|Home | About | Journals | Submit | Contact Us | Français|
The goals of therapy in patients with polycythemia vera (PV) are to improve disease-related symptoms, prevent the incidence or recurrence of thrombosis, and possibly delay or prevent the transformation into myelofibrosis or acute myeloid leukemia (AML). Cytoreductive therapies have been used in older patients and those with a history of thrombosis to achieve these goals. Hydroxyurea (HU) remains the first-line cytoreductive choice; however, up to one in four patients treated with HU over time will develop resistance or intolerance to HU. More importantly, patients who fail HU have a 5.6-fold increase in mortality and a 6.8-fold increase risk of transformation to myelofibrosis or AML; therefore, alternative therapies are needed for these patients. Interferon-α has been used in PV and has shown significant activity in achieving hematologic responses and decreasing JAK2 V617F mutation allele burden. JAK inhibition has also been investigated and recently garnered regulatory approval for this indication. In this review, we will discuss the current treatment options that are available for patients after HU and the novel therapies that are currently under investigation.
The outcomes of PV patients who fail or who are intolerant of hydroxyurea are poor. Although pegylated interferon can be considered in younger patients, currently, ruxolitinib is the only U.S. Food and Drug Administration-approved agent in this setting, representing a viable option, leading to hematocrit control and a reduction in spleen size and constitutional symptoms. Although a small number of patients will achieve a molecular response with continuous treatment, the implications of such response on the clinical outcomes are still unknown. Patients whose disease is not adequately controlled with ruxolitinib, or who lose their response, can be treated with low-dose busulfan or pipobroman; however, they should be encouraged to participate in trials with novel therapies.
Polycythemia vera (PV), a myeloproliferative neoplasm (MPN), is characterized by an increased red cell mass. Clonal proliferation in PV is not restricted to the etymological erythroid, but is often seen in the granulocytic and megakaryocytic precursors as well. The genotypic thread of activating JAK2 mutations strings together PV cases along with a common phenotype, in which approximately 95% of cases harbor the V617F mutation [1–3] and 4% harbor various mutations in exon 12 [4, 5]. Several other recently described mutations are often present with the driver JAK2 mutation, including mutations involving TET2 [6, 7] and EZH2 . Despite these commonalities, clinical presentations vary. Some patients with PV are diagnosed by chance during routine bloodwork, whereas others present with severe thrombotic events or disease-related symptoms (microvascular disturbances, pruritus, or headache) [9, 10]. In the absence of identification and intervention, the natural history of PV predicts a short disease course with a median survival of less than 18 months [11–13]. The most frequent complication of PV is thrombosis, but a minority of patients may see their disease transform to the spent phase, post-PV myelofibrosis, as well as accelerated and blast phase (acute myeloid leukemia; AML) [11, 14–16]. Nonetheless, by enacting strategies to mitigate the risk of these outcomes, median survival for patients with PV can stretch beyond a decade in even the highest-risk patients . Therefore, along with symptom management, the goal of treatment is to prevent the incidence or recurrence of thrombosis and possibly delay or prevent disease progression to myelofibrosis or AML.
Evaluation and modification of well-known cardiovascular risk factors (hypertension, cholesterol, diabetes, and smoking) are the intuitive first steps in attenuating thrombotic risk in patients with PV. The results of the European Collaboration on Low-Dose Aspirin in Polycythemia Vera Investigators study set the efficacy and safety of low-dose aspirin [9, 18]. Aspirin at 100 mg daily was associated with lower risk of arterial and venous thromboembolic events and is recommended for all PV patients who can tolerate it without significant bleeding or gastric side effects. On the basis of reports showing a proportional increase in the rate of thrombotic events with increased hematocrit , phlebotomy, particularly the goal for hematocrit, remained controversial. The randomization between more intensive (target hematocrit, <45%) and less intensive (target hematocrit, 45%–50%) treatment in the Cytoreductive Therapy in PV (CYTO-PV) trial has helped to fill this gap, showing a reduced risk of cardiovascular death and major thrombosis with more hematocrit control . As much as the results of this study provided guidance for therapy, many questions remain. In the CYTO-PV trial, those in the more intensive arm were more likely to start or increase the dose of hydroxyurea, and thus they had a lower white-cell count than in the less intensive arm. This imbalance could confound the effect of lower hematocrit on the primary outcome because leukocytosis has been shown to be a major risk factor for thrombotic events in patients with PV . Furthermore, the study was not powered to see differences in outcome with the use of phlebotomy alone versus cytoreduction alone or in association with phlebotomy to reach the same hematocrit level, and thus some controversy endures.
