolute increase in the red blood cell mass, with an associated leukocytosis and thrombocytosis. Virtually all patients with PV harbor a mutation in the nonreceptor tyrosine kinase JAK2, with the majority of patients harboring the classic JAK2.
John Mascarenhas, MD,
Polycythemia vera (PV) is a myeloproliferative neoplasm characterized by clonal hematopoiesis and an absolute increase in the red blood cell mass, with an associated leukocytosis and thrombocytosis. Virtually all patients with PV harbor a mutation in the nonreceptor tyrosine kinase JAK2, with the majority of patients harboring the classic JAK2 (V617F). This point mutation inactivates the pseudokinase domain, resulting in constitutive enzymatic activity and intracellular signal transduction, ultimately leading to uncontrolled cellular proliferation. The general management of PV is determined by risk assessment based on age and history of thrombosis and attention to modifiable cardiovascular risk factors. All patients lacking a clear contraindication should receive aspirin thromboprophylaxis and therapeutic phlebotomy to maintain the hematocrit below 45%. High-risk patients should be considered for treatment with cytoreductive therapies, of which hydroxyurea is considered first line. For patients with hydroxyurea-resistant disease or those who are unable to tolerate this chemotherapy, the selective JAK1/2 inhibitor ruxolitinib (Jakafi) is an approved second-line option. Ruxolitinib was approved by the FDA for the treatment of hydroxyurea-refractory or intolerant PV based on the results of the randomized, phase III RESPONSE trial. Ruxolitinib therapy was associated with a superior hematologic response compared with best available therapy, and the response was shown to be durable at an 80-week follow-up. Additionally, spleen volume and symptom burden were significantly reduced with ruxolitinib treatment. Trials evaluating the efficacy of other JAK2 inhibitors for PV, such as fedratinib and momelotinib, are also being conducted. Novel agents that target pathways distinct from the JAK-STAT pathway are being tested with early evidence of activity from histone deacetylase inhibitors and MDM2 inhibitors. JAK2 inhibition for PV remains a viable option for a subset of patients that fail hydroxyurea and are particularly in need of symptomatic control. Bone marrow pathologic response and/or molecular response with JAK2 inhibitor therapy is not anticipated and, therefore, the evaluation of other rationally based treatments is under active evaluation with an aim for disease course modification.
Polycythemia vera (PV) is a myeloproliferative neoplasm (MPN) characterized by clonal hematopoiesis with an absolute increase in red blood cell mass and a propensity for leukocytosis and thrombocytosis. It is typically an insidious disease affecting older patients, and it often initially comes to the attention of a hematologist after hematologic abnormalities are noted on routine laboratory studies or, in some cases, in the setting of thrombotic events or microvascular disturbances. The 2016 World Health Organization guidelines for establishing a diagnosis of PV include these major criteria: an increased red cell mass (based on hemoglobin >16.5g/dL or hematocrit >49% in a man, or hemoglobin >16 g/ dL or hematocrit >48% in a woman), a bone marrow biopsy showing hypercellularity for age, and the presence of a JAK2 V617F or JAK2 exon 12 mutation.1The sole minor criterion is a subnormal serum erythropoietin level, which is useful to establish the diagnosis in the absence of a detected mutation in JAK2. The life expectancy of patients with PV is significantly shorter when compared with age- and sex-matched controls.2The risk-adapted treatment approach of PV is based upon the risk of disease-related complications and is essentially aimed at reducing the short-term risk of thrombosis. More recently, an appreciation of symptom burden associated with PV driven by heightened inflammatory cytokines (fatigue, pruritus), microvascular disturbances (headaches, decreased concentration), and splenomegaly (early satiety, abdominal discomfort) has brought attention to therapies such as JAK2 inhibitors that effectively ameliorate these. The focus of current drug development are therapies such as interferon-α intended to alter the natural history of this chronic and progressive myeloid malignancy, preventing its progression, for instance, to myelofibrosis (MF) and acute leukemia (AL).
