Search

Purpose: This global phase I trial investigated the safety, efficacy, pharmacokinetics, and pharmacodynamics of lisaftoclax (APG-2575), a novel, orally active, potent selective B-cell lymphoma 2 (BCL-2) inhibitor, in patients with relapsed or refractory chronic lymphocytic leukemia or small lymphocytic lymphoma (R/R CLL/SLL) and other hematologic malignancies (HMs). Patients and Methods: Maximum tolerated dose (MTD) and recommended phase II dose were evaluated. Outcome measures were safety and tolerability (primary) and pharmacokinetic variables and antitumor effects (secondary). Pharmacodynamics in patient tumor cells were explored. Results: Among 52 patients receiving lisaftoclax, MTD was not reached. Treatment-emergent adverse events (TEAEs) included diarrhea (48.1%), fatigue (34.6%), nausea (30.8%), anemia and thrombocytopenia (28.8% each), neutropenia (26.9%), constipation (25.0%), vomiting (23.1%), headache (21.2%), peripheral edema and hypokalemia (17.3% each), and arthralgia (15.4%). Grade ≥ 3 hematologic TEAEs included neutropenia (21.2%), thrombocytopenia (13.5%), and anemia (9.6%), none resulting in treatment discontinuation. Clinical pharmacokinetic and pharmacodynamic results demonstrated that lisaftoclax had a limited plasma residence and systemic exposure and elicited rapid clearance of malignant cells. With a median treatment of 15 (range, 6–43) cycles, 14 of 22 efficacy-evaluable patients with R/R CLL/SLL experienced partial responses, for an objective response rate of 63.6% and median time to response of 2 (range, 2–8) cycles. Conclusions: Lisaftoclax was well tolerated, with no evidence of tumor lysis syndrome. Dose-limiting toxicity was not reached at the highest dose level. Lisaftoclax has a unique pharmacokinetic profile compatible with a potentially more convenient daily (vs. weekly) dose ramp-up schedule and induced rapid clinical responses in patients with CLL/SLL, warranting continued clinical investigation.

Pedestrian accidents contribute significantly to the high number of annual traffic casualties. It is therefore crucial for pedestrians to use safety measures, such as a crosswalk, and to activate pedestrian signals. However, people often fail to activate the signal or are unable to do so – those who are visually impaired or have occupied hands may be unable to activate the system. Failure to activate the signal can result in an accident. This paper proposes an improvement to crosswalk safety by designing a system that can detect pedestrians and trigger the pedestrian signal automatically when necessary. In this study, a dataset of images was collected in order to train a Convolutional Neural Network (CNN) to distinguish between pedestrians (including bicycle riders) when crossing the street. The resulting system can capture and evaluate images in real-time, and the result can be used to automatically activate a system such as a pedestrian signal. A threshold system is also implemented that triggers the crosswalk only when the positive predictions pass the threshold. This system was tested by deploying it at three real-world environments and comparing the results with a recorded video of the camera’s view. The CNN prediction model is able to correctly predict pedestrian and cyclist intentions with an average accuracy of 84.96% and had an absence trigger rate of 0.037%. The prediction accuracy varies based on the location and whether a cyclist or pedestrian is in front of the camera. Pedestrians crossing the street were correctly predicted more accurately than cyclists crossing the street by up to 11.61%, while passing (i.e., non-crossing) cyclists were correctly ignored more than passing pedestrians, by up to 18.75%. Based on the testing of the system in real-world environments, the authors conclude that it is feasible as a back-up system that can complement existing pedestrian signal buttons, and thereby improve the overall safety of crossing the street. Further improvements to the accuracy can be achieved with a more comprehensive dataset for a specific location where the system is deployed. Implementing different computer vision techniques optimized for tracking objects should also increase the accuracy.

These are all the Supplementary Figures: S1, S2, S3, S4, S5, S6, and S7

Summary of BH3 profiling for PBMC samples collected at baseline.