Conventional treatment approaches recommend adding cytoreduction in those who are at high risk for thrombosis or rethrombosis: those older than 60 years and/or with a history of prior thrombosis . Emerging evidence suggest that younger patients have comparable rates of thrombosis when compared with older patients , making the case for more aggressive treatment in patients with lower-risk PV by traditional risk stratification. Given the results of the CYTO-PV trial, cytoreduction is also indicated for patients in whom it is difficult to maintain a hematocrit below 45%. A final instance where cytoreductive therapy is indicated is for symptomatic control of splenomegaly, microcirculatory and cytokine-mediated symptoms, or progressive leukocytosis or thrombocytosis not otherwise controlled with aspirin and phlebotomy. Hydroxyurea (HU), an oral antimetabolite that inhibits DNA synthesis by inhibiting the enzyme ribonucleoside reductase , has become the de facto first-line cytoreductive therapy in patients with PV. Although, the notion that HU can reduce the risk of thrombosis in PV is extrapolated from data in essential thrombocythemia (ET), patients who are cytoreduced with HU achieve an overall response rate of approximately 90% (complete response 24% and partial response 66% as defined by the European LeukemiaNet [ELN]) . Despite these excellent response rates, 10%–15% of patients will have to reduce the dose or stop HU as a result of peripheral blood cytopenias, mucositis, leg ulcers, skin cancers, dermatitis, or fever, leading to suboptimal treatment . Furthermore, one in four patients on the intensive arm of the CYTO-PV trial failed to reach a hematocrit of 45%, despite adequate cytoreduction (i.e., HU at 2 g/day) . With this in mind, the ELN has attempted to define resistance or intolerance to HU . The number of second-line therapies is limited; moreover, resistance to HU is associated with 5.6-fold increase in death and 6.8-fold increase risk of transformation to myelofibrosis or AML —certainly a call for innovation and development of alternative therapies.
The choice for second-line therapy after HU failure is one to be considered carefully after considering side-effect profile and overall treatment goals. Although no drug to date has been shown to decrease disease progression, some have clearly demonstrated the potential to increase the risk of transformation to leukemia and should be avoided if possible . In the remainder of this review, we will discuss the options that are available as second-line cytoreductive therapies for patients with PV after HU.
Interferon-α (IFN-α) is a class 2 α-helical cytokine that plays an important role in both innate and adaptive immunity , as well as apoptotic pathways [29, 30]. Its biologic profile provides a strong rational for its use in the treatment of myeloproliferative neoplasms specifically. IFN-α has been shown to directly inhibit bone marrow fibroblast progenitor cells and suppress the proliferation of hematopoietic progenitors [31–33], with MPN-derived progenitors having a higher sensitivity to IFN-α than their normal counterparts . Furthermore, a recent analysis of individual colonies grown from erythroid progenitors obtained from PV patients treated with IFN-demonstrated that IFN-α-2a can dramatically decrease the size of JAK2 V617F allele burden without affecting the TET2 mutant clone (a molecular abnormality found in JAK2 V617F independent clones that could be implicated in disease progression to myelofibrosis and AML [35, 36]). This hypothesis is supported by colony assay studies on samples from patients with MPN, in which both JAK2 V6217F positive and negative clones harbored TET2 mutations, suggesting that TET2 may be present in a clone that is distinct from the one that is harboring JAK2 mutation [35, 36]. These observations suggests that IFN-α can change the PV disease course to the preproliferative phase . IFN-α has also been shown to decrease the levels of cytokines such as platelet-derived growth factor, transforming growth factor-β, and others that play a role in the development and clinical symptomology of MPNs .