Arterial thrombosis (occurring in approximately 16% of patients before or at diagnosis) and venous thrombosis (occurring in approximately 7.4% of patients before or at diagnosis)2are common causes of morbidity and mortality in patients with PV. Population-based, prospective cohort studies have demonstrated that advancing age and/or a previous history of thrombosis are the 2 most important independent prognostic factors associated with thrombotic risk in patients with PV.3Therefore, patients 60 years or older or with a history of a thrombotic (arterial or venous) event are categorized as high risk,4while those aged less than 60 years and without a previous thrombotic event are classified as low risk. PV patients can be classified as intermediate risk if they are low risk as defined above but have any classic modifiable cardiovascular risk factor, such as diabetes, hypertension, dyslipidemia, or recent/active tobacco use.
Prognosticating risk of hematologic transformation to MF or AL is less well defined. A longer duration of disease is associated with risk of developing MF, and advanced age may be associated with increased risk of developing AL. Other risk factors for leukemic transformation may include female gender and, possibly, low baseline cholesterol levels.5The association between hydroxyurea (HU) and leukemic transformation is controversial and will be discussed later, but a well-documented association exists between historic therapeutic agents such as pipobroman, busulfan, and phosphorus 32 (32P) and risk of hematologic transformation.5
As previously mentioned, patients with PV have a shorter median survival compared with age- and sex-matched controls. Transformation to AL and thrombotic complications contribute relatively equally to mortality, as does the development of secondary, unrelated malignancies.2 Predictive models have been developed to prognosticate median survival based largely on age, leukocytosis, and thrombotic events.2As in all hematologic malignancies, the roles of genetic and molecular profiling are expanding and being investigated as other ways to prognosticate disease risk and leukemic transformation. As an example, 1 prospective study including more than 1500 patients found an abnormal karyotype to be a risk factor for leukemic transformation in both univariate and multivariate analyses (HR, 3.9; 95% CI, 1.2-13.1).5Furthermore, the JAK2 mutational load has been shown to correlate directly with splenomegaly, leukocytosis, and CD34 count in patients with PV.6
The management of PV is based upon risk assessment and modifiable patient risk factors (Figure). All patients with baseline cardiovascular risk factors such as diabetes and hypertension should have these optimized. In patients in whom antiplatelet therapy is not contraindicated, low-dose daily aspirin can reduce the risk of thrombotic complications and death from cardiovascular causes.7Additionally, all patients with PV should receive therapeutic phlebotomy as needed to maintain the hematocrit below 45% (although some experts advocate for a nonevidence-based target hematocrit below 42% in females).8It is unknown if maintenance of the platelet count at ≤400 x 109/L may reduce the risk of thrombotic or hemorrhagic events,9 although it is included in the most recent European LeukemiaNet criteria for assessing response to therapy.10Additionally, treating thrombocytosis in PV may reduce the risk of hemorrhagic events that are due to acquired type 2 von Willebrand syndrome.9
Traditional Cytoreductive Therapies
Patients with high-risk PV should be treated with cytoreductive therapy in addition to standard management with aspirin and therapeutic phlebotomy. The first major clinical trial to evaluate the optimal management of PV, PVSG-01, randomized 431 patients to phlebotomy alone versus phlebotomy in conjunction with either 32P or chlorambucil. The results indicated that patients in the phlebotomy-only arm had a better overall median survival, but they suffered from excess mortality within the first 4 years post randomization due to an increased frequency of thrombotic complications.11After PVSG-01 came the early termination of a follow-up study, PVSG-05, that evaluated phlebotomy plus aspirin (900 mg/day) and dipyridamole versus phlebotomy and 32P; the cause was excess gastrointestinal hemorrhage and poor outcomes in the aspirin/dipyridamole group.12 Finally, the PSVG-08 study compared patients receiving HU for a median of almost 9 years versus historical controls (who did not require cytoreductive therapy) from the phlebotomy-only arm of PVSG-01.13 The PVSG-08 results demonstrated a significantly lower rate of thrombosis in the HU arm versus the PVSG-01 phlebotomy-only arm (9.8% versus 32.8%) and a nonsignificant increase in rate of leukemic transformation in the HU arm (5.4% vs 1.5%; P = .18).