Purpose:This global phase I trial investigated the safety, efficacy, pharmacokinetics, and pharmacodynamics of lisaftoclax (APG-2575), a novel, orally active, potent selective B-cell lymphoma 2 (BCL-2) inhibitor, in patients with relapsed or refractory chronic lymphocytic leukemia or small lymphocytic lymphoma (R/R CLL/SLL) and other hematologic malignancies (HMs).Patients and Methods:Maximum tolerated dose (MTD) and recommended phase II dose were evaluated. Outcome measures were safety and tolerability (primary) and pharmacokinetic variables and antitumor effects (secondary). Pharmacodynamics in patient tumor cells were explored.Results:Among 52 patients receiving lisaftoclax, MTD was not reached. Treatment-emergent adverse events (TEAEs) included diarrhea (48.1%), fatigue (34.6%), nausea (30.8%), anemia and thrombocytopenia (28.8% each), neutropenia (26.9%), constipation (25.0%), vomiting (23.1%), headache (21.2%), peripheral edema and hypokalemia (17.3% each), and arthralgia (15.4%). Grade ≥ 3 hematologic TEAEs included neutropenia (21.2%), thrombocytopenia (13.5%), and anemia (9.6%), none resulting in treatment discontinuation. Clinical pharmacokinetic and pharmacodynamic results demonstrated that lisaftoclax had a limited plasma residence and systemic exposure and elicited rapid clearance of malignant cells. With a median treatment of 15 (range, 6–43) cycles, 14 of 22 efficacy-evaluable patients with R/R CLL/SLL experienced partial responses, for an objective response rate of 63.6% and median time to response of 2 (range, 2–8) cycles.Conclusions:Lisaftoclax was well tolerated, with no evidence of tumor lysis syndrome. Dose-limiting toxicity was not reached at the highest dose level. Lisaftoclax has a unique pharmacokinetic profile compatible with a potentially more convenient daily (vs. weekly) dose ramp-up schedule and induced rapid clinical responses in patients with CLL/SLL, warranting continued clinical investigation.

GISTs are common malignant mesenchymal tumors that occur in the GI tract, with an estimated incidence rate of 1% to 2%. About 75% to 80% of patients with GIST have mutations in the KIT gene, while 5% to 10% have mutations in the platelet-derived growth factor receptor α (PDGFRA) gene. Tyrosine kinase inhibitors (TKIs) offer patients an improved quality of life but increased pharmacological resistance to TKIs is often observed. BCL-2 is expressed in >80% of GISTs, and amplification of BCL-2 and BCL-xL are common features associated with disease progression. One approach to enhance GIST eradication is to concurrently inhibit oncogenic KIT signaling while actively engaging apoptotic pathways. Olverembatinib (HQP1351) is a new, third-generation TKI that targets BCR-ABL1, KIT and PDGFRA and is currently in development for relapsed or refractory chronic myeloid leukemia and GIST. Lisaftoclax (APG-2575) is a selective BCL-2 inhibitor under development for hematologic malignancies. The purpose of this study was to evaluate whether combining a BCL-2 inhibitor, lisaftoclax, with olverembatinib enhances treatment effect on imatinib-resistant GISTs. Compared to imatinib-sensitive GIST cell lines (GIST882 and GIST-T1), increased BCL-2 protein expression was observed in imatinib-resistant GIST cells, particularly in GIST430, GIST48, and GIST48B lines. Cell-based antiproliferation studies showed additive activity of olverembatinib and lisaftoclax in imatinib-resistant GIST cells. In vivo studies using imatinib-resistant GIST430 cell-derived xenograft models further revealed that coadministration of olverembatinib at 15 mg/kg (every other day for 3 weeks) with lisaftoclax at 50 mg/kg (daily for 3 weeks) resulted in enhanced antitumor effects compared to either agent alone. Tumor growth inhibition (TGI) rates for combined treatment with olverembatinib and lisaftoclax reached 76.8%, which was superior to the TGI rate of either agent alone: 57.6% and 31.2% for olverembatinib and lisaftoclax, respectively. Regarding putative mechanisms, olverembatinib treatment was shown to downregulate BCL-xL and MCL-1 protein expression, possibly via inhibition of KIT signal pathway and downstream signal transducer and activator of transcription 3 (STAT3). Lisaftoclax triggered apoptosis by disrupting the BCL-2:BIM complex. Therefore, selective inhibition of BCL-2, aided by TKI-mediated MCL-1 and BCL-XL inhibition, markedly augmented cleavage of caspase-3 and poly (ADP-ribose) polymerase 1 (PARP-1), thus triggering apoptosis and subsequently enhancing antitumor effects. Our results demonstrate that olverembatinib and BCL-2 inhibitor lisaftoclax have additive antitumor effects in imatinib-resistant GIST. This novel dual approach may have the potential for treating patients with GISTs whose disease has progressed after treatment with TKIs. Citation Format: Ping Min, Guangfeng Wang, Eric Liang, Bingxing Wu, Huidan Yu, Dajun Yang, Yifan Zhai. Combination of olverembatinib (HQP1351) with BCL-2 inhibitor lisaftoclax (APG-2575) overcomes resistance in gastrointestinal stromal tumors (GISTs) [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 1631.

