Preclinical Activity of Abemaciclib Alone Or in Combination with Anti- Mitotic and Targeted Therapies in Breast Cancer

Home , Abemaciclib

Author Manuscript Published OnlineFirst on February 26, 2018; DOI: 10.1158/1535-7163.MCT-17-0290 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Preclinical activity of abemaciclib alone or in combination with anti- mitotic and targeted therapies in breast cancer

Neil O’Brien1, Dylan Conklin1, Richard Beckmann2, Tong, Luo1, Kevin Chau1, Josh Thomas1, Ann Mc Nulty2, Christophe Marchal2, Ondrej Kalous1, Erika Von Euw1, Sara Hurvitz1, Colleen Mockbee2 and Dennis J. Slamon1

1Division of Hematology/Oncology, Department of Medicine, Geffen School of Medicine at UCLA, Los Angeles, CA 2Oncology Discovery Research, Lilly Research Laboratories, Indianapolis, IN

Running Title: Preclinical activity of abemaciclib in breast cancer Financial Support: UCLA funding; Dennis J. Slamon, DOD Innovator Award W81XWH-11-1-0104.

Corresponding Authors: Dennis J. Slamon, M.D., PhD UCLA Translational Oncology, 2825 Santa Monica Blvd, Suite 200, Santa Monica, CA 90404, email: [email protected] Neil O’Brien, PhD UCLA Translational Oncology, 2825 Santa Monica Blvd, Suite 200, Santa Monica, CA 90404, email: [email protected]

Key Words: Breast cancer; Abemaciclib; LY2835219; Rb; HER2

Conflicts of Interest: Richard Beckmann, Ann Mc Nulty, Christophe Marchal, Colleen Mockbee are employees of Eli Lilly and Company. Dennis J. Slamon is a consultant/advisory board member of Eli Lilly and Company and Sara Hurvitz has received research/travel funding from Eli Lilly and Company. The remaining authors declare no conflict of interest.

Downloaded from mct.aacrjournals.org on October 2, 2021. © 2018 American Association for Cancer Research. Author Manuscript Published OnlineFirst on February 26, 2018; DOI: 10.1158/1535-7163.MCT-17-0290 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Abstract The cyclinD:CDK4/6:Rb axis is dysregulated in a variety of human cancers. Targeting this pathway has proven to be a successful therapeutic approach in ER+ breast cancer. In this study, in vitro and in vivo preclinical breast cancer models were used to investigate the expanded use of the CDK4/6 inhibitor, abemaciclib. Using a panel of 44 breast cancer cell lines, differential sensitivity to abemaciclib was observed and was seen predominately in the luminal ER+/HER2- and ER+/HER2+ subtypes. However, a subset of triple negative breast cancer (TNBC) cell lines with intact Rb-signaling were also found to be responsive. Equivalent levels of tumor growth inhibition were observed in ER+/HER2-, ER+/HER2+ as well as biomarker selected TNBC xenografts in response to abemaciclib. In addition, abemaciclib combined with hormonal blockade and/or HER2-targeted therapy induced significantly improved anti-tumor activity. CDK4/6 inhibition with abemaciclib combined with anti-mitotic agents, both in vitro and in vivo, did not antagonize the effect of either agent. Finally, we identified a set of Rb/E2F-regulated genes that consistently track with growth inhibitory response and constitute potential pharmacodynamic-biomarkers of response to abemaciclib. Taken together, these data represent a comprehensive analysis of the preclinical activity of abemaciclib, used alone or in combination, in human breast cancer models. The subtypes most likely to respond to abemaciclib-based therapies can be identified by measurement of a specific set of biomarkers associated with increased dependency on CyclinD:CDK4/6:Rb signaling. These data support the clinical development of abemaciclib as mono-therapy or as a combination partner in selected ER+/HER2-, HER2+/ER+ and TNBCs.

Introduction Advances in our understanding of the molecular diversity of human breast cancer have led to the development of specific, molecularly targeted therapies with subsequent improvement in clinical outcomes for patients with hormone receptor positive (ER+) and human epidermal growth factor receptor-2 positive (HER2+) disease (1-5). Despite these advances, a number of patients with these subtypes who receive appropriate therapies demonstrate either de novo or acquired resistance, particularly in the metastatic setting (6, 7). Investigations into the mechanisms underlying therapeutic resistance have focused on dysregulation of key signaling pathways downstream of ER and/or HER2-alterations, including alterations in PI3K/AKT/mTOR pathway and more recently the cyclinD:CDK4/6:Rb:p16 axis (8-12). The cyclin dependent kinases 4 and 6 (CDK4/6) are key regulators of the G1 to S-phase restriction checkpoint and pharmacologically targeting these proteins in combination with hormonal blockade has been shown to provide significant therapeutic benefit to patients with advanced ER+/HER2- breast cancer (13-15). In ER-dependent tumors, activated ER-signaling leads to increased transcription and synthesis of cyclin D1, promoting formation of activating complexes with CDK4/6 (16). Phosphorylation of the retinoblastoma protein (Rb) by the activated CDK4/6:cyclin D1 complex removes the sequestration of E2F by Rb, allowing transcription of factors that promote progression of the cell cycle through the G1-S restriction point (17). Mitogen activated CDK4/6:Rb signaling can also occur independent of estradiol-mediated ER-signaling. It is known that activated HER2-signaling leads to increased transcription of Cyclin D1 via activation of the PI3K and MAPK signaling pathways (18). Mitogenic control of the CDK4/6:Rb axis is frequently lost through a series of molecular alterations known to occur to a greater or lesser degree, in all subtypes of breast cancer (19, 20). These alterations are often complex and multifactorial, making it difficult to predict which disease subtypes will respond to CDK4/6 inhibition in the absence of biologic data.

Prior preclinical data from our laboratory demonstrated that a majority of luminal ER+ breast cancer cell lines are sensitive both in vitro and in vivo to the CDK4/6 inhibitor, palbociclib, and that a combination of this drug with hormonal blockade is therapeutically synergistic in this subtype. These data led to the clinical development and recent approval of palbociclib for use in ER+/HER2- metastatic breast cancer (13, 21-23) and subsequently, a second CDK4/6 inhibitor, ribociclib in the same breast cancer subtype (15). Abemaciclib (LY2835219; Eli-Lilly) is also an orally available ATP-competitive inhibitor of CDK4 and CDK6, with single digit nanomolar potency against both kinases. Based on early monotherapy clinical data (24) abemaciclib has been given “Breakthrough Therapy” status from the US FDA. This drug has now completed phase II and III clinical development in advanced ER+/HER2- breast cancer resulting in its recent regulatory approval for ER+/HER2- breast cancer either alone or in combination with hormonal blockade (25, 26). In the current study, we wanted to further evaluate the preclinical activity of abemaciclib in cell lines representing each of the major therapeutic molecular subtypes of breast cancer and compare it to other approved drugs in this class. Abemaciclib is known to have increased selectivity for CDK4 over CDK6 as well as some inhibitory activity at other kinases such as CDK9, PIM1, HIPK2 and DYRK2 (27). In addition, we performed genomic and proteomic biomarker analyses in order to identify biomarkers and/or molecular profiles that correlate with sensitivity or resistance to this molecule. Finally we interrogated the potential for therapeutically beneficial dual treatment regimens combining abemaciclib with cytotoxic therapies that are currently approved and/or commonly used in the management of breast cancer. These data provide insight into biomarkers of response, other molecular subtypes of breast cancer that may respond to CDK4/6 therapy and additional potential combination strategies that may be appropriate for clinical development.

Materials and Methods Cell lines, cell culture and reagents The growth inhibitory activity of abemaciclib (LY2835219-mesylate salt, provided by Richard Beckmann at Eli-Lilly), palbociclib (PD-0332991-HCL; Selleck Chemicals) and ribociclib (NVP-LEE011-succinate; MedChem Express) were assessed using a panel of 44 molecularly characterized human breast cancer cell lines representing the known therapeutic subtypes of the disease (22). Cells were cultured in appropriate culture media (e.g RPMI 1640, DMEM, L-15) supplemented with 10-15% heat-inactivated fetal bovine serum (FBS), 2mM glutamine, and 1% penicillin G-streptomycin-fungizone solution (PSF, Irvine Scientific) as previously described (28). Cells were routinely assessed for mycoplasma contamination using a multiplex PCR method and STR profiling by the GenePrint® 10 System (Promega) was used to for cell line authentication. Trastuzumab-resistant (BT-474-TR) cells were established as previously described (12, 29).

In vitro proliferation assays Cells were seeded into 24-well plates in the relevant culture media in duplicate at 5,000 to 50,000 cells per well, as described previously (22) and the following day 10 mol/L of abemaciclib, palbociclib or ribociclib with 2-fold serial dilutions over 6 to 12 concentrations was added to generate dose–response curves

for IC50 determination (Supplementary Procedures). Drug combination studies using docetaxel (Taxotere®, Hospira) and Carboplatin (Hospira) were performed as outlined in the Supplementary Procedures.

