384.Full-Text.Pdf

Published OnlineFirst October 3, 2019; DOI: 10.1158/1535-7163.MCT-19-0135

MOLECULAR CANCER THERAPEUTICS | SMALL MOLECULE THERAPEUTICS

Ibrutinib Potentiates Antihepatocarcinogenic Efﬁcacy of Sorafenib by Targeting EGFR in Tumor Cells and BTK in Immune Cells in the Stroma Cho-Hao Lin1,2, Khadija H. Elkholy1,2, Nissar A. Wani2,3, Ding Li1,2, Peng Hu1,2, Juan M. Barajas1,2, Lianbo Yu2,4, Xiaoli Zhang2,4, Samson T. Jacob2,3, Wasif N. Khan5, Xue-Feng Bai1,2, Anne M. Noonan2,6, and Kalpana Ghoshal1,2

ABSTRACT ◥ Hepatocellular carcinoma (HCC), the most prevalent primary genes promoting differentiation. Moreover, ibrutinib showed liver cancer, is a leading cause of cancer-related death worldwide synergy with sorafenib or regorafenib, a sorafenib congener, by because of rising incidence and limited therapy. Although treatment inducing apoptosis of HCC cells. In vivo, this TKI combination with sorafenib or lenvatinib is the standard of care in patients with significantly inhibited HCC growth and prolonged survival of advanced-stage HCC, the survival benefit from sorafenib is limited immune-deficient mice bearing human HCCLM3 xenograft due to low response rate and drug resistance. Ibrutinib, an irre- tumors and immune-competent mice bearing orthotopic mouse versible tyrosine kinase inhibitor (TKI) of the TEC (e.g., BTK) and Hepa tumors at a dose that did not exhibit systemic toxicity. In ErbB (e.g., EGFR) families, is an approved treatment for B-cell immune-competent mice, the ibrutinib–sorafenib combination þ malignancies. Here, we demonstrate that ibrutinib inhibits prolif- reduced the numbers of BTK immune cells in the tumor micro- þ eration, spheroid formation, and clonogenic survival of HCC environment. Importantly, we found that the BTK immune cells cells, including sorafenib-resistant cells. Mechanistically, ibrutinib were also enriched in the tumor microenvironment in a subset of inactivated EGFR and its downstream Akt and ERK signaling in primary human HCCs. Collectively, our findings implicate BTK HCC cells, and downregulated a set of critical genes involved in cell signaling in hepatocarcinogenesis and support clinical trials of proliferation, migration, survival, and stemness, and upregulated the sorafenib–ibrutinib combination for this deadly disease.

Introduction Sorafenib is a widely used first-line treatment of HCC since 2007 (2). As a pan-kinase inhibitor, sorafenib strongly inhibits Liver cancer is an aggressive disease; there has been a dramatic proangiogenic VEGFRs, PDGFRb, as well as pro-proliferative increase in incidence and mortality in the last several decades world- kinasessuchasRAF,c-KIT,andFLT-3(4).However,patientstreated wide (1, 2). Between 1990 and 2015, there was a 75% increase in liver with sorafenib had a response rate of 2% to 9% and an overall survival cancer incidence globally (3). In 2015, incidence of liver cancer was time of only 10.9 to 12.3 months. Patients also frequently developed 854,000 cases and 810,000 patients died of this disease. Hepatocellular drug resistance. Recently approved targeted therapies, for example, carcinoma (HCC), which accounts for approximately 90% of all lenvatinib in the first-line setting, improved clinical outcomes such primary liver cancers, is the fourth major cause of cancer-related as response rate and progression-free survival, but did not increase death worldwide and the fastest growing cause of cancer-related the median overall survival (5). Regorafenib and cabozantanib mortality in the United States (2). Hepatitis B or hepatitis C virus approved in the second-line setting have both shown improvement infections, alcoholic liver disease, and nonalcoholic fatty liver disease, a in median overall survival (6, 7). The FDA recently approved PD-1 complication of obesity, are the major risk factors for HCC (2). High checkpoint antibodies, nivolumab (8), and pembrolizumab (9) in mortality rate is, in part, due to late diagnosis of this aggressive tumor patients with HCC who have been previously treated with sorafenib. in patients with chronic liver disease. However, only a subset of patients responded to this type of immunotherapy. Furthermore, transplant recipients on immune suppression are not candidates for immunotherapy. More importantly, the 1Department of Pathology, The Ohio State University, Columbus, Ohio. 2Com- first-line trials of the checkpoint inhibitors in patients with HCC prehensive Cancer Center, The Ohio State University, Columbus, Ohio. 3Depart- failed to improve overall survival of patients with HCC compared ment of Cancer Biology and Genetics, The Ohio State University, Columbus, with best supportive care (10). Thus, there is an unmet need to 4 Ohio. Department of Biomedical Informatics, The Ohio State University, Colum- develop better therapeutic options in the first-line setting to improve bus, Ohio. 5Department of Microbiology & Immunology and Sylvester Compre- hensive Cancer Center, University of Miami Miller School of Medicine, Miami, the survival and response rate in HCC. Florida. 6Department of Internal Medicine, The Ohio State University, Columbus, Ibrutinib (PCI-32765) is the first-in-class small-molecule inhib- Ohio. itor targeting Bruton's tyrosine kinase (BTK; ref. 11), a key non- Note: Supplementary data for this article are available at Molecular Cancer receptor tyrosine kinase in the B-cell receptor signaling pathway, Therapeutics Online (http://mct.aacrjournals.org/). and critical for the survival of malignant B cells. The FDA desig- Corresponding Author: Kalpana Ghoshal, The Ohio State University, Columbus, nated ibrutinib as a breakthrough therapy, and it was approved for OH 43210. Phone: 614-292-8865; E-mail: [email protected] the treatment of relapsed and refractory mantle cell lymphoma, relapsed and refractory chronic lymphocytic leukemia and Wal- Mol Cancer Ther 2020;19:384–96 denstrom's Macroglobulinemia (12). Ibrutinib irreversibly inhibits doi: 10.1158/1535-7163.MCT-19-0135 Tyr-223 kinase activity by covalently binding to Cys-481 adjacent to 2019 American Association for Cancer Research. the ATP-binding pocket of BTK (13). Ibrutinib exhibits both

AACRJournals.org | 384

Sorafenib/Ibrutinib Combination Therapy for HCC

antitumorigenic and immunomodulatory and antitumorigenic Clonogenic survival assay functions because of its ability to inhibit tyrosine kinases, for HCC cells were seeded in 12-well plate (1 to 2 104 cells/well). After example, BTK, ITK, BLK, TEC, BMX, and JAK3 expressed in innate 24 hours, cells were treated with sorafenib, ibrutinib, combination of and adaptive immune cells (13) and ErbB family members, for both or vehicle for 5 to 7 days. The culture medium and drugs were example, EGFR/ErbB1, Her2/ErbB2, and ErbB4 (14), predominant- replaced every other day. Cells were fixed in 4% paraformaldehyde and ly expressed in epithelial cells. Several reports demonstrated ibru- colony formation was visualized with 0.05% crystal violet dye. tinib's efficacy in preclinical models of a variety of solid tumors (15). þ Ibrutinib exhibits lasting tumor suppression by activating CD8 T Plasmid transfection cells (16–18) and suppressing myeloid-derived suppressor cells HCC cells were placed in a 6-well plate at 3 105 cells/well. After (MDSC; ref. 19). Currently, ibrutinib in combination with chemo- 24 hours, cells were transfected with 2 mg of Myr-Akt-HA or wild-type or immunotherapy is being assessed in clinical trials in several solid Akt plasmid DNA using the Lipofectamine 3000 reagent (Thermo tumors. Fisher Scientific). In this study, we assessed the antitumorigenic efficacy of ibrutinib alone and in combination with sorafenib in vitro and in vivo RNA interference and determined the underlying molecular mechanisms. We found HCC cells plated overnight in a six-well plate at 3 105 cells/well that ibrutinib cooperates with sorafenib by inactivating its substrate were transfected with siEGFR (catalog no. M-003114-03; Dharmacon) EGFR in tumor cells and BTK in immune cells in the tumor or control siRNA (catalog no. D-001206-13; Dharmacon; final con- þ microenvironment. Our study also demonstrated that the BTK centration, 50 nmol/L) using RNAiMAX (Thermo Fisher Scientific). immune cells are enriched in the tumor stroma in a subset of After 48 hours, cells were treated with the drugs and cell survival was primary human HCCs. measured after 72 hours.

