Novel Therapies for Myelofibrosis

Abstract

Myelofibrosis (MF), including primary, post-essential thrombocythemia and post-polycythemia vera MF, associates with a reduced quality of life and shortened life expectancy. Dysregulation of the Janus kinase (JAK)/signal transducer and activator of transcription (STAT) pathway is prominent, even in the absence of the JAK2^V617F mutation. Therefore, all symptomatic MF patients may potentially derive benefit from JAK inhibitors. Despite the efficacy of JAK inhibitors in controlling signs and symptoms of MF, they do not eradicate the disease. Therefore, JAK inhibitors are currently being tested in combination with other novel therapies, a strategy which may be more effective in reducing disease burden, either by overcoming JAK inhibitor resistance or targeting additional mechanisms of pathogenesis. Additional targets include modulators of epigenetic regulation, pathways that work downstream from JAK/STAT (i.e., mammalian target of rapamycin/AKT/phosphoinositide 3-kinase,) heat shock protein 90, hedgehog signaling, pro-fibrotic factors, abnormal megakaryocytes and telomerase. In this review, we discuss novel MF therapeutic strategies.

INTRODUCTION

Myelofibrosis (MF), including primary MF and that evolving from essential thrombocythemia and polycythemia vera is a heterogeneous stem cell disorder, characterized by constitutional and systemic symptoms, cytopenias including anemia, extramedullary hematopoiesis, which typically manifests as splenomegaly, progressive marrow fibrosis, risk of acute leukemic transformation, and a shortened life expectancy. Historically, treatment needs for MF patients have been unmet, as conventional therapies, such as hydroxyurea, fared no better than placebo in controlling symptoms and/or splenomegaly.(1) While stem cell transplantation can be curative, given the advanced age of most MF patients, only a small proportion may be eligible for the procedure. The discovery of the JAK2^V617F mutation in 2005 (2–5) ushered in a new era for MF treatment with the rapid development of JAK inhibitors; nearly 2 years after its discovery, the first JAK-inhibitor was evaluated in clinical trials.(6) Ruxolitinib the first-in-class JAK-inhibitor, was approved for use in North America and the European Union on the basis of results from two international randomized trials comparing ruxolitinib against placebo (COMFORT-I) or best available therapy (COMFORT-II). (7, 8) Longer term follow-up data have shown that responses are durable and many patients can be safely treated long term. (9–11) In addition, a number of novel JAK-inhibitors are in clinical development, and as a class, JAK inhibitors provide improvements in symptom burden and relief from splenomegaly.

While monotherapy with JAK-inhibitors has improved the treatment landscape, complete remission is not an expected response, myelosuppression is a common side effect, and the development of resistance is a concern. However, complimentary mechanisms of disease pathogenesis suggest numerous additional targets for MF treatment. In this review, we discuss current drug targets, novel JAK-inhibitors and other agents that may be combined to maximize therapeutic potential.

JAK/STAT Dysregulation in MF

While the JAK2^V617F mutation is only identified in 50% to 60% of patients with MF, JAK-STAT signaling is dysregulated in all MF patients, including those with mutations in the MPL and calreticulin (CALR) genes, as well as a minority without any of these founding myeloproliferative neoplasm (MPN) mutations. This concept has been validated by gene expression profiling, single nucleotide polymorphism arrays, and mutational profiling; investigators identified a unique gene expression signature indicative of upregulated JAK-STAT signaling in all MPN patients regardless of their JAK2^V617F mutational status. (12) JAK-STAT pathway activation leads to excess cellular proliferation and resistance to apoptosis. In addition, pro-inflammatory cytokine signaling through the JAK-STAT pathway is responsible for some of MF symptom burden, such as fever, night sweats, and cachexia. (13) Universal dysregulation of JAK-STAT signaling in MF provides the rationale for targeting this pathway and accounts for the similar rates of response to JAK inhibitors in patients with and without the JAK2^V617F mutation. (7, 8) Despite the lack of significant anti-clonal activity by JAK-inhibitors, recent data suggests that MPN cells are dependent on JAK2 signaling for survival, validating the use of JAK inhibitors in MF; however, to more effectively abrogate aberrant signaling in MF, dual therapy may be required. (14)

Calreticulin mutations

CALR gene mutations were recently reported in JAK2^V617F- and MPL-negative MF patients. Klampfl et al. identified CALR frameshift mutations in 88% of JAK2^V617F/MPL-negative MF patients, and patients with a CALR mutation had a better prognosis compared with those with the JAK2^V617F mutation. (15) Interestingly, the most common CALR mutation activated STAT5, resulting in cytokine-independent growth, which may partially explain the response of JAK2^V617F-negative MF to JAK inhibitors. A recent report of significant spleen responses in 2 patients treated with a JAK inhibitor compliments this observation.(16) In a separate report, Nangalia et al identified CALR mutations in 56% of JAK2^V617F/MPL–MF negative patients studied. (17) These mutations appeared to reflect an early step in disease pathogenesis, since they were identified in multipotent progenitors. The prognostic impact of CALR mutations is becoming more clear as well. In a retrospective study of 617 MF patients, ~23% had CALR mutations with a median survival of 17.7 years, which was more favorable compared to those with JAK2 ^V617F mutations, or “triple-negative” MF patients, who lack JAK2/CALR/MPL mutations. (18) More recently, it has been reported that the favorable impact upon prognosis is restricted to those with type 1 CALR mutations (52 base pair deletions). (19) While initial studies suggest that CALR mutations lead to JAK-STAT activation, the mechanism by which these mutations lead to MPN is not yet clear.

