Dysregulation of the JAK pathway plays a role in the development of numerous tumor types; it is particularly central to the pathophysiology of myelofibrosis and has long been recognized as a potentially valuable therapeutic target in that malignancy.
Following the identification of the JAK kinase family in the late 1980s, the novel enzyme group was colloquially known as “just another kinase.”1,2 Since then, these tyrosine kinases have defied that reputation amid abundant evidence showing that they transmit a variety of signals into the cell with many biological consequences.
Dysregulation of the JAK pathway plays a role in the development of numerous tumor types; it is particularly central to the pathophysiology of myelofibrosis (MF) and has long been recognized as a potentially valuable therapeutic target in that malignancy. Ruxolitinib (Jakafi), the first JAK inhibitor to gain FDA approval, has become a centerpiece in the treatment of patients with MF. JAK inhibition has not proved effective in other tumor types, and thus far, no other JAK-targeting therapies have become available.
That picture, however, may be changing. A better understanding of the complexities of JAK signaling in normal and cancerous cells and the design of rational drug combinations, are poised to expand the use of JAK inhibitors in anticancer therapy for hematologic, and possibly solid, tumors. Ruxolitinib continues to be explored in multiple malignancies, and several promising novel agents are being evaluated in clinical trials (TABLE).JAKs are now known as the Janus kinases, after the 2-faced Roman god of the same name, in recognition of their most striking feature: the presence of 2 near-identical kinase domains. There are 4 known members of the JAK family: JAK1, JAK2, JAK3, and TYK2, all sharing 7 conserved JAK homology (JH) domains within their protein structures.
At one end of the protein are JH1 and JH2, which comprise the 2 kinase domains. The first has tyrosine kinase activity, while the second is a pseudokinase domain with no catalytic activity that is thought to negatively regulate the first.3,4 At the other end, 3 of the JH domains collectively make up a larger FERM domain through which the JAKs are recruited to and bound by the cell-surface receptors that activate JAK signaling. A third domain, SH2, facilitates signal transducer and activator of transcription (STAT) signaling (FIGURE 1).3,5
Cytokines and their receptors are among the most significant activators of the JAK pathway. Unlike tyrosine kinase receptors, cytokine receptors have no catalytic activity and are reliant upon the JAKs to perform kinase functions for them. JAKs bind to the part of the cytokine receptor that protrudes into the cell via their FERM domain. When the cytokine receptor is activated, it pairs up with another receptor molecule and this brings 2 JAKs close enough that they are able to phosphorylate and activate one another.
Activated JAKs then phosphorylate other target proteins, including the cytokine receptor to which they are bound, forming a binding platform for proteins containing an SH2 domain, primarily STAT proteins. Once they dock to the cytokine receptor, the STATs themselves are phosphorylated by the JAK proteins and become activated (FIGURE 2).
Activated STATs subsequently disengage from the cytokine receptor, pair up with one another and move into the nucleus, where they attach to specific sequences within the DNA to either activate or, less commonly, repress the transcription of a host of target genes. In this way, the interaction between JAKs and STATs transmits a signal from outside of the cell into the nucleus to coordinate a response to the signal.
Although this at first appears to be a relatively simple signaling cascade, the activity is complicated by the fact that there are 7 STATs. The diversity of possible JAK-STAT pairings, in addition to the range of different ligands and receptors that can activate JAKs upstream, allows for the multiplicity of biological effects coordinated by this pathway.
In addition, the pathway is negatively regulated in several ways. One involves protein tyrosine phosphatases (PTPs), which are enzymes that can deactivate JAK by reversing its phosphorylation. The suppressors of cytokine signaling (SOCS) are a family of proteins that are activated by STATs, and turn off the JAKSTAT pathway, thereby forming a negative feedback mechanism. They do this in several ways: directly inhibiting JAK kinase activity; blocking recruitment of downstream effector proteins; or mediating proteasomal degradation of JAKs. Finally, protein inhibitors of activated STATs (PIAS) block the activity of STATs in the nucleus through mechanisms that remain somewhat unclear.
Further adding to the complexity is the significant cross-talk with other signaling pathways, of which we currently have a relatively poor understanding. The best characterized interaction is with the mitogen-activated protein kinase pathway, which crosses paths with JAKSTAT signaling at multiple points.3,6,7As a result of its myriad roles in essential cellular processes such as growth, proliferation, survival, inflammation, and immunity, JAK pathway signaling is tightly regulated in normal cells by a delicate balance between activators and inhibitors.
Disruption of this balance and aberrant activation of the JAK pathway contributes to the development and progression of numerous cancer types by promoting many of the hallmark abilities of cancer cells, such as unchecked cell growth and proliferation. Because of its important role in regulating cytokine signaling, dysregulation of the JAK pathway can also contribute to inflammation and immunosuppression in the tumor microenvironment.
The mechanisms by which the JAK pathway becomes dysregulated include mutations in the receptors that sit upstream of JAK, which drive persistent activation of JAK, or in downstream effectors, which would bypass the need for JAK activation. Alternatively, the function of molecules involved in deactivation of the pathway, such as PTPs and SOCS, may be lost.