Molecular characterization of the breast cancer cell panel To identify potential predictive mutational biomarkers, whole exome sequencing was performed using Agilent SureSelect hybrid capture technology paired with Illumina HiSeq sequencing according with manufacturer’s protocols (Supplementary Table S1). The deviations from consensus “normal” sequences in

these data were further filtered to enrich for mutations that are most likely to be somatic alterations with functional consequence in cancer using multiple published resources, most notably the COSMIC catalog of somatic mutations in cancer (http://www.sanger.ac.uk/science/tools/cosmic). DNA copy number alterations were determined using comparative genomic hybridization microarray assays (a- CGH) with the Agilent 105K oligonucleotide CGH chip according with

manufacturer’s protocols. Known oncogenes with log2 ratios >1 (2-fold) were

considered ampliﬁed and known tumor suppressor genes log2 ratios <0.8 were considered homozygous deletions. Baseline total and phosphoprotein levels of a list of >280 protein analytes enriched for proteins currently known to be involved in cancer biology, were determined using reverse phase protein array (RPPA) from the core service at the MD Anderson Cancer Center. Cell preparation and analysis were performed in accordance with MD Anderson published protocols. Data are presented here as log2 (intensity) values. Molecular markers commonly tested in breast cancer, such as ER and HER2, and several biomarkers that have previously been hypothesized to play a role in response to CDK4/6 inhibition were examined for association with response to abemaciclib (listed in Supplementary Table S2).

In vivo efficacy studies Xenograft models of ER+/HER2-, HER2+/ER+ and TNBC breast cancer cell lines were established in six-week-old CD-1 athymic nude mice (Charles River Laboratories) as described in Supplementary Materials. For ER+/HER2- and HER2+/ER- studies, 17-ß-estradiol 60-day release pellets (Innovative Research of America) were implanted subcutaneously into the left flank 7 days before tumor inoculation. For in vivo studies, Fulvestrant (Faslodex®, AstraZenica), trastuzumab (Herceptin®, Genentech) and docetaxel (Taxotere®, Hospira) were purchased from the UCLA pharmacy and 4-hydroxytamoxifen was purchased from Sigma-Aldrich. Statistical differences between treatment arms at specific time points were performed using a two-tailed paired Student t-test. Differences between groups were considered statistically significant at p <0.05. All statistics were calculated

using Microsoft Excel. All animal work was carried out under a protocol approved by IACUC and the UCLA Animal Research Committee.

siRNA knockdown of RB1 MDA-231 cells were seeded in 24 well plates at 10,000 cells per well, followed by transfection with 1.25pmol siRNA targeting exon 19 of RB1 gene (Thermofisher Scientific, cat#4390824 - s522), and 1.5uL of lipofectamine (Thermofisher Scientific) in Opti-MEM I Reduced Serum Medium. Silencer Select Negative Control No.1 siRNA (Thermofisher Scientific, cat#4390843) was used as a control in all assays. siRNA knockdown was confirmed by Western blot.

Modaplex RT-PCR assay Total RNA was isolated from homogenized snap-frozen tumor tissue using the RNeasy kit from Qiagen. Specific cDNA was produced from the RNA using SuperScript® VILO™ MasterMix as described by the User’s Guide. mRNA expression of a set of cell cycle targets was quantified using the MODAplex RNA Array multiplex qPCR platform (Qiagen). Data were exported from the MODAplex software and analyzed in Microsoft Excel. Mean expression values for each transcript were generated from at least three replicate tumor samples. Expression values for each target transcript are represented as a ratio of the mean expression in the experimental arm relative to the vehicle control arm of the study. A ‘dose- dependent’ response was defined as a >10% inhibition at low dose (50mg/kg) abemaciclib accompanied by a further increase in inhibition of >10% at higher dose (75mg/kg) of abemaciclib.

Results Activity of abemaciclib in the breast cancer cell line panel The anti-proliferative activity of abemaciclib was assessed in a panel of 44 breast cancer cell lines, representing the known histological subtypes of the disease (see Supplementary Table S2 for full details). The cell lines were differentially

responsive to abemaciclib over a wide concentration range, with IC50s varying between 0.012M–3.73M (Figure 1A). On mechanism activity of abemaciclib was confirmed by a loss of Rb phosphorylation in response to treatment (Figure 1B) followed by induction of G1-cell cycle arrest (Supplemental Figure S1). No loss of pRb or induction of G1-cell cycle arrest was observed in the Rb-deficient MDA-468 cell line (Figure 1B, Supplemental Figure S1). ER+/HER2- and HER2-amplified breast cancer cell lines that express high levels (>median) of ER protein (e.g. HER2+/ER+) (Figure 1A, and Supplementary Table S2) were among the most sensitive to abemaciclib. In addition, despite relatively low levels of ER and HER2, a subset of TNBC cell lines also responded to

abemaciclib at IC50s below 500nM (Figure 1A, Table 1). In general, sensitivity to abemaciclib was associated with high ER levels, Rb wild-type status, high Rb total and phospho protein levels, lack of cyclin E-amplification (normal copy number), low cyclin E protein, p16 deletion and/or low p16 protein levels (Supplementary Table S1). In the same panel of cell lines, strong correlations were identified between response to abemaciclib and two other CDK4/6 inhibitors, palbociclib and ribociclib (Figure 1C-E). Hormone receptor positive cell lines (either HER2 normal or amplified) were commonly sensitive to all three molecules. In comparative analyses abemaciclib was the most potent of the three molecules tested, followed by palbociclib and then ribociclib. In biomarker-positive breast cancer cell lines, the

average (geometric mean) IC50 of abemaciclib was 168nM, compared to 306nm for palbociclib and 913nM for ribociclib.

Abemaciclib in combination with hormone blockade in ER+/HER2- breast cancer cell xenografts Combined activity of abemaciclib and hormone blockade was confirmed in two ER+ breast cancer cell lines xenografts (Figure 2 and Supplementary Figure S2). Significant tumor regressions were observed with single agent abemaciclib in MCF- 7 (ER+/HER2-) xenografts, and combination with either tamoxifen or fulvestrant

induced a marginal increased benefit in anti-tumor response (Figure 2A&B and Supplementary Table S3). However, within the first week of drug-withdrawal, tumors began to progress in the mice treated with 50mg/kg of abemaciclib alone as well in the mice treated with either single agent tamoxifen or fulvestrant. Conversely, sustained inhibition of tumor growth was observed for over 6 weeks post drug-withdrawal in the majority of animals treated with either drug combination or with high dose abemaciclib monotherapy at 75mg/kg (Figure 2C). Molecular analysis of xenograft tissue at day 4 of dosing revealed a greater loss of pRb in response to the combination vs single agent treatments (Figure 2D). This was accompanied by a reduction in mitosis as indicated by a decrease in FOXM1 protein levels (Figure 2D). Western blot analysis of residual xenograft tissue collected at the end of a recovery phase of 6 to 7 weeks revealed that active cell division and Rb- signaling were restored in the 50mg/kg abemaciclib and single agent endocrine treatments, as indicated by the recovery of FOXM1 and pRb protein levels (Figure 2E). In contrast, FOXM1/pRb signal remained low in xenograft tissues collected from combination arms at the same time point (Figure 2E). To investigate if the sustained reduction in Rb/FOXM1 signal in the combination arms was due to induction of permanent growth arrest/senescence or induction of cell death, we measured levels of the human epithelial marker protein keratin-19 (CK-19), a structural protein not directly regulated by abemaciclib treatment (Supplementary Figure S3). In the MCF-7 xenograft tissues, CK-19 protein levels were significantly reduced or lost in the combination treatment arms, particularly in abemaciclib plus tamoxifen treated mice, indicating loss of human epithelial tumor cells (Figure 2E). Using a ß-galactosidase staining assay, in vitro assays of the MCF-7 cells showed increased induction of senescence in response to the combination as opposed to the single agent (Figure 2F).

