Developing strategies to reduce the duration of therapy for patients with myeloproliferative neoplasms

Abstract

Introduction:

All current treatment strategies for myeloproliferative neoplasms (MPN) patients with the exception of allogeneic stem cell transplant (ASCT) are continuously administered. Treatment approaches that reduce the degree of minimal residual disease (MRD) might permit possible drug holidays or potential cures.

Area covered:

Authors discuss the presently available agents and those that are under clinical development that might induce a state of MRD and can be administered intermittently. Data extracted from a comprehensive search of peer review literature performed in Pubmed as well as information presented in scientific meetings.

Expert opinion:

Currently, the only potential curative treatment for MPN is ASCT. ASCT requires a period of intense treatment but ultimately allows the patient to enjoy a period independent of continued treatment. There is evidence that intermittent use of busulfan or prolonged use of IFN-α can induce hematological remissions that are sustained for prolonged periods of time, allowing for drug holidays. The experimental drug Imetelstat is a promising drug that has been reported to prolong survival in very high-risk myelofibrosis patients after a limited period of time of administration. New experimental drugs and drug combinations that target the malignant clone and/or microenvironmental abnormalities have the potential to eliminate MRD, which might allow for drug holidays and reduction in the duration of therapy.

1. Introduction

The chronic myeloproliferative neoplasms (MPN) are a group of hematological malignancies with diverse clinical phenotypes and outcomes. The BCR-ABL1-negative MPNs include polycythemia vera (PV), essential thrombocythemia (ET) and primary myelofibrosis (PMF). PV and ET can each evolve to a clinical syndrome that resembles PMF and each of these MPNs can eventually progress to MPN-blast phase (MPN-BP), which resembles acute myeloid leukemia but is almost universally refractory to intensive chemotherapy. MPNs are debilitating and are frequently progressive. Many patients suffer from constitutional symptoms including fatigue, night sweats, fevers, weight loss, pruritus, and bone pain, as well as symptomatic hepatosplenomegaly that negatively affects their quality of life. Moreover, patients experience thrombotic and bleeding complications as well as vasomotor symptoms. Early stages of PMF can progress to a more overt form of MF characterized by a bone marrow failure state associated with marrow fibrosis and ineffective hematopoiesis leading to anemia and thrombocytopenia. In addition, MPN related hypermetabolism frequently results in hyperuricemia, which leads to gouty arthritis or urate nephropathy. Other common complications include portal hypertension that may lead to variceal bleeding and ascites, non-hepatosplenic extramedullary hematopoiesis with its consequences such as cord compression, ascites, pleural effusion, or pulmonary hypertension.

Universal treatment goals for MPN patients include normalization of blood counts, reduction in the rates of thrombotic and bleeding events, alleviating MPN related symptoms including constitutional symptoms and spleen related symptoms, thereby improving the quality of life of patients, and minimizing the risk of progression to myelofibrosis or transformation to MPN-BP. PMF specific treatment goals also include the reduction of the degree of symptomatic anemia and thrombocytopenia.

Current treatment options include phlebotomy, the use of myelosuppressive therapies (hydroxyurea, oral busulfan), anagrelide, pegylated forms of interferon, and the JAK1/2 inhibitors, ruxolitinib, and fedratinib. None of these approaches substantially alters the risk for progression to more overt forms of MF or MPN-BP. For those MF patients with anemia and thrombocytopenia, transfusion therapy is a core therapeutic approach, which can be supplemented with the use of erythropoiesis-stimulating agents (ESAs), androgens, prednisone, and immune-modulatory agents such as thalidomide and lenalidomide. The only curative treatment for the MPNs is allogeneic stem cell transplantation (ASCT), but its utility is limited by advanced patient age, comorbid conditions, lack of appropriate donor options, and often, patient willingness.

The use of each of these treatment modalities is based on assessing the predicted outcomes of individual patients based on the use of a variety of risk-stratification tools that have been created to predict the risk of a particular patient developing a thrombotic event or evolving to MF or MPN-BP. These tools utilize age, clinical characteristics, cytogenetic abnormalities as well as the number and types of myeloid malignancy gene mutations [1–6]. Although these tools can provide the physician with a means of understanding and anticipating the clinical course of cohorts of patients, their application to individual patients can sometimes be problematic. It is important to emphasize that many of these risk-stratification tools have been created by retrospectively analyzing large cohorts of patients, but none have been validated in a prospective fashion. Most clinicians use these tools to determine the appropriate therapeutic options, which can range from observation to ASCT for an individual patient. This practice usually involves ratcheting up the patient’s particular therapy as the disease progresses over time.

Although most of the treatment options utilized to treat PV and ET patients have used normalization of blood counts as a metric for achieving a “successful“ therapeutic endpoint the validity of such an approach has not been universally validated. The most solid data available indicate that in PV maintenance of hematocrits levels <45% leads to a reduction in the risk of developing a new thrombotic event [7]. Spivak has, however, suggested that in females the target hematocrit should be <42% [8]. He suggests this approach based upon normal women having lower hematocrit levels than their male counterparts. Although this approach is logical, the benefits of using a lower target hematocrit level in females has not been validated in a prospective fashion. Also achieving normalization of leukocyte and platelet counts appears to be intuitively an important goal, but to date has not been associated with improved outcomes. In addition, what leukocyte or platelet count to target with a particular therapeutic approach has not been convincingly documented. The European LeukemiaNet (ELN) established response criteria for ET and PV that include resolution of disease-related signs and symptoms, normalization of counts, and marrow histology. The response criteria for MF include the above as well as a clinical improvement (anemia, spleen, and symptoms response). These criteria are mainly used for standardized assessments of therapeutic outcomes in clinical trials but are not always associated with a reduction of the number of thrombotic episodes or progression to overt MF or MPN-BP [9,10].