MPL mutations

Compared to JAK2^V617F and CALR mutations, MPL mutations that activate the JAK/STAT pathway are less prevalent, but can be identified in up to 10% of MF patients through routine clinical testing. (20, 21) Pikman et al. generated a murine BM transplant model with the MPL^W515L mutation that recapitulated the MF phenotype with splenomegaly, thrombocytosis, and bone marrow fibrosis.(22) In addition, MPL-mutated MF does not appear to have a distinct phenotype or prognosis (20, 23). Other less prevalent mutations found in patients with MF such as CBL and LNK may also account for upregulated JAK-STAT signaling in patients who are JAK^V617F-negative. (21)

JAK Inhibitors

Several JAK inhibitors are in clinical development for MF (Table 1). Below we discuss the most recent efficacy and safety data for each agent.

Table 1.

Selected JAK inhibitors in various stages of development

JAK inhibitor (References)	Phase of development	Spleen and Symptom Response	Adverse events	Unique benefits
Ruxolitinib (7–11)	Approved by the FDA and EMA	Spleen volume (SV) reduction in 41.9 and 28.5% MF symptom improvement in 45.9%	Myelosuppression, particularly anemia and thrombocytopenia Headache, dizziness, easy bruising	Survival benefit Potential improvement or stabilization of bone marrow fibrosis (33)
Momelotinib (Formerly CYT387) (35, 36)	3	Spleen response in 48% Resolution of fever (100%), NS (79%); pruritus (75%) bone pain (63%); appetite loss (40%)	Thrombocytopenia First dose effect (transient dizziness and hypotension) Neuropathy	Anemia response in 59% of patients
Pacritinib (Formerly SB1518) (37–40)	3	Spleen volume reduction in 32% MF symptom improvement noted by month 6	Gastrointestinal symptoms including nausea and diarrhea	Tolerable in MF patients with thrombocytopenia
Fedratinib (Formerly SAR302503) (46–47)	3 Program discontinued	Spleen response in 36 (400mg) and 40% (500mg) Symptom improvement in 37% (400mg) and 34% (500mg)	Gastrointestinal symptoms including diarrhea Anemia Thrombocytopenia Wernicke’s encephalopathy	Spleen and symptoms responses also seen in ruxolitinib intolerant or resistant patients
Gandotinib (Formerly LY2784544) (41, 42)	1 Program discontinued	Spleen and MF symptom response in 56% of patients	Elevated creatinine, tumor lysis	Potential impact on bone marrow fibrosis
BMS911543 (43)	1/2a Program discontinued	Spleen response in 73% (8/11 pts.) >50% with MF symptom improvement	Myelosuppression, nausea, amylase and lipase increase	Not yet known
INCB-39110 (46)	2	Modest splenomegaly and MF symptom improvement	Anemia and thrombocytopenia Fatigue, nausea, constipation	First predominantly JAK1 inhibitor used in MF
NS-018 45	2	Spleen response in 48% Symptom improvement	Myelosuppression Nausea, dizziness	Not yet known

Ruxolitinib

Ruxolitinib is the first commercially available, potent JAK1/JAK2 inhibitor approved for the treatment of MF-related symptoms and splenomegaly. Approval was based on two randomized studies that met a primary endpoint of a ≥35% reduction in spleen volume by magnetic resonance imaging at 24 weeks (COMFORT-I 41.9% vs 0.7%; ruxolitinib vs placebo, P < .0001) and 48 weeks (COMFORT-II 28.5% vs 0%; ruxolitinib vs best available therapy, P < .0001). (7, 8) In addition, ruxolitinib therapy provided clinically meaningful improvements in quality of life and MF-related symptoms, as assessed by both specific and general validated patient-reported outcome tools.

