A CDK4-selective inhibitor puts the brakes on cancer cells

Abstract

In this issue of Cancer cell, Palmer et al. describe the discovery and preclinical testing of the first-in-class CDK4-selective inhibitor atirmociclib. By sparing CDK6, atirmociclib has the potential to ameliorate dose-limiting hematological toxicities that limit drug exposure and treatment continuity, and, by extension, the antitumor efficacy of dual CDK4/6 inhibitors.

A major principle of precision oncology is the development of molecular target-selective inhibitors that allow for a higher therapeutic index and full inhibition of the drug target without dose-limiting toxicities, thereby inducing deeper and more prolonged clinical responses¹. Such agents would be predicted to have better safety profiles and thus enable safer therapeutic combinations. For example, the third-generation EGFR mutant-selective inhibitor osimertinib improves the survival of patients with EGFR-mutant non-small cell lung cancer compared to earlier generation inhibitors, with fewer adverse events². Many other selective inhibitors in oncology have followed suit, including PARP1, KRAS G12C, and mutant-selective PI3Kα inhibitors. In this issue of Cancer Cell, Palmer et al. apply this principle to the development of cyclin-dependent kinase (CDK) inhibitors³.

Uncontrolled proliferation of cancer cells is frequently driven by D-type cyclins in complex with CDK4 and CDK6. These protein complexes phosphorylate and inactivate the tumor suppressor RB, releasing the E2F transcription factors and leading to E2F-dependent gene expression and cell cycle progression from the G1 to S phase (Fig. 1). The combination of anti-estrogens with CDK4/6 inhibitors (CDK4/6i) synergistically blocks the RB-E2F transcriptional axis and improves survival of patients with ER+ breast cancer^4,5. However, the three FDA-approved CDK4/6i—palbociclib, ribociclib, and abemaciclib—induce hematologic toxicities such as neutropenia, leading to dose reductions and treatment interruptions that, in turn, limit the concentration of drug that reaches the tumor. ER+ breast and prostate cancers are highly dependent on CDK4 but usually express low levels of CDK6. In contrast, CDK6 is essential for the differentiation and maintenance of hematopoietic cells, while CDK4 is dispensable⁶.

Fig. 1: A CDK4-selective inhibitor blocks cell cycle progression in ER+ breast cancer and prostate cancer cells, but not in hematopoietic cells.

Cancer cells with high CDK4/Cyclin D1 expression and low CDK6 expression, such as ER+ breast cancers and prostate cancers, are dependent on CDK4 but not CDK6; in contrast, hematopoietic cells are more dependent on CDK6. In tumor cells, CDK4 partners with Cyclin D1 to phosphorylate RB, releasing the E2F transcription factor to enable transcription of genes that drive the transition of the cell cycle from the G1 to S phase. In hematopoietic cells, CDK6 complexed with Cyclin D3 also phosphorylates RB. A CDK4-selective inhibitor, such as atirmociclib, has the potential to ameliorate the hematological dose-limiting toxicities associated with dual CDK4/6i, allowing more complete blockade of CDK4 activity and RB phosphorylation in tumors, potentially leading to deeper and more prolonged responses. CDK2i: CDK2 inhibitor; ADCs: antibody-drug conjugates; ICIs: immune checkpoint inhibitors. Created in BioRender. Hanker, A. (2025) https://BioRender.com/b29t202

Palmer et al. hypothesized that selective targeting of CDK4, while sparing CDK6, can reduce dose-limiting hematological toxicity while maximizing CDK4 target coverage in tumors. This would be expected to more completely block the G1-S cell cycle transition, leading to more complete and sustained inhibition of tumor growth. To this end, the authors sought to identify a CDK4-selective small molecule inhibitor.

Using structure-based design and screening, atirmociclib was identified. The authors performed X-ray crystallography to decipher the co-crystal structures of atirmociclib bound to cyclin D1-CDK4 and -CDK6. They found that the glycine-rich loop (G-loop) is more flexible in CDK4 than in CDK6, thus enabling it to better accommodate the bulky ispoproyl alcohol moiety of atirmociclib. This discrepancy in the G-loop between CDK4 and CDK6 may explain why atirmociclib favors binding to CDK4 over CDK6, despite a 92% consensus sequence in the ATP-binding sites of the two kinases.

To confirm the selectivity of atirmociclib for CDK4, the authors performed a series of proteomic and biochemical analyses. Mass spectrometry-based target engagement profiling revealed that CDK4 was the only kinase associated with atirmociclib. Furthermore, compared to the three dual CDK4/6i, atirmociclib exhibited up to 21-fold higher selectivity for cyclin D1-CDK4 vs. cyclin D3-CDK6 in vitro. As predicted, the CDK4 selectivity of atirmociclib translated to less hematologic toxicity in cell line and animal models. Atirmociclib was 20- and 30-fold less effective than palbociclib at blocking RB phosphorylation and proliferation, respectively, in human primary hematopoietic stem cells. Further, palbociclib treatment resulted in a greater reduction of neutrophil counts in beagle dogs compared to atirmocliclib. Consistently, pharmacokinetic/pharmacodynamic mathematical modeling based on CDK4:CDK6 selectivity and patient data collected from the PALOMA-1/TRIO-18 trial projected that atirmociclib would be expected to cause much less neutropenia than the other dual CDK4/6i in patients.