Fusions, mutations, deletions, and insertions in the JAK1 and JAK3 genes have been noted in intestinal, skin, stomach, and endometrial cancers, among others, whereas mutations in JAK2 have been identified in various hematologic malignancies.8
In particular, the JAK pathway is central to the pathogenesis of myeloproliferative neoplasms (MPNs), including MF. These are cancers of the blood stem cells that cause an overproduction of white or red blood cells or platelets. Three mutually exclusive activating mutations in JAK2 are found in approximately 90% of patients with MF. The most commonly observed mutation is JAK2 V617F, which drives ligand-independent activation of the JAK pathway, and is observed in 40% to 60% of patients.9,10
Activation of STATs is also commonly reported in several cancer types, including prostate cancer, non—small cell lung cancer (NSCLC), glioblastoma, and thyroid cancer, but is particularly prevalent in breast cancer. Upstream of JAK, increased expression of the cytokine interleukin-6 (IL-6) is observed almost universally across tumor types.3,6Ruxolitinib, a small-molecule inhibitor of JAK1 and JAK2, received FDA approval in 2011 for the treatment of patients with intermediateor high-risk MF, including primary MF, or MF that follows polycythemia vera (PV) or essential thrombocythemia.
Approval in this indication was based on data from 2 randomized phase III trials, COMFORT-I and COMFORT-II, which collectively involved more than 500 patients. Ruxolitinib more effectively reduced spleen size and MF symptom burden compared with placebo (COMFORT-I) and best available therapy (BAT; COMFORT-II). These studies also demonstrated an improvement in overall survival (OS) with ruxolitinib treatment. Ruxolitinib was well tolerated; grade 3/4 adverse events (AEs) included thrombocytopenia and anemia but these were manageable and did not often lead to treatment discontinuation.
The survival benefit was confirmed in recently published 5-year follow-up data for both trials, which also showed that long-term treatment with ruxolitinib was safe, with no significant cumulative toxicity. Long-term data from COMFORT-I also suggested that ruxolitinib improved patient quality of life.11-14 Various clinical trials of ruxolitinib are ongoing, including in other MPNs, hematologic malignancies, and solid tumors.Although there are no other FDA-approved JAK inhibitors, other drugs have generated positive data in phase III clinical trials. Momelotinib, also a JAK1/2 inhibitor, has been evaluated in the phase III SIMPLIFY trials in patients with MF.
In SIMPLIFY-1, momelotinib was compared with ruxolitinib in patients with JAK inhibitor—naïve MF. The noninferiority of momelotinib was established in terms of the primary endpoint of a ≥35% reduction in spleen volume at 24 weeks. However, ruxolitinib was superior to momelotinib in the key secondary endpoint of a ≥50% reduction in total symptom score (42.2% vs 28.4%, respectively). There was also a higher rate of infections and treatment-emergent peripheral neuropathy with momelotinib.
Conversely, in the SIMPLIFY-2 trial in patients previously treated with ruxolitinib, participants who received momelotinib for 24 weeks did not experience superior outcomes for splenic response (≥35% reduction in spleen volume) compared with BAT (6.7% vs 5.8%, respectively). Nevertheless, momelotinib was significantly better than BAT at improving disease-related symptoms, defined as a ≥50% reduction in total symptom score (26.2% vs 5.9%) and in the transfusion independence rate (43.3% vs 21.2%). Given these mixed results, momelotinib’s future is uncertain as researchers attempt to determine the best next steps in its evaluation.15,16
Pacritinib is a small-molecule inhibitor of JAK2 and the cytokine receptor FLT3. This drug has shown significant efficacy in a phase III clinical program in MF. In the PERSIST-1 trial, 327 patients were randomized to receive pacritinib or BAT as first-line treatment. At 24-weeks, spleen volume was reduced by ≥35% in 19.1% of patients treated with pacritinib compared with just 5% of those treated with BAT.
In the PERSIST-2 trial, pacritinib (at 2 different dosages—400 mg once daily or 200 mg twice daily) was also compared with BAT in 311 patients with MF and thrombocytopenia. The primary endpoint of a ≥35% reduction in spleen volume at 24 weeks was achieved in a significantly greater number of patients treated with either dosage of pacritinib than in those who received BAT (18.1% vs 2%, respectively).
In terms of this primary endpoint, the twice-daily dose was more effective than the once-daily regimen (21.6% vs 14.7%, respectively). Pacritinib was also superior in terms of total symptom score, with 25% of patients demonstrating a ≥50% reduction compared with 14% for BAT. Additionally, pacritinib reduced transfusion dependence.17,18
Despite this display of substantial efficacy, higher death rates from intracranial hemorrhage, heart failure, and cardiac arrest in the PERSIST-2 trial prompted the FDA to place a clinical hold on pacritinib in February 2016. However, less than a year later, CTI BioPharma, the company developing the drug, reported that the clinical hold had been lifted after submission of final data from the PERSIST trials.19 A new clinical trial examining the safety of pacritinib in patients with primary MF who have previously received ruxolitinib also has been initiated (PAC203; NCT03165734).19
The development of several other promising JAK inhibitors has been discontinued due to severe toxicities. AZD1480 caused neurological AEs, and lestaurtinib (CEP-701) resulted in frequent gastrointestinal toxicity. Fedratinib (SAR302503), a selective JAK2 inhibitor that showed significant efficacy in patients with MF in the phase III JAKARTA-1 trial, was put on a clinical hold by the FDA in 2013 after reports of potential cases of Wernicke encephalopathy. Sanofi subsequently discontinued its development.