Overall, anemia and thrombocytopenia were the most frequently reported grade 3 and 4 adverse events (AEs) in the ruxolitinib arms of both studies, due to inhibition of JAK/STAT signaling. (7, 8) Non-hematological side effects included dizziness, headache, and easy bruising. (7) Recently, ruxolitinib has been shown to impair dendritic cell (DC) function (24); in particular, impaired DC activation and migration were demonstrated along with diminished T cell responses. While this finding may account for the clinical benefits stemming from an anti-inflammatory effect, this property may also lead to infection. In fact, there are case reports of opportunistic infections, such as progressive multifocal leukoencephalopathy (25), retinal toxoplasmosis (26), tuberculosis (27, 28), and Cryptococcal pneumonia (29) in ruxolitinib-treated patients. However, with long term use, adverse effects appear less prevalent over time. (9, 10)

Longer term follow-up with 2- and 3-year data from the COMFORT studies suggest that the response to ruxolitinib is durable as well. (9–11) Furthermore, responses do not depend on the molecular-risk profile, as patients with high-risk mutations, including ASXL1, EZH2, SRSF2, IDH1/2, had a similar response and adverse event profile compared with those without those mutations.(30) Survival benefits have also been reported for ruxolitinib-treated patients compared with those treated with placebo, best available therapy, and untreated historical controls, including from those from the Dynamic IPSS cohort.(9, 10, 31, 32) This view is not universally shared, and it has been suggested that survival benefits should be interpreted with caution based on study design (survival was not a primary endpoint). (33) Additional challenges in interpreting survival data arise from potential selection bias when comparing cohorts via the case-control design.(34) Finally, missing data attributed to discontinuation or loss to follow-up could impact survival differences when comparing cohorts. (34) With regard to other potential benefits, it has been noted that compared with patients treated with hydroxyurea, ruxolitinib-treated patients experienced a greater degree of stability or improvement in bone marrow fibrosis with long term exposure.(35) Ruxolitinib has also been shown to be effective at lower doses (5–10 mg twice daily [BID]) in severely thrombocytopenic patients (50–100 × 10⁹/L). (36)

Momelotinib (CYT387)

Momelotinib is a potent JAK1/JAK2 inhibitor currently being studied in phase 3 trials. Results from a phase 1/2 study, including an analysis of the initial 60 patients, have been published.(37) The spleen response by International Working Group-Myelofibrosis Research and Treatment (IWG-MRT) criteria was 48% (52 evaluable patients), and significant symptom control was reported, including complete resolution of fever (100%), night sweats (79%), pruritus (75%), bone pain (63%), appetite loss (40%), and cough (20%). Unique to current JAK inhibitors, momelotinib appears to improve anemia: 59% of patients experienced improvements in anemia and 70% (n = 33) of transfusion-dependent patients experienced an at least a 12-week period of transfusion independence. Among patients who received momelotinib in an extension study (38); 58 (39%) achieved a spleen response by IWG-MRT criteria, with a median duration of response of 785 days. Fifty-nine (53%) patients achieved an anemia response, with a median duration of 1042 days. Grade 3/4 thrombocytopenia was reported in 30% of patients, and a transient “first-dose effect” of dizziness (16 patients [10%]) and hypotension (8 patients [5%]) was reported. (38) The potentially unique impact on anemia requires confirmation and a randomized phase 3 study with ruxolitinib as a comparator is recruiting participants (NCT01969838).

Pacritinib (SB1518)

In phase 1 studies, pacritinib, a JAK2/FLT3 inhibitor, resulted in a reduction in splenomegaly in 41% of patients.(39) In a subsequent phase 2 study enrolling 34 patients, including 15 with thrombocytopenia (platelets < 100 × 10⁹/L), 88% of patients had a reduction in palpable splenomegaly, and 32% had a ≥35% reduction in spleen volume by MRI. Two patients experienced improvements in Hgb levels, and symptom improvement was noted by month 6. (40) There were 17 discontinuations due to AEs, which were typically related to gastrointestinal toxicities (8 patients), disease progression (5 patients), or lack of efficacy (2 patients). In a safety update of the phase 1/2 studies that included 122 patients with MF, significant myelosuppression was not observed, even in those with baseline thrombocytopenia.(41) Similarly, a pooled analysis of the phase 2 studies demonstrated similar spleen and symptom reduction in patients with baseline platelet counts ≤100 × 10⁹/L and > 100 × 10⁹/L.(42) A randomized phase 3 study (Pacritinib Versus Best Available Therapy in Patients With Primary MF, Post–Polycythemia Vera MF or Post–Essential Thrombocythemia MF 1 [PERSIST-1]) comparing pacritinib with best available therapy is underway (NCT01773187), and a phase 3 PERSIST-2 study will evaluate pacritinib vs best available therapy (including ruxolitinib) in patients with MF and thrombocytopenia (platelets ≤ 100 × 10⁹/L).