Activity of abemaciclib in HER2+/ER+ breast cancer cell lines and xenografts Activity of abemaciclib single agent or combination with HER2 and ER- targeted therapy was assessed in two xenograft models of HER2+/ER+ breast cancer, parental BT474 cells and BT474 cells conditioned through long-term drug

exposure to progress on trastuzumab therapy (BT-474-TR (12)). Single agent abemaciclib induced significant tumor growth inhibition (TGI) in both models (BT- 474; p=0.028, BT-474-TR; p<0.001) (Figure 3A-C & Supplementary Table S4). The combination of abemaciclib with trastuzumab induced significantly improved (p=0.0012) TGIs and tumor regressions in xenografts progressing on trastuzumab alone (Figure 3B & 3C, Supplemental Table S4). The triple combination of abemaciclib, trastuzumab and tamoxifen further improved tumor regressions in both models (Figure 3A-C & Supplementary Table S4). In vitro studies confirmed that triple blockade of CDK4/6, HER2 and ER signaling leads to a greater induction of cell death in both trastuzumab sensitive and resistant HER2+/ER+ breast cancer cell lines (Figure 3D). Abemaciclib was also shown to combine effectively with docetaxel in these HER2-amplified breast cancer cell line xenografts. The addition of docetaxel to abemaciclib did not inhibit the TGI induced by abemaciclib or docetaxel alone. Moreover, the combination of abemaciclib, trastuzumab, tamoxifen and docetaxel was the most efficacious arm in both studies (Figure 3A-C & Supplementary Table S4). Marginal and reversible body weight loss was observed in the mice treated with docetaxel containing arms over the four weeks treatment period (Supplementary Table S4). Western blot analysis of xenograft tissues collected after 4 days of treatment confirmed a dose dependent loss of pRb, TOPOII and pHH3 in mice treated with abemaciclib (Figure 3E). Further reduction in Rb-signaling was observed in xenografts treated with abemaciclib plus trastuzumab or trastuzumab plus tamoxifen. The addition of docetaxel did not block the pRb knockdown induced by abemaciclib or the combination of abemaciclib with trastuzumab plus tamoxifen (Figure 3E).

Identification of a subset of triple TNBC cells sensitive to abemaciclib In TNBC cell lines, where cell growth is independent of ER-status, response to abemaciclib appears to remain dependent on intact Rb-signaling. Those TNBC cell lines that have high baseline levels of total and phosphorylated pRb, accompanied

by low levels of p16 protein are among the most sensitive to abemaciclib (Table 1). Cell lines with copy number gains or amplification of the CCNE1 (Cyclin E1) gene are clearly less responsive (Table 1). Xenografts of TNBC cell lines expressing sensitive and resistant biomarker profiles were measured for response to abemaciclib single agent and combination with docetaxel, a standard of care agent used for TNBC. Consistent with all the preceding data, induction of tumor regressions or stable disease was observed only in the MDA-231 (Figure 4 & Supplementary Table S5) & BT-20 xenografts (Supplementary Figure S4) that have high levels of pRb and low levels of p16 at baseline (Figure 4C. left panels). Conversely, HCC70 xenografts with low levels of pRb and high p16, continued to progress through abemaciclib monotherapy (Figure 4B). On-target activity of abemaciclib was confirmed in the sensitive MDA-231 xenografts by reduction in total and phosphorylated Rb and the cell cycle progression marker, FOXM1, in response to treatment (Figure 4C, right panels). Consistent with our observations in HER2+/ER+ breast cancer xenografts, combination of abemaciclib with docetaxel did not antagonize the activity of either single agent (Figure 4A-B). Finally, the role of pRb in predicting response to abemaciclib in TNBC was confirmed by siRNA knockdown of RB1 gene expression in the MDA-231 cells. Subsequent reduction of baseline pRb levels significantly reduced the sensitivity of the MDA-231 cells to abemaciclib (Figure 4D).

Combined activity of abemaciclib and cytotoxic chemotherapy The potential of abemaciclib to be used in combination with cytotoxic chemotherapy was further investigated in a cell line model of TNBC in vitro. In order to determine if the lack of antagonism observed in our in vivo models could be attributed to an effect of the timing of drug administration, we investigated dosing strategies comparing co-administration versus sequencing of these treatments. Simultaneous treatment of MDA-231 cells with abemaciclib plus docetaxel or carboplatin resulted in increased inhibition of cell proliferation compared to single agent treatments (Figure 5A). Concentrations of abemaciclib below 10nM induced profound (>60%) inhibition of cell proliferation when combined with either anti-

mitotic agent. Pre-treatment of cells with abemaciclib for 2 days reduced cell proliferation rate via induction of G1-cell cycle arrest (Supplemental Figure S5A) without impacting the activity of docetaxel or carboplatin relative to cells pre- treated with vehicle control (Figure 5B and 5C). Pre-treatment with abemaciclib for 24 hours also induced G1-cell cycle arrest (Supplement Figure S5B) without blocking apoptosis induced by high-dose docetaxel treatment. In contrast, a dose dependent decrease in apoptosis induction by docetaxel was observed in response to pre-treatment with palbociclib or ribociclib (Figure 5D). Co-administration of docetaxel with either palbociclib or ribociclib after CDK4/6-inhibitor pre-treatment blocked apoptosis induction, whereas co-treatment with abemaciclib did not (Figure 5E).

Identification of a pharmacodynamic signature of response to abemaciclib treatment To identify potential pharmacodynamic (PD) markers of response to abemaciclib, we used a Modaplex PCR-based platform to measure changes in the expression of a set of 22 cell cycle associated genes in pre- and post-treatment (day 4) samples from tumor tissues responding to abemaciclib-based therapy. In MCF-7 ER+/HER2- xenografts, a range of mRNA expression changes were induced by treatment with abemaciclib, hormonal blockade or a combination of the two (Figure 6A). A subset of 10 transcripts showed a dose dependent reduction in expression in response to low and high-dose abemaciclib. Single agent treatment with either fulvestrant or tamoxifen had little effect on the expression of these 10 candidate PD- biomarkers. However, combination of low dose abemaciclib (50mg/kg) with either fulvestrant or tamoxifen increased the inhibition of the candidate gene set compared to inhibition induced by high dose abemaciclib (75mg/kg) (Figure 6B). Using the same criteria, an abemaciclib dose-dependent gene signature was identified from ER+/HER2+ xenografts. Similar to the ER+/HER2- models, expression of these transcripts was similarly regulated by CDK4/6-inhibition in combination with either trastuzumab or tamoxifen or both (Figure 6C). Transcripts for RRM2, TOPO2A, MKI67, MCM7, CDK2 and CDK4 were all found to be regulated by

abemaciclib in a dose dependent manner in sensitive tumors from MCF-7 (ER+/HER2-), ZR-75-1 (ER+/HER2-) and BT-474-TR (HER2+/ER+) xenografts (Figure 6D). Expression of five of these transcripts, RRM2, TOPO2A, MKI67, MCM7 and CDK2 are directly regulated by the E2F-transcription factor immediately downstream of Rb, indicating a direct on-target regulation of these genes by CDK4/6 inhibition using abemaciclib.

Discussion The cyclinD:CDK4/6:Rb:p16 signaling axis is frequently dysregulated in cancer. Efforts to pharmacologically target this pathway have been validated by the approval of the specific CDK4/6 inhibitor, palbociclib (PD-0332991, Pfizer), ribociclib (LEE-011, Novartis) and most recently abemaciclib (Eli Lilly), in combination with endocrine-based therapies in advanced ER+/HER2- breast cancer (13, 15, 21-23, 25, 26). Abemaciclib is structurally and biologically distinct from palbociclib and ribociclib with greater selectivity for CDK4 over CDK6 (27) and clinically, abemaciclib appears to have superior single agent activity when compared to the other approved CDK4/6 inhibitors (24, 25, 30, 31). It is possible that the CDK4 selectivity of abemaciclib contributes to the improvement in activity by reducing the degree of neutropenia usually associated with inhibition of CDK6, which in turn avoids the need for an interruption in dosing associated with both palbociclib and ribociclib. This difference may improve the chances of forcing tumor cells into permanent growth arrest and ultimately senescence (24). In this study, we present a comprehensive analysis of the preclinical activity of abemaciclib using both in vitro and in vivo preclinical models spanning the known molecular subtypes of breast cancer. Using pre- and post-treatment data, we have identified a set of consistent biomarkers of sensitivity to abemaciclib in ER+, HER2+ and TNBC subtypes. Measurement of the growth inhibitory activity of abemaciclib across a panel of 44 human breast cancer cell lines identified the ER+/HER2- subtype as most sensitive to CDK4/6 inhibition. Despite the wide therapeutic range of this molecule,

each of the 9 cell lines classified as ER+/HER2- had IC50s below 200nM, making further stratification of response biomarkers within this subtype unnecessary. This activity was confirmed in two xenograft models of ER+/HER2- breast cancer by the induction of complete arrest of tumor growth using abemaciclib monotherapy. Inhibition of tumor proliferation was accompanied by dose dependent decreases of phosphorylated Rb and induction of cell cycle arrest as measured by loss of the cell

