Serving as gatekeepers at the entry to the cell cycle, CDK4 and CDK6 make ideal therapeutic targets to block the unchecked proliferation that is a hallmark of cancer cells.
In the oncology sphere, 2 members of the cyclin-dependent kinase (CDK) family have long stood out from the crowd. Serving as gatekeepers at the entry to the cell cycle, CDK4 and CDK6 make ideal therapeutic targets to block the unchecked proliferation that is a hallmark of cancer cells. Case in point: The FDA has approved 3 CDK4/6- targeted drugs for the treatment of a specific subtype of breast cancer.1
Yet the CDK family encompasses 20 proteins2 with important functions beyond cell cycle regulation.3 Growing appreciation of the role of CDKs in transcriptional regulation is fueling renewed interest in developing inhibitors of other CDKs, particularly CDK7 and CDK9.
A better understanding of the role of transcriptional CDKs has refocused the development of drugs such as alvocidib, which may have found a niche in several hematologic malignancies in combination therapies. The drug is an intravenous (IV) inhibitor of multiple CDKs.
Meanwhile, oral inhibitors of transcriptional CDKs also demonstrate promise in hematologic malignancies. Zotiraciclib (TG02), the most potent inhibitor of CDK9 developed to date, has received an orphan drug designation from the FDA for glioma therapy, based on positive phase I data.4
Figure. CDK Proteins in Action1,3 (Click to Enlarge)
Guardians of the Cell Cycle
CDK proteins derive their importance as an anticancer target from the vital role they play in the cell cycle. The cell cycle is divided into 5 phases: growth (G1 and G2 phases), DNA replication (S), mitosis (M), and quiescence (G0). Mitogenic stimulation triggers quiescent cells to enter the cell cycle, and progression from one phase to the next is tightly controlled by numerous proteins to ensure each step occurs at the appropriate time (Figure).1,3
Among these proteins are the CDKs, a family of 20 serine/threonine kinases2 that are fully functional only when bound to a cyclin.1-3,5
CDK activity is also regulated by other activating and inhibitory proteins, including CDK-activating kinases (CAKs) and the INK4 and CIP/KIP families of cyclin-dependent kinase inhibitors (CKIs), respectively.5
Mitogens trigger intracellular signaling cascades, most notably involving the MAPK pathway, that ultimately induce the transcription of D-type cyclins, which form active complexes with CDK4 and CDK6. These CDK4/6-cyclin D complexes drive phosphorylation of the retinoblastoma (RB) protein, leading to the transcription of genes encoding proteins that promote the transition from G1 to S phase, such as cyclin E. Progression to S phase represents a particularly important cell cycle checkpoint—the restriction point—whereupon the cell fully commits to entering the cell cycle.1,3,5
As the cell cycle progresses, the CDK2- cyclin E complex regulates DNA synthesis during S phase. Then, cyclin E levels decrease concurrently with an increase in the levels of cyclin A, which becomes the main cyclin bound to CDK2 to terminate S phase. Subsequently, CDK1 binds to cyclin A to trigger progression into G2 phase and to cyclin B to facilitate the expression of genes involved in mitosis. Cyclin B is degraded at the end of mitosis, switching off CDK1 activity. As RB is dephosphorylated, the cell enters G0 once again.1,3
Deregulation of the cell cycle is a key hallmark of cancer. Mutations in the genes encoding CDKs are rare in cancer, but gene amplification and protein overexpression of CDKs or their cyclin partners are common. Most often, aberrant activation of CDKs occurs indirectly, via alterations in the proteins that regulate them.1,5-7
Amplification of the genes encoding cyclins is particularly common across cancer types and, in fact, is one of the most common alterations in cancer in general. CCND1, which encodes cyclin D1, is amplified at rates of 15% to 40% in many common cancers, such as breast and lung cancers.5,8 Alterations in other CDK regulatory proteins are also commonly observed across the spectrum of cancers.9,10
As a result of their key role in oncogenesis and their status as readily druggable kinases, CDKs have long been recognized as prime targets for therapeutic intervention. The first CDK inhibitor to enter clinical trials was alvocidib (also called flavopiridol). Despite strong preclinical efficacy, alvocidib and other multitargeted CDK inhibitors demonstrated disappointing single-agent activity and were often hindered by low therapeutic indexes and significant toxicity.1,3
