Geoffrey I. Shapiro, MD, PhD, Director of the Early Drug Development Center at the Dana-Farber Cancer Institute, Boston, explained the current research initiatives involving cyclin D–dependent kinase (CDK) 4/6 inhibitors.
Mechanism of Action
How do CDK4/6 inhibitors work at the cellular level in cancer?
CDK4 and CDK6 are cyclin-dependent kinases that control the transition between the G1 and S phases of the cell cycle. The S phase is the period during which the cell synthesizes new DNA and prepares itself to divide during the process of mitosis. CDK4/6 activity is typically deregulated and overactive in cancer cells. There can be amplification or overexpression of the genes encoding cyclins or of the genes encoding the CDKs themselves. Additionally, loss of endogenous INK4 inhibitors, by gene deletion, mutation, or promoter hypermethylation, can also lead to overactivity of CDK4 and CDK6.
A major target of CDK4 and CDK6 during cell-cycle progression is the retinoblastoma protein (Rb). When Rb is phosphorylated, its growth-suppressive properties are inactivated. Selective CDK4/6 inhibitors “turn off” these kinases and dephosphorylate Rb, resulting in a block of cell-cycle progression in mid-G1. This causes cell-cycle arrest and prevents the proliferation of cancer cells. Cancer cells must express wild-type Rb for a selective CDK4/6 inhibitor to be effective.
Although the initial response to a selective CDK4/6 inhibitor is typically cell-cycle arrest, in some cases arrested cells enter a state of senescence. Understanding the determinants of whether a cell undergoes reversible G1 arrest or enters a senescent state is an important research area. Cancer cells entering senescence may undergo gradual regression over time; it is in such cancers that CDK4/6 inhibitors may produce the greatest clinical benefit.
Additionally, there are some cancer cell types in which the biologic outcome is abrupt cell death, as in the case with subsets of hematopoietic malignant cells, such as T-cell leukemia cells.
In Breast Cancer
How does this mechanism of action make CDK4/6 inhibitors such effective drugs in breast cancer, where they are currently approved?
Estrogen receptor–positive breast cancer seems to be especially dependent upon CDK4 for proliferation. The tight linkage of the estrogen receptor signaling pathway to cyclin D1-CDK4 may be one reason why breast cancer cells are highly sensitive to the antiproliferative effects of drugs, particularly when they are combined with endocrine therapy. With the combination of a CDK4/6 inhibitor and antiestrogen therapy, there is enhancement of G1 and cell-cycle arrest and probably enhanced entry into a senescent state. To date, estrogen receptor–positive breast cancer is the malignancy for which this class of drugs has proven most effective and for which we have the most mature data from randomized trials comparing endocrine therapy alone with endocrine therapy combined with CDK4/6 inhibition.
In Other Tumor Types
What other tumor types appear to be good targets for CDK4/6 inhibitors?
Based on research to date, several other tumor types are of great interest, including non–small cell lung cancer (NSCLC), especially the KRAS-mutant subset. In preclinical mouse models, KRAS-driven lung cancer is highly dependent on CDK4, and genetic or pharmacologic ablation of CDK4 activity has effects on both the establishment and maintenance of these tumors.
Selective CDK4/6 inhibitors have also demonstrated activity in mantle cell lymphoma, a malignancy defined by a translocation involving CCND1 resulting in cyclin D1 overexpression, as well as liposarcoma, in which CDK4 is often amplified. The drugs are being looked at preclinically and clinically in a variety of other tumor types as well, including melanoma, glioblastoma, pancreatic cancer, and colorectal cancer.
Early Trial Data in NSCLC
What does early trial data show for these drugs in NSCLC?
We have encouraging results from a phase I study of abemaciclib, recently published in Cancer Discovery.1 In the expansion phase of this trial, we evaluated abemaciclib as a single agent in 68 patients with NSCLC, of whom 29 had tumors harboring KRAS mutation. We saw a clinical benefit as defined by partial response or stable disease (≥ 24 weeks) among both patients with KRAS wild-type tumors and those with KRAS-mutant tumors, but more impressively in the KRAS-mutant subset, where the clinical benefit rate was 55%, vs 39% in the wild-type group. Among 29 patients with KRAS-mutant tumors, 9 had progression-free survival exceeding 24 weeks.