cycle progression markers, TOPOIIα (S-phase), phosphohistone-H3 (G2-M) and FOXM1 (cellular senescence). At the tested doses, abemaciclib did not inhibit CDK-9 signaling, consistent with previous data reported for this molecule (27, 32). Comparison of the activity of abemaciclib, palbociclib and ribociclib across the panel of 44 breast cancer cell lines identified a strong correlation within and between the major histological subtypes of breast cancer for this class of molecule. However, of the three, abemaciclib appeared to be the most potent using our assays. Our studies also confirm that abemaciclib has additive and/or synergistic activity in combination with endocrine-based therapies in cell line xenograft models of ER+/HER2- breast cancer. Continuous daily treatment using clinically achievable doses of abemaciclib (50mg/kg) in combination with tamoxifen or fulvestrant was well tolerated and led to down-regulation of Rb-signaling and significantly improved anti-tumor responses compared to results with either agent alone. While tumors from single agent treated animals began to proliferate within days of treatment cessation, the majority of xenografts from mice treated with combination therapy showed no signs tumor progression, despite withdrawal of treatment for more than 6 weeks. Xenografts collected during the dosing phase of the combination studies indicate an initial induction of cellular senescence as measured by loss of FOXM1 and in vitro assays showing a greater induction of senescence with the combination treatment. Xenograft material collected at end of study, indicate a loss of human epithelial tumor cell population. The marked tumor regressions observed in the combination arms provide further evidence of induction of tumor cell death induced by this regimen. The data reported here are consistent with those recently reported from the MONARCH-2 clinical trial (NCT02107703), that showed significant improvements in progression free survival (PFS) and objective response

rates (ORR) in patients with advanced breast cancer treated with abemaciclib plus fulvestrant versus fulvestrant alone (26). Abemaciclib sensitive cell lines were also identified within the HER2- amplified subgroup. There is considerable preclinical evidence to support targeting CDK4/6 in HER2-amplified breast cancers given that HER2-signaling can drive cell cycle progression through activation of CyclinD1:CDK4/6 signaling (18, 33). Cyclin D1 is required for the formation of HER2-initiated tumors in mouse models and pharmacologically targeting CDK4/6 blocks tumor formation in mice (34, 35). In addition, preclinical studies show that targeting CDK4/6 overcomes resistance to HER2-directed therapy both in vitro and in vivo (9, 36). The data presented here provide further insight into the specific subtypes of HER2-amplified breast cancers most likely to benefit from CDK4/6 targeted therapy. Analysis of abemaciclib response within the panel of 17 HER2-amplified breast cancer cell lines indicate that HER2-amplified cell lines with higher levels of ER protein accompanied by intact downstream Rb-signaling (high total Rb and pRb and no Cyclin E amplification) are most sensitive to abemaciclib treatment. These data indicate that HER2-amplified breast cancer with higher levels of ER (i.e., HER2+/ER+) may be more susceptible to CDK4/6 intervention than those that are HER2+/ER-. The current data demonstrate that single agent abemaciclib induces significant tumor growth inhibition as well as dose dependent inhibition of pRb-signaling and cell cycle arrest in HER2+/ER+ xenografts. In addition, abemaciclib combined effectively with trastuzumab and in triple combination with tamoxifen, induced significant regressions in both trastuzumab-sensitive and resistant HER2+/ER+ xenografts. In vitro studies confirmed that the triple combination of targeting CDK4/6, HER2 and ER leads to a greater induction of cell death in cells when compared to single agent treatment. The specific mechanism by which this triple combination is effective remains to be determined. Finally, abemaciclib activity in HER2-amplified tumors progressing on trastuzumab provides encouragement that targeting CDK4/6 may be efficacious in treatment refractory HER2-amplified disease. These data also support the hypothesis that ER and downstream Rb-signaling are, at least partially, driving the progression of HER2+/ER+ tumors. Clinical data suggest that HER2+/ER+ and

HER2+/ER- breast cancers are clinically distinct subgroups based on their prognosis and response to therapy (37, 38). Furthermore, preclinical studies show that cross talk exists between ER and HER2 signaling and activation of either pathway is associated with resistance to targeting the other pathway (39-41). These data support the design of the MonarcHER trial (NCT02675231), which will evaluate the combination of abemaciclib plus trastuzumab with or without hormonal blockade (fulvestrant) in HER2+/ER+ breast cancer patients. The use of CDK4/6 inhibitors beyond ER+/HER2- breast cancer also requires investigation of combinations with standard-of-care cytotoxic chemotherapies. There are preclinical data to suggest that the mechanism of action of molecules that arrest cell cycle may be antagonistic when used with anti-mitotic agents (35, 42). To address this question, we first investigated the in vivo activity of abemaciclib in combination with chemotherapy (docetaxel) in HER2-amplfied and triple negative breast cancer xenografts. In TNBC, co-treatment with abemaciclib and docetaxel did not antagonize anti-tumor activity of either single agent. Moreover, the addition of docetaxel to a triple (anti-HER2, anti-ER and anti-CDK4/6) combination in HER2+/ER+ xenografts increased the degree of tumor regressions observed. Analysis of tumor tissues confirmed that docetaxel did not block the induction of the cell cycle arrest by abemaciclib. In vitro drug sequencing experiments confirmed that induction of growth arrest by abemaciclib, does not block the inhibition of cell proliferation or the apoptosis induced by docetaxel or carboplatin. In fact, either simultaneous or sequential combination of abemaciclib plus either cytotoxic leads to greater inhibition of tumor cell growth compared to single agent treatment. The combined activity observed between these two classes of molecules could possibly be explained by fact that not all tumor cells will be uniformly arrested by abemaciclib leaving sub-populations of cells sensitive to the mechanism of action of anti-mitotics. It has previously been shown that abemaciclib can be combined with gemcitabine without antagonism in lung cancer xenografts with either sequential or simultaneous administration (27). Our in vitro experiments also show that abemaciclib appears to be unique amongst the CDK4/6 inhibitors tested. In contrast to abemaciclib, both palbociclib and ribociclib pre-treatment significantly reduced

the induction of apoptosis by docetaxel in a dose dependent manner. The specific mechanisms by which abemaciclib differs from palbociclib and ribociclib in this setting requires further investigation, however, these data are encouraging for the development of abemaciclib in indications where combination with cytotoxic chemotherapies might be required. Sensitivity to abemaciclib was also detected in a specific subgroup of TNBC cell lines matching the biomarker profile of intact Rb-signaling, despite expressing low levels of ER and HER2. Of the 17 TNBC cell lines tested, the top 5 most sensitive cell lines had high levels of total and phosphorylated levels of Rb accompanied by

low baseline levels of p16. Abemaciclib IC50s for this subset of TNBC cell lines are similar to those measured for the ER+/HER2- and HER2+/ER+ cell lines. Molecular alterations associated with resistance to CDK4/6 targeted therapy (43, 44) through activation of Rb-signaling independently of CDK4/6, such as high p16 protein levels or amplification of Cyclin E, were found in those TNBC cell lines least sensitive to abemaciclib. Xenograft studies confirmed that TNBC cell lines selected for high baseline levels of pRb and low p16 were sensitive to abemaciclib whereas xenografts with low pRb and high p16 were unresponsive to this therapy. Reduction in pRb levels or RB1 via siRNA knockdown significantly reduced the sensitivity of TNBC cells to abemaciclib. These data support the hypothesis that Rb-dependence goes beyond ER-driven breast cancers and that CDK4/6 inhibition may be effective in other subtypes with the appropriate cyclinD:CDK4/6:Rb activation profile. It is likely that the strong association between ER-status and abemaciclib response in breast cancer models and patients (24) is due to the fact that ER is a useful surrogate marker of downstream dependence on CyclinD:CDK4/6:Rb:p16-signaling. Selection of patients based on measurement of intact Rb pathway signaling as characterized by high tumor levels of pRb, low p16 and an absence of cyclin E1 amplification, may identify additional populations of patients that could benefit from abemaciclib therapy and could provide therapeutic benefit in a subset of TNBCs where there are no current, approved targeted therapies. Of interest, additional factors that associate with response to abemaciclib in this study include the presence of high levels of androgen receptor (AR) and the

presence of PIK3CA activating mutations. It should be noted however that these factors track with the luminal breast cancer subtype and as such may not be independent biomarkers of response (19, 45). High levels of PTEN protein associated with abemaciclib response indicate that intact PI3K-pathway signaling may increase the likelihood of response to CDK4/6 targeted therapy. PD-markers, for the early detection of response to abemaciclib therapy were also identified in this study using a set of cell cycle regulated genes. Expression of a core set of five E2F-regulated genes (RRM2, TOPO2A, MKI67, MCM7 and CDK2), consistently tracked with response to abemaciclib in a dose dependent manner across three independent xenograft studies. Furthermore, the dynamic changes in expression of these transcripts tracked with an increase in anti-tumor response induced by abemaciclib when combined with fulvestrant, tamoxifen or trastuzumab. The ultimate utility of this marker set should now be evaluated in tissues from abemaciclib clinical trials. In this study, we present a comprehensive preclinical profile of abemaciclib used as a single agent or in combination with standard-of-care therapy in each of the known therapeutic molecular subtypes of breast cancer. These data further support the clinical use of abemaciclib as monotherapy as well as a possible combination partner in ER+/HER2-, HER2+/ER+ and some TNBCs. It is likely that beyond ER+/HER2- breast cancer, measurement of an on-mechanism, validated set of biomarkers to confirm dependence on Rb-signaling will likely help better select patients that will benefit from abemaciclib-based therapies.