Largely based on the results described, HU has become the first-line therapy for cytoreduction in patients with high-risk PV. HU is also preferable to other cytoreductive agents because it has not been confirmed to confer leukemogenic risk, as demonstrated by numerous retrospective and prospective studies.2,5,14Although HU is effective in the majority of treated patients, approximately 25% of patients with PV will either manifest HU resistance or develop therapy intolerance.15HU resistance can be defined as a need for continued phlebotomy; ongoing thrombocytosis or leukocytosis; or a suboptimal reduction in splenomegaly or splenomegaly-related symptoms after at least 3 months of the maximum tolerated dose of HU.16Intolerance to HU includes any peripheral cytopenia at the lowest effective dose, or common therapy-limiting adverse effects (AEs) such as gastrointestinal toxicity, fevers, and mucocutaneous toxicity, among others, at any dose of the medication at any duration of therapy.16
Not only do HU resistance and intolerance complicate the management and therapeutic options available to a subgroup of patients with PV, but they also portend a poor prognosis, a retrospective analysis indicates. In a cohort study of 261 patients with PV from a Spanish registry, resistance to HU implied a 5.6-fold increase in mortality and an almost 7-fold increase in risk of leukemic transformation.15Another study by the same researchers demonstrated that HU resistance, based solely upon an ongoing phlebotomy requirement of at least 3 per year, was significantly associated with a higher rate of thrombosis.17Finally, the development of peripheral cytopenia in patients intolerant of HU has been demonstrated to significantly correlate with both risk of myelofibrotic transformation as well as leukemic transformation.18
Cytoreductive options are limited for PV patients with HU-intolerant or -resistant disease. Potential cytoreductive agents utilized in Europe as second-line therapy after HU failure include interferon-α, busulfan, and pipobroma. Interferon has been shown to have considerable efficacy in the treatment of both low-risk patients and those with advanced disease, and it has prospective evidence of molecular remission in a subset of treated patients.19,20 Although interferon use is largely limited by AEs and poor
tolerance, the main limitation of other second-line cytoreductive therapies is their shared potential leukemogenic effect,2,14which has largely relegated their use to individuals with shorter life expectancies.
JAK2 Inhibitor Therapy: Indications and Evidence
Although PV was first described in the late 1800s by Louis Henri Vaquez,21 the clonal nature of PV wasn’t elaborated upon in detail until 1976,22and the JAK2 V617F mutation was not identified until 2005.23 Approximately 95% of patients with PV harbor the JAK2 V617F mutation,24while most of the remaining 5% have somatic mutations affecting exon 12 of JAK2.25 The mutated JAK2 codes for a nonreceptor tyrosine kinase that belongs to the Janus kinase family of proteins, which phosphorylate cytoplasmic targets such as signal transducers and regulators of downstream transcription (STATs). This phosphorylation cascade ultimately results in the upregulation of genes responsible for cellular proliferation and differentiation. The V617F mutation is located within the auto-inhibitory domain of the tyrosine kinase,26which helps explain why the JAK2 V617F mutation is associated with a loss of autoregulation and subsequent uncontrolled cellular proliferation. In contrast to the loss of regulation accompanying the V617F mutation, mutations in exon 12 act as gain-of-function aberrations through a pathway that remains incompletely elucidated.27
Ruxolitinib (Jakafi) is a selective JAK1/2 inhibitor that was first approved in the United States by the FDA for the treatment of intermediate- or high-risk MF, largely based on the results of the phase III COMFORT and COMFORT-II trials.28,29While JAK2 is primarily involved in receptor-mediated signaling pathways for hematopoietic growth factors such as erythropoietin and thrombopoietin, JAK1 plays a larger role in mediating the signaling of proinflammatory cytokines.30Although JAK2 mutations are the major oncogenic signaling driver in the development of MPNs, there may exist important in vivo interactions between JAK1 and JAK2.31Patients with MPNs, particularly MF, display elevated levels of circulating inflammatory cytokines acting primarily through the JAK1 pathway, which may be responsible for a majority of the constitutional symptoms that contribute to morbidity.32
Ruxolitinib was first demonstrated to be active in the treatment of MPNs and PV in preclinical characterization studies.33Subsequently, in a phase II study in patients with PV refractory or intolerant to HU, nearly all patients demonstrated a clinical response by 24 weeks after starting ruxolitinib.34The response seen in this phase II study was durable, with more than 60% of patients attaining phlebotomy independence at 144 weeks after starting ruxolitinib and a similar number achieving a ≥50% reduction in palpable spleen length by the same time.