The present therapeutic strategies utilized to treat each of the MPNs are continuously administered orally or subcutaneously, and are associated with variable degrees of toxicity. Ideally, a treatment approach that would target and deplete the malignant clone and correct microenvironmental abnormalities which allow malignant stem cells to predominate over the reservoir of normal stem cells would be anticipated to favorably alter the natural history of each of the MPNs. Induction chemotherapy for acute myeloid leukemia for instance is capable of creating a state of low or negative minimal residual disease (MRD). Such induction chemotherapy in AML patients is followed by consolidation therapy leading to prolonged remissions or actual cure in a subset of patients eventually permits therapy to be halted without a significant rate of relapse [11,12]. Since such induction type therapy is associated with a significant risk of morbidity and mortality, it is clearly not merited in patients with chronic MPNs that are associated with a survival that can be measured in decades. Only in patients with more overt forms of MF or MPN-BP which are associated with a survival that is less than 2 years have clinicians been willing to treat with highly toxic treatment regimens including ASCT. The patients with less advanced MPNs continue to receive treatments that require the continuous administration of the prescribed drug of choice. The most extreme examples of the need for continued administration of an MPN drug is the JAK1/2 inhibitor ruxolitinib, where rapid withdrawal of the drug is frequently associated with increasing splenomegaly and the rapid return of systemic symptoms and occasional the appearance of pulmonary decompensation due to an apparent cytokine storm. Another example is the use of ESAs to treat MF related anemia that reappears with the cessation of therapy.

Since the MPNs are chronic disorders, one can envision the benefits of delivering a treatment approach that might reduce the burden of the tumor and the degree of MRD to a sufficient degree to permit possible drug holidays, or even potential cures beyond that achieved with ASCT. Recently, for instance, treatment with tyrosine kinase inhibitors has been successfully withdrawn in a number of patients with chronic myeloid leukemia who have achieved a deep molecular remission without relapse occurring [13,14]. In this review, we will focus on presently available therapeutic approaches for PV, ET or MF patients, which might be administered either intermittently or for a limited period of time (Table 1). Although the value of MRD detection in MPN patients remains uncertain, such an approach when successful, will likely require combinations of drugs that will necessitate careful investigation. In this review, we will discuss the presently available agents and those that are under clinical development that might be employed as components of such a therapeutic strategy for MPN patients.

Table 1.

Selected Publications Documenting Intermittent Treatment for MPN Patients.

Agent	Type of clinical trial	Number of patients	Disease	age	regimen	Complete hematologic response/other response	Remission duration/time to next treatment off treatment	ref
Busulfan	Randomized Phase 3	285 (147 in busulfan group)	PV	59.5 (± 12.7)	4–6 mg/d for 4–6 weeks	Not reported	48 m	[16]
Busulfan	Retrospective	36	ET	60–91	4 mg/d for 1 week; 2 mg/d for 3 weeks; 2 mg every other day until plt<400 (average 12 w of treatment)	100%	56 m	[17]
Busulfan	Retrospective	36	PV/ET	77 (61–93)	Continuous 14 mg/week or lower. Median treatment time 36.5 w (10–223)	83% CHR	18 m (2.3–98.7)	[18]
Busulfan	Retrospective	51	Bcr-abl neg MPN	73 (48–91)	Continuous 2 mg/daily/alternate day. Median treatment time per cycle 8.5 weeks Bolusl-2 doses average total 60 mg per cycle	75% CR/PR	Average time between cycles 8 months for continuous 8.5 months for bolus	[22]
Peg-IFNα-2a	Phase 2	37	PV	49 (42–53)	90mcg/week with increase as needed/tolerated	95% CHR	5 patients in CMR who stopped treatment still had undetectable Driver mutation VAF 6 – 18 months after drug discontinuation	[28]
IFNα−2	Cohort	102	Jak V617F mutated MPN	52 (17–75)	variable	95% of ET 68% of PV	5 of 6 patients with deep molecular remission had continued hematological remission off treatment at median follow-up of 45 m (26–56)	[35]
Imetelstat	Phase 2	107	INT2 and high risk MF	68 (31–86)	4.7 mg/kg Q3W (median 17.6 weeks) 9.4 mg/kg Q3W (median 32.7 weeks)	Spleen response (9.4 mg/kg cohort) 10.2% Symptoms response (9.4 mg/kg cohort) 32.2% Reduction in fibrosis (9.4 mg/kg cohort) 40.5%	Median PFS (9.4 mg/kg cohort) 23.2 m (16.8–28.3) Median OS (9.4 mg/kg cohort) 29.9 m (22.8-NE)	Mascarenhas (submitted to publication)

CHR Complete hematologic response; CR complete response; ET essential thrombocythemia; IFN-α Interferon-α; MPN myeloproliferative neoplasms; NE not estimable; PR partial response; PV polycythemia vera