References 1. Howell A, Osborne CK, Morris C, Wakeling AE. ICI 182,780 (Faslodex): development of a novel, "pure" antiestrogen. Cancer. 2000;89(4):817-25. 2. Osborne CK. Tamoxifen in the treatment of breast cancer. The New England journal of medicine. 1998;339(22):1609-18. 3. Osborne CK. Aromatase inhibitors in relation to other forms of endocrine therapy for breast cancer. Endocrine-related cancer. 1999;6(2):271-6. 4. Slamon DJ, Leyland-Jones B, Shak S, Fuchs H, Paton V, Bajamonde A, et al. Use of chemotherapy plus a monoclonal antibody against HER2 for metastatic breast cancer that overexpresses HER2. The New England journal of medicine. 2001;344(11):783-92. 5. Verma S, Miles D, Gianni L, Krop IE, Welslau M, Baselga J, et al. Trastuzumab emtansine for HER2-positive advanced breast cancer. The New England journal of medicine. 2012;367(19):1783-91. 6. Cuzick J, Sestak I, Baum M, Buzdar A, Howell A, Dowsett M, et al. Effect of anastrozole and tamoxifen as adjuvant treatment for early-stage breast cancer: 10-year analysis of the ATAC trial. The Lancet Oncology. 2010;11(12):1135-41. 7. Vogel CL, Cobleigh MA, Tripathy D, Gutheil JC, Harris LN, Fehrenbacher L, et al. Efficacy and safety of trastuzumab as a single agent in first-line treatment of HER2-overexpressing metastatic breast cancer. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2002;20(3):719-26. 8. Berns K, Horlings HM, Hennessy BT, Madiredjo M, Hijmans EM, Beelen K, et al. A functional genetic approach identifies the PI3K pathway as a major determinant of trastuzumab resistance in breast cancer. Cancer Cell. 2007;12(4):395-402. 9. Goel S, Wang Q, Watt AC, Tolaney SM, Dillon DA, Li W, et al. Overcoming Therapeutic Resistance in HER2-Positive Breast Cancers with CDK4/6 Inhibitors. Cancer Cell. 2016;29(3):255-69. 10. Miller TW, Balko JM, Fox EM, Ghazoui Z, Dunbier A, Anderson H, et al. ERalpha-dependent E2F transcription can mediate resistance to estrogen deprivation in human breast cancer. Cancer discovery. 2011;1(4):338-51. 11. Miller TW, Hennessy BT, Gonzalez-Angulo AM, Fox EM, Mills GB, Chen H, et al. Hyperactivation of phosphatidylinositol-3 kinase promotes escape from hormone dependence in estrogen receptor-positive human breast cancer. J Clin Invest. 2010;120(7):2406-13. 12. O'Brien NA, Browne BC, Chow L, Wang Y, Ginther C, Arboleda J, et al. Activated phosphoinositide 3-kinase/AKT signaling confers resistance to trastuzumab but not lapatinib. Mol Cancer Ther. 2010;9(6):1489-502. 13. Cristofanilli M, Turner NC, Bondarenko I, Ro J, Im SA, Masuda N, et al. Fulvestrant plus palbociclib versus fulvestrant plus placebo for treatment of

hormone-receptor-positive, HER2-negative metastatic breast cancer that progressed on previous endocrine therapy (PALOMA-3): final analysis of the multicentre, double-blind, phase 3 randomised controlled trial. The Lancet Oncology. 2016;17(4):425-39. 14. Finn RS, Martin M, Rugo HS, Jones S, Im SA, Gelmon K, et al. Palbociclib and Letrozole in Advanced Breast Cancer. The New England journal of medicine. 2016;375(20):1925-36. 15. Hortobagyi GN, Stemmer SM, Burris HA, Yap YS, Sonke GS, Paluch-Shimon S, et al. Ribociclib as First-Line Therapy for HR-Positive, Advanced Breast Cancer. The New England journal of medicine. 2016;375(18):1738-48. 16. Prall OW, Sarcevic B, Musgrove EA, Watts CK, Sutherland RL. Estrogen- induced activation of Cdk4 and Cdk2 during G1-S phase progression is accompanied by increased cyclin D1 expression and decreased cyclin-dependent kinase inhibitor association with cyclin E-Cdk2. J Biol Chem. 1997;272(16):10882- 94. 17. Harbour JW, Luo RX, Dei Santi A, Postigo AA, Dean DC. Cdk phosphorylation triggers sequential intramolecular interactions that progressively block Rb functions as cells move through G1. Cell. 1999;98(6):859-69. 18. Timms JF, White SL, O'Hare MJ, Waterfield MD. Effects of ErbB-2 overexpression on mitogenic signalling and cell cycle progression in human breast luminal epithelial cells. Oncogene. 2002;21(43):6573-86. 19. Cancer Genome Atlas N. Comprehensive molecular portraits of human breast tumours. Nature. 2012;490(7418):61-70. 20. Foster SA, Wong DJ, Barrett MT, Galloway DA. Inactivation of p16 in human mammary epithelial cells by CpG island methylation. Mol Cell Biol. 1998;18(4):1793-801. 21. Gelbert LM, Cai S, Lin X, Sanchez-Martinez C, Del Prado M, Lallena MJ, et al. Preclinical characterization of the CDK4/6 inhibitor LY2835219: in-vivo cell cycle-dependent/independent anti-tumor activities alone/in combination with gemcitabine. Invest New Drugs. 2014;32(5):825-37. 22. Patnaik A, Rosen LS, Tolaney SM, Tolcher AW, Goldman JW, Gandhi L, 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 discovery. 2016;6(7):740-53. 23. Finn RS, Crown JP, Ettl J, Schmidt M, Bondarenko IM, Lang I, et al. Efficacy and safety of palbociclib in combination with letrozole as first-line treatment of ER-positive, HER2-negative, advanced breast cancer: expanded analyses of subgroups from the randomized pivotal trial PALOMA-1/TRIO-18. Breast Cancer Res. 2016;18(1):67.

24. Finn RS, Dering J, Conklin D, Kalous O, Cohen DJ, Desai AJ, et al. PD 0332991, a selective cyclin D kinase 4/6 inhibitor, preferentially inhibits proliferation of luminal estrogen receptor-positive human breast cancer cell lines in vitro. Breast Cancer Res. 2009;11(5):R77. 25. Turner NC, Ro J, Andre F, Loi S, Verma S, Iwata H, et al. Palbociclib in Hormone-Receptor-Positive Advanced Breast Cancer. The New England journal of medicine. 2015;373(3):209-19. 26. O'Brien NA, McDonald K, Tong L, von Euw E, Kalous O, Conklin D, et al. Targeting PI3K/mTOR overcomes resistance to HER2-targeted therapy independent of feedback activation of AKT. Clin Cancer Res. 2014;20(13):3507- 20. 27. Konecny GE, Pegram MD, Venkatesan N, Finn R, Yang G, Rahmeh M, et al. Activity of the dual kinase inhibitor lapatinib (GW572016) against HER-2- overexpressing and trastuzumab-treated breast cancer cells. Cancer Res. 2006;66(3):1630-9. 28. Tate SC, Cai S, Ajamie RT, Burke T, Beckmann RP, Chan EM, et al. Semi- mechanistic pharmacokinetic/pharmacodynamic modeling of the antitumor activity of LY2835219, a new cyclin-dependent kinase 4/6 inhibitor, in mice bearing human tumor xenografts. Clin Cancer Res. 2014;20(14):3763-74. 29. Infante JR, Cassier PA, Gerecitano JF, Witteveen PO, Chugh R, Ribrag V, et al. A Phase I Study of the Cyclin-Dependent Kinase 4/6 Inhibitor Ribociclib (LEE011) in Patients with Advanced Solid Tumors and Lymphomas. Clin Cancer Res. 2016. 30. DeMichele A, Clark AS, Tan KS, Heitjan DF, Gramlich K, Gallagher M, et al. CDK 4/6 inhibitor palbociclib (PD0332991) in Rb+ advanced breast cancer: phase II activity, safety, and predictive biomarker assessment. Clin Cancer Res. 2015;21(5):995-1001. 31. Dickler MN TS, Rugo HS, Cortes J, Diéras V, Patt DA, Wildiers H, Frenzel M, Koustenis A, Baselga J. Results from a phase II study of abemaciclib, a CDK4 and CDK6 inhibitor, as monotherapy, in patients with HR+/HER2- breast cancer, after chemotherapy for advanced disease. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2016;34(suppl):abstract 510. 32. Sledge GW, Jr., Toi M, Neven P, Sohn J, Inoue K, Pivot X, et al. MONARCH 2: Abemaciclib in Combination With Fulvestrant in Women With HR+/HER2- Advanced Breast Cancer Who Had Progressed While Receiving Endocrine Therapy. Journal of clinical oncology : official journal of the American Society of Clinical Oncology. 2017;35(25):2875-84. 33. Lenferink AEG, Busse D, Flanagan WM, Yakes FM, Arteaga CL. ErbB2/neu kinase modulates cellular p27(Kip1) and cyclin D1 through multiple signaling pathways. Cancer Res. 2001;61(17):6583-91.