Spheroid formation assay Materials and Methods HCC cells (3,000 cells) were plated in 96-well Corning Costar Ultra- Cell culture and drug treatment Low Attachment plates in serum-free DMEM/F-12 medium supple- HCC cell lines HepG2, Hep3B, PLC/PRF/5, SNU-182, SNU-449, mented with containing 2 mmol/L glutamine, 1 mmol/L sodium and BNL 1ME A.7R.1 (BNL) were obtained from the ATCC. Huh-7, pyruvate, 1% MEM nonessential amino acid solution, 1% B-27 sup- Hepa1-6 (Hepa), and HCCLM3 cells were provided by Drs. James plement (Gibco), 100 mg/mL penicillin G, and 100 U/mL streptomycin Taylor (Fox Chase Center), Gretchen Darlington (previously at Baylor supplemented with 20 ng/mL EGF, and 10 ng/mL FGF. After 7 days, College of Medicine, Houston, TX), and Hangxiang Wang (The First spheroids were treated with drugs or vehicle for another 48 hours, and Affiliated Hospital, School of Medicine, Zhejiang University, Hang- cell viability was measured with CellTiter-Glo Luminescent Cell zhou, China), respectively. Cells were maintained in either DMEM or Viability Assay. Minimum Essential Media supplemented with L-glutamine (2 mmol/L), 10% FBS, sodium pyruvate (1 mmol/L), and penicillin/streptomycin/ RNA-sequence analysis fi amphotericin (Thermo Fisher Scienti c) at 37 Cwith5%CO2. Sor- RNA-sequencing (RNA-seq) was performed in PLC/PRF/5 cells afenib-resistant Huh7 (Huh7-SR) cells, previously generated in our treated with sorafenib (2 mmol/L), ibrutinib (4 mmol/L), combination laboratory (20), were grown in the media supplemented with sorafenib of both or vehicle (DMSO) for 12 hours. Total RNA was isolated with (6 mmol/L). Sorafenib was withdrawn from the culture media Huh7-SR TRizol reagent (Invitrogen) following the manufacturer's instructions cells for 2 days prior to performing experiments. Firefly expressing and quantified using Invitrogen Qubit 2.0 Fluorometer (Thermo HCCLM3 (HCCLM3-Luc) and Hepa (Hepa-Luc) cells were generated Fisher Scientific). Samples with RNA Integrity (RIN) scores 7 were by infecting these cells with firefly luciferase lentivirus (GeneCopoeia) used to generate libraries using the Illumina TruSeq Stranded mRNA followed by selection of positive clones with puromycin (5 mg/mL) LT Kit (Illumina). Multiplexed libraries were sequenced on Illumina treatment for 4 weeks. For treatment of cells in culture, sorafenib and HiSeq 4000 Sequencer to achieve 17 to 20 million paired-end 150 bp ibrutinib were dissolved in DMSO. reads per sample. De-multiplexed and quality-filtered reads were mapped to the human genome GRCh38 using Hierarchical Indexing Cell survival assay for Spliced Alignment of Transcripts (HISAT2; ref. 22). Raw read HCC cells seeded into 96-well plates (3,000 cells/well) were allowed counts for each gene were quantified using featureCounts software, to grow overnight followed by treatment with sorafenib (LC Labora- with GENCODE v.27 transcript reference (GENCODE annotation; tories), ibrutinib (Cayman Chemicals and Acorn PharmaTech), or ref. 23). We have submitted RNA-seq raw data to the Geo database combination of both. Cell viability was assessed after 72 hours of drug (Accession no. GSE114184). exposure using CellTiter-Glo Luminescent Cell Viability Assay For statistical analysis of the RNA-seq data, genes with read counts (Promega). Each treatment was done in quadruplicate. below 5 in at least two of three biological replicates in each treatment group were filtered out. Next, the read counts were normalized using Statistical analysis of drug interaction TMM method (23). R package limma with Voom transformation (24) The two drugs (A and B) are considered to act synergistically if the was used to calculate P values for group comparisons under a linear biological response (cell survival in this study) to A (sorafenib) and B model. The P value cutoffs were determined by calculating the mean (ibrutinib) cotreatment is greater than the sum of the response to A and number of false positives (25). B alone. A two-way ANOVA was used to test this hypothesis (mboth mneither) > (mA mneither) þ (mB mneither), where m is the mean Real-time reverse transcription PCR analysis response to each treatment and the vehicle control. P values <0.05 are Total RNA was isolated from cells using TRizol reagent, treated with considered as significant synergistic interactions between the two DNase 1 and reverse transcribed into cDNA using high-capacity drugs (21). cDNA Reverse Transcription Kit (Applied Biosystems). Reverse

AACRJournals.org Mol Cancer Ther; 19(2) February 2020 385

Lin et al.