The field underwent a renaissance with the development of highly selective inhibitors of CDK4/6; the most significant leap forward came in 2009, when the front-runner, palbociclib (Ibrance), was shown to have activity in estrogen receptor (ER)—positive breast cancer cell lines. Combined inhibition of CDK4/6 and ER signaling was particularly effective, likely because of documented interplay between the cyclin D-CDK4/6 pathway and ER activation.11
Palbociclib and 2 other CDK4/6 inhibitors, abemaciclib (Verzenio) and ribociclib (Kisqali), are now approved by the FDA in combination with standard-of-care hormone therapies to treat hormone receptor—positive, HER2-negative, advanced or metastatic breast cancer.1,3
Alvocidib is an inhibitor of CDKs 1, 2, 4, and 9,1 and its antitumor effects were initially attributed to its inhibition of cell cycle CDKs. However, it most potently inhibits CDK9, which is not directly involved in regulation of the cell cycle, suggesting that its mechanism of action is related to another important function of CDKs―the regulation of transcription.12
RNA polymerase II is a multiprotein complex that plays a central role in the transcription of all protein-coding genes. The largest subunit of this complex is the RPB1 protein, which contains a carboxy-terminal domain (CTD) that is recognized by many kinases, phosphatases, and other posttranslational modifiers. CDKs, particularly CDK7 and CDK9, are among the kinases that phosphorylate the CTD at distinct serine residues.3,13
Transcriptional CDKs differ from their cell cycle counterparts. They typically form part of larger multiprotein complexes; CDK7 and cyclin H make up part of the 10-subunit general transcription factor (TFIIH) complex, whereas CDK9 and cyclin T form the catalytic subunit of the positive transcription elongation factor b (P-TEFb) complex. Another important mediator of transcription is CDK8 and its paralogue CDK19, which, along with cyclin C, are reversibly associated with the Mediator complex, a global regulator of RNA polymerase II activity.3,13
Historically, CDK7 (through the TFIIH complex) was thought to primarily control the initiation of transcription. CDK9 and the P-TEFb complex, on the other hand, were believed to function in transcriptional elongation through phosphorylation of serine 2. CDK8 was thought to repress transcription by preventing the Mediator complex from binding to RNA polymerase II.3,13
CDK involvement in transcription is likely much more complex than that, however; evidence has emerged for significant cross talk between transcriptional CDKs, context-dependent roles for each complex, and a potentially transcription-promoting role for CDK8. Furthermore, the identification of 2 additional transcriptional CDKs, CDK12 and CDK13 has further complicated matters.13
Notably, CDK7 also serves as a CAK and indirectly regulates the cell cycle by affecting the activity of cell cycle CDKs, which are its major targets, in addition to CDK9.14
Targeting Transcriptional CDKs
Normal gene transcription is often regulated by distal noncoding regulatory elements termed enhancers. Super-enhancers (SEs) are large clusters of enhancers that are densely occupied by master transcription factors, which drive transcription of genes involved in cell identity.
SEs have also been found to be essential drivers of oncogenesis by facilitating high expression of oncogenes such as MYC. Cancer cells can become transcriptionally “addicted,” or reliant on altered, SE-driven transcriptional programs that developed during tumorigenesis.15,16-18
This has opened new doors for the therapeutic targeting of transcriptional CDKs. Both CDK9 and CDK7 have been implicated in global transcriptional control by promoting the expression of genes regulated by SEs, including MYC.12,19 Thus, the requirement of some SEs for these CDKs has emerged as an important Achilles’ heel in cancer cells that could be targeted by CDK7 or CDK9 inhibition.20 This has also helped assuage concerns that transcriptional CDKs may make poor targets due to insufficient selectivity for cancer cells.
Few of the first-generation, pan-CDK inhibitors are still in development because of their poor showing in clinical trials. Considering the evolving understanding of its transcription activity, alvocidib has been restyled as a “CDK9 inhibitor.” It is being evaluated in early-phase trials that seek to exploit its association with SE-driven transcription factors, which are highly expressed in some hematologic malignancies.