Other JAK inhibitors

Other JAK inhibitors have been evaluated in phase 1 and phase 2 studies, including gandotinib (LY2784544), which is a selective JAK2^V617F inhibitor.(43, 44) In a study of thirty-eight patients (31 with MF), a maximum tolerated dose of 120 mg was established (200-mg doses were associated with grade 3 serum creatinine level increases).(44) At a dose of 120 mg, the most frequently reported drug-related AEs (all grades) included diarrhea (44%), nausea (29%), increased creatinine levels (21%), and anemia, vomiting, and fatigue (9% each). Pooling all dosing cohorts, 15 of 27 evaluable patients (56%) achieved ≥ 50% reduction in palpable spleen length, with a median duration of response of 18.3 weeks. At 12 weeks, 15 of 27 patients (56%) reported a ≥ 50% improvement in Total Symptom Score on the Myeloproliferative Neoplasms Symptom Assessment Form (MPN-SAF). In spite of these findings, this agent is no longer in development.

Another JAK2 inhibitor, BMS911543, (45) reduced splenomegaly and MF-associated symptoms at doses of 160 mg and 200 mg BID; 8/11 (73%) patients had a > 35% reduction in spleen volume and more than 50% had at least a 50% improvement in baseline MPN-SAF score. Thirty-one percent of patients experienced drug-related AE’s, including anemia (1 grade 2 and 2 grade 3), neutropenia (grade 4), grade 3 increases in amylase and lipase, and nausea (grade 3). This JAK-inhibitor is no longer being developed.

INCB-39110, a selective JAK1 inhibitor, resulted in MF-symptom control and modest improvement in splenomegaly at doses of 200 mg BID and 600 mg daily. (46) Adverse effects included fatigue, constipation, diarrhea, upper respiratory tract infection, cough and nausea. In those without transfusion-requirements, mean hemoglobin values were stable, but overall, new or worsening grade 3 anemia and thrombocytopenia occurred in 33% and 24% of patients.

Another JAK inhibitor, NS-018 (JAK2/Src inhibitor) is also in early-stage clinical development. Preliminary results revealed that among 42 patients, 48% had a ≥50% reduction in splenomegaly (11 clinical improvements (CI) in splenomegaly), and there were 3 and 2 CI’s in anemia and thrombocytopenia. (47) The recommended phase 2 dose will be 300mg BID; at this dose, the most frequent adverse events included nausea (18%, grade 1), dizziness (27%, grade 2 or less), thrombocytopenia (18%, grade 2), and anemia, leucopenia, and neutropenia (9% each, grade 3). (47) Despite meeting its primary endpoint in the phase 3 clinical trial, (48) development of fedratinib has been halted due to reports of Wernicke’s encephalopathy (49). It is of interest to note that clinical development of another JAK inhibitor XL019 was discontinued due to neurological toxicity; 7 of 9 patients in the phase 1 study experienced peripheral neuropathy.(50)

Challenges of JAK-Inhibitor Monotherapy

Although JAK inhibitors now have an established role in the treatment of MF, JAK inhibition has certain inherent challenges. First, myelosuppression is a common, though typically manageable AE. Second, splenomegaly and constitutional/systemic symptoms return rapidly after drug discontinuation; while severe withdrawal syndromes were reported in some patients in the phase 2 study, no withdrawal syndrome has been reported in the phase 3 ruxolitinib program (7, 51). Third, JAK inhibitors are not expected to induce disease remission, as they do not eliminate the disease-initiating clone. In fact, the current set of JAK-inhibitors are not specific for mutant-JAK2; while some off target effects may contribute to clinical benefit (anti-inflammation via JAK1 inhibition), other off target effects contribute to adverse effects, including wild-type JAK2 signaling (cytopenias), FLT3 inhibition (diarrhea), and infection (JAK1/JAK2 inhibition resulting in impaired dendritic cell function), as discussed above. (24) Fourth, despite chronic JAK inhibition, persistent JAK/STAT signaling continues due to the ability of JAK2 to form heterodimers with other JAK family members, such as JAK1 or tyrosine kinase 2 (TYK2). (52) Finally, in contrast to CML, whose disease pathogenesis in chronic phase is typically driven by a single molecular abnormality, MF disease pathogenesis involves many additional molecular aberrations, and accordingly, potential therapeutic targets. (Figure 1) (53) As a result, it is not expected that a single agent can address the entirety of the MF disease burden. However, JAK2 is an appropriate and valid target, and JAK-inhibition serves as a foundation for MF treatment. Dual therapy may result in even further alteration of the natural history of the disease. In the next section, a number of agents that may be combined with JAK inhibitors are discussed (Figure 2).

Figure 1.

Figure 2.