transcription PCR (qRT-PCR) was performed using 0.01 to 0.1 mg Lumina (Caliper Life Sciences). Acquisition and analysis of images cDNA with SYBR Green in a thermocycler. The fold difference in were performed by using the Living Image Software. DDC target gene mRNA levels was calculated using the t method and normalized to 18S rRNA or b-actin mRNA. The primer sequences are Analysis of systemic and tissue toxicity in mice provided in the Supplementary Table S1. C57BL/6N mice were treated with the vehicle, sorafenib (15 mg/kg/ alternate day), ibrutinib (20 mg/kg/day), or their combination for Flow cytometry 4 weeks. Serum collected from mice by cardiac puncture and analyzed Cells at distinct phases of cell cycle were distinguished by staining with the Vet Axcel Chemistry Analyzer (Alfa Wasserman) for alanine DNA with propidium iodide (PI) and quantified by flow cytometry. aminotransferase (ALT), aspartate aminotransferase (AST), alkaline HCC cells treated with the drugs alone, in combination, or vehicle for phosphatase (ALP), creatine kinase (CK), blood urea nitrogen (BUN), 24 hours were collected, washed with PBS, and fixed in ice-cold 70% albumin, and globulin. ethanol and stored at 20C overnight. The cells were centrifuged, washed with PBS and resuspended in 0.5 mL PBS. To the single-cell Western blot analysis suspension, 50 mL of RNase A (1 mg/mL in PBS) was added and A total of 7 105 HCC cells were plated overnight in 60-mm dish incubated at 37C for 30 minutes, followed by the addition of 50 mLof and treated with sorafenib and ibrutinib alone or in combination for PI (500 mg/mL in PBS) with gentle mixing and incubation in the dark at 1 hour or 24 hours. Whole-cell extracts were prepared in cell lysis room temperature for 15 minutes and subjected to flow cytometry buffer (CLB; catalog no. 9803; Cell Signaling Technology) containing using a LSRII flow cytometer (BD Biosciences). Cell apoptosis was protease inhibitor cocktails (Sigma). The cell lysates were incubated at determined with PI and Annexin V double staining using Annexin V/ 4C for 10 minutes and centrifuged at 4C for 10 minutes to collect Dead Cell Apoptosis Kit (Invitrogen) following manufacturer's clear supernatants. The tumor lysates from snap-frozen tissues were instruction. prepared by suspension in CLB followed by sonication before centri- fugation. Protein concentrations in the extracts were measured using a Mouse strains, animal husbandry, and treatment Bio-Rad Protein Assay Kit (catalog no. 500-0006) with BSA as the NSG (NOD scid Il2rg / ) mice were purchased from Target standard. Equal amounts of protein (10–50 mg) from whole cell or Validation Shared Resource core facility at the Ohio State University, tissue lysates were separated by SDS-polyacrylamide (10% or 4% to and C57BL/6N mice were purchased from Charles River Laborato- 20% gradient) gel electrophoresis (Bio-Rad), transferred to nitrocel- ries. All animals were housed in a temperature-controlled room lulose membranes (GE Healthcare), incubated using blocking buffer under a 12-hour light/12-hour dark cycle and under helicobacter- (LI-COR) followed by immunoblotting with phospho-Akt (Cell free conditions and fed normal chow diet. All animal studies were Signaling Technology, #4060), total Akt (Cell Signaling Technology, reviewed and approved by the Ohio State University Institutional #9272), phospho-ERK (Cell Signaling Technology, #4370), total ERK Laboratory Animal Care and Use Committee. Both male and female (Cell Signaling Technology, #9102), phospho-EGFR (Cell Signaling mice were used for experiments. For drug treatment, ibrutinib stock Technology, #3777), total EGFR (Santa Cruz Biotechnology, #sc-03) (40 mg/mL in DMSO) was diluted with 10% hydroxypropyl beta- PARP (Cell Signaling Technology, #9532), BTK (Cell Signaling cyclodextrin (HPBCD; CTD Holdings) to the final concentration of Technology, #8547), b-actin (Santa Cruz Biotechnology, 4 mg/mL immediately before injection. Sorafenib was diluted in #sc-47778), tubulin (ProteinTech, #66031), or GAPDH (Cell Signaling Cremophor EL/ethanol/water (1:1:6) to 3 mg/mL before delivery Technology, #5174). Following incubation with appropriate secondary through oral gavage. For subcutaneous xenografts, 10 to 12 weeks old antibody (IRD-680 or IRD-800), the specific immune-reactive bands NSG mice were injected subcutaneously with HCCLM3 (5 106) were visualized using Odyssey CLx Imaging System (LI-COR) and cells into the right flank. When tumor volume reaching 150 mm3, quantified using Image Studio software. mice were randomized into four groups and injected intraperitone- ally with ibrutinib (20 mg/kg/day), sorafenib (15 mg/kg/alternate IHC and tissue microarray day) administrated through oral gavage, combination of both, and After fixation in 4% paraformaldehyde, tumor tissues were pro- respective vehicles (DMSO in HPBCD for ibrutinib and Cremophor cessed, embedded in paraffin, and sectioned at 4-mm thickness. Hepa EL/ethanol/water for sorafenib), respectively. Tumor volumes based tumor and human HCC tissue microarray (TMA; catalog nos. LV242 on digital caliper measurements were calculated by the ellipsoidal and HLivH090BC01; USBiomax) sections were baked at 60C for 30 formula [1/2(length width2)].After4weeksoftreatment,mice minutes, dewaxed in xylene, and rehydrated in graded ethanol. were euthanized and tumor tissues were collected, weighed, and Antigen retrieval was processed in citrate buffer (0.01 mol/L, pH photographed. 6.0) at 95C for 30 minutes followed by incubation with the Ki-67 (Cell For orthotopic model, 10 to 12 weeks old NSG or C57BL/6N mice Signaling Technology, #9027) or BTK antibodies (Cell Signaling were injected with 1 million HCCLM3-Luc or 0.05 million Hepa-Luc Technology, #8547), and color was developed by the DAB method. cells into the left liver lobe, and the mice were monitored by IVIS Images were taken by a phase contrast microscope (BX41TF; OLYM- Lumina (Caliper Life Sciences) every week. Once the luciferase signal PUS) and camera (DP71; OLYMPUS) and the Ki-67–positive tumor was detected, mice were randomized into four groups and treated with cells were counted using ImageJ software. the vehicle, sorafenib (15 mg/kg/alternate day), ibrutinib (20 mg/kg/ day), or their combination as described above. Survival was deter- mining using Kaplan–Meier curve. Results Ibrutinib inhibits HCC cell proliferation and clonogenic survival Bioluminescent imaging by inhibiting EGFR signaling fl fl Mice were anesthetized with 2% iso urane and 100% O2 with a ow To test the antihepatocarcinogenic functions of ibrutinib, a tyrosine rate of 2 L/minute, injected with luciferin (150 mg/kg, i.p.) and imaged kinase inhibitor (TKI), we determined its IC50 in seven human with 1 minute exposure from the dorsal perspective using an IVIS [HepG2, HCCLM3, PLC/PRF/5, Huh7, sorafenib-resistant Huh7

386 Mol Cancer Ther; 19(2) February 2020 MOLECULAR CANCER THERAPEUTICS

Sorafenib/Ibrutinib Combination Therapy for HCC

(Huh7-SR), SNU-182 and SNU-449] and two mouse (Hepa and BNL) (>1.5-fold) of 225 genes DMSO controls. Expression of several genes HCC cell lines. These cell lines were differentially sensitive to ibrutinib that regulate cell proliferation, migration, stemness, and apoptosis m fi as demonstrated by IC50s ranging from 1 mol/L for PLC/PRF/5 cells were signi cantly altered by ibrutinib treatment (Fig. 1F). Effects of to 40 mmol/L for HepG2 cells (Fig. 1A; Table 1). Notably, both the Ibrutinib on expressions of key genes were confirmed by qRT-PCR m m parental Huh7 (IC50 6 mol/L) and Huh7-SR (IC50 7 mol/L) cells (20) (Fig. 1G). These include downregulated genes involved in cell pro- were sensitive to ibrutinib. Clonogenic survival of HCC cells was also liferation (JUN, FOS, EGR1, and S100A9), migration/invasion inhibited by this TKI (Supplementary Fig. S1A). (CXCL5, CEMIP/KIAA1199, and SERPINE1), fibrosis (CTGF), stem- To elucidate the underlying mechanism, we sought to identify the ness (SOX2 and SOX9). Among the upregulated genes are the tumor tyrosine kinase substrate of ibrutinib in HCC cells. Because immune suppressor LHPP that encodes a histidine phosphatase (28), and cell–specificsubstratesofibrutinib,forexample,BTK,ITKarenot differentiation promoting genes (CEBPA and HMGCS2). Intriguingly, expressed in HCC cell lines (Supplementary Figs. S1B and S1C), we CEMIP is an EGFR-interacting protein that stabilizes activated EGFR investigated its effects on ErbB family members that are expressed to promote cell survival, epithelial–mesenchymal transition, and drug in epithelial cells. Among these, EGFR/ErbB1 is activated in pri- resistance (29, 30). mary HCC tissues and cell lines, and its upregulation is associated with poor survival of patients with HCC (26). As shown in Fig. 1B, Ibrutinib cooperates with sorafenib to reduce HCC cell survival ibrutinib (10 mmol/L) treatment for 1 hour reduced EGFR (Y1086) There is a lot of impetus to develop novel combinations to improve phosphorylation in Huh7 by 70%. Notably, p-EGFR was elevated by therapeutic efficacy of sorafenib, a first-line treatment for advanced 50% in Huh7-SR cells, which was reduced by 45% upon ibrutinib HCCs. We, therefore, examined if ibrutinib could sensitize HCC cells exposure. Consistent with a previous report (27), Akt signaling was to sorafenib. To this end, we first treated multiple HCC cell lines with found to be activated in Huh7-SR cells relative to Huh7 cells as both drugs at concentrations ranging from 0.25- to 2-fold of respective demonstrated by 70% increase in Akt (S473) phosphorylation. IC50s. The combination treatment decreased cell viability more sig- Notably, ibrutinib treatment also blocked Akt (S473) phosphory- nificantly compared with individual drugs (Fig. 2A and B; Supple- lation in Huh7-SR cells. Phosphorylation of ERK (T202/Y204), was mentary Figs. S2A–S2F). We determined the degree of synergy – also inhibited in both cell types. Notably, higher IC50sofHepG2, between sorafenib and ibrutinib using the widely accepted Chou HCCLM3, and BNL cells correlated with elevated phospho- and Talalay combination index (CI) method (31). A CI value <1 indicates total-Akt levels (Supplementary Fig. S1D), suggesting a role of Akt that the drugs are acting synergistically; a lower CI value indicates a in ibrutinib resistance. Ibrutinib persistently inactivated EGFR greater degree of synergy. CI values for sorafenib–ibrutinib combina- phosphorylation even after 6 hours of drug removal (Supplemen- tions were <1 in all human and mouse HCC cell lines tested tary Fig. S1E). These results suggest that the suppression of HCC (Fig. 2C; Table 1), suggesting that these two drugs cooperatively cell survival by ibrutinib resulted from inactivation of EGFR suppressed HCC cell survival. Among these cell lines, PLC/PRF/5, function. SNU-449, and HepG2 cells were more sensitive to the cotreatment To identify upstream signaling responsible for activation of Akt and (CI < 0.5; Table 1). Importantly, the two TKIs also inhibited prolif- ERK in HCC cells, we assessed the effects of ibrutinib on EGF-induced eration of sorafenib-resistant Huh7-SR and HCCLM3 cells and ibru- EGFR (Y1086) phosphorylation in HCCLM3 cells. These cells exhib- tinib-resistant HepG2 cells more effectively than individual drugs. ited approximately 4-fold increase in phospho-EGFR level as early as Statistical analysis confirmed significant synergy between the two TKIs 30 minutes of EGF exposure that persisted even after 90 minutes in inhibiting HCC cell survival (data for three cell lines shown in the (Fig. 1C). Importantly, ibrutinib pretreatment not only blocked EGF- Supplementary Table S2). enhanced EGFR (Y1086) phosphorylation by approximately 3- to Next, we tested whether ibrutinib can synergize with regorafenib, 4-fold but also impeded Akt (S473) phosphorylation by approximately recently FDA-approved as a second-in-line therapy for patients with 2-fold in HCCLM3 cells (Fig. 1C), suggesting that ibrutinib inhibited HCC who progress on sorafenib. Notably, ibrutinib exhibited synergy Akt signaling via blocking EGFR activation. in combination with regorafenib in all three cell lines tested (Huh7, fi To con rm that EGFR is a critical target of ibrutinib in HCC cells, PLC/PRF/5, and HCCLM3), as demonstrated by much lower IC50 of we knocked down EGFR expression in HCCLM3 cells by siRNA that the combination compared with individual TKIs (Supplementary reduced phospho-Akt (S473) and phospho-ERK (T202/Y204) levels Figs. S2G–S2I). This is not surprising considering the fact that (Fig. 1D). These cells responded well to siRNA-mediated EGFR regorafenib is structurally identical to sorafenib except for one fluorine depletion without significantly affecting total Akt and ERK levels. As atom in the central phenyl ring (32). expected, ibrutinib treatment further reduced EGFR, Akt, and ERK FACS analysis of PI-stained cells revealed a substantial increase in phosphorylation both in the control siRNA- and EGFR-siRNA trans- sub-G1 cell population in PLC/PRF/5, HCCLM3, and HepG2 cells with fected cells. Notably, the IC50 of ibrutinib increased from 12 to concomitant decrease in G1 population upon ibrutinib-sorafenib m 21 mol/L in EGFR-depleted HCCLM3 cells compared with control cotreatment at their respective IC50s(Fig. 2D; Supplementary Figs. cells (Fig. 1E). Similarly, EGFR knockdown increased ibrutinib IC50 S3A and S3C). In PLC/PRF/5 and HCCLM3 cells, each drug increased m from 4 to 8 mol/L in PLC/PRF/5 cells (Supplementary Fig. S1F). sub-G1 population, however, the effect was more pronounced with the Collectively, these results demonstrate that inactivation of EGFR and drug combination. In contrast, sub-G1 population increased only after its downstream Akt and ERK signaling is a key mechanism by which TKI cotreatment in HepG2 cells (Supplementary Fig. S3C). Intrigu- fi ibrutinib suppresses HCC cell survival. ingly, the population of HepG2 cells at G1 increased without signi cant To gain mechanistic insight at the gene expression level, we queried changes in sub-G1 when treated at half of the IC50 dose (Supplemen- the effects of ibrutinib on the transcriptome of PLC/PRF/5 cells. We tary Fig. S3D), suggesting that the cells were arrested at G1 at lower ¼ chose these cells as they are highly sensitive to ibrutinib (IC50 drug concentrations. Elevated cleaved PARP levels in PLC/PRF/5 cells 1 mmol/L; Fig. 1A). RNA-seq analysis of these cells treated with and Annexin/PI-positive HCCLM3 cells (Fig. 2E; Supplementary ibrutinib (1 mmol/L) for 12 hours showed that significant (P < Fig. S3B) confirmed enhanced apoptosis induced by this drug 0.001) upregulation (>1.5-fold) of 170 genes and downregulation combination.