2. Current drugs utilized to treat MPNs that can be administered intermittently

2.1. Alkylating agents: Busulfan

Busulfan is an alkylating agent that had been commonly used to treat chronic myeloid leukemia for over 50 years. Busulfan treatment was reported to be a safe and effective treatment in a retrospective analysis of 65 PV patients treated between 1962 and 83 without an excess of cases of transformation to MPN-BP [15]. A remarkable phase 3 EORTC clinical trial, conducted from 1967 to 1978, randomized 285 PV patients to radioactive phosphorous (P³²) or busulfan. Busulfan was administered orally (4–6 mg/day) for 4–6 weeks, or was withheld when the platelet count was <120,000/μl. Achievement of hematocrit levels that exceeded the target hematocrit (42–47%) required occasional supplementary phlebotomies. When a relapse occurred the same treatment (P³² or busulfan) was re-administered. Patients treated with busulfan enjoyed a median first remission duration of 4 years, while those patients treated with P³² had a median first remission of 2 years (P < 0001). The 5-year survival in the busulfan group was 86%, and in the P³² group 74%. The 10-year survival was 70% for those randomized to receive busulfan as compared to 55% for the P³² group (P = 0.02). In the P³² group, 25/140 patients died from thrombotic events, as compared with 8/145 patients treated with busulfan. Evolution to PV related MF or MPN-BP occurred at a similar frequency in each treatment group. The authors warned that busulfan should never be administered continuously for more than 4–6 weeks, due to the increased risk of developing severe thrombocytopenia. The reduction of thrombotic events with busulfan is especially intriguing [16]. Subsequently, a study of 36 elderly ET patients (>60) who were treated with intermittent busulfan (median follow up of 72 m) has been reported. Remarkably, a hematological response was observed in 100% of the patients and their responses were long-lasting with half of the patient relapsing after a median of 56 months. Three patients progressed to MF but none transformed to MPN-BP [17]. In a more recent report, 36 elderly patients with PV and ET who were refractory or intolerant to hydroxyurea were treated with busulfan (2 mg/day). A complete hematological remission was achieved in 83%, with a median duration of response of 45.5 months. Almost a third of these patients achieved a partial molecular response, however transformation to MPN-BP or myelodysplastic syndrome (MDS) was observed in three cases, indicating that that hydroxyurea followed by busulfan therapy is associated with a significant rate of transformation to MPN-BP [18]. These data are supported by an earlier report by Nielsen et al., that described a marked increase incidence of transformation to AML or MDS after combination treatment with hydroxyurea and busulfan [19]. In a large international cohort of 1545 PV patients treated with a variety of drugs, intermittent exposure to busulfan, however, was not associated with a higher risk of transformation to MPN-BP [20]. Busulfan use as a second-line treatment was recently reported by Renso et al. 26 elderly (median age 79) ET patients received 4–6 mg/week, followed by a dose deescalation over time up to the minimal dose required for the maintenance of a complete hematological response. 92% of the patients achieved CHR, which was sustained for a prolonged time (43.6 m). Transformation to MPN-BP occurred in two patients [21]. Douglas et al. reported 51 MPN patients who had received busulfan after hydroxyurea intolerance/failure. The majority received continuous low-dose treatment, but six patients received intermittent bolus scheduling. Patients were treated for a median of 8.5 weeks per cycle. The overall hematologic response rate was 75% following at least one busulfan cycle and the median time to require additional therapy was 8 months. Transformation to MPN-BP and MDS was observed in 8% of the patients and correlated with disease duration and cumulative busulfan dose, the median follow-up time, however, was short (39 months) [22]. Begna et al. reported the Mayo Clinic experience with busulfan in patients with MPN in the years between 1970 and 2014. They describe treatment of 75 patients, showing effectiveness with a low transformation rate to MPN-BP [23].

Since busulfan is an alkylating agent, it is anticipated that this therapeutic agent might lead to treatment-related acute myeloid leukemia. Due to such concerns, intermittent busulfan therapy should be carefully considered, and used primarily in elderly patients who have exhausted other therapeutic options. It is preferable to avoid this therapeutic agent in individuals who have been previously treated with HU.

2.2. Interferon-α (IFN-α)

IFN-α has been used to treat PV and ET for over 4 decades. The mechanism by which IFN-α normalizes blood count in MPN patients remains poorly understood. This cytokine has pleiotropic actions and is known to have immunomodulatory, antiproliferative and antiangiogenic properties [24]. Several groups have speculated that IFN-α has direct effects on cancer stem cells. Mullaly and coworkers reported that IFNα treatment-induced hematological responses in a murine model of a JAK2V617F MPN by causing depletion of JAK2V617F MPN-stem cells over time. They demonstrated that IFNα treatment induced cell cycle activation of JAK2V617F mutant long-term hematopoietic stem cells and promoted a predetermined erythroid-lineage differentiation program thereby selectively depleting MPN stem cells [25]. By contrast, Lu and coworkers provided an alternative mechanism of action by which IFN-α depleted MPN stem cells. They provided evidence that IFN-α selectively activated the tumor suppressor P53 causing an increase in MPN CD34+ cells apoptosis [26]. Both of these reports indicate that IFN-α has differential effects on JAK2V617F mutated and normal hematopoiesis.

The immunomodulatory potential of IFN-α was demonstrated by Riley et al. They showed that during IFN-α therapy there is a significant increase in the proportion of CD56^bright NK cells and a reduction of CD56^dim NK-cells in the peripheral blood of patients with JAK2V617F-mutant MPNs. In addition IFN-α treatment increase cytokine-dependent IFN-γ expression by CD56^dim NK cells. This more immunostimulatory profile may enhance the anti-tumor immune response against JAK2-mutated MPN clones [27].