In solid tumors, resistance to checkpoint inhibitors (CPIs) is frequently observed, partially due to upregulation of vascular endothelial growth factor A (VEGFA) and programmed death-ligand 1 (PD-L1). This culminates in an immunosuppressive tumor microenvironment and immune escape. Inhibitors against VEGF and the VEGF receptor (VEGFR) foster tumor vessel normalization and immunostimulatory reprogramming, in turn promoting treatment effects of immunotherapies. In recent years, TKIs, including axitinib, lenvatinib, and cabozantinib, plus immunotherapy have been approved to treat advanced RCC. Currently under clinical development for relapsed or refractory chronic myeloid leukemia and gastrointestinal tumor, olverembatinib (HQP1351) is a new-generation multikinase inhibitor with targets including VEGFR, fibroblast growth factor receptor (FGFR), SRC, BCR-ABL1, c-KIT, and platelet-derived growth factor receptor. The aim of this study was to assess whether olverembatinib combined with immunotherapy can promote inhibitory effects on RCC. In cell-free kinase assays, olverembatinib inhibited VEGFR1, -2, and -3 with IC50 values of 4.2, 6.1, and 4.1 nM, respectively. Compared to lenvatinib, olverembatinib had more potent antiproliferative effects on human umbilical vein endothelial cells. Olverembatinib also had antiproliferative activity in murine RCC lines RANCA and RAG, with IC50 values of 141 and 53 nM, respectively. When olverembatinib was coadministered with an anti-PD-1 antibody in a RANCA-derived syngeneic model, both agents exerted synergistic effects, with tumor growth inhibition rates reaching 60.6%. Mechanistically, olverembatinib influenced cancer cell proliferation directly by inhibiting phosphorylation of FGFR and downstream proteins. Increased cleavage of caspase-3 and poly (ADP-ribose) polymerase 1 were observed, suggesting induction of apoptosis. Olverembatinib also influenced proliferation of vascular endothelial cells by inhibiting phosphorylation of VEGFR, SRC, and downstream proteins Akt and extracellular signal-regulated kinases. Olverembatinib also reduced expression of PD-L1 in RCC cells. In tumor-infiltrating lymphocyte assays, olverembatinib increased numbers of cytotoxic T cells (CTL, CD8+) and natural-killer cells (NK, CD3−/CD49B+) in RANCA tumor tissues. Combined with an anti-PD-1 antibody, olverembatinib increased CTLs, NK cells, dendritic cells (DCs, MHC-II+/CD11C+), and M1 macrophages (F4/80+/CD11B+/CD86+) in RANCA tumor tissues, indicating an immunoregulatory effect of olverembatinib. Taken together, these data suggest that combining olverembatinib with a CPI confers synergistic antitumor effects in an RCC cancer mouse model by targeting tumor growth, angiogenesis, and immune regulation. This novel combination may provide an alternative approach to enhance treatment effects with CPIs in renal cancers. Citation Format: Guangfeng Wang, Eric Liang, Ping Min, Huidan Yu, Bingxing Wu, Dajun Yang, Yifan Zhai. Olverembatinib (HQP1351) enhances antitumor effects of immunotherapy in renal cell carcinoma (RCC). [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 5071.