34. Choi YJ, Li XY, Hydbring P, Sanda T, Stefano J, Christie AL, et al. The Requirement for Cyclin D Function in Tumor Maintenance. Cancer Cell. 2012;22(4):438-51. 35. Roberts PJ, Bisi JE, Strum JC, Combest AJ, Darr DB, Usary JE, et al. Multiple Roles of Cyclin-Dependent Kinase 4/6 Inhibitors in Cancer Therapy. J Natl Cancer I. 2012;104(6):476-87. 36. Witkiewicz AK, Cox D, Knudsen ES. CDK4/6 inhibition provides a potent adjunct to Her2-targeted therapies in preclinical breast cancer models. Genes Cancer. 2014;5(7-8):261-72. 37. Hurvitz SA, Andre F, Jiang Z, Shao Z, Mano MS, Neciosup SP, et al. Combination of everolimus with trastuzumab plus paclitaxel as first-line treatment for patients with HER2-positive advanced breast cancer (BOLERO-1): a phase 3, randomised, double-blind, multicentre trial. The Lancet Oncology. 2015;16(7):816-29. 38. Nahta R, O'Regan RM. Therapeutic implications of estrogen receptor signaling in HER2-positive breast cancers. Breast Cancer Res Treat. 2012;135(1):39-48. 39. Gluck S, Arteaga CL, Osborne CK. Optimizing chemotherapy-free survival for the ER/HER2-positive metastatic breast cancer patient. Clin Cancer Res. 2011;17(17):5559-61. 40. Osborne CK, Shou J, Massarweh S, Schiff R. Crosstalk between estrogen receptor and growth factor receptor pathways as a cause for endocrine therapy resistance in breast cancer. Clin Cancer Res. 2005;11(2 Pt 2):865s-70s. 41. Xia W, Bacus S, Hegde P, Husain I, Strum J, Liu L, et al. A model of acquired autoresistance to a potent ErbB2 tyrosine kinase inhibitor and a therapeutic strategy to prevent its onset in breast cancer. Proc Natl Acad Sci U S A. 2006;103(20):7795-800. 42. Dean JL, McClendon AK, Knudsen ES. Modification of the DNA damage response by therapeutic CDK4/6 inhibition. J Biol Chem. 2012;287(34):29075-87. 43. Konecny GE, Winterhoff B, Kolarova T, Qi J, Manivong K, Dering J, et al. Expression of p16 and retinoblastoma determines response to CDK4/6 inhibition in ovarian cancer. Clin Cancer Res. 2011;17(6):1591-602. 44. Lukas J, Herzinger T, Hansen K, Moroni MC, Resnitzky D, Helin K, et al. Cyclin E-induced S phase without activation of the pRb/E2F pathway. Genes Dev. 1997;11(11):1479-92. 45. Park S, Koo J, Park HS, Kim JH, Choi SY, Lee JH, et al. Expression of androgen receptors in primary breast cancer. Ann Oncol. 2010;21(3):488-92.

Table 1: Biomarkers of response to abemaciclib triple negative breast cancer cell lines RB1 Cell Line IC50 μM ER HER2 PIK3CA PMs RB1 PMs Rb RbpS807S811 CDKN2A PMs CDKN2A CN p16 CCNE1 CN CyclinE1 CN HCC1395 0.310 -2.16 -0.1 Wt Wt GAIN 0.44 3.15 Wt HD -3.01 GAIN 0.27 BT-20 0.328 -1.61 1.52 p.Pro539Arg Wt NC 0.56 2.98 Wt HD -2.81 NC -0.33 CAL-51 0.407 -2.1 1.62 p.Glu542Lys Wt NC 0.89 3.84 Wt NC -2.15 NC 0.07 MDA-MB-231 0.443 -2.26 0.79 Wt Wt NC 0.65 3.43 Wt HD -2.73 GAIN -0.17 HCC1143 0.528 -1.15 0.97 Wt Wt NC 0.36 2.87 Wt LOH -2.55 NC -0.17 COLO-824 0.903 -1.7 1.56 Wt Wt HD -0.31 1.15 Wt GAIN -0.3 NC 0.48 HCC1806 0.904 -1.76 0.68 Wt Wt NC 0.35 3.07 Wt HD -2.96 HIGH AMP 1.98 Hs578T 0.968 -2.18 0.86 Wt Wt NC 0.74 3.56 Wt NC -1.39 NC -0.31 HCC70 0.971 -1.96 1.74 Wt Wt NC -0.08 0.8 Wt AMP -0.31 GAIN 0.21 MDA-MB-436 1.050 -1.98 0.66 Wt Wt GAIN -0.26 0.24 Wt NC -0.69 NC 0.83 MDA-MB-468 1.090 -1.95 -0.53 Wt Wt HD -0.41 1.52 Wt NC -1.02 AMP 0.27 HCC1937 1.180 -1.22 1.08 Wt c.2221*>-G LOH -0.32 1.13 Wt NC -0.51 NC 0.09 MDA-MB-157 1.520 -2.1 0.63 Wt Wt LOH 0.11 1.87 Wt NC -0.23 AMP 1.02 HCC38 2.280 -2.47 2.42 Wt Wt NC 1.48 3.78 Wt NC -3.29 NC -0.58 HCC1187 2.300 -1.92 2.37 Wt Wt NC -0.28 1.37 Wt NC -0.03 GAIN 1.09 DU4475 3.290 -2.17 -0.4 Wt Wt HD 0.28 3.4 Wt NC -2.21 NC 1.54 BT-549 3.730 -2.36 0.91 Wt Wt HD -0.54 0.43 Wt GAIN 0.64 NC 0.23

Figure Legends

Figure 1: Activity of abemaciclib in breast cancer cell lines. A, in vitro IC50s (generational inhibition) for each of the breast cell lines. Data represent mean IC50 +/- 95% confidence interval where available. B, effect of single dose (200 nM) abemaciclib treatment on Rb-signaling, cell lysates were prepared 5 days post- dosing. C, Scattergram comparing the in vitro IC50s of abemaciclib vs palbociclib in matching breast cancer cell lines, D, abemaciclib vs ribociclib, E, palbociclib vs ribociclib, for scattergrams, cell line sub-types are color coded consistent with Figure 1A. R values were calculated in Excel. All experiments were repeated in at least duplicate.

Figure 2: Induction of sustained tumor growth inhibition by abemaciclib plus hormone blockade. A, growth curves for MCF-7 (ER+) breast cancer cell line (BCL) xenografts treated with either single abemaciclib or combination with hormonal blockade (fulvestrant or tamoxifen), 8 mice per arm. For combination arms, 50 mg/kg abemaciclib was used. Data represent mean tumor volume +/- SEM. B, Waterfall plots representing the % progression or tumor regression in each individual tumor within each treatment arm, ≤ 20% progression was considered stable disease, ≤ 50% tumor regression was considered a partial response and 100% regression a complete response. Fulv = Fulvestrant, Tam = Tamoxifen. C, growth curves for individual xenograft tumors post treatment withdrawal relative to mean tumor volumes for the vehicle control. D, Western blot analysis of snap- frozen tumor tissue taken from mice 24 hours after the 3 days of continuous dosing. E, Western blot of representative tumor lysates taken > 6 weeks post drug withdrawal. F, representative images of cells stained for β-galactosidase six days post treatment with the molecules indicated. *Indicates statistically significant difference from vehicle control, p < 0.05.