Overcoming Resistance to JAK inhibitors

Heat shock protein 90 inhibitors

Heat shock proteins (HSPs) are a family of chaperones that stabilize client proteins, including JAK2.(54) In the face of chronic JAK inhibition, JAK2 is persistently activated through heterodimer formation with JAK1 and TYK2 (52), suggesting a mechanism for JAK-inhibitor resistance. This finding has been observed in cell lines, murine models, and patient samples, and appears to be reversible with drug interruption. However, JAK2 inhibitor–resistant cells can also be targeted by the HSP90 inhibitor, PU-H71. Treatment with PU-H71 resulted in the suppression of cell growth and signaling in JAK2-mutant cell lines and primary samples and improved survival in MPN mouse models (ET and PV). (55)

Tanespimycin (17-AAG) was shown to inhibit JAK-STAT signaling in a homozygous JAK2^V617F cell line.(56) Several mutations in the JAK2 kinase domain (G935R, Y931C, and E864K) that are resistant to a panel of JAK inhibitors have been identified in vitro. However, treatment with the HSP90 inhibitor AUY922 led to degradation of both wild-type and mutant JAK2 these mutant cell lines. (57) In another study, AUY922 resulted in depletion of JAK2^V617F and downstream signaling proteins in cell lines and primary samples.(58) When combined with a JAK inhibitor, AUY922 increased apoptosis of the CD34+ cells. In addition, JAK inhibitor–resistant cell lines remained sensitive to both AUY922 and tanespimycin. Yet another HSP90 inhibitor ganetespib resulted in the sustained depletion of JAK2^V617F loss of STAT activity both in vitro and in vivo.(59)

These preclinical data suggest that HSP90 inhibitors may be logical companions to JAK inhibitors, with potential synergy and an ability to overcome resistance mechanisms. Currently, a phase 2 study of AUY922 is recruiting MF patients. (NCT01668173)

Epigenetic Therapies

Modulating epigenetic regulation represents another treatment strategy for MF. Somatic mutations of genes involved in epigenetic regulation, including TET2, DNMT3A, EZH2, ASXL1, and IDH1/2, have been identified in patients with MPN, particularly those with MF. (21) Of these, ASXL1, EZH2 and IDH1/IDH2 mutations appear to independently and adversely influence prognosis and leukemia-free survival.(60) In addition, hypermethylation of genes in the JAK-STAT pathway and transcriptional silencing has been reported, along with increased HDAC expression in MPNs. (61–63) Furthermore, JAK2 translocates to the nucleus, resulting histone H3 modification. (64) These observations, including demonstration of epigenetic deregulating mutations in MF patients, which negatively impact prognosis and evolution, provide a rationale for modulating epigenetic regulation with histone deacetylase inhibitors and hypomethylating agents.

Histone Deacetylase Inhibitors

The histone deacetylase inhibitors panobinostat, givinostat, pracinostat, and vorinostat have been evaluated in MF patients. Perhaps the most promising preliminary results come for early-phase studies with panobinostat. In the phase 1 study, clinical responses were evaluated in 5 of 18 patients with MF receiving more than 6 cycles of therapy.(65) Three of 5 patients had a 100% reduction in splenomegaly (clinical improvement) and stable disease was reported in 2 others. Two patients experienced an improvement in hemoglobin; 1 patient had a near complete response, and another experienced resolution of marrow fibrosis. MF symptom improvement was also reported. Thrombocytopenia was the dose-limiting toxicity: 39% and 17% of patients had grade 3/4 anemia and neutropenia, and 33% had fatigue and musculoskeletal AEs.(65) In another phase 2 study patients were treated with 40 mg 3 times per week. While only 1 patient (3%) had a clinical response, analysis of peripheral blood cells from treated patients revealed decreases in JAK/STAT signaling, JAK2^V617F allele burden, and inflammatory cytokines. (66) Thrombocytopenia and diarrhea were common AEs and only 16 of 35 patients were able to complete more than 2 cycles of therapy. These studies suggest that a dosing regimen of 30 mg 3 times per week or less may be necessary.

In preclinical studies, a combination of ruxolitinib and panobinostat resulted in synergistically reduced clonal growth, decreased marrow hypercellularity, and improved fibrosis.(67) As a result, ruxolitinib and panobinostat in combination is currently being evaluated (NCT01433445 and NCT01693601). Results from the expansion of the phase 1b dose-finding study were recently presented. (68) In this report of 61 patients (38 escalation phase, 23 expansion phase), the recommended phase 2 dose was ruxolitinib 15mg BID plus panobinostat 25mg 3 times per week every other week. Among 34 patients treated at this dose, grade 3/4 anemia and thrombocytopenia were seen in 32% and 24% of patients, respectively, while 12%, 9%, and 9% experienced diarrhea, asthenia, and fatigue. Encouraging spleen responses were observed, including 79% of patients who achieved at least a 50% reduction in palpable splenomegaly and 53% who had a 100% reduction in palpable splenomegaly. (68)

In a multicenter, phase 2 study of the HDAC inhibitor givinostat, gastrointestinal side effects were common (62%), as were grade 1/2 anemia (21%) and thrombocytopenia (10%) and response rates were low in MF (19%).(69) The HDAC inhibitor pracinostat has also been evaluated in MF, but patients only experienced modest responses, with fatigue and cytopenias limiting its use. (70) Vorinostat has also been tested, but resulted in low response rates and frequent toxicity.(71) Thus, low-dose HDAC inhibitors combined with ruxolitinib appear to represent the best strategy for therapy.

