Abstract Bromodomain and extraterminal (BET) proteins recognize acetylated lysine residues for the purpose of transcriptional regulation of genes, including those involved in stem cell renewal and oncogenes such as MYC. Inhibition of BET proteins BRD2, BRD3, BRD4, and BRDT represents a promising treatment option for patients with cancer. However, little is known about the consequence of BET inhibition on normal stem cell renewal processes, such as hematopoiesis. This study explored the mechanistic effects of BET inhibition on bone marrow (BM) hematopoiesis. Rats treated with BET inhibitors JQ1, BMS-X, or BMS-986158 for 4 days in vivo showed dose-dependent pan-cellular BM atrophy and reduction of hematopoietic progenitors of myeloid and erythroid lineage (combination of anti-rat CD45, CD11b, anti-granulocyte, CD71, anti-erythroid, and CD90 via flow cytometry) and consequential reductions in circulating platelet and reticulocyte counts, with complete reversibility within 10 days of stopping treatment with BET inhibitors. Primary rat BM stem and progenitor cells treated with BET inhibitors in vitro were evaluated with the colony-forming unit assay and resulted in dose-dependent reduction of multiple lineage progenitor colonies, especially the erythroid and megakaryocyte lineages. To further elucidate pathways involved in BET-related BM atrophy, erythropoiesis and thrombopoiesis genes regulated by GATA1, a BRD-associated transcription factor, from rat BM, as well as rat and human whole-blood samples exposed to BET inhibitor(s) were evaluated via RNAseq and RT-PCR. Dose-dependent responses in genes involved in erythropoiesis (Alas2, ABCme, PBG-D, HMBS) and thrombopoiesis (NFE2, PF4, GP1Bb, MPL) were observed after 4 days of treatment with BMS-986158. In a clinical trial (NCT02419417), patients with solid tumors treated with BMS-986158 demonstrated reversible thrombocytopenia and downregulation of NFE2, PF4, and HMBS expression, similar to that observed in rats. GATA1 was also downregulated in rat BM, with target engagement in rat and human demonstrated by the induction of HEXIM1 transcription, a pharmacodynamic (PD) biomarker of growth inhibition and apoptosis induced by BRD4 inhibition. Overall, our results suggest inhibition of BET signaling causes target-related, dose-dependent repression of hematopoietic progenitors through alterations of GATA1-associated erythropoiesis and thrombopoiesis regulation in rat BM and human blood samples, and these effects are reversible on cessation of BET inhibitor treatment. This is the first in vivo study demonstrating the mechanism of BET inhibition resulting in GATA1-associated repression of hematopoietic progenitors that is correlated to clinical pharmacokinetics and PD (Chen X, et al. AACR 2020) and is translatable from preclinical evaluation to clinical experience. Citation Format: Cindy Zhang, Ke Xu, Julie Panzica-Kelly, Jennifer Price, Denise Bounous, Shodeinde Coker, Kezi Unsal-Kacmaz, Danielle Greenawalt, Ashvinikumar Gavai, Ronald Fleming, Karen Augustine-Rauch, Richard Westhouse. Inhibition of BET signaling leads to reversible GATA1-associated repression of hematopoietic progenitors: translation from preclinical assessment to clinical development [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 1742.

