On the Rise of Cyclin-Dependent Kinase Inhibitors in Breast Cancer: Progress & Ongoing Challenges

Given marked improvements in progression free and overall survival in the metastatic setting, along with a favorable toxicity profile, the cyclin-dependent kinase 4/6 inhibitors (CDK4/6i’s) have become standard of care in frontline therapy for hormone receptor positive metastatic breast cancer (HR+ MBC). Despite initial durable responses, however, virtually all patients with MBC will eventually progress; as such, rigorous studies into the mechanisms of resistance are ongoing, as is an effort to develop rational sequencing strategies and identify subgroups that will retain benefit upon progression. Herein we briefly discuss the grounds for initial experimentation on the CDK pathway and the early successes to which it gave rise. We then focus on what we know about key resistance mechanisms and discuss correlative biomarker studies as a vital strategy for advancing the field of targeted therapy in MBC.

Preclinical Rationale and Early Phase Development

Proliferation, a consequence of dysregulation of the cell cycle, has long been a therapeutic target in cancer. In the era of targeted therapy, however, attention focused on selectively targeting key cell cycle proteins. CDK4 and CDK6, activated by cyclin D1 (CCND1), are key regulators of cell cycle progression from G1 to S phase (point of DNA replication) via phosphorylation of the Retinoblastoma protein (Rb), a tumor suppressor that protects the cell from entering S-phase prematurely. Following studies showing that cyclin D1 is overexpressed in up to 45% and amplified in up to 15% of breast cancers[1, 2], Finn et al demonstrated that preclinically, PD 0332991 (Palbociclib) led to G1 arrest in luminal ER+ and HER2-enriched breast cancer cell lines. Moreover, PD 0332991 displayed synergy with tamoxifen and trastuzumab, respectively, in halting cellular proliferation[3]. Concurrently, dose, schedule and single agent activity of palbociclib in heavily pre-treated HR+ MBC patients was established[4, 5]. In parallel to palbociclib, two other CDK4/6i’s were developed (Ribociclib, Abemaciclib), and all three were advanced to phase 2 and then confirmatory phase 3 randomized trials of CDK4/6i’s in combination with endocrine therapy for HR+ MBC.

Early Successes

As a class, the three CDK4/6i’s demonstrated remarkably similar and impressive benefits in progression free survival in postmenopausal patients with HR+ MBC, nearly doubling the time to progression in the frontline setting when combined with aromatase inhibitors (AI) or, for those who had progressed on an AI, with Fulvestrant. This effect was replicated in MONALEESA-7, testing Ribociclib in a purely pre-menopausal population. Updated analyses of Ribocilclib and Abemaciclib also showed improvements in overall survival[6, 7]. These data, along with a more comprehensive overview of the major CDK4/6 clinical trials in metastatic breast cancer, are reviewed elsewhere[8]. Importantly, correlative biomarker analyses of these phase II/III clinical trials failed to demonstrate any markers predictive of response[4, 9] with the exception of cyclin E1 mRNA levels, where higher levels were associated with shorter PFS[10], thus supporting the notion that most patients stand to benefit from a CDK4/6 inhibitor as part of a frontline strategy in HR+ MBC.

Understanding Mechanisms of Resistance at Progression

Despite early success, however, virtually all patients receiving a CDK4/6i will progress. Intriguingly, unlike many standard chemotherapies whereby progression signals an end to their utility, CDK4/6i’s maintain a complex and dynamic relationship with the cell cycle and, consequently, may still retain value in the context of progressive disease. As such, there has been a significant effort in uncovering clinically relevant mechanisms of resistance that may help inform who may continue to derive benefit. By studying CDK4/6i-resistant cell lines, investigators have discovered several activated pathways upon which intervention may be realistic.

For instance, Herrera-Abreu and colleagues described BC cells’ early adaptation to palbociclib in vitro[11]. Seventy-two hours of palbociclib exposure to CDK4/6i-naïve BC cells led to a rise in the levels of CCND1, which then complexed with CDK2 in a non-canonical fashion, bypassing CDK4/6 entirely. They also found that the PI3K/AKT pathway sustained expression of CDK2 and entry from G1 to S phase. Treatment with Ribociclib, Fulvestrant and a PI3Kalpha inhibitor (BYL719) in a patient-derived xenograft (PDX) mouse model induced significantly greater tumor regressions than doublet Ribociclib and BYL719. Further investigation into late acquired resistance identified Rb loss and CDK2 gain as the predominant mechanisms. In a PDX mouse model, adding a PIK3CA inhibitor to palbociclib was able to prevent acquired resistance. Additional work has highlighted the role of CDK6 amplification through suppression of FAT1 and the Hippo signaling pathway, and CDK7 (an activator of CDK2 and CDK9), as major mechanisms of resistance; importantly, targeting CDK6 and CDK7 has shown activity in CDK4/6-resistance cell lines [12–14]. Finally, withdrawing CDK4/6i (i.e. a drug holiday) for 28 days led to re-sensitization of resistant cell lines and tumor regression in mouse xenografts, with corresponding reduction in measured CDK6 and CCND1 levels, suggesting that re-exposure to CDK4/6i after initial resistance may be a valid strategy[15].