Hypomethylating Agents

The hypomethylating agents, 5-azacytidine and decitabine have both been studied in MF. In one study of 34 MF patients, 5-azacytidine (7-day schedule) resulted in global hypomethylation but few clinical responses that were short lived (24%; 1 with partial remission and 7 with clinical improvement).(72) In addition, myelosuppression was relatively common (grade 3/4 neutropenia, 29%). In another study of 10 patients with MF, no patient improved when 5-azacytidine was administered on a 5-day schedule, and discontinuations were frequent.(73) The best results with 5-azacytidine have been reported in accelerated phase MPNs transforming to myelodysplastic syndromes (MDS) or acute myeloid leukemia: an overall response rate of 52% was reported, with disease often reverting back to the chronic-phase MPN.(74)

When decitabine was evaluated in 21 patients with MF, 7 of 19 evaluable patients responded (1 CR, 2 PR, and 4 with hematologic improvement), although reduction in spleen size was not reported. (75) However, grade 3/4 neutropenia and thrombocytopenia occurred in 95% and 52% of patients, respectively, limiting its use. Case reports have also reported efficacy in treating high-risk MF with decitabine.(76, 77) A recent case report reported the use of hypomethylating agents in combination with ruxolitinib.(78) Ruxolitinib in combination with 5-azacytidine will be formally studied in a phase 3 trial. (NCT01787487) In addition, decitabine and ruxolitniib will be studied in those with MPN-blast phase/accelerated disease (clinical trials identifier: NCT02076191).

Targeting Additional Signaling Abnormalities

Inhibition of Mammalian target of rapamycin, PI3K, and AKT

Phosphatidylinositol 3-kinase (PI3K)/AKT/mammalian target of rapamycin (mTOR) pathway activation represents an additional target for MF treatment. This pathway is downstream from JAK/STAT and plays a role in cell growth, proliferation, and survival, and has been shown to be activated in MPN cells. (79) Initially, RAD001 (everolimus), an mTOR inhibitor, was shown to inhibit proliferation of MPN cell lines and primary samples. (80) Subsequently, everolimus was studied in 39 (30 evaluable) MF patients across multiple centers.(79) Six patients had a clinical improvement (spleen reduction in 5 patients, Hgb increase in 1 patient), and 1 patient had a partial remission; 23 patients had stable disease. Twenty-seven percent of patients experienced grade 2/3 anemia and 7% and 3% experienced grade 2 neutropenia and thrombocytopenia, respectively. Mild stomatitis occurred in 70% of patients.(79)

The combination of PI3K/mTOR inhibitors and JAK1/2 inhibitors has been reported to result in synergistic inhibition of MPN cell growth, suggesting that combining PI3K/mTOR inhibitors and JAK inhibitors may be a potential therapeutic combination. (81, 82) Ruxolitinib combined with the PI3K inhibitor buparlisib (BKM120), is currently being evaluated in a study now recruiting patients. In a preliminary report of 33 patients encouraging spleen responses (≥50% reduction) were noted 70% and 54% of JAK-inhibitor naïve and resistant patients, respectively. (83) Grade 3/4 anemia was seen in 30 and 31%; grade 3/4 thrombocytopenia was seen in 40% and 62% of JAK-inhibitor naïve and resistant patients, respectively. Non-hematological toxicities were typically grade 2 or less. The AKT inhibitor MK-2206 reduced hepatosplenomegaly and megakaryocyte burden in MPL-driven murine model, as well as megakaryocyte colony formation in patient samples. (84) In addition, when combined with ruxolitinib, MK-2206 synergistically inhibited growth of JAK2^V617F-SET2 cells, suggesting another novel treatment combination. (84)

Hedgehog inhibitors

The Hedgehog (Hh) pathway plays an important role in the maintenance of stem cell precursors, as well as proliferation and differentiation of hematopoietic stem cells. Activated Hh signaling can result in aberrant hematopoiesis, and may contribute to the pathogenesis of hematological malignancies. (85) Increased expression of Hh signaling pathway mediators have been demonstrated in the granulocytes of MPN patients, and therefore, inhibition of this pathway offers a potential strategy for MF treatment (14, 85) In a study with the Hh inhibitor PF-04449913, 5 MF patients achieved stable disease and 1 achieved clinical improvement in splenomegaly.(86) A phase 2 trial has been initiated to investigate the efficacy of this agent in MF patients and a variety of other hematologic malignancies. (NCT00953758) A phase 2 clinical trial of saridegib in MF was recently reported, but did not suggest efficacy as a single agent. (87) Similarly, erismodegib, another inhibitor of Hh signaling, did not show efficacy alone, but when combined with ruxolitinib resulted in a reduction in bone marrow fibrosis.(88) The combination of the smoothened (SMO) inhibitor, sonidegib (LDE225) with ruxolitinib was recently presented in 23 patients, and 65% achieved ≥ 50% reduction in palpable splenomegaly, with 9 achieving complete resolution. (89) Among three dosing cohorts (ruxolitinib 10mg BID/sonidegib 400mg daily; ruxolitinib 15mg BID/sonidegib 400mg daily; and ruxoltinib 20mg BID/sonidegib 400mg daily), fatigue was the most common adverse event, and other serious events included facial edema, hyponatremia, pyrexia, right ventricular failure, and increased creatine kinase. Grade 3/4 anemia and thrombocytopenia were noted in 22% and 4%, respectively. (89)