Biotherapeutics are highly efficacious, but the pain and inconvenience of chronic injections lead to poor patient compliance and compromise effective disease management. Despite innumerable attempts, oral delivery of biotherapeutics remains unsuccessful due to their degradation in the gastrointestinal (GI) environment and poor intestinal absorption. We have developed an orally ingestible robotic pill (RP) for drug delivery, which protects the biotherapeutic drug payload from digestion in the GI tract and auto-injects it into the wall of the small intestine as a safe, pain-free injection since the intestines are insensate to sharp stimuli. The payload is delivered upon inflation of a balloon folded within the RP, which deflates immediately after drug delivery. Here we present results from two clinical studies demonstrating the safety, tolerability and performance of the RP in healthy humans. In the first study, three versions of the RP (A, B and C) were evaluated, which were identical in all respects except for the diameter of the balloon. The RP successfully delivered a biotherapeutic (octreotide) in 3 out of 12 subjects in group A, 10 out of 20 subjects in group B and 16 out of 20 subjects in group C, with a mean bioavailability of 65 ± 9% (based on successful drug deliveries in groups A and B). Thus, reliability of drug delivery with the RP ranged from 25 to 80%, with success rate directly related to balloon size. In a separate study, the deployment of the RP was unaffected by fed or fasting conditions suggesting that the RP may be taken with or without food. These promising clinical data suggest that biotherapeutics currently administered parenterally may be safely and reliably delivered via this versatile, orally ingestible drug delivery platform. Graphical abstract

Query optimizers rely on accurate cardinality estimates to produce good execution plans. Despite decades of research, existing cardinality estimators are inaccurate for complex queries, due to making lossy modeling assumptions and not capturing inter-table correlations. In this work, we show that it is possible to learn the correlations across all tables in a database without any independence assumptions. We present NeuroCard, a join cardinality estimator that builds a single neural density estimator over an entire database. Leveraging join sampling and modern deep autoregressive models, NeuroCard makes no inter-table or inter-column independence assumptions in its probabilistic modeling. NeuroCard achieves orders of magnitude higher accuracy than the best prior methods (a new state-of-the-art result of 8.5$\times$ maximum error on JOB-light), scales to dozens of tables, while being compact in space (several MBs) and efficient to construct or update (seconds to minutes)., Comment: VLDB 2021

Task-based distributed frameworks (e.g., Ray, Dask, Hydro) have become increasingly popular for distributed applications that contain asynchronous and dynamic workloads, including asynchronous gradient descent, reinforcement learning, and model serving. As more data-intensive applications move to run on top of task-based systems, collective communication efficiency has become an important problem. Unfortunately, traditional collective communication libraries (e.g., MPI, Horovod, NCCL) are an ill fit, because they require the communication schedule to be known before runtime and they do not provide fault tolerance. We design and implement Hoplite, an efficient and fault-tolerant collective communication layer for task-based distributed systems. Our key technique is to compute data transfer schedules on the fly and execute the schedules efficiently through fine-grained pipelining. At the same time, when a task fails, the data transfer schedule adapts quickly to allow other tasks to keep making progress. We apply Hoplite to a popular task-based distributed framework, Ray. We show that Hoplite speeds up asynchronous stochastic gradient descent, reinforcement learning, and serving an ensemble of machine learning models that are difficult to execute efficiently with traditional collective communication by up to 7.8x, 3.9x, and 3.3x, respectively., Comment: SIGCOMM 2021