AACRJournals.org Mol Cancer Ther; 19(2) February 2020 387

Lin et al.

Figure 1. Effect of ibrutinib (IBT) on HCC cell survival and EGFR, Akt, and ERK phosphorylation. A, HCC cell lines were incubated with ibrutinib at concentrations ranging from 0.5 to 60 mmol/L for 72 hours, and cell viability was determined using CellTiter-Glo Kit, which measures cellular ATP level (n ¼ 4/each cell line, data are mean SD). B, Huh7 and Huh7-SR cells were treated with ibrutinib for 1 hour, and the lysates were subjected to Western blotting with specific antibodies. C, HCCLM3 cells were treated with EGF (20 ng/mL) along with ibrutinib in DMSO or DMSO alone for the indicated times, and phospho-/total- EGFR and Akt levels were determined by immunoblotting. The signals were quantified using Image Studio software. D, HCCLM3 cells were transfected with EGFR-siRNA (siEGFR) or control siRNA (siCon) for 48 hours followed by ibrutinib (24 mmol/L) or vehicle treatment for 1 hour, and the lysates were subjected to immunoblotting, and cell viability of siRNA-transfected cells treated with ibrutinib at indicated concentrations for 72 hours was measured (, P < 0.05, relative to siCon, n ¼ 4, t test, data are mean SD; E). F, The heatmap depicting expression of several critical genes whose expression was significantly altered in IBT-treated PLC/PRF/5 cells, as identified by RNA-seq analysis. G, qRT- PCR validation of selected genes in PLC/PRF/5 cells treated with ibrutinib (2 mmol/L) for 24 hours (, P < 0.05, relative to DMSO, n ¼ 3, t test, data are mean SD).

Investigation of the molecular mechanism underlying synergy cells by transfecting the wild-type (WT-Akt) or constitutively active, revealed that the ibrutinib–sorafenib combination reduced the phos- myristoylated Akt (Myr-Akt) expression vector, followed by drug phorylation of Akt in HCC cells when compared with ibrutinib only treatment. As expected, Myr-Akt was stabilized and the phospho-/ (Fig. 2F; Supplementary Fig. S4A). Akt signaling, critical for chemo- total-Akt ratio increased 3-fold compared with WT-Akt in Huh7 cells resistance in HCC (33), is highly activated in sorafenib resistant (Supplementary Fig. S4B). Assessment of cell viability after 72 hours of cells (27, 34). To conﬁrm the role of Akt pathway in mediating the drug exposure showed that Myr-Akt expressing cells became resistant cooperation of these two TKIs, Akt signaling was activated in Huh7 individual drugs (IC50 of each TKI could not be reached at the drug

