Chemotherapy and CDK4/6 inhibitors: Unexpected bedfellows

Abstract

Cyclin-dependent kinases 4 and 6 (CDK4/6) have emerged as important therapeutic targets. Pharmacological inhibitors of these kinases function to inhibit cell cycle progression and exert other important effects on the tumor and host environment. Due to their impact on the cell cycle, CDK4/6 inhibitors (CDK4/6i) have been hypothesized to antagonize the anti-tumor effects of cytotoxic chemotherapy in tumors that are CDK4/6 dependent. However, there are multiple preclinical studies that illustrate potent cooperation between CDK4/6i and chemotherapy. Furthermore, the combination of CDK4/6i and chemotherapy is being tested in clinical trials to both enhance anti-tumor efficacy and limit toxicity. Exploitation of the non-canonical effects of CDK4/6i could also provide an impetus for future studies in combination with chemotherapy. Thus, while seemingly mutually exclusive mechanisms are at play, the combination of CDK4/6 inhibition and chemotherapy could exemplify rational medicine.

CDK4/6 in cell-cycle progression:

Cyclin dependent kinases (CDKs) are serine/threonine kinases that regulate the sequential progression of the cell cycle in eukaryotic organisms. The molecular functions of these kinases in different phases of the cell cycle have been well characterized (1, 2). The cell cycle machinery in higher eukaryotes is tightly regulated by the presence of more than 10 proteins in the CDK family that can have overlapping and distinct functions (2). Cell-cycle initiation occurs in G1 phase, which is conventionally governed by the activation of CDK4 and CDK6 kinases that are downstream of mitogenic signals (3–5). The catalytic activity of CDK4 and CDK6 is positively regulated by the binding of D-type cyclins (D1, D2 and D3). Expression of D-type cyclins is induced in response to mitogenic stimuli and remains high as the cells progress to the G1/S phase boundary (6). Therefore, unlike other cyclins and CDKs that are regulated by other components of the cell-cycle machinery, the expression of D-type cyclins --and by extension CDK4/6 associated kinase activity--largely depend on mitogenic signaling pathways (7, 8). Transcription of D-type cyclins is intimately linked to multiple pathways that coalesce to lead to the accumulation of transcripts (7, 9, 10). Mitogenic signaling pathways also regulate the stability and localization of these proteins (11, 12). Importantly, a host of growth inhibitory mechanisms also impact CDK4/6 activity, including the induction of endogenous CDK4/6-specific inhibitors with specific stresses (e.g. CDKN2A which encodes p16INK4A), and active mechanisms of cyclin D1 degradation (13, 14). Thus, CDK4/6 activity acts as a sensor linking multiple-signaling pathways to the initiation of the cell cycle (15–17).

CDK4/6 regulates the cell cycle through phosphorylation of key substrates. Unlike the prototypical CDK1 and CDK2, which can phosphorylate many substrates, CDK4/6 has a very limited repertoire of targets (18). CDK4 and CDK6 selectively phosphorylate the RB tumor suppressor protein and additional members of the RB family (18–21). RB-family proteins function as transcriptional co-repressors and limit the expression of E2F target genes that include multiple genes required for cell cycle progression, DNA replication, and mitotic progression (22, 23). The phosphorylation of RB, which is initiated by CDK4 or CDK6 serves to limit transcriptional repression and enable progression through latter phases of the cell cycle defining the canonical CDK4/6-RB pathway (Fig 1A).

Figure 1. Different cell cycle states in cancer.

(A) The Canonical G1/S regulatory circuit: CDK4/6 kinase activity is stimulated down-stream of mitogenic/oncogenic signals to initiate the phosphorylation of RB and related proteins. Phosphorylation facilitates the de-repression of E2F-family of transcription factors that drive the expression of many genes required for DNA replication, mitosis, and cell division. (B) CDK4/6 dependent cells: In cells, tumors, or tissues that are dependent on CDK4/6 activity treatment with pharmacological inhibitors yields the robust activation of RB. This event limits other CDK activities and represses the expression of essential genes for cellular division resulting in a G1/G0 like arrest. (C) CDK4/6-independent cells: There are clearly two distinct states that yieldCDK4/6 independent proliferation. (1) Loss of RB as occurs in a subset of human tumors removes the down-stream target and as such inhibition of CDK4/6 has minimal efficacy in controlling cell cycle. (2) through various mechanisms RB phosphorylation can remain during pharmacological inhibition. This cell cycle plasticity can be generated through either CDK4/6 or CDK2 complexes and is prevalent in a number of tumor types that retain the RB tumor suppressor.

The requirement for CDK4/6 in cell division has been interrogated utilizing multiple approaches and has illustrated important features of the cell cycle. The inhibition of CDK4/6 by the expression of endogenous inhibitors (e.g. p16INK4A) potently arrests cells that contain a functional RB protein and subsequently limits gene expression controlled by RB/E2F (Fig 1A). Multiple experimental methods (e.g. antibody injection, RNAi, etc.) have further suggested that D-type cyclins and/or CDK4/6 activity are generally important for progression from G1/S in normal cells as well as multiple cancer models (24). These findings contrast with studies in mouse models that clearly demonstrate that the cell cycle can proceed with genetic deletion of CDK4 and 6 or deletion of all D-type cyclins (25, 26). In this context, adaptation occurs in many tissues by enabling CDK2 or CDK1 activity to drive cell cycle entry. However, genetic suppression of CDK4/6 activity can limit or block tumor development in select models (27–30). This was clearly shown in the context of HER2-driven breast cancer where CDK4/6 activity is required both for tumor etiology and maintenance (31).