Notably, these transcription factors include the MCL1 protein, a negative regulator of apoptosis that is overexpressed in about half of cases of relapsed/refractory (R/R) acute myeloid leukemia (AML). Overexpressed MCL1 inhibits apoptosis and sustains the survival of leukemic blast cells, which drives relapse and resistance to therapy.21 Preclinical studies revealed increased sensitivity to alvocidib in patients whose tumors display MCL1 dependence.22
Table. Novel CDK Inhibitors in Clinical Development (Click to Enlarge)
Results of the ongoing phase I Zella 101 study (NCT03298984) of alvocidib followed by 7 + 3 induction chemotherapy in patients with newly diagnosed AML showed that alvocidib was well tolerated. Among 22 patients, the most common grade ≥3 nonhematologic TEAEs were diarrhea, hypophosphatemia, hypokalemia, and TLS. Among 18 patients evaluable for response, 14 achieved CR or CRi.23
The clinical development of alvocidib is somewhat limited by its IV route of administration. For this reason, an oral prodrug of alvocidib, TP-1287, is also in development; a phase I study (NCT03604783) in patients with advanced solid tumors began in January 2019.24,25
CYC065 is another multitargeted CDK inhibitor that includes CDK9 among its targets, and ongoing phase I clinical trials in AML or myelodysplastic syndrome (MDS) and chronic lymphocytic leukemia are focusing on a combinatorial strategy with the BCL-2 inhibitor venetoclax (Venclexta). Upregulation of MCL1 has been shown to drive resistance to inhibitors of antiapoptotic proteins such as venetoclax, which could be overcome by indirectly targeting MCL1 through CDK9 inhibition.26
Preliminary data show that the combination of CYC065 and venetoclax was well tolerated, with no dose-limiting toxicities (DLTs) and no TLS. In the CYC065-03 study (NCT04017546), 3 of 9 patients with AML or MDS treated at doses ranging from 64 to 150 mg/m2 achieved a reduction in leukemia blast cells in their peripheral blood. In the CYC065-02 study (NCT03739554), 2 patients who had failed the Bruton tyrosine kinase inhibitor ibrutinib (Imbruvica), 1 of whom had also failed chimeric antigen receptor T-cell therapy, achieved shrinkage of enlarged lymph nodes by computed tomography scan at a dose of 64 mg/m2 every 2 weeks; the latter patient was also negative for minimal residual disease.27
Zotiraciclib is the most potent CDK9 inhibitor developed to date, but it also strongly inhibits CDKs 1, 2, 3, and 5. The FDA recently granted the agent orphan drug designation for patients with glioma,28 a tumor type that often exhibits MYC messenger RNA expression.29 The designation was awarded based on an ongoing trial evaluating zotiraciclib in combination with temozolomide.4,30
In intertim phase I results, 40 patients were treated with zotiraciclib at a starting dose of 200 mg orally combined with temozolomide on either a dose-dense (125 mg/m2/day, 7 days on, 7 days off) or metronomic (50 mg/m2/day) dosing schedule. In all, 38 patients were evaluable, 18 in the dose-dense arm and 20 in the metronomic arm.
In the dose-dense arm, DLTs included grade 3 diarrhea (200-mg dose) and grade 4 neutropenia, grade 3 elevated alanine aminotransferase (ALT), and grade 3 fatigue (250-mg dose). In the metronomic arm, DLTs included recurrent grade 3 neutropenia (200-mg dose); grade 3 elevated ALT, grade 3 fatigue, and grade 4 neutropenia (250-mg dose); and grade 4 elevated ALT, grade 4 elevated AST, and grade 4 febrile neutropenia (300-mg dose). The investigators reported that 4 patients completed 12 cycles of therapy with “prolonged disease control” but did not provide further details.4
Pharmaceutical companies are also eyeing more selective inhibitors of CDK9 and CDK7. AZD4573 is a highly selective IV CDK9 inhibitor with a half-maximal inhibitory concentration (IC50) of 14 nM and over 10-fold selectivity for CDK9 compared with 13 of the other 14 kinases tested. AZD4573 is being evaluated in a phase I clinical trial (NCT03263637) in R/R hematologic malignancies after preclinical studies demonstrated its ability to induce apoptosis in cancer cells across tumor cell lines and hematologic cancer models.30
SY-5609 is an oral, noncovalent inhibitor of CDK7 with an IC50 of 60 nM. It is 49,000-, 16,0000-, and 13,000-fold less selective for CDK2, 9, and 12, respectively.31 Syros Pharmaceuticals was originally developing an IV CDK inhibitor, SY-1365, but halted in 2019 to focus on SY-5609, which was more potent and selective and demonstrated greater antitumor activity in preclinical studies. The first patient in a phase I study in advanced solid tumors with RB pathway alterations was dosed in January 2020.32
CT7001 is another orally bioavailable CDK7 inhibitor, with an IC50 of 40 nM and high selectivity over other kinases tested. Formerly known as ICE0942, CT7001 has demonstrated antitumor activity in preclinical models across a range of tumor types, including AML, small cell lung cancer, and hormone-sensitive triple-negative breast cancer. A phase I clinical trial was initiated at the end of 2017.33,34
Finally, SEL120 is a first-in-class selective inhibitor of CDK8. In preclinical AML xenograft models, it induced CR and demonstrated synergy with apoptosis-inducing drugs and chemotherapy.35 A phase I trial of SEL120 began in September 2019 (NCT4021368).36
In the ongoing phase II Zella 201 study (NCT02520011), alvocidib is administered before cytarabine and mitoxantrone in patients with MCL1-dependent R/R AML. Among 25 patients enrolled in stage 1, 21 were evaluable for response; the overall rate of complete remission (CR) or CR with incomplete hematologic recovery (CRi) was 62%. The most common grade 3 or greater nonhematologic treatment-emergent adverse events (TEAEs) were tumor lysis syndrome (TLS), diarrhea, increased aspartate aminotransferase (AST), sepsis, and peripheral edema.22
Efforts to recapitulate the success of CDK4/6 inhibitors are also driving the pursuit of more selective inhibitors. Promising preclinical data propelled several drugs into early stages of clinical testing amid a potential new era of CDK-targeted therapy (Table).