Initially, short-acting forms of IFN-α that required subcutaneous administration several times a week were used to treat MPN patients. This form of therapy was heralded by several investigators as a means of normalizing blood counts in PV and ET patients without the need to expose patients to potentially leukemogenic forms of chemotherapy. Patient compliance, however, was a serious limitation due to the need for repeated injections and potentially serious side effects including flu-like symptoms, hematological toxicity, elevated transaminases, nausea, fatigue, psychiatric sequelae including disabling depression, reversible left-sided heart failure as well as precipitating flares of a variety of autoimmune disorders including hemolytic anemia, hypothyroidism, and inflammatory bowel diseases. Lack of approval by regulatory agencies also limited access of IFNs to MPN patients. Interest in IFN therapy was re-awakened almost 20 years ago when pegylated forms of IFN-α which had been developed to treat patients with hepatitis C were used to treat MPN patients. The pegylated forms of IFN possessed more favorable pharmacokinetic properties which allowed them to be administered weekly or even at less frequent intervals and to have a more favorable toxicity profile.

Interest in pegylated forms of IFN therapy to treat MPN patients was heightened by the seminal report of a phase −2 trial of pegylated IFN-α−2a (Pegasys) of 37 PV patients communicated by Kiladjian and his colleagues in France. In this report, 95% of the patients achieved hematologic complete responses. Remarkably a reduction in the JAK2 V617F variant allele frequency was observed in 90% of the patients, with 24% achieving a complete molecular remission which was sustained for 6 to 18 months after drug discontinuation [28]. This paper provided evidence that IFN therapy was accompanied by meaningful reductions in MRD and clearly demonstrated that IFN therapy could be administered intermittently with prolonged periods of normal blood counts persisting. These patients frequently experienced molecular relapses prior to the re-appearance of the characteristic MPN blood count abnormalities. The ability of INF therapy to induce meaningful molecular remissions have been confirmed in several other phase 2 trials and retrospective studies [29–33]. Importantly, INF treatment is also effective in JAK2-negative MPN, and can lead to molecular remission with other driver mutations [33,34].

The durability of hematological responses to IFN was further illustrated by a report from Danish investigators which showed that prolonged rIFN-alpha2 (2a and 2b) therapy was capable of inducing complete hematological responses in 95% of ET patients and 68% of PV patients. 74% of these patients had reductions in JAK2V617F VAFs, and 10% achieved a reduction in their VAFs to <2% [35]. Remarkably 2 out of 10 patients treated by this same group, that received long-term treatment with IFN alpha2 (>3.5 years) achieved, hematological, molecular, and histopathological remissions, that were long-lasting and were sustained up to 5–6 years after discontinuation of IFN therapy [36]. Although these sustained remissions are experienced by the minority of treated patients they do serve as a testimony to the occasional-sustained beneficial effects of IFN therapy in PV and ET patients.

Although pegylated IFN-α−2a (Pegasys) has been extensively evaluated for the treatment of PV and ET, this drug has never been approved by the regulatory agencies for this indication. Recently another form of pegylated IFN, Ropeginterferon alfa-2b, has been approved by the regulatory agencies in Europe as first-line monotherapy in adults for the treatment of PV without symptomatic splenomegaly. In contrast to other pegylated IFNα compounds, ropeginterferon alfa-2b consists of a single positional isomer resulting in an extended elimination half-life, enabling less frequent dosing (every other week, or monthly during maintenance therapy) and improved tolerability, supporting long-term patient compliance. Ropeginterferon alfa-2b is administered once every 2 weeks, but dosing is based on efficacy rather than tolerability. A phase −1/2 clinical study of ropeginterferon alfa-2b was first performed in 51 PV patients and demonstrated a cumulative overall response rate of 90%, with 47% of patients attaining a complete response and 43% a partial response; in addition a complete molecular response was achieved in 21% of the patients and a partial response in 47% [37]. With these exciting data in hand, the efficacy of ropeginterferon alfa-2b was compared to hydroxyurea in a phase 3 randomized trial of 257 PV patients (PROUD-PV, and the extension CONTINUATION-PV). This trial demonstrated that while initially similar hematologic responses were observed with both drugs, ropeginterferon alfa-2b responses increased with time, and after 3 years of treatment, the proportion of patients who achieved a hematological response alone or hematological response with improved disease burden was greater in the ropeginterferon alfa-2b group (53% Vs 38%). In addition, molecular responses were significantly more common in the ropeginterferon alfa-2b group. The mean JAK2 V617F allele burden was reduced from 42.8% to 19.7% at 36 m as opposed to 42.9% to 39.3% in the HU group [38]. However, follow up for 36 months of patients participating in this randomized-controlled trial, indicated that the numbers of thrombotic events were similar between the two arms, despite the superior molecular responses achieved with ropegintereferon therapy. As expected for this cohort of patients with recently diagnosed PV managed optimally in the setting of a prospective trial, very few major vascular events were recorded, and thus a relationship between molecular response and the long-term benefits in terms of reduction of vascular events could not be made. Based on these data one can conclude that ropeginterferon alfa-2b is an effective agent in treating PV patients that can be administered less frequently than other forms of IFN. Although the clinical and molecular responses with this drug are encouraging, it continues to remain uncertain if this drug will alter the natural history of PV. It is anticipated that intermittent therapy might also be possible with this form of therapy thereby allowing patients to experience prolonged drug holidays.