Pharmacological inhibitors of CDK4/6—mechanisms of action and resistance:

Due to the function of CDK4/6 in coordinating cell division, pharmacological inhibitors have been developed as anti-cancer drugs. There are five selective CDK4/6 inhibitors (CDK4/6i); palbociclib (PD0332991), ribociclib (LEE011), abemaciclib (LY2835219), trilaciclib, (G1T28) and lerociclib (G1T38) (32–38). Currently, three of these drugs are FDA-approved for the treatment of ER+ metastatic breast cancer based on multiple randomized clinical trials (palbociclib, ribociclib, abemaciclib). While all of these compounds are selective for CDK4/6, palbociclib, ribociclib, abemaciclib, and lerociclib are formulated for oral long-term dosing. Trilaciclib was formulated specifically for intravenous delivery and short half-life with the intended goal of preventing chemotherapy-induced host toxicities. Consistent with their mechanism of action, all CDK4/6i have cytostatic activity that is associated with RB-dependent suppression of the G1/S transition (32, 36). Pharmacological CDK4/6i mimic the effect of RB activation (Fig 1B) and suppress the expression of genes that are conventionally regulated by the E2F-family of transcription factors (39, 40). Since many of these genes are involved in core functions of DNA replication and mitotic progression, and are considered essential for proliferation, the magnitude of transcriptional repression downstream from CDK4/6 inhibition is critical for cytostatic activity.

Multiple determinants of response to CDK4/6 inhibition are being elucidated through both preclinical investigation and the analysis of clinical specimens (Fig 1C). This work has illustrated that there are multiple cell cycle related alterations present in models or tumors (e.g. RB loss or overexpression of cyclin E) that are associated with resistance to CDK4/6 inhibitors (41–44) (Fig 1C). Conversely, a number of oncogenic signaling pathways (e.g. RAS/MAPK, PTEN/PI3K, or HIPPO) have emerged as contributing to resistance (45–47). These derangements enable escape from CDK4/6 inhibition by facilitating the inactivation/phosphorylation of RB even in the presence of the pharmacological CDK4/6i. This is believed to occur due to “plasticity”, which is associated with either incomplete inhibition of CDK4/6 or the ability of CDK2 to initiate the phosphorylation of RB. While inhibition of CDK4/6 can arrest cells, dual inhibition of CDK4 and CDK2 has been shown to be required for durable responses in preclinical models (48). In addition, the hyperactivation of CDK2 kinase in breast cancer cells due to overexpression of cyclin E1/cyclin E2 drives resistance to CDK4/6i (49). Conversely, RNAi mediated knockdown of CCNE1 or CDK2 along with CDK4/6 inhibition can reverse acquired resistance to CDK4/6 inhibition in select models (50–52). Thus, the level of CDK2 activity during response is an important determinant and potential biomarker for the efficacy of CDK4/6i.

The spectrum of CDK4/6i tumor sensitivities and theoretical intersection with chemotherapy:

From preclinical and clinical studies, emerging data indicate that there are tumors that are sensitive to CDK4/6i, but that many of these sensitive tumors develop adaptive resistance mechanisms. In these contexts, CDK4/6i combination therapies can enhance the efficacy and durability of the tumor response. In contrast, there are a subset of cancers that are intrinsically CDK4/6-independent (e.g. as a consequence of RB loss) (33, 53). The consequences of the addition of CDK4/6i to chemotherapy must be considered within this framework. In patients with CDK4/6-independent tumors, the anticipated clinical benefit would be to protect normal cells from chemotherapy as the normal cells are sensitive to CDK4/6 inhibition and the tumor cells are insensitive to CDK4/6 inhibition (Fig 2A). In patients with CDK4/6-dependent tumors, there may be opportunities to enhance anti-tumor efficacy; however, there is a theoretical risk that CDK4/6 inhibition in this setting may antagonize the intended cytotoxicity of the chemotherapy (Fig 2B). Preclinical and clinical data suggest that the risk of chemotherapy antagonism by CDK4/6i is not as well understood as initially thought. In addition, the molecular determinants of CDK4/6-independence and -dependence are complex, such that it can be difficult to identify those tumors that truly rely upon CDK4/6 for proliferation. We discuss these approaches using the “theoretical” binary tumor classification of CDK4/6-independent and -dependent and acknowledge that the spectrum of dependence may actually be continuous.

Figure 2. Differential response to CDK4/6 inhibition determines mode of interaction with chemotherapy.

(A) CDK4/6 independent tumors: In CDK4/6 independent tumor models, by definition the pharmacological inhibitors will have minimal effect. Small cell lung cancer tumors exhibit near universal loss of the RB tumor suppressor, while triple negative breast cancer exhibits multiple mechanisms that render CDK4/6 independence. In contrast, the host-tissues will be responsive to the CDK4/6 inhibitor and the reduced cell division could be expected to mitigate the toxicity associated with chemotherapy in specific tissue settings. (B) Conceptual frame-work for antagonism: Chemotherapy represents a diverse class of drugs that exploit the rapid division of tumor cells that impinges on DNA replication or mitotic division. This sensitivity is often enhanced as a result of cell cycle checkpoint deficits in tumor cells. Ultimately, apoptosis, necrosis, or mitotic catastrophe lead to cytotoxic activity. Since CDK4/6 inhibitors can slow cell cycle progression they could negatively impact on the efficacy of chemotherapy.

CDK4/6-independent Tumors - host protection:

The use of CDK4/6i to arrest cells in the G1 phase in cancer patients who are being treated with chemotherapy may not seem to be intuitive. However, this biological phenomenon can be exploited to prevent chemotherapy-induced cellular damage of normal cells that harbor an intact RB pathway (54, 55) (Fig 2A). One of the common side effects of chemotherapy is myelosuppression, that can lead to the exhaustion of hematopoietic stem and progenitor cells (HSPCs) (56–58).