Toxicity has been a significant barrier to widespread use of IFN-α therapy. Approximately 25%–40% of patients who were treated with IFN-α in PV trials discontinued their therapy due to side effects, with the majority of these discontinuations occurring during the first year of treatment.
Although little data exist specifically in the post-HU setting, a number of studies have evaluated the role of IFN-α in PV by using different commercial preparations in the frontline setting. It should be noted, however, that neither the U.S. Food and Drug Administration (FDA) nor any other health agencies have granted regulatory approval for any form to treat PV. The most commonly used preparations were IFN-α 2a and IFN-α 2b (allelic versions of IFN-α 2a), and, due to similar formulations, no clinically significant differences in the efficacy and side effects have been detected between them. The first study of IFN-α in PV was conducted in 1988 and showed promising activity in controlling the peripheral blood counts and improving disease-related symptoms . Subsequent studies have been conducted in PV; however, the majority of these trials included small numbers of patients with varying drug formulas and dosages and different response criteria, thus limiting the ability to summarize the results in a meta-analysis. Furthermore, the majority of these trials did not report the number of patients who discontinued IFN-α because of side effects [40–50]. Overall, approximately 47%–95% of the patients included in these trials had reduction in the number of phlebotomies, and 6%–85% required no phlebotomies during their treatment period . In one study with long-term follow-up, of the 38 PV patients treated with IFN-α  at dose of 9 × 106 IU weekly, 11 patients (28.9%) achieved a complete response (maintaining a normal hematocrit without phlebotomy), and 8 patients (21%) achieved partial response (>50% reduction in phlebotomy requirement) . Responses were durable with a 40-month median duration of treatment (range, 12–52 months), but toxicities were observed in 36.7% of the patients requiring treatment discontinuation in all such cases .
Toxicity has been a significant barrier to widespread use of IFN-α therapy. Approximately 25%–40% of patients who were treated with IFN-α in PV trials discontinued their therapy due to side effects, with the majority of these discontinuations occurring during the first year of treatment . A wide variety of side effects have been reported with IFN-α. Flu-like symptoms occurred in almost all patients and usually appear after 1–3 hours of administration, but can be prevented with acetaminophen. More importantly, these symptoms are dose-dependent and can be improved or avoided by decreasing the starting dose . Other side effects associated with chronic use of IFN include fatigue, depression, musculoskeletal pain, changes in mood, anxiety, injection site reaction, and gastrointestinal toxicity, including nausea, vomiting, diarrhea, and weight loss. These side effects often lead to dose reduction of IFN-α or discontinuation in severe cases.
Therefore, pegylated forms of IFN-α (PEG-IFN) have been developed. Pegylation results in several advantageous properties, including enhanced dug stability and solubility, as well as a longer half-life (can be administered on weekly basis instead of every 24–48 hours) with the goal of leading to a more favorable toxicity profile . Although the efficacy of pegylated forms of IFN-α is at least similar, or in some trials even higher, than nonpegylated IFN-α, the rate of side effects requiring discontinuation of PEG-IFN has not been consistent across studies [49, 55, 56]. In a phase II trial of PEG-IFN-α 2b, Samuelsson et al. treated 21 patients with PV and 21 patients with ET . At 6 months, 29 of 42 patients (69%) had achieved a complete remission (CR) and were still on therapy. Of the 13 nonresponder patients, 4 patients went off the study early because of side effects, and another 2 patients were subsequently taken off the study at 6 months because of side effects. A total of 23 patients (12 with PV) continued on therapy at 2 years. Side effects were the main cause for discontinuation of therapy in 16 of 23 patients . However, two subsequent trials of PEG IFN-α-2a in PV patients have shown a discontinuation rate of 8%–10% [49, 56]. The majority of discontinuation happened in the first year of treatment, suggesting that patients who are able to tolerate its initial side effects may continue on treatment with fewer side effects later. In a long-term follow-up of a phase II trial of PEG IFN-α-2a in patients with PV and ET, complete hematologic response was observed in 76% and 77% of patients with PV and ET, respectively. Furthermore, 18% of patients with PV, and 17% of patients with ET, had undetectable levels of JAK2 V617F with treatment, thus reaching complete molecular response . Interestingly, patients who did not achieve complete molecular response (CMR) were more likely to have disease that harbored mutations outside the canonical JAK-STAT pathway and were more likely to acquire new mutations during therapy .
