MTAP Deletion in Oncogenesis: A Synthetic Lethality Scenario

Abstract

Homozygous deletions in the gene encoding methylthioadenosine phosphorylase (MTAP) occur in ∼10% of patients with cancer, including up to 45% in some tumor types, and may be associated with poor prognosis. MTAP deficiency causes accumulation of its catabolic target methylthioadenosine (MTA) that outcompetes S-adenosyl methionine (SAM) for binding to protein arginine methyltransferase 5 (PRMT5), partially inhibiting PRMT5 activity as a posttranslational regulator of a variety of critical cellular functions. Prior anticancer treatments developed to target PRMT5 exhibited high rates of dose-limiting hematologic toxicities because of a lack of selectivity for tumor cells. More recently, several agents have been developed that exploit the vulnerability of MTAP-deleted cancer cells to further inhibition of the PRMT5 pathway, selectively inducing synthetic lethality in those cancer cells. MTA-cooperative PRMT5 inhibitors such as BMS-986504/MRTX1719 and AMG 193 target the PRMT5–MTA complex, whereas inhibitors of the SAM synthetase methionine adenosyl transferase 2A, such as IDE397, deprive PRMT5 of its methyl donor SAM. In this review article, we summarize the mechanisms of action, preclinical data, and clinical data available thus far for these novel classes of oncology precision medicine and discuss potential future directions relevant to MTAP deletion as a promising synthetic lethal vulnerability for cancer therapy.

Introduction

Although targeted therapy and immunotherapy have greatly improved survival and quality of life compared with traditional chemotherapy treatments in cancer, the cancer burden continues to grow worldwide, with 20 million new cases and 9.7 million cancer-related deaths in 2022 (RRID: SCR_025451). As such, there is a continual pursuit for new anticancer agents that can balance efficacy with tolerability. A key approach in developing these therapies is identifying cancer-specific biomarkers and their role in oncogenesis.

MTAP, encoding the enzyme methylthioadenosine phosphorylase, lies in the same gene cluster as CDKN2A, a tumor suppressor gene commonly deleted in cancer cells, and several IFN genes on the chromosome 9p21 locus (1). Homozygous deletions in MTAP (MTAP-del) occur in approximately 10% of all patients with cancer, presumably because of its proximity to CDKN2A (RRID: SCR_003193; ref. 2). Tumor types with particularly high rates of MTAP-del include glioblastoma (∼45% of patients), mesothelioma (∼36%), urothelial carcinoma (∼26%), malignant peripheral nerve sheath tumor (∼25%), pancreatic cancer (∼22%), and non–small cell lung cancer (NSCLC; ∼15%; RRID: SCR_003193; refs. 2, 3). MTAP deficiency (i.e., MTAP-del and/or MTAP protein loss) affects various cellular pathways, including metabolic, epigenetic, and immune-related signaling, as well as cell migration and invasion (4–10). Studies suggest that MTAP acts as a tumor suppressor and that its deletion contributes to tumorigenesis, tumor growth, and a poor prognosis by disrupting these critical processes, including increased expression of the stem cell regulator PROM1/CD133 and the kinase RIOK1, promoting the proliferation of glioblastoma and pancreatic cancer, respectively (6–14). However, our understanding of the complete spectrum of MTAP deficiency and its impact on specific cancers continues to evolve.

Several early-stage compounds have shown promising treatment outcomes in patients with MTAP-del tumors by exploiting protein arginine methyltransferase 5 (PRMT5) and methionine adenosyl transferase 2A (MAT2A) as synthetic lethal targets. Here, we review the impacts of MTAP deficiency on PRMT5 activity, summarize preclinical and clinical data for PRMT5- and MAT2A-targeted therapies in MTAP-del tumors, and discuss future directions for these novel classes of synthetic lethal drugs.