A preplanned interim analysis of a phase 2 randomized trial (Low-PV) was recently reported. In this trial, 100 low-risk PV Patients were treated with a stringent monthly phlebotomy policy with or without Ropeginterferon alpha-2b 100 μg sub-cutaneously every 2 weeks. The primary composite endpoint defined by the percentage of patients maintaining the median hematocrit values ≤45% during 12 months, in the absence of progressive disease (i.e., thrombosis, bleeding, progressive leukocytosis, symptomatic or extreme thrombocytosis, symptomatic splenomegaly, or other uncontrolled symptoms) was reached in 84% of the cases on Ropeginterferon vs. 60% in the standard arm (p = 0.008). The total number of phlebotomies after 1 year was lower in the Ropeginterferon arm (43%) than in the standard arm (57%) (p = 0.024). Improvements in symptoms, were seen in 7 out of 10 patients (mean change −21%) in the Ropeginterferon arm whereas a worsening in half of items was noted in the standard arm (mean change +10%) (p = 0.033). In addition, Ropeginterferon was associated with significant reduction of splenomegaly, leukocyte, and platelet counts. There was no difference in grade ≥3 AEs, (6% and 8% of patients treated with Ropeginterferon or standard therapy, respectively) [39]. Further follow up of these treatment arms will be required in order to evaluate the value of Ropeginterferon therapy in low-risk patients. This approach should be presently limited to clinical trials and not generally applied to clinical practice.

Of note, some data exist which indicate that response to INF may be influenced by the patient’s cytogenetic status as well as driver mutational status [29,34,40] which leads us to conclude that further research is needed to identify the sub-group of patients who might benefit most from this treatment.

Reported experience of INF treatment in MF patients is less robust. A retrospective analysis of 62 MF patients treated with Peg-IFNa-2a in France and Belgium showed symptom improvement in 82% of patients, reduction in splenomegaly in 46%, and anemia response in 72% (some of whom also received ESA). The majority of the patients in this cohort had low-risk MF [32]. Pizzi et al. reported a significant positive bone marrow histology effect of IFN-α treatment in 12 MF patients (e.g. lower cellularity and fibrosis) in parallel with clinical improvement [41]. A prospective study of 30 low-risk MF patients (DIPSS low- or intermediate-1) treated with IFN-α revealed an overall response rate of 50% (including CR, PR, and clinical improvement). Forty percent of patients experienced a ≥ 50% decrease in spleen size, and 5 out of the 25 patients with available bone marrow biopsies showed a reduction in reticulin fibrosis [42]. Data supporting the use of IFN-α treatment in patients with more overt forms of MF is presently unavailable and the long-term benefits of its use in earlier stages of MF requires follow-up for substantially longer periods of time before widespread implementation of this strategy can be considered.

3. Experimental drugs targeting stem cell/microenvironment

3.1. Imetelstat

Although there are numerous drugs and drug combinations which affect a variety of targets (e.g. stem cell, inflammatory microenvironment, apoptosis, reviewed in [43,44]) which are being evaluated for treating MPN patients their ability to be administered intermittently has yet to be studied. In this review, we will focus on the experimental agents where there is some evidence that intermittent therapy for the underlying MPN might be possible.

Imetelstat, a telomerase inhibitor, is an agent that currently has the potential to reduce the duration of therapy. Telomeres are protein-bound repetitive double-strand hexanucleotides (TTAGGG in humans). They constitute the ends of linear chromosomes, and their role is to protect coding DNA from genetic damage or degradation. Telomere length is shortened after each cycle of cell division.

In neoplastic cells telomeres are continuously stabilized by aberrantly activated telomerase [45]. Telomerase activity is increased in MPN patient cells and telomere length is reduced [46]. These findings suggested that telomerase inhibition might be effective in selective eliminating MPN-mutated cells as compared to normal hematopoietic cells. Imetelstat, a 13-mer oligonucleotide, targets the RNA template of human telomerase reverse transcriptase and competitively inhibits telomerase enzymatic activity and cell proliferation.

Tefferi et al. reported a pilot study of the use of Imetelstat in 33 Intermediate II-and high-risk myelofibrosis patients. Complete or partial responses were reported in 21% of the patients. The four patients with CR had a reversal of their BM fibrosis and three had molecular remission. A greater number of responses were observed in patients harboring U2AF1 or SF3B1 mutations [47].

Baerlocher reported a phase 2 clinical trial of Imetelstat in 18 patients with ET. Complete hematological response was achieved in 89% of the patients. Partial molecular responses were observed in 88% of the JAK2V617F mutated patients with a median mutant allele burden reduction of 71% after 3 months. Partial molecular response was observed also in CALR and MPL mutated patients [48]. Of note, three patients on this trial progressed to MF. Grade 3 neutropenia however occurred in 32%, which diminished enthusiasm for further development of this drug in ET patients who have survivals that are frequently measured in decades.

Further clinical development in MF patients was pursued in a phase 2 multicenter trial of Imetelstat in 107 MF patients that had relapsed or were refractory to prior ruxolitinib therapy. These patients reported by Mascarenhas and coworkers almost uniformly had poor-risk disease with 25% being triple negative and 68% having greater than 1 high-risk myeloid gene mutations (ASXL1, EZH2, SRSF2, or IDH1/2). The patients were randomized to receive either 9.4 or 4.7 mg/kg of Imetelstat intravenously every 3 weeks for a total of 24 weeks. The median numbers of cycles received were 10. The median follow-up for these patients participating was 41.7 months thereby indicating that many patients did not continuously receive Imetelstat. The 4.7 mg/kg dose was found to not be clinically effective however the benefits of the higher dose were remarkable. 32% of the patients receiving the higher dose experienced an improvement in symptoms and 10.2% achieved spleen responses (>35% spleen volume reduction). In this trial, the VAF of the driver mutation was reduced in 42% of the patients, and 35% had a reduction in marrow fibrosis. Most importantly high dose Imetelstat therapy prolonged survival in this high-risk MF population, an effect that has not previously reported with other agents. The median OS was 28.1 months for patients receiving 9.4 mg/kg and 19.9 months for the 4.7 mg/kg. Most impressively, the patients with triple-negative disease, a population with a particular poor prognosis, had not reached their median overall survival at the time of this report while the overall survival for patients with a driver mutation was 24.6 months. The median overall survival for patients receiving the 9.4 mg/kg dose was double that previously reported for this high-risk patient population with other therapeutic agents [49,50].