Ruxolitinib was approved for the treatment of HU-refractory or -intolerant PV based on the results of the randomized, phase III RESPONSE trial.35RESPONSE enrolled 222 phlebotomy-dependent PV patients who had resistance or intolerance to HU as well as palpable splenomegaly of at least 5 cm, and they were randomized to either ruxolitinib (starting dose: 10 mg twice daily) versus best available therapy (BAT). Nearly 60% of patients in the BAT arm continued HU, reflecting the limited options available for patients with PV after failing first-line therapy. The remainder of the therapies prescribed in the BAT arm included interferon, anagrelide, immunomodulators, and pipobroman. Patients were followed for 32 weeks after randomization, with crossover to ruxolitinib allowed after the 32-week primary endpoint follow-up.
At 32 weeks post randomization, compared with patients in the BAT arm, patients in the ruxolitinib arm had superior, statistically significant hematocrit control, reduction in splenomegaly, reduction in PV-related symptoms, and decreased phlebotomy requirements. Specifically, the primary endpoint (a composite endpoint of hematocrit control and ≥35% reduction in spleen volume as assessed radiographically) was achieved by 21% of patients in the ruxolitinib arm versus 1% in the BAT arm. At least 1 component of the primary endpoint was achieved in 77.3% of patients who received ruxolitinib, versus only 22.7% of patients who received HU and 30.8% of patients who received interferon. The mean change in JAK2 V617F variant allele frequency from baseline to week 32 was 12.2% in the ruxolitinib treatment group versus 1.2% in BAT.
The response to ruxolitinib was shown to be durable as well, as an extended follow-up analysis showed that the probabilities of maintaining the primary composite endpoint, and specifically hematocrit control, at the 80-week follow-up point were 92% and 89%, respectively.36At 80-week follow-up, thromboembolic events had occurred in 1.8% of the original ruxolitinib treatment arm versus 8.2% of the original BAT arm, although this number did not reach statistical significance. Ruxolitinib was fairly well tolerated, with just 4.5% of the ruxolitinib arm discontinuing the medication due to AEs.
The most commonly cited treatment-emergent AEs were new or worsening cytopenias (anemia, 27.2 vs 47.6; thrombocytopenia, 14.9 vs 29.9; lymphopenia, 27.2 vs 78.8), headache (10.5 vs 28.5), diarrhea (9.7 vs 12.2), fatigue (8.3 vs 23.1), and pruritus (9.7 vs 32.6); all data were reported as rate per 100 patient-years of exposure in ruxolitinib versus BAT arm.
Because ruxolitinib is a potent immunomodulatory agent, concern exists regarding increased infection risk. Multiple cases of opportunistic infections and viral reactivations have been reported.37-40Notably, the number of infections within the first 80 weeks of treatment was lower in the ruxolitinib arm versus the BAT arm (rate per 100 patient-years of exposure, 29.4 versus 58.4), though 12 herpes zoster infections occurred in the former versus none in the latter. Additionally, higher rates of nonmelanoma skin cancer were observed at the 80-week follow-up analysis (rate per 100 patient-years of exposure, 4.4 versus 2.7). More research needs to be done to investigate the possible link between oncogenesis and the immunomodulatory mechanism of ruxolitinib.
An ongoing follow-up study (RESPONSE-2) randomized 149 PV patients with HU resistance or intolerance, but lacking palpable splenomegaly, to ruxolitinib versus BAT. It demonstrated safety and efficacy results similar to those of the original REPONSE trial.41Additionally, the ongoing Prospective Observational Study of Patients With Polycythemia Vera In US Clinical Practices (REVEAL) trial, a phase IV, multicenter, noninterventional, prospective, observational study, will help further elucidate the role of JAK2 inhibition in PV.