388 Mol Cancer Ther; 19(2) February 2020 MOLECULAR CANCER THERAPEUTICS

Sorafenib/Ibrutinib Combination Therapy for HCC

Table 1. IC50 and CI when fraction affected (FA) is 50%. mentary Fig. S4E). After 4 weeks of treatment, tumor volume was significantly reduced by the TKI combination compared with the Ibrutinib (IC50) Sorafenib (IC50) CI at 0.5 vehicle controls or single drugs (Fig. 4B and C). Although decrease in m m Cell line ( mol/L) ( mol/L) FA tumor weight was significant in mice treated with sorafenib (35%) or PLC/PRF/5 1 1.4 0.36 ibrutinib (36%), cotreatment showed more pronounced decrease Hepa 1.7 1.4 0.53 (60%) in tumor weight (Fig. 4D). Scoring Ki-67 positive cells showed SNU-449 3 4 0.39 significant decrease in proliferating tumor cells in combination ther- Huh7 6 3 0.64 apy compared with the monotherapy (Fig. 4E and F). Immunoblot Huh7-SR 7 12 0.73 analysis showed that the treatment with each drug reduced phospho- SNU-182 16 5 0.83 EGFR/Akt/ERK levels in tumors, which was more prominent in HCCLM3 24 12 0.69 tumors developed in mice treated with both drugs (Supplementary BNL 32 9 0.64 Figs. S4F and S4G). HepG2 40 4 0.43 Monitoring NSG mice bearing orthotopic HCCLM3-Luc xenografts showed that tumor growth was significantly (P ¼ 0.011) inhibited by the ibrutinib–sorafenib combination compared with the vehicle (Fig. 5A and B). Notably, the median survival of mice increased concentrations tested; Fig. 2G and H). Notably, CI values at different from 27 to 40 days (P ¼ 0.0007) upon treatment with both drugs concentrations of the two drugs in combination increased in Myr-Akt compared with vehicle controls (Fig. 5C). compared with WT-Akt expressing cells (Fig. 2I), suggesting that the constitutively active Akt abrogated drug synergy. Ibrutinib and sorafenib act in concert to increase median survival of immune-competent mice bearing orthotopic Hepa Ibrutinib and sorafenib combination inhibits survival of liver tumors cancer stem cells in vitro Because ibrutinib exhibits immunomodulatory functions, we next Suppression of SOX2 and SOX9 expression (Fig. 1F and G) led us to tested therapeutic efficacy of these drugs in immunocompetent hypothesize that the ibrutinib–sorafenib combination could inhibit C57BL/6N mice bearing orthotopic tumors developed by transplant- cancer stem cell (CSC) properties of HCC cells. The tumor spheroid ing Hepa-Luc cells. These cells form tumors within 1 to 2 weeks when model has been widely used to elucidate the mechanism of chemore- inoculated in the livers (Fig. 6A; ref. 41). When the tumor volume sistance (35, 36), and to uncover CTC characteristics that are not reached 150 mm3, we started treating mice with the drugs as manifested in monolayer cultures. To determine the effect of the described in Fig. 5A and monitored their survival. Notably, tumor- sorafenib–ibrutinib combination on CSC properties of HCC cells, we bearing mice treated with the TKI combination significantly (P ¼ first established HCC spheroid cultures in ultra-low attachment plates 0.0004) prolonged survival compared with the vehicle control (Fig. 3A) and measured expressions of CSC marker genes, for example, (Fig. 6B); the median survival increased from 8 to 38 days in mice CD13, CD44, CD133, SOX2, SOX9, EPCAM, and KLF4 in self- treated with both drugs. renewing spheroids. As expected, compared with monolayers, expres- Next, we determined systemic toxicity of this treatment regimen sions of these genes were highly upregulated in PLC/PRF/5 spheroids (Fig. 4A) by monitoring body weights and assessing organ functions in (Fig. 3B). Survival analysis revealed increased resistance of spheroids C57BL/6N mice. Body weights of these mice were not significantly to both drugs compared with monolayers (Fig. 3C and D). Never- affected by these drugs alone or in combination (Supplementary theless, both drugs cooperatively (CI < 1) inhibited the growth of HCC Fig. S5A). The serum levels of liver enzymes, for example, ALP, ALT, spheroids (Fig. 3E), which correlated with marked decrease in expres- AST, and cardiac enzyme, for example, CK were not significantly sions of CD13, CD44, KLF4, SOX2, and SOX9 by this TKI combination affected by the TKIs alone or in combination (Supplementary Figs. compared with the vehicle controls (Fig. 3F). This drug combination S5B–S5E). Only BUN level was slightly decreased with single drugs; also inhibited survival of Hepa and HCCLM3 spheroids (Supplemen- however, it was not further affected by cotreatment (Supplementary tary Figs. S4C and S4D). These results suggest that this combination Fig. S5F). Unaltered serum albumin and globulin levels indicate that therapy could potentially reduce HCC tumor recurrence by suppres- the protein synthesis capacity of the liver was not compromised by sing CTC survival. these TKIs in mice (Supplementary Figs. S5G and S5H). Microscopic examination of hematoxylin and eosin (H&E)-stained tissue sections Ibrutinib and sorafenib combination inhibits the growth of did not detect any significant difference in the structure of liver, lung, human HCC xenografts and increases median survival of NSG spleen, and kidney among the drug-treated and control mice mice (Supplementary Figs. S5I–S5L). To validate our in vitro findings, we tested the efficacy of the Given that the main function of BTK is in the immune cells, we ibrutinib–sorafenib combination therapy in NSG (immunodeficient) assessed the influence of ibrutinib on the immune cells in the tumor mice bearing subcutaneous HCCLM3 xenograft tumors. We chose microenvironment. Indeed, IHC analysis of Hepa orthotopic tumor þ HCCLM3 cells because these cells form tumors within a short time sections showed enrichment of BTK cells in the tumor stroma þ with 100% penetrance. When the tumor volume reached 150 mm3, (Fig. 6C and D). Importantly, the BTK cell population was pro- we started treating mice with the drugs or vehicle and measured body foundly decreased in mice subjected to combination therapy compared weights and tumor volumes weekly. To determine drug synergy, we with monotherapy or vehicle. Immunoblot results of the representative treated mice with sorafenib at 15 mg/kg/alternate day and ibrutinib at tumor lysates showed that ibrutinib treatment reduced BTK protein 20 mg/kg/day (Fig. 4A), which are much less than usual doses of level, which was more significant in mice treated with both drugs sorafenib (37, 38) and ibrutinib (39, 40) used to treat tumor-bearing (Fig. 6E and F). These results indicate that ibrutinib and sorafenib mice. The loss of body weight in mice due to tumor burden was <5% combination therapy is effective in suppressing HCC growth in the and it was not significantly affected by the drug treatment (Supple- liver microenvironment of immune-competent mice.

AACRJournals.org Mol Cancer Ther; 19(2) February 2020 389

Lin et al.

Figure 2. Effect of ibrutinib and sorafenib cotreatment on HCC cell survival. Huh7 (A) and Huh7-SR (B) cells were incubated with various concentrations of sorafenib and ibrutinib for 72 hours, and cell viability was assessed using CellTiter-Glo assay (, P < 0.05, relative to Combo, n ¼ 4, one-way ANOVA, data are mean SD). C, CI analysis of multiple HCC cell lines treated with sorafenib and ibrutinib at different concentrations. Trend lines indicate CI values at any given effect (FA, fraction affected), and a CI value <1 indicates synergism. The dotted line indicates a reference point of a CI value of 1. D, PLC/PRF/5 cells treated with sorafenib (2 mmol/L) and ibrutinib (4 mmol/L) for 24 hours were subjected to ﬂow cytometry after staining with PI (, P < 0.05, relative to DMSO, n ¼ 3, one-way ANOVA, data are mean SD). E, PLC/PRF/5 cells were treated with sorafenib (2 mmol/L) and ibrutinib (4 mmol/L) for 24 hours, and PARP levels were analyzed by immunoblotting. F, Immunoblotting of phospho- and total-EGFR, Akt, and ERK in HCCLM3 cells treated with either sorafenib (12 mmol/L) or ibrutinib (24 mmol/L) alone or in combination (Combo) for 1 hour. G and H, Huh7 cells transfected with wild-type Akt (WT-Akt) or myristoylated Akt (Myr-Akt) for 48 hours followed by treatment with IBT and SOR at the indicated concentration for 72 hours, and cell viability was determined (, P < 0.05, relative to Combo, n ¼ 4, one-way ANOVA, data are mean SD). I, The CI plots were generated in WT-Akt– and Myr-Akt–transfected Huh7 cells. The dotted line indicates a reference point of a CI value of 1.