Like all kinase inhibitors, CDK4/6 inhibitors will not be curative, and cells will engineer a way around them. Defining the mechanisms of resistance will be critical for designing future strategies.— Geoffrey I. Shapiro, MD, PhD
Tweet this quote
KRAS-mutated lung cancer may be subdivided further into groups including KRAS alone, KRAS with concomitant TP53 mutation, and KRAS with concomitant LKB1 (STK11) loss. Concomitant mutation of KRAS and loss of LKB1 predict an especially aggressive phenotype. Tumor regressions were noted across all these subtypes of KRAS-mutant lung cancer. Although more work needs to be done to identify the lung cancer population most likely to benefit from abemaciclib monotherapy, the data in KRAS-mutant disease are promising.
We hope to glean additional information from the phase III JUNIPER trial. It is comparing abemaciclib and erlotinib in 550 patients with stage IV NSCLC whose tumors have a KRAS mutation and whose disease progresses after platinum-based chemotherapy and one additional therapy. Erlotinib was chosen as the control arm as a therapy often used for second- or third-line therapy in advanced NSCLC. The study was designed when erlotinib was approved by the U.S. Food and Drug Administration for all lung adenocarcinomas; however, it has since become clear that erlotinib works only in tumors with epidermal growth factor receptor (EGFR) mutations. The label for the drug is now restricted to EGFR-mutant NSCLC. Therefore, the use of erlotinib as a comparator in a trial for patients with tumors harboring KRAS mutations may not be meaningful. Nonetheless, the outcome of patients enrolled to the abemaciclib arm will still be of interest and will help define the monotherapy response rate and progression-free survival (see page 9).
Study Data in Other Solid Tumors
What did your phase I study in Cancer Discovery reveal for patients with other solid tumors?
Besides the results in NSCLC, clinical benefit was seen in several other tumor types. Among 36 patients with estrogen receptor–positive breast cancer, 11 had a partial response, and 18 achieved stable disease. Among the 11 responders, 7 received abemaciclib without continuing prior endocrine therapy, so these were true monotherapy responses. The median duration of response was 13.4 months, and median progression-free survival was 8.8 months.
In addition, among 26 patients with melanoma, 1 had a partial response and 6 had stable disease. Finally, among 17 patients with glioblastoma, 3 achieved stable disease, 2 of long term. Interestingly, the results among patients with glioblastoma were consistent with the ability of abemaciclib to cross the blood-brain barrier, as demonstrated in preclinical models and by levels of drug in the cerebrospinal fluid of patients that approximated those observed in plasma. (This study is described in more detail in an accompanying article in this supplement. See page 10.)
How useful will CDK4/6 inhibitors be as single agents?
Among patients with estrogen receptor–positive breast cancer, single-agent activity has been seen with abemaciclib, more so than with palbociclib or ribociclib. At the 2017 American Association of Cancer Research (AACR) Annual Meeting, Dr. Hope Rugo updated the results of the phase II MONARCH-1 study in 132 heavily pretreated patients with estrogen receptor–positive breast cancer.2 The response rate was 19.7%, which is higher than had been seen with palbociclib or ribociclib as monotherapy. Median progression-free survival was about 6 months, and overall survival was 22 months.
There are several possibilities to explain the single-agent activity of abemaciclib compared with the other agents in the class. Because abemaciclib produces less neutropenia, possibly because it hits CDK4 “harder” than CDK6 (CDK6 inhibition is more associated with neutropenia), this drug can be given continuously. It is possible that this administration schedule does a better job of driving cells into permanent G1 arrest and senescence, with possible regression over time. Additionally, although abemaciclib has been classified as a selective CDK4/6 inhibitor, it also hits other targets in biochemical assays, including CDK9. This promiscuity may contribute to its ability to produce responses as a single agent, although this remains a point of significant debate. In any case, there is potential value of this drug as a single agent; for example, in heavily pretreated patients, monotherapy may be preferable.