In addition to acquired CDK/Rb-axis alterations, several other signaling pathways have been identified that exhibit cross talk with the CDK pathway. For instance, FGFR1 was found to be overexpressed in a large kinome screen of a CDK4/6i-resistant ER+ BC cell line[16]. Notably, stimulation of FGFR1 led to a rise in CCND1 levels; conversely, suppression of CCND1 in an FGFR1 over-expressed cell line sensitized the cells to CDK4/6 inhibition, highlighting FGFR1 as a potentially actionable target. Most strikingly, analysis of 32 post-CDK4/6i progression tissue samples showed that 41% of patients harbored some form of FGFR1 alteration, suggesting this may be an important bypass mechanism at time of progression.

Other studies have implicated the PI3K/AKT/mTOR pathway as a tumor escape mechanism. Phosphoinositide-dependent kinase-1 (PDK1), a kinase required for full AKT activation, is upregulated upon exposure to CDK4/6i, and PDK1-knockout demonstrated synergy with CDK4/6 inhibition in resistant models[17]. Separately, PTEN (a negative regulator of AKT) loss was identified in a series of patients who progressed on Ribociclib, and downstream analysis revealed activation of AKT and rise in CDK2 levels through inhibition of p27, a negative regulator of CDK2. In support of this mechanism, inhibition of AKT in a PTEN-knockout mouse xenograft restored sensitivity to Ribociclib and led to tumor regression[18].

Finally, studies aimed at describing the genomic landscape in CDK4/6 resistant models have identified several candidate genes and pathways, including Aurora Kinase A, RAS, cyclin E2, and the JAK/STAT pathway [19, 20]. Importantly, several of those identified have corresponding therapeutic inhibitors, with promising preclinical results.

Clinical Trials for Biomarker Discovery and Testing of New Approaches after Progression

Harnessing these insights from the lab, several studies seek to demonstrate the clinical validity and utility of the resistance mechanisms discussed above. For example, in SIDEOUT-3, ER+ MBC patients who progress within 12 months of initiation of CDK4/6i are offered a tissue biopsy for multi-omic investigation of actionable resistance alterations, and this data is provided to the treating physicians along with possible treatment options[21]. On follow up, clinical endpoints are then compared among those subsequently treated with resistance-specific therapy and with traditional next-line therapy. In an effort to minimize the need for invasive tissue biopsies, investigators are also looking at circulating tumor cells and circulating tumor DNA as an attractive alternative with the additional benefit of representing a more comprehensive analysis of multi-site disease resistance. These liquid biopsies have proved capable of detecting acquired resistance mutations in patient’s that progressed on CDK4/6i’s [22, 23]. Finally, clinical trials informed by the biologic rationale established in preclinical studies are testing the continuation of CDK4/6i’s after progression (e.g. NCT03147287, NCT03238196, NCT02632045), and will provide invaluable genetic material for correlative biomarker investigation into acquired resistance mechanisms.

Conclusion

Several key themes emerge as we look ahead to the future of targeted therapy against the cell cycle in breast cancer. Ongoing preclinical studies in cell lines and mouse models are generating hypotheses for novel treatment approaches, which are undergoing formal testing in clinical trials. Incorporated into these next generation trials are pre-planned correlative biomarker studies, where the aim is to uncover mechanisms for treatment failure and thus advise future treatment approaches. Based on the encouraging results from the lab thus far, one can envision a future whereby patients progressing on a CDK4/6i will undergo custom panel testing for all the known major resistance pathways. These results will facilitate a tailored treatment strategy to re-sensitize cells to CDK4/6 inhibition or to target a different CDK altogether. Furthermore, combinatorial frontline therapy for MBC may evolve to target interdependent cell cycle pathways, such as PI3K/AKT and CDK4/6, as a means of preventing the onset of resistance, thereby extending the duration of benefit of these therapies.

It has been four years since Palbociclib received accelerated approval for HR+ MBC. Many of our patients who initially enjoyed stellar responses are now returning to clinic with disease progression. Hence, it is more important than ever to employ a ‘bench to bedside’ paradigm of utilizing discoveries in the lab to test novel treatment strategies in patients, while also building in thoughtful correlative studies to further advance our understanding of resistance mechanisms. This iterative process will continue to expand our knowledge of disease as well as therapeutic options for patients.