To determine the relevance of these findings to human patients with recruited in the benign inflamed liver tissues adjacent to the HCCs HCC, we probed two human primary HCC TMAs with a BTK-specific (Supplementary Fig. S6C), enrichment of these cells was significantly þ antibody. Results showed that BTK-expressing (BTK ) cells were higher (P ¼ 0.29) in the tumor microenvironment (Fig. 6H). Scoring enriched in the stroma of a subset of HCCs, whereas few such cells two TMAs revealed that 27% of HCCs (23 of 86) and only 4% (1 of 24) were detected only in the sinusoids of normal livers (Fig. 6G; benign livers exhibited score 3. These results clearly demonstrate that þ þ Supplementary Figs. S6A and S6B). Although, some BTK cells were BTK immune cells are recruited in the tumor stroma of a subset of

390 Mol Cancer Ther; 19(2) February 2020 MOLECULAR CANCER THERAPEUTICS

Sorafenib/Ibrutinib Combination Therapy for HCC

Figure 3. Sorafenib and ibrutinib combination treatment regulates CSC properties in HCC spheroids. A, PLC/PRF/5 cells were cultured in ultra-low attachment plate for 9 days to form spheroids. The images were captured with microscope at 200-fold magnification. B, PLC/PRF/5 cells were cultured in ultra-low attachment plate for 9 days to form spheroids, followed by qRT-PCR analysis of cancer stem cell markers (, P < 0.05, relative to monolayer, n ¼ 3, t test, data are mean SD). C, Spheroid or monolayer cultures were treated with the drugs at the indicated concentrations and cell viability was measured after 48 hours (, P < 0.05, relative to Combo, n ¼ 4, one-way ANOVA, data are mean SD). D, Images of PLC/PRF/5 spheroids after drug treatment for 48 hours (magnification: 100). E, CI plot in drug-treated PLC/ PRF/5 spheroid and monolayer cultures. F, qRT-PCR analysis of cancer stem cell markers in drug-treated PLC/PRF/5 cells (, P < 0.05, relative to DMSO spheroids, n ¼ 3, one-way ANOVA, data are mean SD). patients with HCC. Collectively, these results implicate BTK expres- study (5). Targeted therapy directed towards crucial survival pathways sing immune cells in HCC pathogenesis and provide rationale for in both cancer cells and tumor microenvironment is an emerging ibrutinib–sorafenib combination therapy. treatment strategy in cancer. In this study, we have demonstrated that ibrutinib exhibits antitumorigenic effects on both sorafenib sensitive and resistant HCC cell lines by irreversibly inactivating EGFR signal- Discussion ing. Ibrutinib also inactivates BTK in immune cells, and thereby Although systemic treatment with sorafenib is a standard of care modulating the tumor microenvironment. EGFR, a receptor tyrosine option in patients with advanced-stage HCC, its overall survival kinase (RTK), is stimulated by multiple ligands resulting in the benefit is limited, with <3 months improvement in the SHARP activation of various signaling pathways that primarily control cell study (42) and median overall survival of 12.3 months in the REFLECT proliferation, differentiation and survival. EGFR signaling plays a key

AACRJournals.org Mol Cancer Ther; 19(2) February 2020 391

Lin et al.

Figure 4. Effect of ibrutinib, sorafenib, and their combination on subcutaneous HCCLM3 xenograft growth in NSG mice. A, Schematic of the treatment protocol. NSG mice were injected subcutaneously with 5 106 HCCLM3 cells (n ¼ 9–10 mice/group, data are mean SD) and tumor growth was monitored weekly. B, The volume of tumors following ibrutinib (IBT), sorafenib (SOR), and their combination (Combo) therapy for 4 weeks. The tumor images (C) and tumor weights (D) after drug treatment for 4 weeks (n ¼ 9–10, one-way ANOVA, data are mean SD). E, Photographs of Ki-67–stained tumor sections [magniﬁcation: 200 and 400 (inset)]. F, Quantiﬁcation of Ki-67–positive cells using ImageJ software (n ¼ 9–10, one-way ANOVA, data are mean SD).

role in liver generation after acute and chronic liver damage. Also, it cells by ibrutinib treatment holds great therapeutic potential because plays a key role in cirrhosis and HCC. Furthermore, HCC metastasis activation of these signaling pathways is associated with the tumor- and production of inﬂammatory cytokines have been shown to be igenic potential of various cancers including HCC (26). regulated by EGFR-regulated pathways (43). Unlike other clinically Activation of Akt signaling is involved in sorafenib resistance (27). approved EGFR inhibitors directed towards its kinase domain, ibru- Ability of ibrutinib to suppress EGFR, Akt, and ERK pathways suggests tinib can irreversibly inactivate both the wild-type and kinase-mutant a great potential for ibrutinib to overcome sorafenib resistance. More EGFR (44, 45). Inhibition of EGFR, Akt, and ERK signaling in HCC importantly, ibrutinib and sorafenib combination therapy inhibited

392 Mol Cancer Ther; 19(2) February 2020 MOLECULAR CANCER THERAPEUTICS

Sorafenib/Ibrutinib Combination Therapy for HCC

Figure 5. Effect of ibrutinib, sorafenib alone, and their combination on orthotopic HCCLM3 xenograft growth and survival of mice. A, NSG mice were injected orthotopically with 106 HCCLM3-Luc cells (n ¼ 9–10 mice/group), and tumor growth was monitored weekly by bioluminescent imaging. Representative images of each treatment group at indicated times are shown. B, Luciferase activity (photons/second) of each group was quantiﬁed using Living Image software (n ¼ 9–10 mice/group, one- way ANOVA, data are mean SD). C, The Kaplan–Meier survival curves of HCCLM3-Luc tumor bearing mice following treatments (n ¼ 9–10 mice/ group, log-rank test).

HCC cell survival of highly aggressive and sorafenib-resistant vironment is expected to be a rewarding strategy in the treatment of HCCLM3 cells in NSG mice. This was accomplished at a dose much HCC. Indeed, although not significant (P ¼ 0.059), ibrutinib alone lower than the dose of each drug alone without causing notable increased median survival of immune-competent C57BL/6N mice systemic toxicity to the mice. These data obtained in preclinical models bearing orthotopic Hepa tumors from 8 to 21 days, which was suggest that ibrutinib–sorafenib combination therapy will be effective increased to 38 days in combination with sorafenib (P ¼ 0.004). þ as a first line of treatment for patients with sorafenib resistance often IHC analysis showed that infiltration of BTK immune cells in associated with activated EGFR/Akt/ERK signaling. The improved human primary HCCs and mouse Hepa orthotopic tumors. Pre- þ survival (27 to 40 days) of the immune cell-deficient NSG mice bearing cipitous reduction in BTK cells in the Hepa tumor stroma upon orthotopic HCC-LM3 xenografts with this combination therapy indi- combination therapy implicates their role in modulating HCC þ cates that the inhibition of EGFR signaling in the tumor cells is a key to growth. However, the role of BTK immune cells in hepatocarci- the mechanism of action of ibrutinib. nogenesis has not been elucidated. Nonetheless, the efficacy of Currently, ibrutinib is undergoing clinical trials in several solid ibrutinib in potentiating anti-HCC action of sorafenib both in NSG tumors, because of its immunomodulatory effects targeting the and C57BL/6N mouse models suggests that it functions by inhibit- tumor microenvironment. Several reports underscore the relevance ing distinct tyrosine kinase signaling both in the tumor cells and of understanding the role of tumor-infiltrating immune cells includ- immune cells in the tumor microenvironment. Studies are ongoing ing B cells in hepatocarcinogenesis (46–48). Ibrutinib inhibits the to determine the cell type–specific role of BTK in HCC development function of all BTK expressing hematopoietic cells, for example, B using cell type–specific knockout mice. It has been reported that cells,basophils,monocytes,MDSCs,andNKcells.Infact,itinhibits infiltration of PD-1high B cells in the stroma is associated with poor T helper cells as well due to its action on ITK (13, 16). Thus, prognosis of patients with HCC (50). It would be of interest to þ ibrutinib has broad immune modulatory effects (19, 49). Immune monitor PD-1 expression in BTK cells before and after the drug microenvironment plays a critical role in the pathogenesis of HCC treatment. due to its prototypical inflammation-associated nature where pro- There is an unmet need to develop better therapeutic options in the longed hepatitis accounts for approximately 90% of the HCC first-line setting when patients are most fit to receive therapy, to burden. Therefore, targeting immune cells in the tumor microen- improve the survival and response rate in HCC. Although data with

AACRJournals.org Mol Cancer Ther; 19(2) February 2020 393

Lin et al.

Figure 6. Effect of ibrutinib–sorafenib combination on survival of immune-competent mice bearing orthotopic Hepa-Luc tumors. A, C57BL/6N mice were injected orthotopically with 5 104 Hepa-Luc cells (n ¼ 7 mice/group), and tumor growth was monitored weekly by imaging. Representative picture and bioluminescence image of the Hepa-Luc tumor are shown. B, The Kaplan–Meier survival curves of Hepa-Luc tumor bearing mice following drug treatment (n ¼ 7 mice/group, log-rank test). C, C57BL/6N mice were injected orthotopically with 5 104 Hepa-Luc cells and treated with drugs for 14 days. BTK IHC of Hepa-Luc tumor sections is shown (magnification: 200 and 400). D, Quantification of BTKþ cells using ImageJ software (n ¼ 6, one-way ANOVA, data are mean SD). Immunoblotting (E)and quantification (F) of BTK protein expression in Hepa-Luc tumor (n ¼ 56, one-way ANOVA, data are mean SD). G, BTK IHC of human HCC TMA (magnification: 100 and 400). H, Distribution and quantification of human liver (n ¼ 24) and HCC (n ¼ 86) TMA samples with IHC scores (t test, data are mean SD).