Combination Therapies in NSCLC
Could combinations of CDK4/6 inhibitors with targeted agents improve treatment benefit in NSCLC?
In NSCLC, CDK4/6 inhibitor monotherapy has produced more stable disease than responses. There may be better efficacy if we combine CDK4/6 inhibitors with other agents, especially those that target downstream components of RAS signaling pathways. In many instances, CDK4/6 inhibitors will be drugs that we will use in combination. We already do this with hormonal therapy in estrogen receptor–positive breast cancer, with excellent results.
At the 2017 AACR Annual Meeting, I presented a study in which we combined palbociclib and a MEK inhibitor (PD-0325901).3 The idea was to attack KRAS signaling pathways at multiple nodes and to use combinations that will inhibit downstream proteins. For KRAS-mutant NSCLC, such combinations will be the way forward.
We found the combination to be tolerable at doses of palbociclib that engaged the CDK4/6, demonstrated by a variety of assays. We also demonstrated promising progression-free survival among patients with KRAS-mutant tumors. Among 32 patients with NSCLC, 1 had a partial response, and 7 were progression-free for more than 6 months. Three patients remained on study for more than 1 year, and one patient has continued on study for more than 2 years. These outcomes were observed even among patients who had received prior immune checkpoint blockade and in those who had loss of various tumor suppressors, such as TP53, which can increase the aggressiveness of the tumor.
In mantle cell lymphoma, both palbociclib and abemaciclib have demonstrated single-agent activity. Perhaps more exciting are the data for CDK4/6 inhibition in combination with the BTK [Bruton’s tyrosine kinase] inhibitor ibrutinib. It has been shown that the G1 arrest afforded by CDK4/6 inhibition changes the gene-expression profile of cells in a permanent way, which may increase sensitivity to ibrutinib or other inhibitors of signal transduction. There is a lot of very interesting work going on investigating the mechanisms by which CDK4/6 inhibitors will synergize with other targeted agents.
Assessing Therapeutic Response
Is there any indication that a biomarker or some other means of gauging response to these drugs may emerge?
In the study presented at the 2017 AACR Annual Meeting of palbobiclib and the MEK inhibitor PD-0325901, we also evaluated novel assays of activity, including an assay of phosphorylation of Rb in skin biopsies and a noninvasive assay for cell-cycle arrest.4 The assay for cell-cycle arrest showed a reduction in thymidine kinase (TK) protein in the blood. The TK enzyme is produced during the S phase of the cell cycle and is secreted by tumor into the blood stream. Expression of TK occurs in actively proliferating cells and is downregulated after CDK4/6 inhibitor–mediated G1 arrest.
In the study, we observed a marked decrease in serum TK activity at day 21 of palbociclib treatment. This assay has also been incorporated into neoadjuvant studies in breast cancer. Ultimately, serial assessment of the TK assay may help predict response or disease progression to these agents.
Other experiments have shown that the RAS allelic fraction is decreased in plasma as a result of treatment with combined palbociclib and PD-0325901, suggesting that allelic burden of RAS in plasma could also potentially be a biomarker of treatment response.
Use in Combination With Immunotherapy
How about using CDK4/6 inhibitors in combination with immunotherapies?
The checkpoint inhibitors have come very far in the treatment of NSCLC. There are rapidly evolving data indicating that CDK4/6 inhibitors may afford favorable modulation of the immune microenvironment and thus may be good partners for checkpoint blockade. Lung adenocarcinoma is a good place to study these combinations, which are also being evaluated in trials of squamous cell lung and breast cancers. Such combinations may be schedule-dependent and may rely on discontinuous dosing to avoid reducing the proliferation of cytotoxic T cells.
With two approved agents in this class and another on the horizon, how will they be positioned?
All the CDK4/6 inhibitors are meeting critical endpoints in large phase II and III breast cancer trials, suggesting they are reasonable additions to the armamentarium. Right now, it’s difficult to see how they will all fall into place, and it is not completely clear whether one might work after another has failed.
