BG-89894 (SYH2039): BeOne pulls a MAT2A inhibitor before it can report

What was tried

BG-89894, also known as SYH2039, is an oral methionine adenosyltransferase 2A (MAT2A) inhibitor. BeOne Medicines (formerly BeiGene) licensed it from CSPC Pharmaceutical's subsidiary CSPC Zhongqi in December 2024 for global rights outside Greater China. The Phase 1a/1b dose-escalation and expansion study, NCT06568614, opened in late 2024 in adults with advanced or metastatic MTAP-deleted solid tumors. The trial enrolled 29 patients before the sponsor terminated it on 13 April 2026. The reason posted on ClinicalTrials.gov reads, "The study was discontinued following an internal decision by the sponsor."

The protocol followed the standard architecture for synthetic-lethal Phase 1 oncology trials. Eligibility required histologically confirmed MTAP-deleted advanced solid tumors that had progressed on standard therapy. Primary endpoints in Phase 1a were adverse events, maximum tolerated dose, and a recommended dose for expansion. Phase 1b would have read out objective response rate at the recommended Phase 2 dose. None of that happened in public view. There is no Phase 1a recommended dose, no exposure-response curve, no pharmacodynamic readout on plasma S-adenosylmethionine (SAM).

The trial closed in the same week BeOne disclosed in Q1 2026 earnings that four of five cancer assets were being discontinued, alongside the EGFR program BG-60366, the pan-KRAS inhibitor BGB-53038, and the CDK2 inhibitor BG-68501.

The biological hypothesis

MTAP (methylthioadenosine phosphorylase) sits on the same 9p21 locus as CDKN2A. Co-deletion of both genes is one of the most common copy-number losses in cancer, present in roughly 10 to 15% of all solid tumors and enriched in glioblastoma, pancreatic adenocarcinoma, mesothelioma, and squamous NSCLC. When MTAP is lost, its substrate methylthioadenosine (MTA) accumulates. MTA binds the protein arginine methyltransferase PRMT5 and partially inhibits it. MTAP-deleted cells are therefore running PRMT5 at reduced capacity, which gives any additional PRMT5 stress disproportionate effect. This is the synthetic lethal vulnerability described in two simultaneous 2016 papers (Kryukov et al. and Marjon et al.)).

MAT2A is the metabolic feeder. It catalyzes the synthesis of SAM from methionine and ATP, and SAM is the methyl donor used by PRMT5 to symmetrically dimethylate arginine residues on substrates like SmD1, SmD3, and SF3B1. Inhibiting MAT2A starves PRMT5 of its substrate, and Kalev and colleagues showed that this preferentially kills MTAP-deleted cells by collapsing PRMT5-dependent splicing and inducing DNA damage (Kalev et al., Cancer Cell 2021). The clinical strategy unifies under one mechanistic logic. Block MAT2A to lower SAM. Lower SAM hurts PRMT5. PRMT5 was already half-broken in MTAP-deleted cells, so they tip over before normal cells notice.

The first oral MAT2A inhibitor, AG-270, defined the class. Konteatis and colleagues described its discovery and selectivity (Konteatis et al., J Med Chem 2021). The Phase 1 trial of AG-270/S095033 was reported in 2025 and showed pharmacodynamic engagement, with plasma SAM reductions of up to 70%, but only two partial responses in 38 efficacy-evaluable patients and a disease control rate of 17.5% at 16 weeks (Konstantinopoulos et al., Nat Commun 2024). Ideaya's IDE397 has reported more encouraging numbers, with a monotherapy ORR around 33% at the 30 mg recommended Phase 2 dose, though in small cohorts. BG-89894 was designed to enter that landscape.

What actually happened

The trial enrolled 29 patients and was terminated before any pharmacodynamic or efficacy data were released. The April 2026 record update lists the primary completion date as 13 April 2026 and the why-stopped field as an internal sponsor decision. The Q1 2026 earnings deck, published on 6 May 2026, listed the program among four cancer assets being shelved. BeOne explicitly framed the change as a strategic shift toward a smaller, higher-conviction oncology portfolio, with a new emphasis on protein degraders and continued investment in BTK inhibition, the PD-1 antibody tislelizumab, and selected hematology assets.

The decision is informative even without efficacy numbers. BeOne paid to acquire BG-89894 in late 2024. Sixteen months later, after 29 enrolled patients, the company concluded that continuing was less attractive than redeploying capital. Two readings fit. The harsher reading is that early signal was unimpressive enough to lose the option value rather than fund the readout. The kinder reading is that competitive dynamics changed faster than the asset matured. By May 2026, IDE397 had a monotherapy ORR of about 33% and had moved into multiple combination cohorts. AMG 193 reported an ORR of 21.4% (95% CI 10.3 to 36.8) in 42 evaluable patients (Rodon et al., Annals of Oncology 2024). MRTX1719 had a 33% confirmed response rate in its dose-escalation evaluable population (Bryant et al., Cancer Discovery 2023). The bar for a fifth-to-market MAT2A inhibitor was no longer Phase 1 proof of concept.