PD-1 inhibitors show improved results for HCC treatment in the Disclosure of Potential Conflicts of Interest second-line setting, in recently completed clinical trials, PD1 inhibitors Anne M. Noonan is a member of the advisory board of Exelixis, QED failed to improve HCC patient survival in the first-in-line setting. In Therapeutics, and Helsinn. No potential conflicts of interest were disclosed by addition, not all patients are candidates for immune therapies, either the other authors. due to history of orthotopic liver transplant or autoimmune disorders. Authors’ Contributions Thus, developing novel combination therapy to improve survival fi Conception and design: C.-H. Lin, X.-F. Bai, K. Ghoshal bene t of sorafenib is important and relevant. Our data in preclinical Development of methodology: C.-H. Lin, N.A. Wani, D. Li, P. Hu models clearly show that ibrutinib may potentiate anti-HCC thera- Acquisition of data (provided animals, acquired and managed patients, provided peutic efficacy of sorafenib. facilities, etc.): C.-H. Lin, K.H. Elkholy

394 Mol Cancer Ther; 19(2) February 2020 MOLECULAR CANCER THERAPEUTICS

Sorafenib/Ibrutinib Combination Therapy for HCC

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, and bioluminescence imaging of tumors. This work was supported, in part, by computational analysis): C.-H. Lin, K.H. Elkholy, J.M. Barajas, L. Yu, X. Zhang, grants R01CA193244 (to K. Ghoshal) and R01CA086978 (to S.T. Jacob and K. W.N. Khan Ghoshal) from the NCI, a seed grant from Translational Therapeutics program of Writing, review, and/or revision of the manuscript: C.-H. Lin, K.H. Elkholy, OSUCCC and Intramural Funding Program from Sylvester Comprehensive J.M. Barajas, L. Yu, X. Zhang, S.T. Jacob, W.N. Khan, X.-F. Bai, A.M. Noonan, Cancer Center, the University of Miami (to W.N. Khan). J.M. Barajas was a K. Ghoshal Howard Hughes Medical Institute Gilliam fellow. K.H. Elkholy was supported by Administrative, technical, or material support (i.e., reporting or organizing data, a PhD scholarship from the Ministry of Higher Education and National Research constructing databases): C.-H. Lin, K.H. Elkholy, N.A. Wani, D. Li, P. Hu Center in Egypt.

Acknowledgments The costs of publication of this article were defrayed in part by the payment of page We thank Dr. Amy Johnson and Dr. Hui-Lung Sun for kindly providing charges. This article must therefore be hereby marked advertisement in accordance ibrutinib and Akt plasmids, respectively. We thank Jianmin Zhu and Pietrzak with 18 U.S.C. Section 1734 solely to indicate this fact. Maciej for their assistance with ﬂow cytometry and RNA-seq analysis, respectively. We thank analytic cytometry, target validation, and small-animal imaging shared resources at the Ohio State University Comprehensive Cancer Center Received February 9, 2019; revised August 12, 2019; accepted September 27, 2019; (supported by P30CA016058 from NCI) for providing FACS analysis, NSG mice, published ﬁrst October 3, 2019.

References 1. Dhanasekaran R, Nault JC, Roberts LR, Zucman-Rossi J. Genomic medicine and 17. Masso-Valles D, Jauset T, Serrano E, Sodir NM, Pedersen K, Affara NI, et al. implications for hepatocellular carcinoma prevention and therapy. Gastroen- Ibrutinib exerts potent antifibrotic and antitumor activities in mouse models of terology 2019;156:492–509. pancreatic adenocarcinoma. Cancer Res 2015;75:1675–81. 2. Kanwal F, Singal AG. Surveillance for hepatocellular carcinoma: current best 18. Gunderson AJ, Kaneda MM, Tsujikawa T, Nguyen AV, Affara NI, Ruffell B, et al. practice and future direction. Gastroenterology 2019;157:54–64. Bruton tyrosine kinase-dependent immune cell cross-talk drives pancreas 3. Akinyemiju T, Abera S, Ahmed M, Alam N, Alemayohu MA, Allen C, et al. The cancer. Cancer Discov 2016;6:270–85. burden of primary liver cancer and underlying etiologies from 1990 to 2015 at the 19.StiffA,TrikhaP,WesolowskiR,KendraK,HsuV,UppatiS,etal. global, regional, and national level: results from the Global Burden of Disease Myeloid-derived suppressor cells expressBruton'styrosinekinaseandcan Study 2015. JAMA Oncol 2017;3:1683–91. be depleted in tumor-bearing hosts by ibrutinib treatment. Cancer Res 4. Wilhelm SM, Adnane L, Newell P, Villanueva A, Llovet JM, Lynch M. Preclinical 2016;76:2125–36. overview of sorafenib, a multikinase inhibitor that targets both Raf and VEGF 20. Reyes R, Wani NA, Ghoshal K, Jacob ST, Motiwala T. Sorafenib and 2-deox- and PDGF receptor tyrosine kinase signaling. Mol Cancer Ther 2008;7:3129–40. yglucose synergistically inhibit proliferation of both sorafenib sensitive and 5. Kudo M, Finn RS, Qin S, Han KH, Ikeda K, Piscaglia F, et al. Lenvatinib versus resistant HCC cells by inhibiting ATP production. Gene Expr 2017;17:129–40. sorafenib in first-line treatment of patients with unresectable hepatocellular 21. Slinker BK. The statistics of synergism. J Mol Cell Cardiol 1998;30:723–31. carcinoma: a randomised phase 3 non-inferiority trial. Lancet 2018;391:1163–73. 22. Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory 6. Bruix J, Qin S, Merle P, Granito A, Huang YH, Bodoky G, et al. Regorafenib for requirements. Nat Methods 2015;12:357–60. patients with hepatocellular carcinoma who progressed on sorafenib treatment 23. Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program (RESORCE): a randomised, double-blind, placebo-controlled, phase 3 trial. for assigning sequence reads to genomic features. Bioinformatics 2014;30:923– Lancet 2017;389:56–66. 30. 7. Abou-Alfa GK, Meyer T, Cheng AL, El-Khoueiry AB, Rimassa L, Ryoo BY, et al. 24. Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, et al. limma powers Cabozantinib in patients with advanced and progressing hepatocellular carci- differential expression analyses for RNA-sequencing and microarray studies. noma. N Engl J Med 2018;379:54–63. Nucleic Acids Res 2015;43:e47. 8. El-Khoueiry AB, Sangro B, Yau T, Crocenzi TS, Kudo M, Hsu C, et al. Nivolumab 25. Gordon A, Glazko G, Qiu X, Yakovlev A. Control of the mean number of false in patients with advanced hepatocellular carcinoma (CheckMate 040): an open- discoveries, Bonferroni and stability of multiple testing. Ann Appl Stat 2007;1: label, non-comparative, phase 1/2 dose escalation and expansion trial. Lancet 179–90. 2017;389:2492–502. 26. Ito Y, Takeda T, Sakon M, Tsujimoto M, Higashiyama S, Noda K, et al. 9. Zhu AX, Finn RS, Edeline J, Cattan S, Ogasawara S, Palmer D, et al. Pembro- Expression and clinical significance of erb-B receptor family in hepatocellular lizumab in patients with advanced hepatocellular carcinoma previously treated carcinoma. Br J Cancer 2001;84:1377–83. with sorafenib (KEYNOTE-224): a non-randomised, open-label phase 2 trial. 27. Chen KF, Chen HL, Tai WT, Feng WC, Hsu CH, Chen PJ, et al. Activation of Lancet Oncol 2018;19:940–52. phosphatidylinositol 3-kinase/Akt signaling pathway mediates acquired resis- 10. Finn RS, Ryoo BY, Merle P, Kudo M, Bouattour M, Lim HY, et al. Results of tance to sorafenib in hepatocellular carcinoma cells. J Pharmacol Exp Ther 2011; KEYNOTE-240: phase 3 study of pembrolizumab (Pembro) vs. best supportive 337:155–61. care (BSC) for second line therapy in advanced hepatocellular carcinoma (HCC). 28. Hindupur SK, Colombi M, Fuhs SR, Matter MS, Guri Y, Adam K, et al. The J Clin Oncol 2019;37:4004. protein histidine phosphatase LHPP is a tumour suppressor. Nature 2018;555: 11. Advani RH, Buggy JJ, Sharman JP, Smith SM, Boyd TE, Grant B, et al. Bruton 678–82. tyrosine kinase inhibitor ibrutinib (PCI-32765) has significant activity in patients 29. Rizzolio S, Battistini C, Cagnoni G, Apicella M, Vella V, Giordano S, et al. with relapsed/refractory B-cell malignancies. J Clin Oncol 2013;31:88–94. Downregulating Neuropilin-2 triggers a novel mechanism enabling EGFR- 12. Awan FT, Byrd JC. New strategies in chronic lymphocytic leukemia: shifting dependent resistance to oncogene-targeted therapies. Cancer Res 2018;78: treatment paradigms. Clin Cancer Res 2014;20:5869–74. 1058–68. 13. Honigberg LA, Smith AM, Sirisawad M, Verner E, Loury D, Chang B, et al. The 30. Shostak K, Zhang X, Hubert P, Goktuna SI, Jiang Z, Klevernic I, et al. NF-kB- Bruton tyrosine kinase inhibitor PCI-32765 blocks B-cell activation and is induced KIAA1199 promotes survival through EGFR signalling. Nat Commun efficacious in models of autoimmune disease and B-cell malignancy. 2014;5:5232. Proc Natl Acad Sci U S A 2010;107:13075–80. 31. Chou TC. Drug combination studies and their synergy quantification using the 14. Rauf F, Festa F, Park JG, Magee M, Eaton S, Rinaldi C, et al. Ibrutinib inhibition of Chou-Talalay method. Cancer Res 2010;70:440–6. ERBB4 reduces cell growth in a WNT5A-dependent manner. Oncogene 2018;37: 32. Wilhelm SM, Dumas J, Adnane L, Lynch M, Carter CA, Schutz G, et al. 2237–50. Regorafenib (BAY 73-4506): a new oral multikinase inhibitor of angiogenic, 15. Masso-Valles D, Jauset T, Soucek L. Ibrutinib repurposing: from B-cell malig- stromal and oncogenic receptor tyrosine kinases with potent preclinical anti- nancies to solid tumors. Oncoscience 2016;3:147–8. tumor activity. Int J Cancer 2011;129:245–55. 16. Sagiv-Barfi I, Kohrt HE, Czerwinski DK, Ng PP, Chang BY, Levy R. Therapeutic 33. Ma S, Lee TK, Zheng BJ, Chan KW, Guan XY. CD133þ HCC cancer stem cells antitumor immunity by checkpoint blockade is enhanced by ibrutinib, an confer chemoresistance by preferential expression of the Akt/PKB survival inhibitor of both BTK and ITK. Proc Natl Acad Sci U S A 2015;112:E966–72. pathway. Oncogene 2007;27:1749–58.