Anti-fibrotic agents

Another novel strategy for MF treatment involves the targeting the genes involved in the formation of bone marrow fibrosis, such as transforming growth factor β, and the enzyme lysyl oxidase, which is involved in megakaryocyte proliferation and fibrosis.(90) Fresolimumab (GC1008; NCT01291784), a monoclonal antibody directed against transforming growth factor β, is in early phase testing. A monoclonal antibody directed against lysyl oxidase, simtuzumab (GS-6624) is currently being evaluated, either alone or in combination with ruxolitinib (NCT01369498). In addition, PRM-151, an anti-fibrotic agent which was first studied in idiopathic pulmonary fibrosis,(91) is under evaluation both alone or in combination with ruxolitinib (NCT01981850). Preliminary phase 2 study results were recently presented showing that PRM-151 is tolerable either alone or in combination with ruxolitinib; significant myelosuppression was not seen, 9 of 26 evaluable (35%) responded, including 6 with improvement in bone marrow fibrosis. (92) Seventy-seven percent of patients experienced stable disease, and there were trends toward improvement in anemia (6 patients/40%), thrombocytopenia (8 patients/62%), blast counts (3 patients/21%), symptoms (10 patients/38%) and spleenomegaly (5 patients/26%). From this preliminary report, most AE’s were grade 1/2 and unrelated to the study regimen; 3 grade 3 and 5 potentially serious SAE’s were noted. (92)

Targeting megakaryocyte atypia

Megarkaryocyte proliferation and atypia is a central histological feature in MF that also contributes to disease pathogenesis. Recent data showed that inhibiting aurora kinase A with dimethylfasudil and MLN8237 (alesertib) resulted in polyploidization, arrested proliferation, and apoptosis in a JAK2^V617F megakaryocytic SET2 cell line and in megakaryocytes derived from primary patient samples. (93) Synergy was shown with ruxolitinib, and these polyploidy inducers resulted in growth arrest of JAK-inhibitor persistent cells. Polyploid induction also reduced liver and spleen weight, leukocytosis/thrombocytosis, bone marrow fibrosis, and TGF-β levels in murine MF models. (93)

Additional novel strategies

Another intriguing strategy being tested is to combine JAK-inhibitors with definitive therapy, such as allogeneic stem cell transplantation (ASCT) (NCT01790295). (94) A study of 22 patients who received ruxolitinib prior to ASCT showed that patients experienced improvements in symptoms (86%) and splenomegaly (41%) before proceeding to transplant, and after discontinuation of ruxolitinib no severe rebound syndrome was noted. Immune reconstitution, as assessed by CD3, CD4, and CD8 levels at day 100 did not differ between ruxolitinib-treated (N=10) and untreated patients (N=11). The authors (94) concluded that early outcomes were not negatively influenced by this strategy, though toxicity (tumor lysis in 3 patients and cardiogenic shock in 3 patients) from another study were cited. (95)

Finally, telomerase inhibition with imetelstat may also represent a novel approach to MF treatment (NCT01731951), as 4 CRs (4 with reversal of fibrosis, 3 with complete molecular response) and 3 PRs were recently reported among 33 patients treated with the drug. (96) Four of 13 transfusion-dependent patients experienced anemia responses and 9/23 (39%) experienced ≥ 50% improvement in splenomegaly. Interestingly, responders were more likely to have JAK2 mutations and lack ASXL1 mutations; further, CR rates were noted in those was spliceosome mutations (SF3B1/U2AF1). Myelosuppression was the most common AE; grade 4 neutropenia and thrombocytopenia was seen in 18% and 27% of patients, respectively. Grade 5 intracranial bleeding was noted in 1 patient, as was grade 5 GI bleeding (not related). Liver function abnormalities were grade 1/2, including abnormalities in bilirubin (46%), alkaline phosphatase (52%), and AST/ALT (55%, 24%, respectively). (96)