Figure 3: Activity of abemaciclib in combination with HER2 and ER targeted therapies in HER2+/ER+ breast cancer. A, B, growth curves for BT-474 and trastuzumab conditioned BT-474-TR cell line xenografts treated with either single agent abemaciclib (‘Abe’) or combination with HER2-targeting (trastuzumab, ‘Tz’), ER-targeting (tamoxifen, 'tam') or SOC chemotherapy (docetaxel, ‘Dtx’), 8 mice per arm. Fifty mg/kg abemaciclib was used in combination arms. C, Change in tumor volume from baseline for each experimental arm. D, Induction of apoptosis in cell line in 2D cultures treated with combination CDK4/6, HER2/ER combination therapy. Histograms represent % Annexin V positive (AN V+) cells after 5 days of treatment. Data represent mean +/- SD. All experiments were repeated in at least duplicate. E, lysates from snap frozen BT474-TR tumor tissue, collected after 4 days of dosing, were subject to immunoblotting with the antibodies indicated. *Indicates statistically significant difference compared to vehicle control, #indicates statistically significant difference compared to single agent trastuzumab, p < 0.05.

Figure 4: Abemaciclib has activity in xenograft models of triple negative breast cancer. A, B, growth curves for triple negative breast cancer cell line xenografts treated with either single abemaciclib (‘Abe’) or combination with SOC chemotherapy (docetaxel ‘dtxl’), 8 mice per arm. Fifty mg/kg abemaciclib was used in combination arms. C, Left panels, Western blot analysis of snap frozen tumor tissue taken from baseline vehicle control arms of each xenograft study. Right panels, Western blot analysis of snap frozen tumor tissue, collected from mice in each arm of the MDA-231 study, 24 hours after the 3 days of continuous dosing. D, siRNA knockdown of Rb1 confirms Rb1 signaling is required for abemaciclib response. Histogram shows % generational inhibition in response to 5 days abemaciclib treatment in MDA-231 with and without Rb1 siRNA knockdown. Western blots confirm knockdown of Rb1 by siRNA at the time points indicated. #Indicates statistically significant vs scrambled non-targeting control, p-value according to Student t-test.

Figure 5: Abemaciclib combined with cytotoxic chemotherapy induces increased inhibition of cell proliferation and induction of cell death. A, MDA- 231 TNBC cells were treated simultaneously with a range of concentrations of abemaciclib plus docetaxel or carboplatin for 5 days. B, C, cells were pre-treated for 48 hours with or without 200nM abemaciclib followed by treatment with 1nM docetaxel (dtx) or 25μM carboplatin (carb) or the combination of abemaciclib plus docetaxel or carboplatin for the remaining 72 hours. Total cell counts were measured at day 0, day 2 (washout) and day 5. D, Annexin V staining of cells pre- treated with or without increasing concentrations of abemaciclib, palbociclib or ribociclib for 24 hours followed by 24 hours of treatment with 100nM docetaxel, left panels show representative examples, summary histogram on the right. E, Annexin V staining of cells pre-treated for 24 hours with a range of concentrations of either abemaciclib, palbociclib or ribociclib followed by switch to 100nM docetaxel plus abemaciclib, palbociclib or ribociclib. Veh indicates 'vehicle control/untreated'. Data represent mean +/- SD. All experiments were repeated in at least duplicate. *Indicates statistically significant difference compared to vehicle control, #statistically significant difference compared to vehicle/docetaxel arm, < 0.05, p- values according to Student t-test.

Figure 6: Identification of PD-biomarkers of abemaciclib response in vivo. A, percent inhibition of mRNA expression of 22 cell cycle regulated genes, relative to vehicle control, in each of the treatment arms in tumor collected snap-frozen 24 hours after the 3 days of continuous dosing in the MCF7 xenograft study. B, percent inhibition of a subset of genes that show abemaciclib dose dependence and response to abemaciclib plus hormone blockade therapy. C, abemaciclib dose dependent genes in HER2+/ER+ breast cancer cell line xenografts are also responsive to HER2/ER targeted therapy. D, subset of genes commonly altered by abemaciclib treatment across three breast cancer cell line xenograft studies. Values represent mean percent inhibition of gene expression relative to vehicle control in each study.

Fig.1 A B HER2-/ER+ HER2+/ER- 10 EFM-19 ZR-75-1 MDA-361 MDA-468 HER2+/ER+ TNBC Abema 200 nM - + - + - + - +

1 S807/811 μM pRb

0.1 Total-Rb

Log IC Cyclin D1 0.01 CK-19

0.001 β-Actin BT-20 T-47D MCF-7 HCC70 HCC38 JIMT-1 CAL-51 BT-474 BT-549 Hs578T EFM-19 ZR-75-1 SK-BR-3 HCC202 DU4475 CAMA-1 ZR-75-30 HCC1500 HCC1419 HCC2218 HCC1395 HCC1143 HCC1806 HCC1954 HCC1937 HCC1569 HCC1187 COLO-824 UACC-732 UACC-812 UACC-893 EFM-192A SUM-190PT MDA-MB-453 MDA-MB-415 MDA-MB-361 MDA-MB-231 MDA-MB-436 MDA-MB-468 MDA-MB-157 SUM-225CWN MDA-MB-134-VI MDA-MB-175-VII C D E

1.00E+01 1.00E+01 1.00E+01

μM μM

μM R² = 0.7606

50 R² = 0.6447 50 R² = 0.5956 50 1.00E+00 1.00E+00 1.00E+00 1.00E-02 1.00E-01 1.00E+00 1.00E+01 1.00E-02 1.00E-01 1.00E+00 1.00E+01 1.00E-02 1.00E-01 1.00E+00 1.00E+01 Log IC

Log IC

1.00E-01 1.00E-01 1.00E-01 Ribociclib Ribociclib Palbociclib Log IC Palbociclib

1.00E-02 1.00E-02 1.00E-02

Abemaciclib Log IC50 μM Abemaciclib Log IC50 μM Palbociclib Log IC50 μM Downloaded from mct.aacrjournals.org on October 2, 2021. © 2018 American Association for Cancer Research. Author Manuscript Published OnlineFirst on February 26, 2018; DOI: 10.1158/1535-7163.MCT-17-0290 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Fig.2

A B 300 Abe Abe F 1,200 MCF-7: ER+ BCL Vehicle Abe Fulv Tam Abe DAY 6: Vehicle Abema Fulvestrant 50mg/kg 75mg/kg + + 200x ) 250 Fulv Tam 3 1,000 200 800 * 150 600 100

400 50 Tamoxifen Abema+Fulvestrant Abema+Tamoxifen

200 0

Tumor Volume (mm Volume Tumor -50

0 (%) Regressions and ΔT/C 0 5 10 15 20 25 30 35 -100 Vehicle Abema 50 mg/kg Abema 75 mg/kg Progressive Disease Stable Disease Partial Response Fulvestrant Abema+Fulv Tamoxifen -150 Complete Response Abema+Tam C

) Abema 50mg/kg Abema 75mg/kg 3 1,600 1,600 1,600 Tamoxifen 1,600 Fulvestrant 1,600 Abema + Tamox 1,600 Abema + Fulv

1,200 1,200 1,200 1,200 1,200 1,200

800 800 800 800 800 800

400 400 400 400 400 400

Tumor Volume (mm Volume Tumor 0 0 0 0 0 0 Days 0 15 30 45 60 75 90 0 15 30 45 60 75 90 0 15 30 45 60 75 90 0 15 30 45 60 75 90 0 15 30 45 60 75 90 0 15 30 45 60 75 90 Treatment withdrawn Treatment withdrawn Treatment withdrawn Treatment withdrawn Treatment withdrawn Treatment withdrawn D E On treatment (Day 3) > 6 weeks post drug withdrawal Vehicle Abema Fulvestrant Abe + fulv Tamox Abe + tam Vehicle Abema Fulvestrant Abema + fulv Tamox Abema + tam 1R 1B 2L 2R 4N 4B 4N 5N 5R 6B 6N 7N 7B 1N 1R 1L 1B 2L 2N 2B 4R 4B 4N 5R 5B 5L 6L 6R 6N 7R 7L 7N pRbS807/811 pRbS807/811 Total-Rb Total-Rb FOXM1 Cyclin-D FOXM1 Keratin-19 β-actin Downloaded from mct.aacrjournals.org on October 2,β -2021.actin © 2018 American Association for Cancer Research. Author Manuscript Published OnlineFirst on February 26, 2018; DOI: 10.1158/1535-7163.MCT-17-0290 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Fig.3

A BT-474, HER2+/ER+, Trastuzumab Sensitive B BT-474-TR, HER2+/ER+, Trastuzumab Acquired Resistant C 600 1,000 550

) BT-474 ) 3

3 450 500 ) 3 800 BT-474-TR 400 * 350 600 300 * 250

# 400 # 200 150 Tumor Volume (mm Volume Tumor 200 100 (mm Volume Tumor

in Tumor Volume (mm Volume in Tumor 50

Δ 0 0 -50 0 4 8 12 16 20 24 28 Days 0 4 8 12 Days16 20 24 28 Vehicle Abe 50mg/kg Trastuzumab Vehicle Abe 50mg/kg Abe 75mg/kg -150 Abe+Tz Abe+Dx Abe+Tz+Tam Trastuzumab Abe+Tz Abe+Dx Abe+Tz+Dx+Tam Abe+Tz+Tam Abe+Tz+Dx+Tam