Natural killer/T-cell lymphoma (NK/TCL) is one of the most common subtypes (10.4%) of peripheral T-cell lymphoma, which in turn accounts for 10% to 15% of all cases of non-Hodgkin lymphoma. NK/TCL occurs more frequently in Asia and Latin America than other regions, and although associated with Epstein-Barr virus (EBV) infection, NK/TCL has an unclear pathogenesis of genetic and molecular alterations. Asparaginase-based chemotherapy regimens are frequently used, typically with unsatisfactory outcomes. Novel targeted therapies, such as histone deacetylase (HDAC) inhibitors and programmed death-1 antibodies, are reported effective either as single agents or in combination with other agents. Studies have also shown that EBV-positive NK/TCL cell lines are B-cell lymphoma-extra large (BCL-xL) dependent. Other preclinical experiments have shown that BH3-mimetic drugs targeting BCL-xL-induced cell death in NK/TCL cell lines were effective in NK/TCL xenograft models. BCL-xL inhibitors have shown narrow therapeutic windows in clinical trials because of dose-limiting on-target thrombocytopenia. Pelcitoclax (APG-1252) is a novel dual BCL-2/BCL-xL inhibitor under clinical development for solid tumors. As a result of its unique prodrug design, APG-1252 can overcome the undesired platelet toxicity. This study evaluated the potential antitumor effect of APT-1252 in preclinical models of NK/TCL. Cell-based antiproliferation studies showed activity of APG-1252 and its more potent metabolite APG-1252-M1 toward NK/TCL cell lines that overexpressed BCL-xL. Half-maximal inhibitory concentrations (IC 50) for APG-1252 in SNK-1, SNK-6, and SNK-8 (EBV-positive NK/TCL) cell lines were 2.652 ± 2.606 µM, 1.568 ± 1.109 µM, and 0.557 ± 0.383 µM, respectively. Corresponding values for APG-1252-M1 were 0.133 ± 0.056 µM, 0.064 ± 0.014 µM, and 0.020 ± 0.008 µM, respectively. Mechanistic studies demonstrated that APG-1252 and APG-1252-M1 disrupted the complex of BCL-xL/BCL-2-associated X protein (Bax) and BCL-xL/BCL-2 homologous antagonist killer protein (Bak) in SNK-6 cells, thereby liberating these proapoptotic proteins and further activating downstream apoptosis pathways by cleaving poly-ADP ribose polymerase-1 (PARP-1) and caspase-3. In an SNK-6 xenograft model, administration of APG-1252 at 65 mg/kg and 100 mg/kg either twice or once weekly resulted in significant antitumor effects, with tumor growth rate (T/C%) values ranging from 13.7% to 30.7%. Furthermore, the combination of APG-1252 with HDAC inhibitor chidamide or DDGP (dexamethasone, cisplatin, gemcitabine, and pegaspargase) chemotherapy demonstrated synergistic effects. Pharmacokinetic assessment in mice showed that APG-1252 had a long half-life in plasma (127 hours) and tumor tissues (25.2 hours), justifying intermittent dosing schedules used in vivo. Importantly, the transformation of APG-1252 to APG-1252-M1 was 16 times higher in tumor tissues compared to plasma (22% vs. 1.3%) after administration of APG-1252, thereby suggesting that APG-1252 can reduce platelet toxicity caused by APG-1252-M1 in plasma. In conclusion, APG-1252 has promising antitumor effects in NK/TCL, either as a single agent or in combination with an HDAC inhibitor or chemotherapy. These findings provide evidence to further evaluate APG-1252 as a potential treatment for NK/TCL. Disclosures Wang: Ascentage Pharma Co., Ltd., Shanghai: Current Employment, Current equity holder in publicly-traded company. Liang: Ascentage Pharma Group Inc.: Current Employment, Current equity holder in publicly-traded company. Ming: Ascentage Pharma Co., Ltd., Shanghai: Current Employment, Current equity holder in publicly-traded company. Rui: Ascentage Pharma Co., Ltd., Taizhou: Current Employment, Current equity holder in publicly-traded company. Tang: Ascentage Pharma Co., Ltd., Taizhou: Current Employment, Current equity holder in publicly-traded company. LV: Ascentage Pharma Co., Ltd., Shanghai: Current Employment, Current equity holder in publicly-traded company. Ge: Ascentage Pharma Co., Ltd., Shanghai: Current Employment, Current equity holder in publicly-traded company. Zhang: Ascentage Pharma Co., Ltd., Taizhou: Current Employment, Current equity holder in publicly-traded company. Wang: Ascentage Pharma Co., Ltd., Taizhou: Current Employment, Current equity holder in publicly-traded company. Shang: Ascentage Pharma Co., Ltd., Shanghai: Current Employment, Current equity holder in publicly-traded company. Yang: Ascentage Pharma (Suzhou) Co., Ltd.: Current Employment, Current equity holder in publicly-traded company, Other: Leadership and other ownership interests, Patents & Royalties, Research Funding. Zhai: Ascentage Pharma Group Inc.: Current Employment, Current equity holder in publicly-traded company, Other: Leadership and other ownership interests, Patents & Royalties, Research Funding; Ascentage Pharma (Suzhou) Co., Ltd.: Current Employment, Current equity holder in publicly-traded company, Other: Leadership and other ownership interests, Patents & Royalties, Research Funding.