Trilaciclib (G1T28) has been developed to specifically prevent chemotherapy-induced myelosuppression (38). Trilaciclib maintains a selective and reversible G1-arrest in the RB proficient HPSCs and prevents or mitigates the acute and long-term hematopoietic toxicity of the cytotoxic chemotherapeutic agent, 5-fluorouracil, when administered concurrently (38, 55). Similarly, the cytostatic effect of CDK4/6i can prevent or mitigate the hematopoietic toxicity of ionizing radiation by preventing HSPCs from immediately entering the cell cycle when the cells sense the radiation-induced DNA damage (54).

To translate these findings to the clinic, three randomized, placebo-controlled, double-blind clinical trials designed to evaluate the myelopreservation effects of trilaciclib versus placebo in combination with chemotherapy have been completed in small cell lung carcinoma (SCLC) (59–62)(Table 1). SCLC was chosen as the first clinical setting to test the myelopreservation benefit of trilaciclib because: (1) standard of care chemotherapy regimens are myelosuppressive, (2) SCLC replicates independently of CDK4/6 due to obligate loss of RB (63), thereby minimizing theoretical concerns related to chemotherapy antagonism, and (3) SCLC treated in the 1^st line setting is a chemosensitive tumor which provided an optimal background upon which to demonstrate that trilaciclib does not antagonize chemotherapy efficacy. In all three studies, trilaciclib demonstrated consistent clinical benefit across myelosuppression endpoints, including highly statistically significant improvements for both primary endpoints: the duration of severe neutropenia [SN] in Cycle 1 (a surrogate for febrile neutropenia and infections) and the percentage of patients with SN. Furthermore, integrated analysis from the three studies demonstrated statistically significant improvement across multiple hematopoietic lineages; including neutrophils (duration of SN in Cycle 1, percentage of patients with SN), red blood cells (RBC) (percentage of patients with Grade 3 or 4 anemia, percentage of patients receiving RBC transfusions on or after 5 weeks, the rate of RBC transfusions on or after 5 weeks), and platelets (percentage of patients with Grade 3 or 4 thrombocytopenia) (59). Consistent with the improvement in chemotherapy safety, patient reported outcome (PRO) measures demonstrated an improved experience for patients receiving trilaciclib, including improved measures of fatigue (59). Importantly, the addition of trilaciclib to chemotherapy did not have an adverse effect on anti-tumor efficacy (60–62).

Table 1:

Clinical studies of CDK4/6 inhibitors with chemotherapy

Study Title	CDK4/6 Inhibitor	Chemotherapy	Phase	Description	NCT#
Phase 2 Study of Carboplatin, Etoposide, and Atezolizumab With or Without Trilaciclib in Patients With Untreated Extensive Stage Small Cell Lung Cancer	Trilaciclib	Carboplatin, Etoposide, and Atezolizumab	2	This is a study to investigate the potential clinical benefit of trilaciclib (G1T28) in preserving the bone marrow and the immune system, and enhancing antitumor efficacy when administered with carboplatin, etoposide, and atezolizumab (E/P/A) therapy in first line treatment for patients with newly diagnosed extensive-stage SCLC.	NCT03041311
Phase 2 Study of the Safety, Efficacy, and Pharmacokinetics of G1T28 in Patients With Metastatic Triple Negative Breast Cancer Receiving Gemcitabine and Carboplatin Chemotherapy	Trilaciclib	Gemcitabine and Carboplatin	2	his is a study to investigate the potential clinical benefit of trilaciclib (G1T28) in preserving the bone marrow and the immune system, and enhancing chemotherapy antitumor efficacy when administered prior to carboplatin and gemcitabine (GC therapy) for patients with metastatic triple negative breast cancer.	NCT02978716
Phase 1b/2a Safety and Pharmacokinetic Study of G1T28 in Patients With Previously Treated Extensive Stage Small Cell Lung Cancer (SCLC) Receiving Topotecan Chemotherapy	Trilaciclib	Topotecan	1/2	This is a study to investigate the potential clinical benefit of trilaciclib (G1T28) in preserving the bone marrow and the immune system, and enhancing chemotherapy antitumor efficacy when administered prior to topotecan in patients previously treated for extensive-stage SCLC.	NCT02514447
Phase 1b/2a Safety and Pharmacokinetic Study of G1T28 in Patients With Extensive Stage Small Cell Lung Cancer (SCLC) Receiving Etoposide and Carboplatin	Trilaciclib	Combination and Etoposide	1/2	This is a study to investigate the potential clinical benefit of trilaciclib (G1T28) in preserving the bone marrow and the immune system, and enhancing chemotherapy antitumor efficacy when administered prior to carboplatin and etoposide in first line treatment for patients with newly diagnosed extensive-stage SCLC.	NCT02499770
A Phase 1 Study Of Palbociclib, A CDK 4/6 Inhibitor, In Combination With Chemotherapy In Children With Relapsed Acute Lymphoblastic Leukemia (ALL) Or Lymphoblastic Lymphoma (LL)	Palbociclib	Cytarabine, Methotrexate, Hydrocortisone, Doxorubicin, Prednisolone, Vincristine, Pegaspargase, Prednisone	1	AINV18P1 is a Phase 1 study where palbociclib will be administrated in combination with a standard re-induction platform in pediatric relapsed Acute Lymphoblastic Leukemia (ALL) and lymphoblastic lymphoma (LL).	NCT03792256
A Phase I Study of the CDK4/6 Inhibitor PD-0332991, 5-Fluorouracil, and Oxaliplatin in Patients With Advanced Solid Tumor Malignancies	Palbociclib	5-Fluorouracil and Oxaliplatin	1	The purpose of this study is to test the safety and effectiveness of a new combination of drugs, palbociclib and 5-Fluorouracil and Oxaliplatin for patients with advanced solid tumor malignancies.	NCT01522989
Phase 1B Study of PD-0332991 in Combination With T-DM1 in the Treatment of Patients With Advanced HER2 (Human Epidermal Growth Factor Receptor 2)-Positive Breast Cancer	Palbociclib	T-DM1	1	This is a phase 1B inter-patient dose escalation study of PD-0332991 in combination with T--DM1 in patients with recurrent or metastatic HER2-positive breast cancer after prior trastuzumab or other HER2-directed therapies.	NCT01976169
A Phase 1 Trial of PD0332991 and Paclitaxel in Patients With Rb-Expressing Advanced Breast Cancer	Palbociclib	Paclitaxel	1	This study is a phase I, single arm, open-label trial of PD0332991 in combination with Paclitaxel in patients with Rb-expressing metastatic breast cancer.	NCT01320592
An Open-Label Phase 1B Study of Palbociclib (Oral CDK4/6 Inhibitor) Plus Abraxane (Nab-Paclitaxel) In Patients with Metastatic Pancreatic Ductal Adenocarcinoma	Palbociclib	Nab-Paclitaxel	1	This is a Phase 1, open label, multi center, multiple dose, dose escalation, safety, pharmacokinetic and pharmacodynamic study of palbociclib in combination with nab-P, in sequential cohorts of adult patients with mPDAC, with MTD expansion cohort(s).	NCT02501902
A Phase 1 Study of Palbociclib in Combination With Cisplatin or Carboplatin in Advanced Solid Malignancies	Palbociclib	Carboplatin or Cisplatin	1	This phase I trial studies the side effects and best dose of palbociclib with cisplatin or carboplatin in treating patients with solid tumors that have spread to other places and usually cannot be cured or controlled with treatment.	NCT02897375
A Phase I Study of the CDK4/6 Inhibitor PD-0332991, 5-Fluorouracil, and Oxaliplatin in Patients With Advanced Solid Tumor Malignancies	Palbociclib	5FU Oxaliplatin	1	The purpose of this study is to test the safety and effectiveness of a new combination of drugs, PD-0332991 and 5-Fluorouracil and Oxaliplatin for patients with advanced solid tumor malignancies.	NCT01522989
A Phase 1b/2 Study of the Oral CDK4/6 Inhibitor LEE011 (Ribociclib) in Combination With Docetaxel Plus Prednisone in Metastatic Castration Resistant Prostate Cancer	Ribociclib	Docetaxel, Ribociclib, Prednisone, Filgrastim	1/2	his is a Phase Ib/II open label clinical trial in patients with metastatic castration resistant prostate cancer. The objective of the phase Ib portion of the study is to establish the maximum tolerated dose (MTD) and dose limiting toxicities (DLT) of docetaxel (75 mg/m2 IV q21 days) and prednisone (5mg orally BID) in combination with ribociclib in escalating oral daily doses in patients with metastatic CRPC with prior resistance to abiraterone and/or enzalutamide who have not undergone prior chemotherapy for metastatic disease.	NCT02494921
A Phase I Study of CDK4/6 Inhibitor LEE011 Combined With Gemcitabine in Patients With Advanced Solid Tumors or Lymphoma	Ribociclib	Gemcitabine	1	This phase I trial studies the side effects and best dose of ribociclib and gemcitabine hydrochloride in treating patients with solid tumors or lymphoma that have spread to other places in the body and usually cannot be cured or controlled with treatment	NCT02414724
Phase I Study of CDK4/6 Inhibitor Ribociclib (LEE011) Combined With Gemcitabine in Patients With Advanced Solid Tumors	Ribociclib	Gemcitabine	1	This phase I trial studies ribociclib and gemcitabine hydrochloride in treating patients with solid tumors that have spread to other places in the body.	NCT03237390
Study of Ribociclib With Everolimus + Exemestane in HR+ HER2- Locally Advanced/Metastatic Breast Cancer Post Progression on CDK 4/6 Inhibitor. (TRINITI-1)	Ribociclib	Everolimus Exemestane	1/2	The purpose of this study is determine if the triplet combination of ribociclib, everolimus and exemastane is effective in the treatment of locally advanced/metastatic breast cancer following treatment with a CDK 4/6 inhibitor	NCT02732119
Phase Ib Trial of LEE011 With Everolimus (RAD001) and Exemestane in the Treatment of Hormone Receptor Positive HER2 Negative Advanced Breast Cancer	Ribociclib	Everolimus Exemestane	1	This study evealuates the safety and tolerability of the triplet combination of LEE011 + everolimus + exemestane in patients naïve or refractory to CDK4/6 inhibitor-based therapy, and the safety and tolerability of the doublet combination of LEE011 + exemestane in patients refractory to CDK4/6 inhibitor-based therapy.	NCT01857193
An Open-Label, Phase Ib/II Clinical Trial Of Cdk 4/6 Inhibitor, Ribociclib (Lee011), In Combination With Trastuzumab Or T-Dm1 For Advanced/Metastatic Her2-Positive Breast Cancer	Ribociclib	T-DM1 Trastuzumab Fulvestrant	1/2	This study tests the combination of ribociclib in combination with T-DM1, trastuzumab, or trastuzumab plus fulvestrant in patients with advanced/metastatic Her2+ breast cancer.	NCT02657343
A Phase I Trial of Ribocilcib (LEE011) and Weekly Paclitaxel in Patients With Rb+ Advanced Breast Cancer	Ribociclib	Paclitaxel	1	This is a Phase I study to assess the safety and MTD of paclitaxel + ribociclib (LEE011) in patients with Rb+, advanced breast cancer. Dose escalation will be performed using standard 3 + 3 dosing strategy.	NCT02599363
A Phase Ib/II Study of LEE011 and Chemoembolization In Patients With Advanced Hepatocellular Carcinoma	Ribociclib	TACE	2	The purpose of this study is determine whether the combination therapy with LEE011 and chemoembolization in patients with locally advanced Hepatocellular Carcinoma not amenable to curative therapies will provide greater efficacy than chemoembolization alone with a tolerable safety profile.	NCT02524119
A Phase 1b/2 Study of the Oral CDK4/6 Inhibitor LEE011 (Ribociclib) in Combination With Docetaxel Plus Prednisone in Metastatic Castration Resistant Prostate Cancer	Ribociclib	Docetaxel Prednisone	1/2	This is an open-label study of ribociclib (dosed at the RP2D) in combination with docetaxel and prednisone to determine the efficacy and safety of the treatment combination in patients with metastatic castration resistant prostate cancer.	NCT02494921
A Phase 1b Study of LY2835219 in Combination With Multiple Single Agent Options for Patients With Stage IV NSCLC	Abemaciclib	Pemetrexed Gemcitabine	1	The main purpose of this study is to evaluate the safety and tolerability of abemaciclib in combination with another anti-cancer drug in participants with NSCLC that is advanced or has spread to other parts of the body (stage IV).	NCT02079636
A Phase 2 Study of Abemaciclib in Patients With Brain Metastases Secondary to Hormone Receptor Positive Breast Cancer, Non-small Cell Lung Cancer, or Melanoma	Abemaciclib	Pemetrexed Gemcitabine	2	The main purpose of this study is to evaluate the safety and effectiveness of the study drug known as abemaciclib in participants with hormone receptor positive breast cancer, non-small cell lung cancer (NSCLC), or melanoma that has spread to the brain. Some cohorts allow concurrent pemetrexed and/or gemcitabine.	NCT02308020
A Phase II Trial Program Exploring The Integration Of Novel HER2-targeted Tyrosine Kinase Inhibitor Pyrotinib and CDK4/6 Inhibitor SHR6390 Into Current Chemotherapy/Endocrine Therapy Regimes For Prior Trastuzumab-treated Advanced HER2-positive Breast Cancer	SHR6390	Capecitabine	2	Patients previously failing transtuzumab therapy are randomized to Pyrotinib and SHR6390 plus capcitabine, letrozole, or placebo.	NCT04095390