Investigational Agents and Future Directions
Trials examining the efficacy of other JAK2 inhibitors for PV are ongoing. A phase II trial investigating fedratinib, a selective JAK2 inhibitor, for patients with PV or essential thrombocythemia (ET) resistant or intolerant to HU, was recently completed.42Other investigational agents such as lestaurtinib43(a JAK2/FLT3/TRK inhibitor), momelotinib44(a JAK1/ JAK2 inhibitor), and gandotinib45(a JAK2/STAT3 inhibitor) are being tested in early-phase studies, mostly involving patients with MF or post PV/post ET MF. However, a phase II study involving momelotinib for the treatment of PV or ET was recently terminated early due to limited efficacy (overall response rate: 5.1%).46
Novel agents that target pathways distinct from the JAK-STAT pathway are also being evaluated in clinical trials for PV. Agents such as givinostat (a histone deacetylase inhibitor), TGR-1202 (a PI3K-δ inhibitor), ropeginterferon (which uses a different pegylation process than interferon-α 2b), mirabegron (a ß-3-sympathomimetic), and RG7388 (an MDM2 inhibitor) are all currently being studied in clinical trials.
These agents will need to be carefully assessed for their ability to not only adequately control the hematologic parameters, but also to reduce thrombotic complications and ideally reduce the malignant cell burden and favorably adjust the natural history of PV. In order to achieve this goal, well-designed multicenter clinical trials will need to be executed with meaningful biomarker correlates of response and outcome. Molecular profiling at baseline and at the time of response, or of treatment failure, will also be crucial to determine predictive mutational signatures and to determine if therapies are adequately extinguishing malignant hematopoiesis.
PV is a clonal hematopoietic stem cell disorder that results in an increased red cell mass and predisposes affected individuals to thrombotic and hemorrhagic complications. Additionally, individuals with PV are at risk for progression to MF or evolution to AL. Advanced age and a history of thrombosis help risk-stratify patients with PV and guide modern management. HU remains the first-line agent for PV patients who qualify for cytoreductive therapy based on their risk score. At present, clear evidence of the leukemogenicity of HU is lacking and the majority of PV patients tolerate this agent with adequate response. However, a subset of patients treated with HU fail due to either issues surrounding intolerance or resistance, and these patients may in fact have worse clinical outcomes. Ruxolitinib, a JAK1/JAK2 inhibitor, was recently approved for the treatment of HU-refractory PV and serves as the only mechanism-based therapy for this MPN. Although initial results from trials evaluating ruxolitinib in PV demonstrate clear benefits in terms of reduction in phlebotomy rate, spleen volume diminution, and symptom improvement, long-term AEs and safety remain unknown with this agent, which also has potent immunomodulatory effects. Infectious complications such as viral reactivation, pneumonia, and the emergence of secondary malignancies must be considered on a case-by-case basis when starting ruxolitinib therapy. Importantly, the main reason for adding cytoreduction to the treatment approach of patients with high-risk PV is to reduce the thrombotic risk, and this has not been definitively demonstrated with ruxolitinib. Although it is reassuring that thrombosis rates were lower in the ruxolitinib treatment arm of the RESPONSE trial, insufficient numbers of patients followed over a relatively short period limit the ability to clearly evaluate this important endpoint. Undoubtedly, ruxolitinib has proven to be an important agent for those PV patients with otherwise unmanageable symptoms such as pruritus, fatigue, bone pains, and headaches, and this subgroup of patients appears to be the most appropriate to consider for JAK2 inhibition. Whether other agents from this class will contribute more benefit than ruxolitinib for PV remains uncertain, given the significant time it has taken to introduce another JAK inhibitor into the armamentarium of MF. Novel agents that effectively target the clonal hematopoietic origin of PV and ultimately favorably affect the disease course are still needed, and they are the focus of current translational research efforts.