Treatment with tyrosine kinase inhibitors (TKIs) directed against the ATP-binding site of BCR-ABL promotes recovery of Ph+ leukemia. However, emergence of gatekeeper mutation T315I and compound mutants confer resistance to these TKIs. The allosteric inhibitor, asciminib, effectively inhibits BCR-ABL kinase through binding to the myristoyl-binding site. Combining asciminib with ponatinib can overcome only a subset of the resistance caused by BCR-ABL compound mutants. Olverembatinib (HQP1351) is a new generation TKI targeting BCR-ABL and currently in development for r/r CML. The purpose of this study is to evaluate whether a novel combination of olverembatinib and asciminib, targeting both ATP pocket and allosteric region of BCR-ABL protein, can promote the inhibitory effect on the kinase harboring compound mutations. A series of cell lines expressing BCR-ABL single or compound mutations were constructed based on BaF3 murine pro-B cell line. Effect of olverembatinib as a single agent or in combination with asciminib were analyzed using antiproliferation assay, Western blot and FACS assays in vitro. In vivo efficacy was evaluated using the syngeneic mouse model derived from BaF3 cells with T315I or compound mutations. The cell based antiproliferation studies demonstrated superior activity of olverembatinib toward BCR-ABL single or compound mutations with IC50 values ranging between 6-300 nM. In particular, olverembatinib was more effective than ponatinib, asciminib and other TKIs against those compound mutations. Moreover, the combination of olverembatinib with asciminib was highly effective against BCR-ABL compound mutations, especially those containing T315I. In vivo studies further revealed that co-administration of olverembatinib with asciminib resulted in significant prolongation of survival compared to single agents. Importantly, the anti-tumor effect was more potent than that of ponatinib plus asciminib in models harboring compound mutations. In terms of mechanism, the combined treatment synergistically downregulated phosphorylation of BCR-ABL and the downstream proteins CRKL and STAT5, and augmented cleavage of Caspase-3 and PARP-1, thus triggered apoptosis and subsequently enhanced the antitumor effects. Our results demonstrated that the combination of ATP binding site inhibitor olverembatinib and allosteric inhibitor have synergistic anti-tumor effect on tumor cells harboring single or compound mutations in BCR-ABL. This novel strategy may help to overcome the secondary compound mutations post the treatment with TKIs. Citation Format: Guangfeng Wang, Jing Lv, Chunyang Tang, Ping Min, Li Rui, Fei Zhang, Lvcheng Wang, Yangfeng Ge, Kejie Lian, Eric Liang, Dajun Yang, Yifan Zhai. ATP-site inhibitor olverembatinib, HQP1351, enhanced the effect of allosteric inhibitor on the resistance conferred by the compound mutations of BCR-ABL [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 1463.