Abstract Background: BMS-986158 is a potent, selective, and orally bioavailable small-molecule inhibitor of the bromodomain and extraterminal (BET) family of transcription modulators being evaluated in patients with various tumors in a phase 1 clinical study (NCT02419417). An integrated PK/PD analysis leveraging PK, biomarker, and safety data from the phase 1 study was performed to support dose and schedule (Sch) selection further clinical studies with BMS-986158. Methods: Five doses (0.75-4.5 mg) and 3 dosing regimens (Sch A: 5 days on, 2 days off over 21 days; Sch B: 14 days on, 7 days off; Sch C: 7 days on, 14 days off) were evaluated. Reversible thrombocytopenia (TTP) was the primary safety signal observed, and peripheral gene expression modulation was a PD biomarker indicative of drug target engagement. BMS-986158 serum PK was characterized with a 2-compartment population PK (PPK) model. A semimechanistic PK/PD model describing the platelet-reducing effect with BMS-986158 was developed to characterize platelet profiles for individual patients and to simulate and project the incidence of TTP at different doses and dosing regimens. The association of BMS-986158 exposure with expression modulation of select peripheral BET target genes, including HEXM1 and CCR2, was also examined. Results: BMS-986158 exhibited linear PK with rapid oral absorption (Tmax ≈ 2-4 h) and a terminal half-life of ≈ 60 h over the dose range of 0.75-4.5 mg. PPK modeling and simulation suggested that at the same dose level, Sch A led to comparable Cmax, with a higher Ctrough and Cavg at steady state compared with Sch B and C across the dosing interval. The semimechanistic PK/PD model predicted a higher incidence rate of TTP with Sch A than Sch C at the same dose level, and the predictions agreed with observed primary safety data from the phase 1 study. The model predicted grade 4 TTP incidence rates of 24% (95% CI, 21%-26%) and 43% (95% CI, 38%-45%) at 4.5 and 6 mg, respectively, with Sch A. This suggests that 4.5 mg would be the maximum tolerated dose, given that a 6 mg dose was predicted to exceed the target dose-limiting toxicity (DLT) rate of grade 4 TTP at 27%. The exploratory analysis of BMS-986158 exposure and expression of selected peripheral genes associated with the BET pathway suggested a direct association between BMS-986158 exposure and the magnitude of peripheral gene expression modulation. Conclusions: An integrated PK/PD analysis of BMS-986158 incorporating incidence of reversible TTP and modulation of peripheral BET target genes identified a maximum dose of 4.5 mg at Sch A for further clinical development. With the assumption that sustained gene expression modulation is desirable for BMS-986158 efficacy, Sch A, which provides sustained drug exposure, is recommended to enable continuous BET target gene expression modulation. Citation Format: Xi (Cindy) Chen, Lora Hamuro, Shodeinde Coker, John Hilton, Jennifer R. Diamond, Capucine Baldini, Mark Voskoboynik, Mihaela Cristea, William Edenfield, Kezi Unsal-Kacmaz, Donald Jackson, Abraham Apfel, Ke Xu, Li Zhu, Amit Roy, Akintunde Bello, Ronald Fleming, Paul Statkevich. Integrated pharmacokinetic (PK)/pharmacodynamic (PD) modeling leveraging PK, biomarker, and safety data to support dose and schedule selection for the BET inhibitor BMS-986158 [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3021.

Abstract Background: Mesothelin, a GPI-anchored cell surface protein, is highly expressed in several tumor types and can be targeted for antibody (ab)-based cancer therapy. BMS-986148 is a fully human IgG1 anti-mesothelin monoclonal ab conjugated to tubulysin to promote selective cytotoxic delivery to tumor cells. In preclinical models, combination of anti–mesothelin-tubulysin with anti–PD-1 promoted a synergistic antitumor response and influx of tumor-infiltrating lymphocytes. Here, we present initial data for BMS-986148 ± nivolumab (NIVO; anti–PD-1) from a phase 1/2a trial in a biomarker-defined population of patients (pts) with select advanced solid tumors (NCT02341625). Methods: During dose escalation (ESC), pts received BMS-986148 0.1 to 1.6 mg/kg IV Q3W, BMS-986148 0.4 or 0.8 mg/kg IV Q1W, or BMS-986148 0.8 mg/kg + NIVO 360 mg IV Q3W. During monotherapy (mono) dose expansion (EXP), pts with mesothelin-selected mesothelioma (ie, H score ≥ 100 for tumor mesothelin expression), non-small cell lung cancer (NSCLC), or ovarian cancer received BMS-986148 1.2 mg/kg mono IV Q3W, and during combination (combo) dose EXP, pts with mesothelin-selected mesothelioma or pancreatic cancer received BMS-986148 0.8 mg/kg + NIVO 360 mg combo IV Q3W. The primary