However, fedratinib may be making a comeback. The rights for global development have been acquired by Impact Biomedicines. The company recently announced that the clinical hold was lifted based on a review of additional data from 18 studies in more than 800 patients, including JAKARTA-1 and the follow-up JAKARTA-2 study, which recruited patients who were resistant to all available therapies including ruxolitinib.
In JAKARTA-2, 55% of patients treated with fedratinib experienced a ≥35% reduction in spleen size. The most common AEs were anemia, nausea, diarrhea, and vomiting, and there were no cases of encephalopathy. No clinical trials of fedratinib are ongoing, but Impact Biomedicines plans to commercialize the drug for patients with MF and PV.20-22The complexity of the JAK pathway, the pleiotropic nature of its biological effects, and substantial amount of cross-talk between it and other important signaling pathways presents a significant challenge to the effectiveness of JAK inhibitors. The development of rational combinations of drugs could help to improve outcomes and these are starting to be studied in clinical trials.
Preclinical study results have suggested synergy between JAK inhibitors and epigenetic modifiers such as histone deacetylase and DNA methyltransferase inhibitors (DNMTs); agents targeting the Hedgehog (Hh) pathway, which plays a critical role in hematopoietic cell development and stem cell survival; inhibitors of PI3K, a key downstream effector of JAK signaling; and immune checkpoint inhibitors targeting PD-1 and its ligands.
The results of an ongoing phase II trial of ruxolitinib plus azacitidine, a DNMT inhibitor, in patients with MF, MPNs, and myelodysplastic syndrome (MDS) were recently presented at the 2017 American Society of Clinical Oncology Annual Meeting. Among the 39 patients evaluable for response, the objective response rate was 72%, and 79% of patients had >50% reduction in spleen size. Results from an ongoing phase I/II study of ruxolitinib plus decitabine, also a DNMT inhibitor, in accelerated phase MPN or post-MPN acute myeloid leukemia demonstrated a complete response (CR) or CR with incomplete blood count recovery in 33% of patients.23,24
A phase I/II study of ruxolitinib in combination with the Hh inhibitor sonidegib in patients with MF is ongoing, but not actively recruiting participants. The regimen has demonstrated a ≥35% reduction in spleen size at 24 weeks in 44% of patients and was well tolerated.25 Meanwhile, the combination of JAK and PI3K inhibition has also been investigated in early-stage clinical trials. The phase Ib HARMONY trial showed promise for the combination of ruxolitinib plus buparlisib, and although studies of that combination are no longer ongoing, ruxolitinib is being evaluated in combinations with TGR-1202 and INCB050465, which are inhibitors of the delta isoform of PI3K. Preliminary data from the TGR-1202 study showed that the combination was well tolerated, with a clinical benefit rate of 89%.26
Despite initial promise, JAK inhibitors have suffered several high-profile failures in clinical trials in patients with solid tumors. Incyte has halted its development of ruxolitinib in solid malignancies, following lack of efficacy in patients with pancreatic and colon cancers. However, combination therapy also could help to boost efficacy in this instance and several trials involving other JAK inhibitors are underway.
For example, there is evidence that activation of the JAK pathway could be a contributing factor to EGFR inhibitor resistance in EGFR-mutant NSCLC. A phase I study of ruxolitinib in combination with afatinib in patients with acquired resistance to EGFR inhibitors is ongoing, and preliminary data were recently presented. The combination demonstrated promising antitumor activity and was well tolerated.27 Also in this setting, itacitinib, a JAK1 inhibitor, is being evaluated in combination with osimertinib.There is also some interest in targeting other components of the JAK pathway, with the cytokine IL-6 emerging as the most promising target thus far. Monoclonal antibodies have been developed against IL-6, primarily for the treatment of inflammatory diseases such as rheumatoid arthritis. Siltuximab (Sylvant), an IL-6 inhibitor, is being evaluated in multiple myeloma. As the principal downstream target of JAK signaling, the STAT proteins also make rational targets, but have proved to be a therapeutic challenge thus far. Several types of STAT inhibitors, predominantly aimed at STAT3 and STAT5, have been developed, including small-molecule inhibitors and, more recently, antisense oligonucleotides. The latter are short stretches of DNA or RNA that can bind to specific stretches of messenger RNA, and prevent its translation into a protein. AZD1950 is one example of this technology that has demonstrated single-agent antitumor activity in patients with treatment-refractory lymphoma or NSCLC.