Failure mechanism, best guess

A clinical trial that stops at 29 patients with no posted data is not a failure of biology. It is a failure of portfolio math. Three forces probably converged. First, the synthetic lethal premise of MTAP-deleted cancers has been validated, but the cleanest clinical signals are coming from MTA-cooperative PRMT5 inhibitors that act one step downstream of MAT2A, not from MAT2A inhibitors themselves. PRMT5 inhibition with MTA cooperativity (AMG 193, MRTX1719, BGB-58067) avoids the systemic SAM reduction that MAT2A inhibition causes, and that may matter for tolerability over months of dosing. The pharmacology favors the downstream node.

Second, BeOne already owned its own MTA-cooperative PRMT5 inhibitor, BGB-58067, in Phase 1. Sustaining a MAT2A program in the same portfolio meant running the combination in-house, which doubles complexity, or competing for trial slots with the wholly-owned asset. The cleanest answer is to keep BGB-58067 and drop BG-89894. Press coverage from BeOne's Q1 call described the decision as a deliberate move toward degraders (Fierce Biotech, 6 May 2026).

Third, MAT2A inhibitors have a structural disadvantage at the methionine-restriction end of the dose-response curve. SAM is required by every methyltransferase, including those maintaining genomic methylation in normal hematopoietic cells. AG-270's safety package showed reversible reductions in hemoglobin and thrombocytopenia at active doses. The therapeutic window for deep PRMT5 inhibition in tumor while sparing bone marrow is narrow.

What this means for similar programs

The class is not dead but the center of gravity has shifted. Ideaya's IDE397, now being combined with trastuzumab deruxtecan and gemcitabine, is the surviving MAT2A frontrunner. MTA-cooperative PRMT5 inhibitors look more likely to define the registration path. Programs entering this space need either differentiated selectivity, a brain-penetrant profile, or a clear combination story that justifies the systemic SAM cost. A new MAT2A asset in 2026 carries a fast follower's burden of proof.

Open questions

• Why did BeOne acquire BG-89894 in late 2024 if 16 months of in-house preclinical and early clinical work could not preserve the program?
• Did BG-89894 show pharmacodynamic SAM reduction comparable to AG-270 and IDE397 in its first 29 patients?
• Will BGB-58067 advance alone, or will BeOne seek another MAT2A partner for combination?
• Is the apparent efficacy gap between MAT2A inhibitors (5 to 33% ORR) and MTA-cooperative PRMT5 inhibitors (21 to 33% ORR) real or a function of small-cohort variance?
• What is the long-term hematologic safety profile of sustained MAT2A inhibition, and is it compatible with combination immunotherapy regimens?

Sources

1. ClinicalTrials.gov. A Phase 1 Study to Evaluate the Safety, Tolerability, Pharmacokinetics, Pharmacodynamics, and Preliminary Antitumor Activity of BG-89894 (SYH2039) Tablets in Patients With Advanced Solid Tumors. NCT06568614.

2. Kalev P, Hyer ML, Gross S, Konteatis Z, Chen CC, Fletcher M, et al.. MAT2A Inhibition Blocks the Growth of MTAP-Deleted Cancer Cells by Reducing PRMT5-Dependent mRNA Splicing and Inducing DNA Damage. Cancer Cell. 2021;39(2);209-224.e11. PMID: 33450196.

3. Konteatis Z, Travins J, Gross S, Marjon K, Barnett A, Mandley E, et al.. Discovery of AG-270, a First-in-Class Oral MAT2A Inhibitor for the Treatment of Tumors with Homozygous MTAP Deletion. J Med Chem. 2021;64(8);4430-4449. PMID: 33829783.

4. Gounder M, Johnson M, Heist RS, Shapiro GI, Postel-Vinay S, Wilson FH, et al.. MAT2A inhibitor AG-270/S095033 in patients with advanced malignancies: a phase I trial. Nat Commun. 2025;16(1);423. PMID: 39762248.

5. Engstrom LD, Aranda R, Waters L, Moya K, Bowcut V, Vegar L, et al.. MRTX1719 Is an MTA-Cooperative PRMT5 Inhibitor That Exhibits Synthetic Lethality in Preclinical Models and Patients with MTAP-Deleted Cancer. Cancer Discov. 2023;13(11);2412-2431. PMID: 37552839.

6. Belmontes B, Slemmons KK, Su C, Liu S, Policheni AN, Moriguchi J, et al.. AMG 193, a Clinical Stage MTA-Cooperative PRMT5 Inhibitor, Drives Antitumor Activity Preclinically and in Patients with MTAP-Deleted Cancers. Cancer Discov. 2025;15(1);139-161. PMID: 39282709.

7. Bedard GT, Gilaj N, Peregrina K, Brew I, Tosti E, Shaffer K, et al.. Combined inhibition of MTAP and MAT2a mimics synthetic lethality in tumor models via PRMT5 inhibition. J Biol Chem. 2024;300(1);105492. PMID: 38000655.

8. Kryukov GV, Wilson FH, Ruth JR, Paulk J, Tsherniak A, Marlow SE, et al.. MTAP deletion confers enhanced dependency on the PRMT5 arginine methyltransferase in cancer cells. Science. 2016;351(6278);1214-1218. PMID: 26912360.

9. BeiGene. Press release, 12 December 2024. BeiGene announces global licensing agreement for MAT2A inhibitor. businesswire.com.

10. Inman P. BeOne shelves 5 cancer programs, halts phase 2 rheumatoid arthritis trial. Fierce Biotech, 6 May 2026. fiercebiotech.com.