The superior CDK4 selectivity of atirmociclib should be more tolerable than the dual CDK4/6i, enabling dose intensification to maximize CDK4 inhibition in tumors. Indeed, the authors observed comparable neutrophil suppression in vivo, even when the plasma concentration of atirmociclib was 30-fold higher than that of palbociclib. Moreover, clinically relevant doses of atirmociclib resulted in a significantly greater reduction ER+ breast cancer spheroid formation and growth of ER+ and prostate cancer cell line- and patient-derived xenografts compared to palbociclib. Ribociclib and abemaciclib displayed a selectivity of CDK4:CDK6 in between that of palbociclib and atirmociclib, which translated to superior efficacy vs. palbociclib but inferior efficacy vs. atirmociclib in breast cancer spheroid assays. This raises the intriguing possibility that the enhanced selectivity of CDK4:CDK6 of ribociclib and abemaciclib vs. palbociclib may, in part, underly the apparent clinical superiority of the two former agents on overall survival relative to palbociclib⁵.

Acquired drug resistance is a major cause of cancer treatment failure. The authors generated atirmociclib-resistant ER+ breast cancer xenografts and identified upregulated expression of cyclin E1 (which partners with CDK2 to phosphorylate RB) and FGFR1, as well as upregulation of gene signatures of estrogen response and the TNFα-NFkB in resistant tumors. Further, they found compensatory upregulation of CDK6 following atirmociclib treatment in a subset of ER+ breast cancer cell lines and demonstrated that knockdown of CDK6 enhanced sensitivity to atirmoclib, while overexpression promoted resistance. These findings are in line with previous studies⁵, suggesting that bypassing the essentiality of CDK4 and reinforcing cyclin D1/CDK4/6 activity are common resistance mechanisms shared between dual CDK4/6i and atirmociclib. These findings provide insights into potential therapeutic combinations of atirmociclib with agents to counter these resistance mechanisms. To prove this concept, the authors tested the antitumor effects of atirmociclib-based therapeutic combinations with a CDK2-selective inhibitor (tagtociclib) or an ER degrader (vepdegestrant). These combinations exhibited superior inhibitory effects on spheroid formation and/or tumor growth compared to atirmociclib alone, including in models that were resistant to palbociclib. Further, the authors noted that atirmociclib treatment enhanced cell surface expression of HER2; the combination of atirmociclib with a HER2 antibody-drug conjugate more effective than either agent alone at reducing the growth of breast cancer spheroids.

Dual CDK4/6i are also known to augment antitumor immunity⁷. The authors found that atirmociclib treatment more strongly induced upregulation of PD-L1 as well as genes involved in antigen presentation and pro-inflammatory cytokine signaling than palbociclib at clinically relevant doses. Furthermore, co-treatment of mouse syngeneic tumor models with atirmociclib and an immune checkpoint inhibitor synergistically suppressed tumor growth, suggesting a potential therapeutic combination to exert antitumor immunity.

Based on these compelling preclinical data, several clinical trials of atirmociclib are underway. Initial results from the first-in-human studies have been reported, showing evidence of clinical efficacy in advanced ER+ breast cancer patients who progressed on dual CDK4/6i, with lower rates of neutropenia compared to other CDK4/6i—although low grade neutropenia remained a commonly reported adverse event^8,9. Ongoing clinical trials are testing atirmociclib + ER inhibitors in patients with advanced ER+ breast cancers that have progressed on prior CDK4/6i. The combination with the ERα degrader vepdegestrant (NCT06206837) is particularly attractive given that many patients in this setting have cancers harboring ESR1 mutations, which are resistant to fulvestrant but benefit from ER degraders¹⁰. Atirmociclib is also being tested in combination with a CDK2-selective inhibitor in advanced tumors (NCT05262400). Other potentially attractive combinations with atirmociclib include an immune checkpoint inhibitor or a PI3Kα or AKT selective inhibitor, which have shown synergy with CDK4/6i.

Perhaps the most direct confirmation of whether a more CDK4-selective inhibitor can improve antitumor efficacy with fewer hematological toxicities will come from the Phase 3 trial NCT06760637, a head-to-head comparison of atirmociclib + letrozole (aromatase inhibitor) compared to dual CDK4/6i + letrozole in first-line advanced ER+ breast cancer. This trial will also shed light on whether CDK6 upregulation is a common mechanism of escape from CDK4 selective inhibition that could limit long-term efficacy in the clinic. If results of these trials align with the preclinical data presented by Palmer et al., atirmociclib—or future CDK4-selective inhibitors—could transform the care of patients with advanced ER+ breast cancer.