In order to understand how this drug might modify the natural history, Wang and coworkers explored the mechanism by which a short course of Imetelstat might prolong survival of MF patients [51]. Treatment of CD34⁺ cells with Imetelstat reduced the numbers of MF but not normal blood hematopoietic progenitor cells irrespective of the patient’s mutational status. Moreover, Imetelstat treatment resulted in depletion of mutated HPCs from JAK2V617F⁺ MF patients. Furthermore, treatment of immunodeficient mice that had been previously transplanted with MF splenic CD34⁺ cells with Imetelstat at a dose of 15 mg/kg, 3 times per week for 4 weeks had a limited effect on the degree of human hematopoietic cell chimerism achieved by normal stem cells but resulted in a significant reduction in the degree of human MF cell chimerism as well as the proportion of mutated donor cells. These effects were sustained for at least 3 months after Imetelstat treatment was discontinued. These actions of Imetelstat on MF HSCs/HPCs were associated with inhibition of telomerase activity and the induction of apoptosis. These findings indicate that the effects of Imetelstat therapy observed in MF patients are likely attributable to the greater sensitivity of Imetelstat against MF HSCs as compared with to normal HSCs as well as the intensity of the Imetelstat dose schedule, resulting in the depletion of MF myeloid progenitors and HSCs, as compared to normal CD34+ cells.

The persistence of the depletion of the MF HSCs for at least 3 months after therapy in these mouse model experiments resembles the prolongation in patient survival after administration of a relatively brief course of Imetelstat that was observed in the multicenter phase 2 trial. Additional insight into the mechanisms underlying the surprisingly promising phase 2 studies of Imetelstat therapy in MPN patients that were refractory to or relapsed following ruxolitinib therapy was provided by studies performed by Dahlstrom and co-workers [52]. They reported that treatment of a JAK2V617F cell line with a JAK2 inhibitor resulted in telomere lengthening due to a telomerase independent mechanism leading to an accumulation of mutated cells. These investigators, however, showed that combination therapy with a JAK2 inhibitor and a telomerase inhibitor blocked this telomere lengthening thereby eliminating the accumulation of CD34+ JAK2V617F+ stem cells. These authors suggest that resistance to JAK2 inhibitor therapy is a consequence of this lengthening of telomeres that is independent of telomerase which results in a progressive accumulation of mutated stem cells that can be eliminated by administration of a telomerase inhibitor. Clinical evaluation of this combination of agents administered either simultaneously or sequentially merit further clinical testing. Such a combination might provide a strategy for more effective intermittent therapy. These remarkable effects of Imetelstat on MF patient survival clearly requires a more careful evaluation in a phase 3 trial before the drug can be considered for approval.

4. Allogeneic stem cell transplant

Allogeneic stem cell transplant (ASCT) is currently the only curative treatment, which results in reversal of the clinical, molecular, and histopathological consequences of MF [53–55]. Since advanced forms of MF are associated with a survival that ranges from 1 to 3 years, ASCT is usually reserved for this population of patients due to the risk associated with this modality of therapy. PV and ET patients who have survived that are measured in decades are not presently candidates for ASCT due to the unacceptable transplant-related morbidity and mortality. ASCT is currently the only treatment approach capable of eliminating MRD, which allows the patient to be free of continued therapy for the remainder of their lives. There is no debate that the year following ASCT is challenging, but those patients that survive this year are able to return to a full productive life. The decision of when and if to offer a high-risk MF patient ASCT has led to a plethora of risk-stratification schema based on clinical, histopathological, cytogenetic, and mutational patterns to aid in this decision-making process [56,57]. These schema are of great use in helping to identify patients who are in imminent danger of progressing to life threatening forms of MF or MPN-BP. It is irrefutable that the outcomes with ASCT are far better when patients are transplanted in chronic stages of MF rather than an accelerated phase of MF or MPN-BP [58–60]. The individual patient and his or her treatment team must determine the appropriate time when the risk of the disease outweighs the known risk of the transplant. Ruxolitinib, the current standard of care for high-risk MF patients, is an extremely useful agent that reduces MF related symptoms but does not prevent disease progression [61–65]. A frequent error that leads to missing the “window of opportunity” during which patient remains a viable transplant candidate is to continue patients on ruxolitinib for prolonged periods of time due to dramatic improvement of symptoms in the face of objective signs that their disease is progressing. These issues are best highlighted in a report by Kremyanskaya and co-workers where a patient refused ASCT due to ruxolitinib relief of MF related symptoms and the patient’s own fear of ASCT. By the time she had decided to actually move forward with the procedure she had developed an aggressive form of MPN-BP and was unable to be transplanted [66]. This patient’s fears were not unrealistic but it is the physician’s responsibility to educate the patient about the risks and benefits of ASCT and to make sure that the patient considers the performance of this procedure at a center which is familiar with the obstacles to successful ASCT in MF patients. One must be careful and not be coercive when dealing with such patients, since ultimately the decision to proceed with this procedure should be made by the patients and their families trying to develop a consensus decision. Frequent family conferences with the treating physicians make the emotional stresses that are encountered during the post-transplant period easier to deal with. Respect for the patient’s concerns about quality versus quantity of life should be openly discussed before proceeding with ASCT. In the past, one of the limiting factors in deciding whether a patient could proceed with transplant was donor availability. Since MF patients are frequently older, matched sibling donors are frequently unavailable. This problem has been overcome recently with the use of grafts from alternative donors including matched unrelated donors or more commonly at our center haploidentical donors, which can include siblings or children [67]. There are numerous reports describing the outcomes of ASCT for MF which are beyond the scope of this report [53,68–74] (Table 2). Unfortunately, many physicians treating MF patients are either strongly pro ASCT or anti ASCT for MF. One can conclude that although such polarization might academically be challenging to the physician community, it leads to confusion and frequent paralysis in the patient community. This conflict can only be resolved by a randomized study comparing the standard of care versus ASCT in patients with advanced forms of MF which has been repeatedly been proposed by Vikas Gupta and his coworkers (Personal Communication). Although this is a critical trial to do, the obstacles to its performance have proven insurmountable. The decision to proceed to transplant therefore must be made based on data generated from patient registries of MF patients that have undergone ASCT. These are imperfect tools but are the best data that is currently available.