Background: Pelcitoclax (APG-1252), a novel dual inhibitor of Bcl-2/Bcl-xL, is active as monotherapy in patients with advanced solid tumors and well tolerated up to 240 mg twice weekly (NCT03387332). Preclinical data suggest that cells with Janus-associated kinase-2 (JAK2) mutations, including those associated with bone marrow fibrosis, are dependent on Bcl-2/Bcl-xL for survival and that addition of BH3 mimetics targeting Bcl-2/Bcl-xL induces apoptosis. Furthermore, in JAK2‒mutated cell models, apoptotic synergy is demonstrated when a JAK2 inhibitor and Bcl-2/Bcl-xL inhibitor are combined, as inhibition of Bcl-xL overcomes resistance to JAK2 inhibitors. Taken together, APG-1252 could overcome resistance to JAK2 inhibitors, and the combination could augment clinical benefit in patients with suboptimal responses to JAK2 inhibitor‒based therapy. Study Objectives: The primary objective of this open-label trial is to evaluate the safety and efficacy of APG-1252, as monotherapy and when combined with ruxolitinib, in adults with histologically or cytologically confirmed MF who require therapy and are ineligible for JAK2 inhibitors (and can receive single-agent APG-1252) or have had inadequate responses to ruxolitinib-based therapy (and can receive this treatment plus APG-1252). Secondary objectives include APG-1252 pharmacokinetics, time to response, and duration of response. Exploratory objectives include changes in cytogenetics and molecular mutations, bone marrow fibrosis, and cytokines on treatment. Study Design: The study is divided into Part 1 (APG-1252 monotherapy) and Part 2 (APG-1252 plus ruxolitinib). For Part 1, the key inclusion criterion is ineligibility for JAK2 inhibitors and for Part 2, inadequate responses to prior ruxolitinib-based therapy. A standard 3+3 dose-escalation design is being implemented to determine the maximum tolerated dose (MTD) of APG-1252 monotherapy in Part 1 and APG-1252 combined with ruxolitinib in Part 2. APG-1252 will initially be administered at 160 mg intravenously by 30-minute injection once weekly in a 28-day cycle. The dose can be escalated to a maximum of 240 mg or reduced to a minimum of 80 mg, depending on tolerability. Part 2 will begin once the MTD and recommended phase 2 dose (RP2D) of APG-1252 monotherapy have been determined. In Part 2, ruxolitinib will be administered orally twice daily per the package insert. After the MTD for APG-1252 monotherapy has been determined, no additional patients will be enrolled in Part 1; however, up to 15 to 30 additional patients can be enrolled in Part 2, to further evaluate the safety and anticancer activity of the combination at MTD or RP2D. Patients will continue treatment until disease progression or unacceptable toxicity. Clinical responses are being assessed every 12 weeks according to criteria from the International Working Group‒Myeloproliferative Neoplasms Research and Treatment and European LeukemiaNet panels, while optimal clinical benefit will be evaluated at 24 weeks. Enrollment will be from September 2020 and preliminary results estimated in October 2022. For further information, contact: yzhai@ascentage.com. Registration: ClinicalTrials.gov Identifier NCT04354727. Disclosures Pemmaraju: Pacylex Pharmaceuticals: Consultancy; Roche Diagnostics: Honoraria; LFB Biotechnologies: Honoraria; Stemline Therapeutics: Honoraria, Research Funding; Celgene: Honoraria; AbbVie: Honoraria, Research Funding; MustangBio: Honoraria; Affymetrix: Other: Grant Support, Research Funding; Cellectis: Research Funding; Daiichi Sankyo: Research Funding; Plexxikon: Research Funding; Samus Therapeutics: Research Funding; DAVA Oncology: Honoraria; Blueprint Medicines: Honoraria; Novartis: Honoraria, Research Funding; Incyte Corporation: Honoraria; SagerStrong Foundation: Other: Grant Support. Mudenda:Ascentage Pharma Group Inc.: Current Employment, Current equity holder in publicly-traded company. Wang:Ascentage Pharma Group Inc.: Current Employment, Current equity holder in publicly-traded company. JI:Ascentage Pharma (Suzhou) Co., Ltd.: Current Employment, Current equity holder in publicly-traded company. Lu:Ascentage Pharma Group: Current Employment, Current equity holder in publicly-traded company. Fu:Ascentage Pharma Group Inc.: Current Employment, Current equity holder in publicly-traded company. Liang:Ascentage Pharma Group Inc.: Current Employment, Current equity holder in publicly-traded company. McClain:Ascentage Pharma Group Inc.: Current Employment, Current equity holder in publicly-traded company. Sheladia:Ascentage Pharma Group Inc.: Current Employment, Current equity holder in publicly-traded company. Verstovsek:Novartis: Consultancy, Research Funding; Sierra Oncology: Consultancy, Research Funding; Blueprint Medicines Corp: Research Funding; PharmaEssentia: Research Funding; ItalPharma: Research Funding; AstraZeneca: Research Funding; Protagonist Therapeutics: Research Funding; Promedior: Research Funding; Celgene: Consultancy, Research Funding; NS Pharma: Research Funding; Genentech: Research Funding; CTI Biopharma Corp: Research Funding; Incyte Corporation: Consultancy, Research Funding; Roche: Research Funding; Gilead: Research Funding. Yang:Ascentage Pharma (SuZhou) Co., Ltd: Current Employment, Current equity holder in publicly-traded company, Other: Leadership and other ownership interests. Zhai:Ascentage Pharma (SuZhou) Co., Ltd: Current Employment, Current equity holder in publicly-traded company, Other: Leadership and other ownership interests.