Similar to SCLC, triple negative breast cancer (TNBC) is thought to be a mostly CDK4/6-independent tumor, based on both tumor genetics and the relatively poor response of such tumors to CDK4/6i therapy in preclinical studies (64, 65). Trilaciclib has been tested in combination with gemcitabine and carboplatin in patients with metastatic TNBC (66). In this study the addition of trilaciclib to chemotherapy generally did not improve myelosuppression endpoints; however, there were positive trends for RBC and platelet measures, and patients in the trilaciclib arms received significantly more chemotherapy than the control. In contrast, the anti-tumor efficacy results demonstrated a clinically meaningful survival benefit in both combination groups compared to the chemotherapy alone control group. Median progression-free survival was 5.7 months (95% CI: 3.4–9.2) for the control group compared to 9.4 (6.1–13.0; hazard ratio [HR]: 0.60, p=0.13) and 7.3 (6.2–12.19; HR:0.59, p=0.12) for the two trilaciclib groups. Median overall survival was 12.6 months (6.3–15.6) for the chemotherapy control group compared to 20.1 (10.2-not reached; HR: 0.33, p=0.028) and 17.8 (12.9-not reached; HR: 0.34, p=0.0023) for the two trilaciclib + chemotherapy groups. While patients receiving trilaciclib received more chemotherapy, it is unlikely that this can explain the magnitude of survival benefit achieved with transient CDK4/6 inhibition. Instead, an alternative mechanism of action related to enhanced anti-tumor immunity is more likely as discussed in more detail below.

Antagonism of chemotherapy-mediated cytotoxicity in CDK4/6-dependent pre-clinical models:

Based on the intra-cellular targets of many chemotherapeutic agents, it is evident that dividing cells are more chemo-sensitive than arrested cells which underlies the therapeutic index of such agents (67). Considering the mechanism of action of CDK4/6i in inhibiting cell division through activation of the RB pathway, it is hypothesized that concurrent CDK4/6 inhibition may antagonize the cytotoxic effects of chemotherapeutic agents in tumors that are CDK4/6 dependent (Fig 2B). Indeed, a number of preclinical reports have described the antagonistic effects of combining a CDK4/6i with chemotherapy (50, 68–70) (Table 2). In breast cancer cell lines, xenografts, and GEMM models, treatment with CDK4/6i can limit the acute induction of tumor-specific toxicity with taxanes, anthracyclines, and platinum-agents (68, 70–72). These effects are RB-dependent and link the antagonism of chemotherapy cytotoxicity with the cell cycle pause induced by CDK4/6 inhibition (68, 70, 71). In considering these data, it is important to appreciate that most studies utilized conditions where CDK4/6 inhibition elicits profound cell cycle inhibition and that effects measured on antagonism were relatively short-term; however, it should not be discounted that such antagonism could be clinically relevant and caution should be taken to evaluate whether these effects will be seen in specific clinical settings.