Table 2.

Selected Publications Documenting Outcome for SCT for MF Patients.

	Type of clinical trial	Number of patients	Disease	age	regimen	DIPSS	5y OS BSC Vs SCT	5y OS
SCT	Retrospective data bases	−438	PMF	18–65	BSC	Low risk	95% Vs 69%		[73]
					SCT	INT-1	77% Vs 52%
						INT-2	41% vs 50%
						High risk	11% Vs 32%
SCT	Retrospective registry	2224	MF	52.9 (18–74)	MAC			53% (49.1–56.9)	[74]
				57.5 (21–76)	RIC			51% (48.3–53.7)	[74]

BSC best supportive care; MAC myeloablative conditioning; MF myelofibrosis; RIC reduced-intensity conditioning SCT stem cell transplant

ASCT has been shown to improve OS in a subset of high-risk MF patients [73]. In a retrospective analysis of 438 patients age less than 65, 190 received ASCT and 248 patients received conventional treatment. The relative risk (RR) of death among patients receiving ASCT vs those receiving conventional therapies was 0.55 and 0.37 for DIPSS int-2 and high-risk MF patients, respectively. The 5-year survival in the transplant and nontransplant cohorts were 69% and 95% for low-risk patients, 52% and 77% for int-1 patients, 50% and 41% for int-2 patients, and 32% and 11% for high-risk patients, respectively [73]. Of note, this cohort did not include ruxolitinib-treated patients. These authors, however, concluded that MF patients with intermediate-2 or high-risk MF but not low risk benefited from ASCT.

A recent retrospective study by the European Society for Blood and Marrow Transplantation has shed additional light on the role of ASCT in MF [74]. This retrospective analysis summarized the outcomes of 2,224 MF patients who underwent ASCT from 2000 to 2014 at multiple centers in Europe. 35% of the patients received myeloablative conditioning (MAC) and 65% received reduced-intensity conditioning preparative regimens (RIC). Importantly the median age of the participants in this study was 52.9 years (range, 18 to 74 years) for those patients receiving MAC and 57.5 years (range, 21 to 76 years) for those patients receiving RIC.

In 40% of patients receiving each conditioning regimen, the grafts were from matched sibling donors and 60% from unrelated donors. The 5 years overall survival was 53% for the patients receiving MAC and 51% for those receiving RIC. Graft versus host disease (GVHD) remained a serious cause of morbidity and mortality; The cumulative incidence of non-relapse mortality at 1, 3, and 5 years was 25.5%, 32.2%, and 34.6% for MAC patients and 26.3%, 32.8%, and 34.4% for RIC patients. There was a trend toward a higher relapse rate in patients receiving RIC as compared with patients receiving MAC. Multivariable analysis indicated an inferior outcome for older recipients, recipients receiving grafts from unrelated donors or mismatched donors and those with recipients with a poor performance status. RIC appeared to be better tolerated in patients older than 50 years of age which represents the majority of MF patients that we encounter at our center. The EBMT recommended that MAC be used for younger individuals while RIC allo-SCT was felt to be a better fit for older, MF patients. This report of over 2,000 patients conclusively confirms that ASCT is capable of changing the natural history of MF and actually resulting in cure and the elimination of the need for continuous therapy. Although the morbidity and mortality associated with ASCT for MF is significant, the EBMT registry study demonstrated that this form of therapy can be performed at numerous centers and currently provides the most promising treatment option for a carefully selected patient population. Until drugs or combinations of drugs are identified that can equal or improve these outcomes, ASCT should be considered the treatment of choice for higher-risk MF patient population. The recent recommendations of the ELN stated that allogeneic stem cell transplant be considered for all transplant-eligible patients with IPSS/DIPSS/DIPSS plus high or intermediate-2 risk. The Panel also recommended consideration of allogeneic stem cell transplantation for transplant-eligible patients with intermediate-1 risk score who present with either refractory, transfusion-dependent anemia, a percentage of blasts in peripheral blood >2%, the presence of adverse cytogenetic abnormalities, or high-risk myeloid malignancy gene mutations [64].