AACRJournals.org Mol Cancer Ther; 19(2) February 2020 395

Lin et al.

34. ZhaiB,HuF,JiangX,XuJ,ZhaoD,LiuB,et al. Inhibition of Akt reverses the 42. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Sorafenib in acquired resistance to sorafenib by switching protective autophagy to advanced hepatocellular carcinoma. N Engl J Med 2008;359:378–90. autophagic cell death in hepatocellular carcinoma. Mol Cancer Ther 2014; 43. Huang P, Xu X, Wang L, Zhu B, Wang X, Xia J. The role of EGF-EGFR signalling 13:1589–98. pathway in hepatocellular carcinoma inflammatory microenvironment. J Cell 35. Durand RE, Olive PL. Resistance of tumor cells to chemo- and radiotherapy Mol Med 2014;18:218–30. modulated by the three-dimensional architecture of solid tumors and spheroids. 44. Wang A, Yan XE, Wu H, Wang W, Hu C, Chen C, et al. Ibrutinib targets mutant- Methods Cell Biol 2001;64:211–33. EGFR kinase with a distinct binding conformation. Oncotarget 2016;7:69760–9. 36. Perche F, Torchilin VP. Cancer cell spheroids as a model to evaluate chemo- 45. Gao W, Wang M, Wang L, Lu H, Wu S, Dai B, et al. Selective antitumor activity of therapy protocols. Cancer Biol Ther 2012;13:1205–13. ibrutinib in EGFR-mutant non–small cell lung cancer cells. J Natl Cancer Inst 37. Tovar V, Cornella H, Moeini A, Vidal S, Hoshida Y, Sia D, et al. Tumour 2014;106:dju204. initiating cells and IGF/FGF signalling contribute to sorafenib resistance in 46. Garnelo M, Tan A, Her Z, Yeong J, Lim CJ, Chen J, et al. Interaction between hepatocellular carcinoma. Gut 2017;66:530–40. tumour-infiltrating B cells and T cells controls the progression of hepatocellular 38. Huynh H, Ngo VC, Koong HN, Poon D, Choo SP, Thng CH, et al. Sorafenib and carcinoma. Gut 2017;66:342–51. rapamycin induce growth suppression in mouse models of hepatocellular 47. Rohr-Udilova N, Klinglmuller F, Schulte-Hermann R, Stift J, Herac M, Salzmann carcinoma. J Cell Mol Med 2009;13:2673–83. M, et al. Deviations of the immune cell landscape between healthy liver and 39. Kim E, Hurtz C, Koehrer S, Wang Z, Balasubramanian S, Chang BY, et al. hepatocellular carcinoma. Sci Rep 2018;8:6220. Ibrutinib inhibits pre-BCR(þ) B-cell acute lymphoblastic leukemia progression 48. Foerster F, Hess M, Gerhold-Ay A, Marquardt JU, Becker D, Galle PR, et al. The by targeting BTK and BLK. Blood 2017;129:1155–65. immune contexture of hepatocellular carcinoma predicts clinical outcome. 40. Ruella M, Kenderian SS, Shestova O, Fraietta JA, Qayyum S, Zhang Q, et al. The Sci Rep 2018;8:5351. addition of the BTK inhibitor ibrutinib to anti-CD19 chimeric antigen receptor T 49. Dubovsky JA, Beckwith KA, Natarajan G, Woyach JA, Jaglowski S, Zhong Y, et al. cells (CART19) improves responses against mantle cell lymphoma. Clin Cancer Ibrutinib is an irreversible molecular inhibitor of ITK driving a Th1-selective Res 2016;22:2684–96. pressure in T lymphocytes. Blood 2013;122:2539–49. 41. Li X, Yao W, Yuan Y, Chen P, Li B, Li J, et al. Targeting of tumour-infiltrating 50. Xiao X, Lao XM, Chen MM, Liu RX, Wei Y, Ouyang FZ, et al. PD-1hi identifies a macrophages via CCL2/CCR2 signalling as a therapeutic strategy against novel regulatory B-cell population in human hepatoma that promotes disease hepatocellular carcinoma. Gut 2017;66:157–67. progression. Cancer Discov 2016;6:546–59.

396 Mol Cancer Ther; 19(2) February 2020 MOLECULAR CANCER THERAPEUTICS

Ibrutinib Potentiates Antihepatocarcinogenic Efficacy of Sorafenib by Targeting EGFR in Tumor Cells and BTK in Immune Cells in the Stroma

Cho-Hao Lin, Khadija H. Elkholy, Nissar A. Wani, et al.

Mol Cancer Ther 2020;19:384-396. Published OnlineFirst October 3, 2019.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-19-0135

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2019/10/03/1535-7163.MCT-19-0135.DC1

Cited articles This article cites 50 articles, 19 of which you can access for free at: http://mct.aacrjournals.org/content/19/2/384.full#ref-list-1

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 Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/19/2/384. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.