Ropeginterferon α-2b, a next-generation PEG-IFN, has a longer half-life and is dosed every 2 weeks. In a phase I/II study including 51 PV patients, dose-limiting toxicities were not defined in the phase I portion. Overall, 10 patients (20%) permanently discontinued study drug treatment because of an adverse event. The cumulative overall response rate was 90% (47% and 43% complete and partial responses, respectively) .These results are the basis for the ongoing pivotal, randomized, phase III PROUD-PV (Pegylated Interferon Alpha-2b Versus Hydroxyurea in PV) trial.
A unique feature of treatment with IFN-α is its ability to induce cytogenetic and molecular remissions in PV patients, suggesting that IFN-α can eradicate the malignant clone, a feature that has never been shown with all other available therapies for PV. Furthermore, these molecular responses appear to be durable and may last several months after treatment discontinuation. In a phase II multicenter study of PEG IFN-α-2a, 35 of 37 evaluable patients achieved a CR. Among the 29 patients with molecular monitoring available, 89.6% had reduction in their JAK2 V617F allele burden, 24.1% in CMR (undetectable mutation) . Some of these CMRs lasted for more than 18 months after treatment discontinuation . In another trial of PEG IFN-α-2a in 79 patients with PV and ET , complete hematologic responses were observed in 80% of PV patients. Furthermore, 54% of patients experienced reduction in JAK2 V617F allele burden (15% in CMR), and over the course of the study, the JAK2 V617F mutant allele burden continued to decrease after a median of 21 months with no clear evidence of a plateau . Overall, the drug was tolerated, with 69% of patients developing some toxicity, but generally grade 1 or 2. The most common grade 3 or 4 toxicity was neutropenia, occurring in 20% of the patients .
A unique feature of treatment with IFN-α is its ability to induce cytogenetic and molecular remissions in PV patients, suggesting that IFN-α can eradicate the malignant clone, a feature that has never been shown with all other available therapies for PV.
JAK2 signaling, via cytokine receptors—such as erythropoietin receptor, thrombopoietin receptor, and others—plays a critical role in regulating normal myelopoiesis [1, 2]. Thus, when activating mutations in JAK2 were discovered more than a decade ago, it became clear that the clinical phenotype of MPNs is rooted in continuous activation of JAK signaling independent of cytokine receptor activation, resulting in enhanced cell proliferation, differentiation, and survival [2, 59]. This is most evident in PV because approximately 95% of PV patients harbor JAK2 V617F mutation . In turn, this groundbreaking discovery prompted the rapid development of therapeutics that inhibit the JAK signaling pathway, providing a therapeutic option in patients with MPNs.
Ruxolitinib, an oral JAK1/2 inhibitor, was the first in its class to be approved by the FDA for treatment of primary myelofibrosis [60–62], based on the results of two randomized, phase III trials pitting ruxolitinib against placebo (COMFORT-1) or best available therapy (BAT; COMFORT-2), demonstrating that treatment with ruxolitinib was associated with significant reduction in splenomegaly and constitutional symptoms, as well as improvement in quality of life [62–65]. With extended follow-up, patients randomized to ruxolitinib had a longer overall survival, even despite crossover from the placebo and BAT arms [65, 66]. This success in primary myelofibrosis has been translated to patients with PV. The randomized, phase III, open-label, RESPONSE-1 trial, compared ruxolitinib to BAT in patients with PV who failed or were intolerant to HU . A total of 60% of patients treated with ruxolitinib achieved hematocrit control (≤45% in the absence of phlebotomy), as compared with 20% who received BAT (58.9% of which continued HU, 11.6% received interferon, and 15.2% received supportive care alone). In addition, ruxolitinib was found to be superior in reducing spleen size (38% vs. 1% with at least a 35% reduction in pretreatment size) and symptom burden (49% vs. 5% with at least a 50% reduction in their total symptoms score at 32 weeks) .