Table 2:

Preclinical Studies Demonstrating Some Evidence of Antagonism Between CDK4/6 Inhibitors and Chemotherapy

Author	Tumor Type	Context	Chemo	Outcome
Franco et al. Oncotarget 2014	Human PDA	In vitro	Gemcitabine 5FU	Observed antagonism when palbociclib was added to gemcitabine but cooperation between palbociclib and 5FU. Palbociclib reduced TS, the target of 5FU. Differential cell cycle sensitivity to palbociclib despite uniform suppression of RB phosphorylation.
McClendon et al. Cell Cycle 2012	Human TNBC	In vitro In vivo	Doxorubicin	In RB-proficient TNBC cell lines, co-treatment yielded cooperative cytostatic effects but reduced doxorubicin mediated cytotoxicity.
Dean et al. JBC 2012	Human TNBC	In vitro In vivo	Doxorubicin Paclitaxel Radiation	Mostly a biochemical demonstration that CDK4/6 inhibition blocks cells in G1 and is associated with reduced E2F gene expression even in the presence of chemotherapy. Pre-treatment and continuous palbociclib reduced paclitaxel effects in outgrowth assays, but pre and concurrent synchronization experiments enhanced activity in vivo. Showed CDK4/6 inhibition pushes DNA repair from HR to NHEJ and palbociclib leads to reduction of Ku70 needed for NHEJ, which implies that CDK4/6 inhibition may impair a cancer cells ability to repair chemotherapy induced DNA damage.
Roberts et al. JCI 2012	TNBC HER2	In vitro In vivo	Carboplatin Doxorubicin, Etoposide Camptothecin,Paclitaxel	Protection of “normal” cell line and bone marrow from chemotherapy. Concurrent palbociclib did not antagonize carboplatin in TNBC model. Concurrent palbociclib plus carboplatin treatment led to tumor regression in Her2 model, but depth of response was not as deep as single agent carboplatin.
Konecny et al. Clin Ca Re. 2011	Human Ovarian	In vitro	Carboplatin	Concurrent palbociclib treatment with carboplatin or paclitaxel resulted in synergism, however pre-treatment resulted in antagonism.
Cretella et al. Scientific Reports 2019	Human TNBC	In vitro	Paclitaxel	Pre-treatment with palbociclib sensitized cells to paclitaxel (sequential) and there was no difference in efficacy when paclitaxel was added 0, 4, 8 hours after palbociclib. Simultaneous treatment showed an antagonistic effect.

Potential cooperation between CDK4/6i and chemotherapy in cancer therapy in preclinical models:

In contrast to the above reports, a number of preclinical studies suggest that the combination of chemotherapy and CDK4/6 inhibition can have cooperative anti-tumor effects; similar observations are beginning to emerge from clinical studies. Palbociclib, ribociclib, and abemaciclib have been shown to enhance, rather than antagonize, chemotherapy cytotoxicity when combined with camptothecin, carboplatin, cisplatin, docetaxel, doxorubicin, 5FU, gemcitabine, irinotecan, paclitaxel, and temozolomide (73–86). These effects were shown in RB-proficient in vitro and in vivo models of non-small cell lung carcinoma (NSCLC), ovarian cancer, gastric cancer, TNBC, atypical teratoid rhabdoid tumors, Ewing sarcoma, pancreatic cancer and glioblastoma using both sequential and concurrent dosing schedules (73–86) (Table 3).

Table 3:

Preclinical Studies Demonstrating Enhanced Efficacy with CDK4/6 Inhibitors + Chemotherapy Treatment

Author	Inhibitor	Tumor Type	Context	Chemotherapy	Outcome
Hamilton et al. Molecules 2014	Palbociclib	SCLC	In vitro	Camptothecins	Palbociclib reduced TOPO1 expression and enhanced camptothecin activity
Gelbert et al. Inv New Drug 2014	Abemaciclib	Lung cancer	In vivo	Gemcitabine	Abemaciclib plus gemcitabine sequentially or concurrently was better than either agent alone.
Hashizume et al. Neuro-Oncology 2016	Palbociclib	ATRT and glioblastoma	In vitro In vivo	Radiation	Palbociclib treatment reduced phospho-RB levels and pre-, concurrent and post-palbociclib treatment enhanced gH2AX formation, enhanced tumor response and improved overall survival.
O’Brien et al. MCT 2018	Abemaciclib	TNBC	In vitro In vivo	Docetaxel, Carboplatin	No antagonism seen in either phospho-RB high/p16 low (MDA-231) or phospho-RB low/p16 high (HCC70) TNBC mouse models when abemaciclib was combined with docetaxel (concurrent treatment). Additionally, in MDA-231 cells, abemaciclib enhanced activity with concurrent treatment for 5 days with either docetaxel or carboplatin.
Iyengar et al. Oncotarget 2018	Ribociclib	Ovarian Ca	In vitro In vivo	Cisplatin	Ribociclib + cisplatin improved efficacy in vitro and in vivo compared to cisplatin alone in RB1 competent, CDK4/6 sensitive ovarian models.
Dowless et al. Clin Ca Re 2018	Abemaciclib	Ewing sarcoma	In vitro In vivo	Doxorubicin, Etoposide, Cisplatin, Temozolomide Irinotecan	Abemaciclib increased IC50 of various chemotherapies (doxorubicin, etoposide, cisplatin, paclitaxel) in vitro, but enhanced anti-tumor efficacy of doxorubicin and temozolomide/irinotecan in vivo.
Chou et al. Gut 2018	Palbociclib	Pancreatic Cancer	In vitro In vivo	Gemcitabine Paclitaxel	Palbociclib improved efficacy of gemcitabine and gemcitabine/paclitaxel in RB-high patient derived PDAC models. Improved endpoints include synergy, apoptosis, reduced metastasis (in vitro and in vivo), primary and recurrent (2nd-line therapy) tumor growth.
Raub et al. Drug Metab Dispos 2015	Abemaciclib	Glioblastoma	In vivo	Temozolomide	Abemaciclib improved anti-tumor efficacy and survival when added to temozolomide in U87MG xenograft model and rat orthotopic model.
Wang et al. Int J Mol Med. 2018	Palbociclib	Gastric Ca	In vitro	5FU	Palbociclib produced a dose dependent G1 arrest in gastric cancer cells and enhanced 5FU mediated cytotoxicity. Proteomic analysis demonstrated that palbociclib altered expression of proteins involved in the regulation of cell death, cell cycle, cell growth, proliferation, and cell migration.
Gao et al. Cell Oncol. 2017	Palbociclib	Ovarian	In vitro	Paclitaxel	Inhibition of CDK4(and 6) by palbociclib or CDK4-specific siRNA increased the sensitivity of both RB-positive and -negative ovarian cancer cells to paclitaxel. Combination treatment enhanced apoptosis as measured by increased PARP cleavage and reduced Bcl-xL and Survivin. Demonstrated a role for MDR in paclitaxel resistance and showed that palbociclib and CDK4 siRNA may act in part to reduce MDR activity.
Zhang et al. Cancer Biology & Therapy. 2013	CINK4	NSCLC	In Vitro	Paclitaxel	CDK4 siRNA significantly increased paclitaxel sensitivity in KRAS mutation-positive H23 cells. CINK4 demonstrated concentration- and time-dependent anti-proliferative activity in 5 NSCLC cell lines. Combined CINK4 and paclitaxel produced synergistic anti-proliferative activity and increased apoptosis through reduced cyclin D1 and Bcl-2 in KRAS mutation-positive cancer cells.
Cao et al. Oncogene 2019	Palbociclib	Squam Lung	In vitro In vivo	Paclitaxel	Palbociclib enhanced in vitro and in vivo efficacy through RB-dependent E2F mediated alterations in senescence, G2/M spindle checkpoint, and angiogenesis. Note: this series of experiments used sequential treatment schedule. In vivo Pac D1 Palbo D2–6 for ~4 weekly cycles
Kumarasamy et al. Oncogene 2019	Palbociclib, Ribociclib	PDAC	In vitro In vivo	Gemcitabine Docetaxel	CDK4/6i prevented the outgrowth of tumor cells following the release from gemcitabine by inhibiting the re-entry of cells to cell-cycle through the downregulation of DNA replication and repair genes. CDK4/6i in combination with docetaxel resulted in a cooperative inhibitory effect on cell-cycle progression by inhibiting the CDK2 kinase activity
Salvador-Barbero et al. Cancer cell 2020	Palbociclib	PDAC	In vitro In vivo	Taxol	Sequential administration of palbociclib following taxol treatment cooperatively prevented cell proliferation in PDAC cells and PDX models. Palbociclib-mediated repression of proteins involved in homologous recombination prevented the ability of cells to recover from chromosomal damage.induced by chemotherapy