A number of clinical trials are studying these agents in various sequences and in various tumor types. For example, Pfizer is studying patients who progress on palbociclib and letrozole, randomizing them to more palbociclib with a different endocrine agent, fulvestrant (Faslodex), vs fulvestrant alone. Additionally, because estrogen receptor–positive breast cancer frequently demonstrates PIK3CA mutation, PI3K inhibitors in combination with CDK4/6 inhibition plus hormonal therapy is currently under investigation, with trials underway or planned for palbociclib, ribociclib, and abemaciclib.
Further Areas of Study
What data are you most interested in seeing reported?
In breast cancer, these inhibitors are being evaluated in patients outside of the estrogen receptor–positive subtype, in particular HER2-positive patients. It will be very interesting to see whether they are active when added to anti-HER2 therapy, and whether among patients with HER2-amplified tumors, estrogen receptor–positive and estrogen receptor–negative populations both benefit. Additionally, the activity of these agents will also be defined in Rb wild-type triple-negative breast cancer. In lung cancer and perhaps other tumor types, data regarding their use in combination with MAP kinase inhibitors and also with immune checkpoint blockade will be critically important.
In hematologic malignancies, there is much work to be done. There is tremendous interest in the degree to which these agents might enhance the efficacy of ibrutinib in mantle cell lymphoma. An untouched area of interest as well are the T-cell leukemias, where these drugs have elicited apoptotic responses in preclinical models.
What are the main challenges as the field evolves?
A key challenge is defining biomarkers to identify which patients may benefit most. Right now, CCND1 amplification and CDKN2A/B loss have not sorted out as biomarkers useful for this purpose.
As we discussed previously, a burgeoning field of research is the effort to understand more about those patients whose tumors undergo transient G1 arrest, vs those whose tumors go into a deep senescent state. Understanding these differences may help identify patients most likely to benefit from these treatments.
Lastly, a research area of great importance is the study of mechanisms of acquired resistance to these drugs. Like all kinase inhibitors, CDK4/6 inhibitors will not be curative, and cells will engineer a way around them. Defining the mechanisms of resistance, such as alterations in expression of Rb, the target kinases, or cyclin E, will be critical for designing future strategies. ■
Study Title: JUNIPER: A randomized phase III study of abemaciclib plus best supportive care versus erlotinib plus best supportive care in patients with stage IV non–small cell lung cancer with a detectable KRAS mutation who have progressed after platinum-based chemotherapy
Study Type: Phase III interventional
Study Sponsor: Eli Lilly and Company
Purpose: To evaluate how safe and effective abemaciclib is in participants with lung cancer
Primary Outcome Measure: Overall Survival [time frame: baseline to date of death from any cause (estimated up to 47 months)]
Contact: Study Director:
1-877-CTLILLY (1-877-285-4559) or 1-317-615-4559
Clinical Trials Identifier:
1. Patnaik A, Rosen LS, Tolaney SM, et al: Efficacy and safety of abemaciclib, an inhibitor of CDK4 and CDK6, for patients with breast cancer, non-small cell lung cancer, and other solid tumors. Cancer Discov 6:740-753, 2016.
2. Rugo HS, Tolaney SM, Cortés J, et al: MONARCH 1: Final overall survival analysis of a phase 2 study of abemaciclib, a CDK4 and CDK6 inhibitor, as monotherapy, in patients with HR+/HER2- breast cancer, after chemotherapy for advanced disease. 2017 American Association of Cancer Research. Abstract CT044. Presented April 3, 2017.
3. Shapiro GI, Hilton J, Gandi L, et al: Phase I dose escalation study of the CDK4/6 inhibitor palbociclib in combination with the MEK inhibitor PD-0325901 in patients with RAS mutant solid tumors. 2017 American Association of Cancer Research Annual Meeting. Abstract CT046. Presented April 3, 2017.
4. Shapiro G, Neumuller M, Loof S, et al: Thymidine kinase activity as a response marker for CDK4/6 inhibition. 2017 American Association of Cancer Research Annual Meeting. Abstract 2340. Presented April 3, 2017.