endpoint was safety. Secondary endpoints included pharmacokinetics (PK) and antitumor activity. Results: As of April 1, 2019, 126 pts have been treated in this trial; 96 pts received BMS-986148 mono (Q3W, n = 84; Q1W, n = 12), and 30 pts received BMS-986148 + NIVO combo. PK analysis in pts treated during ESC demonstrated that total antibody and active ADC exposures increased in a dose-proportional manner. Unconjugated tubulysin concentrations were low but sustained over the dosing interval. The maximum tolerated dose (MTD) for mono was 1.2 mg/kg Q3W, and the 0.8 mg/kg dose was determined to be tolerable for the combo. Across all cohorts, any-grade and grade 3/4 treatment-related AEs (TRAEs) were reported in 87% and 44% of all pts, respectively. Overall, the most common TRAEs (any grade; grade 3-4) reported in ≥ 15% of pts were AST increased (43%; 17%), ALT increased (41%; 16%), fatigue (37%; 5%), nausea (29%; 0%), decreased appetite (22%; 1%), and blood alkaline phosphatase increased (18%; 4%). The frequency of TRAEs and dose-limiting toxicities related to liver function tests increased with increasing dose of BMS-986148. Most TRAEs resolved with dose interruption, dose reduction, or treatment discontinuation. Serious TRAEs were reported in 18% of pts with mono and 23% of pts with combo. Fifteen percent of pts discontinued due to TRAEs, and 1 pt died due to a TRAE (pneumonitis; 1.2 mg/kg Q3W mono). The ORR was 6% (0 CR, 3 PR) with mono in EXP and 20% (0 CR, 6 PR) with the combo in ESC and EXP. Among pts with mesothelioma, the ORR was 4% with mono in EXP and 31% with combo in ESC and EXP. Among pts with ovarian carcinoma, the ORR was 9% with mono in EXP. Of note, durable responses lasting up to ≈ 20 mo in pts with ovarian cancer with mono and ≈ 9 mo in pts with mesothelioma with mono and combo were observed. Conclusions: BMS-986148 ± NIVO was tolerable and demonstrated a clinically manageable safety profile. Preliminary clinical activity was observed with BMS-986148 ± NIVO in select patient populations, including pts with mesothelioma. Citation Format: Jeffrey Clarke, Siu-Chung Chu, Lillian L Siu, Jean-Pascal Machiels, Benjamin Markman, Kimberley Heinhuis, Michael Millward, Martijn Lolkema, Sandip Pravin Patel, Paul de Souza, Giuseppe Curigliano, Armando Santoro, Michelle Brown, Ronald Fleming, Heather Vezina, Chunsheng He, Sylvie Rottey. BMS-986148, an anti-mesothelin antibody-drug conjugate (ADC), alone or in combination with nivolumab demonstrates clinical activity in patients with select advanced solid tumors [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference on Molecular Targets and Cancer Therapeutics; 2019 Oct 26-30; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2019;18(12 Suppl):Abstract nr B057. doi:10.1158/1535-7163.TARG-19-B057

BACKGROUND: Focal adhesion kinase (FAK) is overexpressed in glioblastoma (GBM) and is associated with poor prognosis. The FAK inhibitor GSK2256098 inhibits growth, migration, and invasion and induces apoptosis in a subset of GBM cell lines. Pharmacokinetic (PK) studies in mice and rats with an intact blood brain barrier indicate that the penetration of GSK2256098 into the CNS is poor. The purpose of this study was to determine the safety, systemic PK, and distribution of GSK2256098 into the brain and tumor tissue in patients (pts) with GBM (NCT01138033). METHODS: Pts with recurrent GBM received 750-1000mg of GSK2256098 orally, twice daily until disease progression or withdrawal due to adverse events (AEs). Serial PK samples were collected in blood on Days 1 and 15. On a single day between Days 9-20, pts received a microdose (≤10 µg) of 11C-GSK2256098 (≤ 500MBq) by IV administration and scanned by PET over 90 minutes. Pts were assessed for safety throughout the study and disease assessment scans were performed every six weeks. RESULTS: As of the March 15, 2015 data cutoff, 10 pts were enrolled (7 males, 3 females; median age 48.5 y, range 33-65 y) and 4 and 6 pts received 750 mg and 1000 mg BID of GSK2256098 respectively. AEs (≥25%) regardless of causality were fatigue (50%), diarrhea (40%), vomiting (40%), headache (30%), and somnolence (30%). Based on radioactivity data obtained during PET in 4 pts (3 at 1000mg, 1 at 750mg BID), GSK2256098 is estimated to achieve normal brain and tumor concentrations of 30-55% and 75-170% respectively of blood concentrations at equilibrium. No structural radiologic responses have been seen to date. One pt remains on study for 6 months. CONCLUSIONS: GSK2256098 achieves concentrations in tumor exceeding those associated with preclinical activity. Evaluation of lower doses of GSK2256098 is ongoing.