Ca, cancer; NSCLC, non-small cell lung cancer; SCLC, small cell lung cancer; Squam lung, squamous lung cancer; TNBC, triple negative breast cancer; PDAC, Pancreatic ductal adenocarcinoma

While the CDK4/6-RB-E2F axis is responsible for controlling expression of genes required for cell cycle progression, DNA replication, and mitotic progression (23, 39), the unexpected observations of cooperation described above may be due to other less well understood mechanisms. One mechanistic explanation of the enhanced, rather than antagonistic activity of combination CDK4/6i plus chemotherapy regimens, has been the reduced expression of specific E2F-regulated genes, whose products are targeted by chemotherapy (Fig. 3A). Palbociclib treatment reduces thymidylate synthase (TS; 5FU target), Topoisomerase 1, and Topoisomerase 2 alpha expression. These effects on gene expression could potentially enhance the response to select chemotherapies by limiting the threshold needed for efficacy of chemotherapy (70, 78, 87). Similarly, E2F regulates the expression of multiple genes required for DNA damage repair and thus would limit the ability of tumor cells to recover from chemotherapy-mediated damage. Consistent with the impact on DNA-repair machinery, it has been shown that CDK4/6i can cooperate with PARP inhibitors ostensibly by limiting the ability of damaged cells to carry out HR-mediated repair (88) (Fig. 3A). Conversely, Gao et al demonstrated a role for MDR (multidrug resistance; P-glycoprotein) in paclitaxel resistance, an effect counteracted by both CDK4 siRNA and palbociclib treatment (76). Finally, multiple studies have demonstrated that CDK4/6 inhibition can enhance chemotherapy-induced apoptosis (76, 80, 86), and that CDK4/6 are upstream regulators of transcription factors that control global gene expression leading to changes in metabolism, DNA repair and cell plasticity, all of which can render a cancer cell more susceptible to chemotherapy cytotoxicity (89). Collectively, these results suggest that the net effect of concomitant CDK4/6 inhibition during chemotherapy exposure in patients with CDK4/6-dependent tumors will provide cooperation rather than antagonism (Table 3).

Figure 3. Cellular response to CDK4/6 inhibition in combination with chemotherapy.

(A) Downstream from the activation of the RB-pathway are the repression of multiple genes involved in different pathways that could impinge on different features of response to distinct chemotherapeutic agents. (B) There are multiple mechanisms through which chemotherapy impinges on the activity of CDK2 and CDK4/6 complexes that would be expected to enhance the efficacy of pharmacological CDK4/6 inhibitors. For example, CDC25A is rapidly degraded or the CDK2-inhibitor P21CIP1 is rapidly induced as a specific response to chemotherapy. Conversely, degradation of cyclin D1 and inhibition of CDK4/6 complexes is a consequence of different chemotherapy agents that induce S-phase block.

