Review of novel agents for myeloproliferative neoplasms

Common “Philadelphia chromosome”-negative myeloproliferative neoplasms (MPN) include essential thrombocythemia (ET), polycythemia vera (PV) and myelofibrosis (MF), which are a clinically diverse group of diseases. In the last 15 years, there have been rapid advances in understanding the importance and therapeutic potential of the Janus kinase/signal transducers and activators of transcription (JAK/STAT) signaling pathway in MPN. In their recent review,¹ Donal McLornan and Claire Harrison, both from the Department of Haematology, Guys and St Thomas’ NHS Foundation Trust, London, UK, aimed to discuss the therapeutic challenges, diagnostic issues, molecular characterization, monitoring and management, novel therapies and treatment failure definition in MPN.

Issues and challenges

In their discussion of MPN-related issues and challenges, Dr McLornan and Prof Harrison highlighted a study by Grinfield et al. (2018)¹ which indicated that MPN outcomes are better predicted by genomic profiling than by traditional clinical criteria. However, they felt that as genomic profiling did not predict outcome with therapy, nor thrombosis, these assessments were unlikely to be adopted at present. Grinfield et al. showed that patients with ET that lack driver mutations (triple negative ET) have the best outcomes amongst ET patients and may therefore benefit from de-escalation of therapy to prevent iatrogenic harm. Further, in an aspirin study by Alvarez-Larran et al. (2016)¹, calreticulin (CALR) mutated ET was associated with a bleeding rather than a thrombotic phenotype, suggesting that selected CALR mutated patients may also benefit from de-escalation of therapy.

Studies of novel oral anticoagulants (NOAC) for the prevention of thrombosis suggest that they present a safe and effective treatment, but longer-term studies are needed.

MAJIC-ET, a phase II study of ruxolitinib (a JAK inhibitor; rux) vs best available therapy (BAT) found no difference in efficacy between the treatments but did show that a subgroup of patients with hydroxycarbamide (HC) resistant/intolerant ET have significantly worse outcomes. These were related to the presence of tumor protein p53 (TP53) and splicing factor mutations that led to reduced transformation-free survival [splicing factor 3b subunit 1 (SF3B1), zinc finger CCCH-type RNA binding motif and serine/arginine rich 2 (ZRSR2) and serine and arginine rich splicing factor 2 (SRF2)]. Ten-eleven translocation 2 (TET2) mutations were not associated with decreased survival.

When comparing HC and interferon (IFN) treatment in PV, studies such as the PROUD study have demonstrated that IFNs (specifically pegylated interferon alpha 2b) are considered key players in the treatment of PV. Indeed, the results of the PROUD trial led to the approval of IFN compounds for the treatment of patients with PV without splenomegaly. The RESPONSE phase III trial, which led to the approval of rux for the treatment of hydroxyurea resistant/intolerant PV, suggested that rux may also be effective after IFN treatment. The ongoing phase III study MITHRIDATE will assess rux vs HC or IFN as first line therapy in high-risk PV in terms of thrombosis, hemorrhagic events and disease progression.

Dr McLornan and Prof Harrison go on to discuss how most available MPN studies have not focused on disease progression as an outcome, largely due to the low event rate of progression hence requiring large patient numbers and long term follow-ups. Surrogate markers of disease progression are needed as many studies use markers that have not been shown to correlate with progression (with the exception of hematocrit). Interestingly, the analysis of the MAJIC-PV trial showed that patients achieving a European LeukemiaNET (ELN) partial molecular response by one year (≥50% decrease in JAK2 V617F allelic burden) have improved progression-free survival and no thrombotic events.

Studies of MF include the COMFORT I and II studies, which showed that rux therapy in intermediate-2 or high-risk patients (as defined using the International Prognostic Scoring System [IPSS]) improved overall survival. In addition, the studies found that spleen volume reduction (SVR), and reduction in total symptom score (TSS) may be good surrogate markers of long-term benefit for MF.

Unanswered questions remain regarding the level of SVR and TSS reduction that would be acceptable for adoption of novel therapies, and whether there are other endpoints that may be important indicators of successful treatment. In addition, there are questions around how to define failure and outcomes after rux failure in patients with MF. These are not standardized and vary between trials, and although transplantation remains the best option for a cure, when to transplant patients on JAK inhibitors, the role of these in lower-risk patients and how JAK inhibition may contribute to a reduction in graft-versus-host disease (GvHD) remain largely unknown.

Patient perspectives

In the MPN Landmark survey,¹ fatigue was highlighted as the most common and significant symptom that patients were most keen to address. Fatigue is an elusive side effect that cannot be easily addressed by current treatments and is in fact a side effect of many drugs used for MPN treatment. Patients also reported that interference with family and social life, heavy emotional burden, frequent cancellation of planned activities, reduced productivity, financial hardships and impacts on employment, were significant factors that impacted them, with only a very few of those being assessed in trials. According to the survey, patients also felt that one of the most important treatment outcomes was the prevention of disease progression.