In general, treatment with ruxolitinib was well tolerated. Rates of grade-3 or -4 cytopenias and infections were infrequent in both groups, but, notably, herpes zoster infection (grade 1 or 2 in all cases) was reported in 6% of patients in the ruxolitinib arm as compared with 0% of those in the BAT group. With patient crossover in mind, adverse event rates were adjusted for cumulative exposure, and the rate of grade-3 or -4 adverse events per 100 patient-years was 28.8 in the ruxolitinib group as compared with 44 in the BAT group. Although not adequately powered to detect a difference, thromboembolic events occurred in one patient treated with ruxolitinib and in six patients receiving standard therapy . Based on these results, the FDA granted approval for ruxolitinib in patients with PV who fail or are intolerant to HU in 2014. The ongoing RESPONSE-2 trial looks to confirm these results in patients lacking splenomegaly who have failed or are intolerant of HU (ClinicalTrials.gov Identifier: NCT02038036).
Busulfan, is an alkylating agent that has purine analog activity in one molecule, which has been used as a treatment for several hematologic malignancies . In patients with PV, busulfan, used at lower dose (2–4 mg/day), can result in reasonable hematologic control .
Pipobroman (PB) is a piperazine derivative with activity as alkylant that has demonstrated efficacy in PV [70–72]. In one study with long-term follow-up of 163 PV patients treated with PB (1 mg/kg per day until hematologic response, followed by a maintenance dose of 0.3–0.6 mg/kg per day), 94% of patients achieved a hematologic response, with median time to response of 13 weeks . The 12-year cumulative incidence of thromboembolic events was 20%, and 10-year cumulative risk of leukemia, myelofibrosis, and solid tumors was 5%, 4%, and 8%, respectively .
Imatinib, a potent tyrosine kinase inhibitor of the fusion protein BCR-ABL1 and other protein tyrosine kinases including platelet-derived growth factor receptors, KIT, and ARG (ABL1-related gene), has revolutionized the care of patients with chronic myeloid leukemia and has become the archetype for targeted therapy. However, clinical trials of imatinib in patients with PV did not show any significant activity of this agent in this setting [74–76]. Several newer agents are currently under investigation in patients with PV (Table 1).
For Further Reading: Amy E. DeZern, Mikkael A. Sekeres. The Challenging World of Cytopenias: Distinguishing Myelodysplastic Syndromes From Other Disorders of Marrow Failure. The Oncologist 2014;19:735–745.
Implications for Practice: Over the past decade, our understanding of the molecular mechanisms underlying bone marrow failure (BMF) conditions has expanded, especially with respect to molecular testing for somatic mutations in myelodysplastic syndromes (MDS). Applications of genomics include correctly diagnosing the BMF condition and incorporating prognostic markers to ensure correct therapy. This review aims to clarify the current state of diagnostic modalities in BMF and to examine the potential utility of an algorithmic approach using molecular markers distinguish MDS from other BMF conditions. In addition, the prognostic value of molecular markers in predicting the risk of disease aggressiveness, recurrence, and mortality is discussed.
Conception/Design: Aziz Nazha
Manuscript writing: Aziz Nazha, Aaron T. Gerds
Final approval of manuscript: Aziz Nazha, Aaron T. Gerds
Aaron T. Gerds: Incyte, Roche, Astra-Zeneca (C/A). The other author indicated no financial relationships.
(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board