Practical Approaches: Reducing Anemia Associated with JAK-Inhibitors

With the possible exception of momelotinib,(37) anemia is an expected consequence of JAK-inhibitor therapy, as these agents also typically inhibit wild-type JAK2 signaling. Therefore, a practical combination would include anemia-directed therapies and a JAK-inhibitor. One such combination involves danazol and ruxolitinib, (NCT01732445), based on retrospective studies showing modest anemia responses. (97, 98) Modest results from this combination were reported, including stable disease in 80% and CI in 10%. (99) Erythropoietin stimulating agents (ESA) have also resulted in modest improvements in MF-associated anemia, in part influenced by the baseline epo level and severity of anemia.(100, 101) A theoretical concern with these agents includes progression of splenomegaly; however, an analysis of 13 patients treated with a combination of ruxolitinib and ESAs did not report show any impairment in splenic response to ruxolitinib, though it should be noted that there was no change in transfusion rates in that study.(102) Finally, the addition of immunomodulatory drugs (IMIDs) represents another potential strategy. In a cooperative group study of lenalidomide with prednisone, only 19% of patients met criteria for an anemia response.(103) In a pooled single-center analysis, response rates varied between 34–48%. (104) Ruxolitinib in combination with lenalidomide is currently under investigation (NCT01375140), but does not appear tolerable together due to myelosuppression. (105) Unfortunately, phase 3 data comparing pomalidomide to placebo for the treatment of anemia has tempered enthusiasm for this agent, since the anemia responses were similar between the two treatment groups.(106) Pomalidomide in combination with ruxolitinib is currently under investigation (NCT01644110) and results from 6 patients have been presented. (107)

Conclusion

The discovery of the JAK2^V617F mutation in MF has ushered in a new era marked by a better understanding of disease pathogenesis, and the development of the first medication specifically approved for MF. Recent work has confirmed a central role for JAK-STAT dysregulation in MPNs, even in patients without the JAK2^V617F mutation. (12) In addition, recent investigations suggest that JAK2 remains a valid and appropriate therapeutic target.(14) Clinical experience with JAK inhibitors thus far shows that they are effective in relieving disease-related symptoms and splenomegaly. Longer term follow-up from the COMFORT 1 and 2 studies show that the responses are durable and adverse event rates decline over time. (9–11) Unique benefits have been reported with many of the JAK inhibitors discussed in this review, including a potential to stabilize or improve bone marrow fibrosis, improve anemia, and prolong survival. However, to maximize the potential of this class of medications, perhaps they are best used as a foundation for combination therapies. The identification of JAK inhibitor resistance and contributions of epigenetic regulators and JAK-STAT-related pathways (i.e, mTOR/PI3K/AKT) to disease pathogenesis suggests a number of companions for JAK inhibitors. A practical approach includes the partnering of JAK inhibitors with drugs previously used to treat MF-associated anemia, including IMiDs, ESAs, and androgens. (Table 2) Along these lines, a novel erythropoietic agent, sotatercept (ACE-011) is being evaluated in MF (NCT01712308). The role of JAK-inhibitors in low-risk, early MF has not yet been defined, but if this class of medications can delay or stabilize fibrosis with long-term exposure, perhaps future studies will evaluate this concept. Similarly, the role of interferon early on in the MF disease course is under investigation (NCT01758588), based on a prior study suggesting its potential to modify bone histology in MF. (108) One novel idea is to combine JAK inhibitors with interferon in an effort to improve disease signs and symptoms while also potentially preventing or reversing BM fibrosis. (109) Another intriguing concept involves the use of JAK-inhibition prior to transplant, as discussed above.

Table 2.

Clinical trials with JAK-inhibitors in combination with other therapies

Combination	Clinical trials identifier(s)
Ruxolitinib with Panobinostat (HDAC inhibitor)	NCT01433445 and NCT01693601
Ruxolitinib with 5-azacytidine (Hypomethylating agent)	NCT01787487
Ruxolitinib with decitabine	NCT02076191
Ruxolitinib with buparlisib (BKM120, a PI3K inhibitor)	NCT01730248
Ruxolitinib with erismodegib (hedgehog inhibitor)	NCT01787552
Ruxolitinib with PRM-151 (anti-fibrotic agent)	NCT01981850
Ruxolitinib with lenalidomide (IMID)	NCT01375140
Ruxolitinib with pomalidomide (IMID)	NCT01644110
Ruxolitinib with danazol	NCT01732445
Ruxolitinib prior to stem cell transplantation	NCT01790295

Many of these clinical trials are already recruiting and preliminary results have been presented. (Table 2) While it is hoped that such combinations can lead to disease-modifying effects, to date, though results are preliminary, most of the reported clinical benefits from JAK-inhibitor clinical trials revolve around improvement of splenomegaly. Alternatively, some strategies have shown only modest benefit (ruxolitinib with danazol), or no benefit (ruxolitinib with lenalidomide). Last, though combination therapy is intriguing, the preliminary reports described in this review may underestimate or underplay potential overlapping or unique toxicities, which will require careful consideration. Taking everything together, with the rapid pace of discovery, emergence of novel agents and development of logical therapeutic combinations, there is optimism that MF patients may achieve not only better disease control, but also disease remission, akin to the experience in CML.