5. Conclusion

The only curative treatment for MPN is ASCT. This treatment modality is reserved for higher risk MF patients and has been shown to lead to cure and prolonged survival in a subset of patients.

Busulfan and pegylated forms of IFN-α are the only agents that have been shown to have the ability to obtain hematological remission that can be sustained for a prolonged periods of time allowing for drug holidays. In addition, the use of pegylated forms of IFN has been associated with reductions in the driver mutation VAF and occasional molecular remissions but patients almost always relapse months to years after discontinuation of therapy. Imetelstat is an experimental drug that has shown some exciting results in early phase clinical trial. It has been reported to lead to a significant prolongation of survival for very high-risk MF patients after a limited period of treatment.

6. Expert opinion

Current treatment options for MPN are mainly targeted at symptom relief and reduction in the rate of thrombotic and bleeding events. The majority of the approved treatments are not known to alter the rate of disease progression. Moreover, the benefit to the patients from each of these treatments is usually transient and the disease phenotype reappears upon cessation of treatment. A treatment strategy that would target and deplete the malignant clone or dampen the tumor-promoting microenvironment would be anticipated to reduce tumor burden to a state of MRD, which would allow intermittent treatment or even treatment cessation without disease recurrence.

There is evidence that intermittent use of busulfan or prolonged use of pegylated forms of IFN-α can induce hematological remissions that can be sustained for prolonged periods of time allowing for drug holidays. The use of pegylated forms of IFN has been associated with reductions in the driver mutation VAF and occasional molecular remissions but patients almost universally relapse months to years after discontinuation of therapy. Surprisingly, intermittent busulfan therapy but not IFN therapy has been reported to be associated, at least in PV patients, with a reduction in the number of thrombotic events. Busulfan is an alkylating agent that likely contributes to the development of MPN-BP. Although many proponents of IFN have assumed that ropeginterferon alfa-2b therapy would reduce the risk of evolution to MPN-BP due to its ability to reduce the VAFs for MPN driver mutations, such claims have not been substantiated by appropriately powered long term randomized clinical trials. Proof of such claims is necessary and not solely of academic interest due to the known adverse effects associated with IFN therapy and a lack of knowledge of what adverse events might occur in MPN patients that would be anticipated to require decades of IFN therapy. When choosing a therapy that can be administered intermittently ropeginterferon alfa-2b should be considered to be a more acceptable choice than busulfan due to the likely leukemia promoting effects associated with the use of an alkylating agent. Imetelstat, a telomerase inhibitor, is a promising agent which is currently under clinical development which after a limited period of treatment has been reported to lead to a significant prolongation of survival for extremely high-risk MF patients. These promising results with Imetelstat, however, require validation by the performance of a properly powered phase 3 randomized clinical trial. ASCT, however, remains the most promising form of therapy that is capable of improving the natural history of MF and which provides a patient with an opportunity to return to a high-quality life after 6–12 months of intensive inpatient and outpatient treatment. During the initial year post-transplant, many of these ASCT recipients require transfusion therapy, immunosuppressive therapy, and repeated hospitalizations. The patients that are able to survive this gantlet of challenges that are a consequence of transplant-related events or relapse represent that fortunate group who can resume their lives. Unfortunately, ASCT is not without substantial risks that both the patient and their care providers must accept before proceeding with such an approach. The possibility of identifying agents to treat MPN patients based on their unique mutational pattern has now become a reality. Chifotides and coworkers recently reported treating MPN-BP patients with the IDH1/2 mutations with IDH1/2 inhibitors alone or in combination with ruxolitinib and a number of other drugs. Remarkably, several of the patients achieved complete remissions with undetectable IDH1/2 mutations by NGS [in press]. This achievement of MRD negativity is encouraging and hopefully additional patients with other MPN mutations can be similarly treated in the future with such mutation-specific drugs.

The MPNs have become a focus for the development of new therapeutic approaches including next-generation JAK2 inhibitors, epigenetic modifiers (such as bromodomain inhibitors), anti-cytokine therapies (such as TGFβ inhibitors), P53 activators (such as MDM2 inhibitors), BCL2, and BCL-xL antagonists as well as immunotherapeutic approaches. Each of these modalities of therapy represent an incremental step forward which holds a promise that combinations of such agents might be useful in eliminating MRD which would allow for longer drug holidays for patients who suffer from these challenging chronic hematological malignancies.g

Article highlights box.

• MPNs are incurable diseases without ASCT. Treatment is tailored to an individual patient’s symptoms and risk for disease progression as determined using a variety of stratification tools. MPNs almost uniformly necessitate prolonged continuous treatment with a variety of agents to maintain disease control.

• Treatment approaches that target and deplete the malignant clone and correct microenvironmental abnormalities might allow the reservoir of normal stem cells to predominate over malignant stem cells which would likely reduce the degree of MRD enabling drug holidays or potential cures.

• Busulfan and pegylated forms of IFN-α are two drugs currently available that can induce hematological remissions that can be sustained for prolonged time periods, thereby allowing for drug holidays.

• The use of pegylated forms of IFN has been associated with reductions in the driver mutation variant allele frequencies (VAFs) and occasional molecular remissions, which can be sustained even after drug interruption.

• Imetelstat, a telomerase inhibitor under clinical development, has been shown to lead to a significant prolongation of survival in very high-risk MF patients after a limited period of treatment.

• Novel therapeutic agents and combinations are being developed and evaluated with the goal of eliminating MRD and allowing prolonged drug holidays and possible cures.