Strategies for incorporating CDK4/6i into chemotherapy regimens:

Based on the available preclinical and clinical data, there are several therapeutic strategies by which CDK4/6i can be incorporated into standard chemotherapy regimens to provide therapeutic benefit to patients:

Protection of Normal Tissues:

As described above, protection of HSPCs to reduce dose-limiting myelosuppression has been demonstrated preclinically and clinically (38, 60–62, 68, 90). While myelosuppression is recognized as a common complication of chemotherapy, damage to other normal tissues including the gastrointestinal track, kidney and hair follicles also occurs. CDK4/6 inhibition has been shown to ameliorate kidney injury in preclinical models following both cisplatin treatment and acute renal ischemia, and to provide intestinal radioprotection (39, 91, 92). Additionally, alopecia, while not life threatening, is one of the most distressing side effects of chemotherapy. Similar to other tissues, transient CDK4/6 inhibition has been shown to protect hair follicles from taxane-induced damage in preclinical models (93). While clinical benefit for this approach has been shown in a CDK4/6-independent setting (SCLC), the question remains whether it can be employed in CDK4/6-dependent tumors without antagonizing chemotherapy anti-tumor efficacy. In vivo evaluation of trilaciclib with chemotherapy in CDK4/6 sensitive breast cancer models has not shown antagonism at the tumor level, and with some models, the combination shows enhanced anti-tumor efficacy (94). Interestingly, subset analysis of patients in the trilaciclib metastatic TNBC (mTNBC) study using PAM50 and other molecular stratification approaches revealed no antagonism, and demonstrated improved PFS and OS across all groups (55, 66, 95, 96).

Concurrent interactions, maintenance therapy, and staggered strategies to enhance anti-tumor efficacy:

Since the mechanisms of CDK4/6i and chemotherapy action are distinct, there could be drug interactions that would enhance the efficacy of each class of agent. Chemotherapy is well known to impact CDK-biology at multiple points that would be expected to enhance the cytostatic response to CDK4/6 inhibition (73). For example, chemotherapy can impact CDK2 activity via the induction of the endogenous CDK inhibitor p21 or loss of the CDC25a protein phosphatase which would yield increased the inhibitory phosphorylation on CDK2 (Fig. 3B). This cooperation has been illustrated in models of pancreatic cancer and other tumor models that do not show robust response to CDK4/6i (97, 98). Regarding resistance to chemotherapy, as discussed above, CDK4/6i can impact the expression of genes associated with DNA-repair and dNTP-metabolism. These effects could be broadly relevant to chemotherapy that induces DNA damage or is associated with nucleotide metabolism or function (Fig. 3A). Whether these interactions manifest clinically remains unclear, although multiple clinical trials are interrogating CDK4/6i and chemotherapy combinations where there is clear overlap in the treatment.

Given the canonical action of chemotherapy and CDK4/6i, it is appealing to separate/stagger the dosing of the chemotherapy and the CDK4/6i. In this context, the chemotherapy can have the desired impact of killing the tumor cells, while the CDK4/6i prevents the expansion of cells that are not killed by the chemotherapy (99). In ovarian cancer models, following the release from cisplatin-mediated S-phase arrest, tumor cells undergo normal cell-cycle progression and proliferation, which is significantly blocked by CDK4/6 inhibition, indicating a positive interaction (81). Similarly, the targeted microtubule poison trastuzumab-emtansine (T-DM1) displayed a cooperative anti-tumor effect where CDK4/6i could block the recovery of residual cells following T-DM1 treatment (100). Presently there are a number of clinical trials that explore the interaction of CDK4/6 inhibition and chemotherapy using dosing strategies to enhance durability of response (Table 1). In one of the first reported trials, the combination of paclitaxel with palbociclib appeared to have efficacy in heavily pre-treated breast cancer (101).

As CDK4/6i are not associated with cumulative toxicity, which is a common feature of chemotherapy, they can be given for long periods of time. Therefore, chronic administration of a CDK4/6i in the maintenance treatment setting, after the tumor has been de-bulked and chemotherapy discontinued could lead to improved patient outcomes by delaying tumor progression and or allowing the host immune system to eliminate the residual disease (81).

Enhancing the response to immunotherapy:

There is significant preclinical data demonstrating that CDK4/6i can enhance immune checkpoint inhibitor (ICI) efficacy through enhanced T-cell activation, increased antigen presentation, increased expression of PD-L1, and reduced T-cell exclusion and immune evasion gene signature (89, 102, 103) In the clinic, chemotherapy has successfully been used to enhance ICI efficacy (104–109) through induction of immunogenic cell death, enhancement of immunosurveillance and T cell activity, and reduction of immunosuppressive cell types (110–113). Despite these benefits, chemotherapy induced myelosuppression and immunosuppression may limit the full benefit of combinatorial treatments with ICIs. Given that intratumor immune cells are highly proliferative, one strategy to further enhance chemotherapy/ICI combinations is through transient CDK4/6 inhibition during chemotherapy exposure. Trilaciclib has been shown to favorably alter the tumor immune microenvironment through transient T cell inhibition (114, 115). Treatment of immunocompetent tumor models with trilaciclib plus chemotherapy/ICI combinations significantly improved anti-tumor efficacy and survival compared to chemotherapy/ICI combinations (114, 115).

Summary and under-explored areas:

In summary, while it was originally hypothesized that CDK4/6i combined with chemotherapy could potentially result in antagonistic anti-tumor effects (at least in CDK4/6-dependent tumors), emerging data suggest that these agents can be safely combined using different clinical strategies to enhance anti-tumor efficacy and/or reduce chemotherapy-induced toxicity. There are a number of unanswered questions whose answers could help guide implementation of these strategies in the clinic. Understanding which of these clinical strategies would be best employed in tumors that are “truly” CDK4/6-dependent is still needed. However, identifying “truly” CDK4/6-dependent tumors remains elusive as predictive biomarkers have not been validated in clinical practice. Identification and validation of such biomarkers would allow clinical testing of CDK4/6i + chemotherapy combinations in a homogeneous CDK4/6-dependent tumor population to definitively determine whether CDK4/6 inhibition during chemotherapy exposure interferes with the intended anti-tumor efficacy of chemotherapy. Additionally, understanding the contribution of anti-tumor effects arising from cell cycle inhibition in tumor cells versus non-tumor cells (e.g. cancer associated fibroblasts and immune cells) and through non-canonical biological processes controlled by CDK4/6 will further aid the rational design of novel CDK4/6i plus chemotherapy and/or immunotherapy regimens.