Genomics – classification and prognostics

The Grinfield et al. study (2018), which was mentioned previously, included 2,035 patients (1,321 ET, 356 PV, 309 MF, and 49 other MPN) and by utilizing Bayesian modelling of genomic profiles, revealed eight different MPN genomic subgroups that correlated with clinical phenotype and outcome and may enable treatment stratification once further validated.

Overview of selected agents

JAK/STAT

• Rux is a JAK 1/2 inhibitor that leads to SVR and TSS reduction with quality of life benefits in MPN patients (COMFORT I and II trials). Nevertheless, rux has no impact on acute myeloid leukemia (AML) transformation risk and only a small benefit on bone marrow (BM) fibrosis. Rux-associated adverse events (AE) included anemia, thrombocytopenia, immunosuppression, risk of non-melanoma skin cancer, and a potential link to B-cell lymphoma. The average response lasts for ~168 weeks, with little known about why response fails
• Fedratinib is a JAK2 inhibitor with anti-FMS-like tyrosine kinase 3 (FLT3) properties. It has been approved for first and second line treatment of intermediate-2 and high risk MF. There were previous concerns over an increased risk of Wernicke encephalopathy, which has now been revised and withdrawn, but monitoring for thiamine levels and risk of encephalopathy is still required. FREEDOM 1 and 2 are two currently ongoing trials assessing fedratinib in patients previously treated with rux
• Momelotinib is a JAK1/2 inhibitor that also inhibits iron sensing bone morphogenic protein-activin A receptor type 1 (BMP-ACVR1) suggestive of its potential to improve anemia. As a first line treatment it is non-inferior to rux in terms of SVR but not TSS response (SIMPLIFY 1 trial). As a second line treatment it is superior to rux in terms of TSS and transfusion independence but not SVR (SIMPLIFY 2 trial). Of particular concern was the treatment-emergent AE of peripheral neuropathy. The ongoing MOMENTUM trial is assessing momelotinib vs danazol as second line in anemic/transfusion independent patient with MF .
• Pacritinib is a JAK2, FLT3, interleukin/receptor-associated kinase 1 (IRAK1) and colony stimulating factor 1 receptor (CSF1R) inhibitor. The PERSIST trial was previously placed on hold due to concerns over pacritinib-associated cardiac and hemorrhage events, but these concerns were later withdrawn. Pacritinib is further investigated as a potential treatment for MF in the ongoing PAC203 and PACIFICA trials
• Itacitinib is a JAK1 inhibitor and has been found to have minimal hematological toxicity, acceptable TSS and SVR control rates
• PU-H71 inhibits the epichaperone heat shock protein 90 (HSP90) resulting in the disruption of HSP90 and degradation of JAK2. Studies into this agent are ongoing.
• Methotrexate inhibits JAK signaling. There are limited case reports and pre-clinical studies into this for MPN currently

Other signaling pathways

Despite studies of hedgehog signaling pathway inhibitors, and phosphatidylinositol 3 (PI3) kinase pathway inhibitors showing initial promise for MPN, these haven’t being investigated further as of yet.

Epigenetics

Due to mutations seen in MPN that affect epigenetic processing, the authors highlighted this as an area of growing interest.

• Bromodomain and extra-terminal motif (BET) inhibitors are thought to decrease cell proliferation and increase apoptosis. Mouse model studies suggest synergy between BET inhibition and JAK inhibition. The MANIFEST trial is assessing the BET inhibitor CPI-0610 as monotherapy and in combination with rux for the treatment of MF. Preliminary data presented at the American Society of Hematology (ASH) annual meeting in 2019 (Orlando, US) suggested that combination therapy with CPI-0610 may exceed rux monotherapy in terms of SVR and TSS benefit.
• Demethylating agents:
  • Azacytidine (aza) in combination with rux achieved an Interanional Working Group-Myelofibroris Research and Treatment (IWG-MRT) response in 72% of patients and 25% of these responses only occurred after aza was administered. There was a >50% reduction in spleen length at 24 weeks in 62% of patients, of which 95% maintained this at 48 weeks. A total of 9% of patients discontinued treatment due to cytopenias
  • Decitabine has mostly been studied in accelerated phase (AP) or blast phase (BP) MPN setting, where it has been trialed in combination with rux. The overall response rate (ORR) was 53%, the maximum tolerated dose was not reached, and discontinuation of treatment was due to AEs (37%) or disease progression (21%)
  • Lysine-specific demethylase 1 (LSD1) inhibition in preliminary data from ASH 2019 reduced spleen volume in 50% of patients and reduced TSS in 79% of patients with no dose limiting toxicity, progression to AML or death
• Histone deacetylase (HDAC) inhibitors:
  • Vorinostat seemed promising but is too toxic
  • Givinostat has shown promising activity as monotherapy or in combination with HC for patient with PV and is currently being evaluated in a phase III trial
  • Panobinostat in combination with rux had interesting responses, but studies have since were halted