Aim: We developed a platform to painlessly deliver by mouth drugs that cannot be administered orally. This ingestible, capsule-like device deploys inside the small intestine and injects into its wall a biotherapeutic payload sealed inside a dissolvable microneedle. Results: We tested the delivery and examined the pharmacokinetics of several biopharmaceuticals, including recombinant human insulin, human immunoglobulin G, octreotide and teriparatide in anesthetized and conscious pigs and dogs. The bioavailability of these compounds delivered by the capsules was similar to that after subcutaneous delivery. For instance, the mean peak serum concentration of insulin after capsule delivery in 8 anesthetized pigs was 517 ± 109 pmol/l, versus 342 ± 50 pmol/l after subcutaneous injection in 9 other, sham-operated pigs. In vivo safety studies were conducted in 23 conscious beagles, which received between 2 and 18 capsules. All devices transited painlessly through the GI tract and were excreted within 96 h. The mean gastric residence time (GRT) of the capsules was 93 min, and subsequent intestinal deployment time (IDT) was 28 minutes. A pilot study in 10 fasting and 10 postprandial human volunteers examined the tolerability and safety of the capsule devoid of microneedle and payload. The mean GRT of the capsule was 217 ± 36 min in the postprandial and 100 ± 79 min in the fasting state, while the IDT were closely similar (100 ± 40 vs. 97 ± 30 min) in both groups. No subject perceived the device transit, deployment or excretion. Conclusions: The studies completed thus far in anesthetized or conscious animals have confirmed that this ingestible device can deliver drugs in sufficient amounts and rate to obtain therapeutic blood concentrations. Its oral administration was well tolerated by the human subjects. First-in-man studies are being planned to test the ability of this ingestible device to deliver pharmaceuticals in healthy human volunteers and patients. Disclosure R. Ruffy: Employee; Self; RANI Therapeutics. M. Hashim: Board Member; Self; Rani Therapeutics. Employee; Self; InCube Labs LLC. A. Dhalla: Employee; Self; Rani Therapeutics, LLC. R. Korupolu: None. P. Karamchedu: Employee; Self; Incube labaoratories, Rani Therapeutics. B. Syed: None. R. Abdul Wahab: Employee; Self; Rani Therapeutics. S. Beraki: Employee; Self; InCube Laboratories, RANI Therapeutics. A. Yamaguchi: Employee; Self; Rani Therapeutics. L. Fusaro: None. A. Garapaty: None. A. Dasari: None. L. Fung: Employee; Self; Rani Therapeutics. Z.T. Al-Shamsie: Employee; Spouse/Partner; Google. Employee; Self; Rani Therapeutics, InCube Labs LLC. A. Toledo: Employee; Self; InCube Labs LLC, Rani Therapeutics. V. Salgotra: Employee; Self; Incubelabs, Rani Therapeutics. S. Sharma: None. E. Liang: None. D. Gratta: Employee; Self; InCube Labs, Rani Theraputics. M. Imran: Other Relationship; Self; InCube Labs, InCube Ventures, Rani Therapeutics.

Cardinality estimation has long been grounded in statistical tools for density estimation. To capture the rich multivariate distributions of relational tables, we propose the use of a new type of high-capacity statistical model: deep autoregressive models. However, direct application of these models leads to a limited estimator that is prohibitively expensive to evaluate for range or wildcard predicates. To produce a truly usable estimator, we develop a Monte Carlo integration scheme on top of autoregressive models that can efficiently handle range queries with dozens of dimensions or more. Like classical synopses, our estimator summarizes the data without supervision. Unlike previous solutions, we approximate the joint data distribution without any independence assumptions. Evaluated on real-world datasets and compared against real systems and dominant families of techniques, our estimator achieves single-digit multiplicative error at tail, an up to 90$\times$ accuracy improvement over the second best method, and is space- and runtime-efficient., Comment: VLDB 2020. Updates since version 1: new title and new/revised content