D E 35 Vehicle Abe (50) Abe (75) Trast (Tz) Abe (50)+Tz Abe+Dtx Abe+Tz+Tam Abe+Tz+Tam+Dtx 30 1 2 3 1 2 3 1 2 3 1 2 3 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 25 pRbS807/811 V + Cells 20 TOPO IIα 15 PHH3

Annexin 10

% total HER2 5 0 B-actin BT474 BT-474-TR Vehicle Abema 200nM Traz 15μg/ml Tam 500nM Traz+Tam Abema+Tam Abema+Traz Abema+Traz+Tam Downloaded from mct.aacrjournals.org on October 2, 2021. © 2018 American Association for Cancer Research. Author Manuscript Published OnlineFirst on February 26, 2018; DOI: 10.1158/1535-7163.MCT-17-0290 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Fig.4 Vehicle Abema 50 Abema 75 C HCC70 MDA231 A MDA-231, High pRb, Low p16 MDA-231: Day 21 500 200 1N 1R 1L 1N 1L 1N 2L 2R 3L 3N Vehicl Abe Abe Docetaxe Abema

) e + S807/811 S807/811

3 50mg/kg 75mg/kg l pRb 400 pRb 150 Dtxl p16 Total-Rb 300 100 β-actin FOXM1 200 50 Cyclin D 100 0

Tumor Volume (mm Volume Tumor β-actin 0 ΔT/C and Regressions (%) Regressions and ΔT/C 0 3 6 9 12 15 18 21 -50 Vehicle Control Abemaciclib 50 mg/kg QD D Abemaciclib 75 mg/kg QD Docetaxel 10 mg/kg QW -100 100 Progressive Disease Stable Disease Partial Response Abemaciclib + Docetaxel 90 Untransfected Complete Response 76.4 75.8 80 Scrambled Control 70 Rb1 Knockdown B 60 46.4 49.4 150 1,200 HCC70, Low pRb, High p16 Vehicl Abe Abema 50 43.5 Abe Docetaxe e 50mg/kg 75mg/kg l + 40 ) Dtxl

3 23.7 1,000 23.9 100 30 20.6 15.0 800 20 10

50 Inhibition % Generational 600 0 1000nM Abema 200nM Abema 40nM Abema 400 0 Day 0 Day 2 Day 5 200 Tumor Volume (mm Volume Tumor

0 (%) Regressions and ΔT/C -50 0 3 6 9 12 15 18 21 Abema 200 nM Vehicle Control Abemaciclib 50 mg/kg QD ------+ + + - - - + + + -100 S807/811 Abemaciclib 75 mg/kg QD Docetaxel 10 mg/kg QW Progressive Disease Partial Response pRb Abemaciclib + Docetaxel tRb

β-Actin

Fig.5

A Abemaciclib (Abe) B C 120 12000 12000

Docetaxel (Dtx) Growth Day 2 to Day 5 Growth Day 2 to Day 5

Carboplatin (Carb)

100 10000 Growth Day 0 to Day 2 10000 Abe+Dtx Growth Day 0 to Day 2 80 Abe+Carb 8000 8000

* 6000 6000 60 * *

4000 Cell Number Cell Cell Number 4000 40 2000 2000 20 % Generational Inhibition % Generational 0 0 0 Abema (nM) 1 0.2 0.04 0.008 0.0016 0.00032 Dtx (nM) 10 2.0 0.40 0.080 0.0160 0.00320 Veh/Veh Veh/Abe Abe/Veh Veh/Dtx Abe/Abe Abe/Dtx Veh/Veh Veh/Abe Abe/Veh Abe/Abe Veh/Carb Carb (μM) 25 12.5 6.25 3.125 1.5635 0.78125 Abe/Carb Veh/Abe+Dtx Abe/Abe+Dtx Veh/Abe+Carb Abe/Abe+Carb D E Vehicle/Vehicle Vehicle/Dtx 100nM 20 18 18

16 14 14 # # 12 # 12 #

V + Cells 10 Iodide V + Cells

10 # # 8 Ribo/Dtx 100nM 8 # Abema/Dtx 100nM Palbo/Dtx 100nM # # 6 6 4 4 Annexin Annexin Propidium

% 2 % 2 0 0 Veh/Veh Annexin-V FITC Veh/Veh Plb 200nM/Dtx Plb 500nM/Dtx Plb Rib 200nM, Dtx 200nM, Rib Dtx 500nM, Rib Veh/Dtx 100nM Veh/Dtx Abe 200nM/Dtx Abe 500nM/Dtx Abe Veh/Dtx 100nM Veh/Dtx Plb 1000nM/Dtx Plb Rib 1000nM, Dtx 1000nM, Rib Abe 1000nM/Dtx Abe Plb 200nM/Plb+Dtx Plb 500nM/Plb+Dtx Plb Rib 200nM, Rib+Dtx 200nM, Rib Rib+Dtx 500nM, Rib Downloaded from mct.aacrjournals.org on October 2, 2021. © 2018 American Association for Cancer Research. Plb1000nM/Plb+Dtx Rib 1000nM, Rib+Dtx 1000nM, Rib Abe 200nM/Abe+Dtx Abe 500nM/Abe+Dtx Abe Abe 1000nM/Abe+Dtx Abe Author Manuscript Published OnlineFirst on February 26, 2018; DOI: 10.1158/1535-7163.MCT-17-0290 Author manuscripts have been peer reviewed and accepted for publication but have not yet been edited.

Fig.6

A MCF-7 (ER+) Xenografts B MCF-7 (ER+) Xenografts 50 mg/kg abemaciclib 75 mg/kg abemaciclib CCNB1 RRM2 TOPO2A MKI67 E2F4 E2F1 CDKN1A MCM7 100 Abemaciclib (50 mg/kg) + fulvestrant Abemaciclib (50 mg/kg) + Tamox CDK4 PTEN E2F2 CDK2 RB-1 EX17-18 MYC Cyclin E E2F3 RBL1 CCND3 RB-1 EX25-26 CCND1 E2F5 RBL2 90 100 90 80 80 70 70 60 60 50 50 40 40 30 30 20 20 10 10 0 0 % Inhibition of gene expression of gene % Inhibition % Inhibition of gene expression of gene % Inhibition 50 mg/kg abemaciclib 75 mg/kg abemaciclib fulvestrant Abemaciclib (50 mg/kg) Tamoxifen Abemaciclib (50 mg/kg) CCNB1 RRM2 TOPO2A MKI67 E2F1 MCM7 CDK4 PTEN E2F2 CDK2 + fulvestrant + Tamox C BT474-TR (HER2+/ER+) Xenografts D 100 50 mpk Abemaciclib MCF-7 ZR-75-1 BT474-TR

90 75 mpk Abemaciclib 50 mg/kg 75 mg/kg 50 mg/kg 75 mg/kg 50 mg/kg 75 mg/kg Abemaciclib Abemaciclib Abemaciclib Abemaciclib Abemaciclib Abemaciclib 50-AB + Trastuzumab 80 CCNB1 74.52 92.05 0.50 13.45 88.73 94.00 50-AB + Tamox 70 RRM2 67.20 96.56 52.28 91.20 31.53 93.37 50-AB + Trastuzumab + Tamox TOPO2A 61.15 84.21 76.94 94.43 78.73 94.33 60 MKI67 55.43 88.23 70.66 94.22 74.50 94.50 50 E2F1 32.82 84.09 0.00 55.28 0.00 65.70 40 MCM7 29.38 76.18 25.37 74.84 28.30 77.37

30 CDK4 27.21 38.17 17.10 35.95 26.87 44.57 PTEN 15.86 57.55 31.22 30.10 0.00 0.00 20 E2F2 14.04 84.61 18.53 64.21 -1.93 14.77 10 % Inhibition of gene expression of gene % Inhibition CDK2 11.69 46.86 20.32 57.23 11.03 64.63 0 RBL1 0.00 13.25 24.16 52.63 1.13 0.00 TOPO2A MKI67 FOXM1 RRM2 MCM7 CDK4 RB-1 EX25-26 RB-1 EX17-18 CDK2

Preclinical activity of abemaciclib alone or in combination with anti-mitotic and targeted therapies in breast cancer

Neil O'Brien, Dylan Conklin, Richard Beckmann, et al.

Mol Cancer Ther Published OnlineFirst February 26, 2018.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-17-0290

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2018/02/24/1535-7163.MCT-17-0290.DC1

Author Author manuscripts have been peer reviewed and accepted for publication but have not yet been Manuscript edited.

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/early/2018/02/24/1535-7163.MCT-17-0290. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.