Apoptosis targets

• B-cell leukemia/lymphoma 2 (BCL-2) inhibitor in combination with rux is being studied in an ongoing phase II trial for the treatment of MF. Preliminary data was presented at ASH 2019 showing that 42% of patients achieved SVR of ≥35% at any time during the study. There was a decrease in driver mutation allelic burden of >5% in 42% of patients and of these 25% had BM fibrosis improvement of grade ≥1. The study reported a grade ≥3 AE rate of 77% that included thrombocytopenia and anemia. The serious AE (SAE) event rate was 15%, including anemia, upper abdominal pain, vomiting, chest pain, increased C-reactive protein, and abnormal liver function tests. There were no significant bleeding events or treatment-related deaths
• Mouse double minute 2 homolog (MDM2) inhibition restores functional p53 and preliminary data of its use in PV and MF is promising. Studies are ongoing with the MDM2 inhibitors idasanutlin (monotherapy and in combination with IFN), KRT-232 (for rux failure/intolerant MF and poorly controlled PB), and with an MDM2 antagonist in combination with BET inhibition in patients with MF
• Second mitochondria-derived activator of caspase (SMAC) mimetics are thought to lead to apoptotic cell death and have shown promise in patients with heavily pre-treated MF, leading to a reduction in anemia
• Aurora kinase A (AURKA) targeting can ultimately lead to apoptosis and a phase I trial of AURKA inhibition in MF has found it to be well tolerated, with 29% of patients showing a spleen response and 32% demonstrating symptom response. Further studies in combination with rux are planned

Inflammation and bone marrow

The authors discussed how the link between BM fibrosis and clinical outcome is unclear and highlighted how long term rux treatment correlates with a greater odds of fibrosis improvement when compared to BAT.

• Rux treatment is also associated with a better spleen size reduction
• Similar findings have also been reported with fedratinib treatment
• PRM-151, a recombinant human pentraxin-2 analogue, led to a reduction in BM fibrosis (presented at the European Hematology Association [EHA] Annual Congress 2019, here) in 28% of patients with MF, and a ≥50% reduction in the need for transfusion or an improvement of >10g/l hemaglobin for 12 consecutive weeks after treatment (16-29%).
• Vascular endothelial growth factor (VEGF) targeting with bevacizumab or vatalanib has demonstrated very little benefit
• Antifibrotic agents (which may be important in reducing fibrosis) have also been assessed:
  • Established ones such as pirfenidone show no clinical benefit
  • A study of fresolimumab (a monoclonal antibody to transforming growth factor β; TGFβ) has also not demonstrated any benefit in terms of fibrosis
  • Galunisertib (an inhibitor of TGFβ receptor 1 kinase) has shown promise in MPN mouse models
  • Lysyl oxidase-like-2 (LOXL2) antibodies have shown a limited effect in a phase II MF trial
  • Interim data, reported at ASH 2019, from a phase II trial of luspatercept (TGFβ ligand that reduces Smad2/3 signaling)in patients with MF and anemia) showed that 10% of the monotherapy patients with 21% of patients treated with combination luspatercept plus rux had increased hemoglobin of ≥1.5 g/dl, and 10% monotherapy and 32% combination achieved transfusion independence. Treatment-related AEs occurred in ≥3% of patients, including hypertension, bone pain, and diarrhea. There were 9% of patients who discontinued treatment due to AEs. Further studies of luspatercept are planned
• Studies of potential microenvironment targets include:
  • A phase II trial of sympathetic innervation signaling disruption with the oral β₃ adrenergic agonist mirabegron, which increased nestin-positive mesenchymal stem cells, a slight reduction in fibrosis and changes in megakaryocyte clustering in JAK2 V617F mutated MPN patients. Primary outcomes of JAK2 allelic burden reduction and splenomegaly reduction were not reached
  • The ongoing TAMARIN trial is assessing inhibition of estrogen signaling by tamoxifen in patients with MPN
• The authors also highlighted other targets of interest for MPN, such as hypoxia-inducible factor 1α (HIF1α), inhibition with bortezomib, angiotensin II production disruption by captopril that may lead to fibrosis modulation, and oligonucleotide telomerase inhibition that may reduce fibrosis

Looking ahead

In their review, Dr McLornan and Prof Harrison highlight the lack of mutation-specific therapies for MPN and pinpoint mutant CALR as a particular area of interest for future targeting. They also refer to JAK inhibition as the main area of focus  and mention that combination therapies with rux could increase the benefits of this treatment, but also suggest exploring sequential treatments. There are a number of highly ambitious combination studies ongoing, but the authors point out that they all focus on slightly different cohorts thus potentially impacting the applicability of the results, encouraging careful annotation of patient population in future studies. They also revisit the need to determine appropriate endpoints for studies in the MPN setting and suggest that this is “a pressing clinical need”.