Biological Significance of MTAP and PRMT5 in Cells

Role of MTAP and PRMT5 in MTAP wild-type cells

MTAP is the sole phosphorylase that metabolizes methylthioadenosine (MTA) into adenine and 5-methylthioribose-1-phosphate, a precursor of methionine, in mammalian cells (RRID: SCR_004426). As such, it plays a vital role in both the adenine and methionine salvage pathways and is responsible for salvaging almost all adenine (a critical building block of DNA and RNA) in the human body (15). ATP (derived from adenine) and methionine can be recycled into the universal methyl donor S-adenosyl methionine (SAM) by MAT2A (RRID: SCR_004426). MTA competes with SAM for binding to PRMT5, a methylosome protein 50 cofactor–dependent type II methyltransferase.

By reducing cellular levels of MTA and thereby regulating MTA:SAM ratios, which is thought to ensure a high concentration of the active PRMT5–SAM complex (Fig. 1; refs. 4, 5, 16, 17), MTAP plays a critical role in fine-tuning the balance between metabolic and epigenetic pathways. PRMT5 utilizes SAM to add symmetric dimethylarginine (SDMA), a unique posttranslational modification, to histones and other target proteins involved in a variety of essential cellular functions, including pathways related to cell proliferation and differentiation (RRID: SCR_004426). PRMT5-mediated histone methylation promotes gene expression (e.g., FOXP1 expression via H3R2 methylation) or repression (e.g., differentiation-related pathways via H2R3 methylation and genes such as c-Myc, Cyclin D1, FGFR3, and PTCH1 via H4R3 and H3R8 methylation; refs. 18–20). Additionally, PRMT5 directly methylates proteins such as small nuclear ribonucleoprotein proteins, which play crucial roles in RNA splicing (21–23). Other proteins involved in cell cycle progression, transcription, translation, and DNA repair have also been identified as PRMT5 targets, including EGFR and RAF proteins, which downregulate EGF signaling through the MAPK pathway (24, 25), as well as p53 (26). Although the PRMT5-mediated methylation of splicing factors is now well established, EGFR, RAF proteins, and p53 have not been reported as substrates in multiple studies using pull-down proteomics to identify PRMT5 or arginine methylation targets (27–30). This discrepancy highlights the need for future studies to clarify the direct methylation targets of PRMT5, their role in PRMT5-mediated cellular functions, and whether these targets change in response to stimuli or in different cell types.

Figure 1.

PRMT5 pathway in MTAP wild-type cells, MTAP-deleted cells, MTAP wild-type cells with SAM-cooperative PRMT5 inhibition, MTAP-deleted cells with MTA-cooperative PRMT5 inhibition, MTAP wild-type cells with MTA-cooperative PRMT5 inhibition, or MTAP-deleted cells with MAT2A inhibition. IMP, inosine monophosphate; MAT2Ai, MAT2 inhibitor; MTR-1P, methylthioribose-1-phosphate; PRMT5i, PRMT5 inhibitor; WT, wild type.

PRMT5 overexpression is associated with poor prognosis in several cancers (31–35), suggesting PRMT5 blockade as an attractive target for the development of new anticancer therapies. PRMT5 depletion triggers cell cycle arrest and apoptosis and is lethal early in embryogenesis in mouse models (26, 36, 37). There has been a long-standing interest to target PRMT5 activity specifically in tumor cells, with growing attention on MTAP-del tumors in recent years.

Impact of MTAP deficiency on PRMT5 activity in tumor cells

In MTAP-deficient cell lines, the inability to metabolize MTA can increase intracellular MTA levels by an average of 15-fold compared with MTAP-intact cell lines (5). Accumulated MTA outcompetes SAM for binding to PRMT5 and is thought to lead to the enrichment of PRMT5–MTA complexes (Fig. 1). Importantly, MTA partially inhibits PRMT5 activity as evidenced by reduced SDMA levels (often used as a pharmacodynamic marker of PRMT5 activity) in MTAP-del cells (4, 5, 17). An in vitro biochemical screen assessing the activity of 33 different methyltransferases and follow-up Ki assays indicated that the addition of MTA strongly and selectively inhibited PRMT5 (17). Finally, estimations of PRMT5 activity using the Ki and Km values for MTA and SAM, respectively, along with their measured cellular concentrations, found that PRMT5 was 26% inhibited in HT-29–cultured cells under normal physiologic conditions, 73% inhibited following treatment of the cells with an MTAP inhibitor, and 93% inhibited following treatment with both MTAP and MAT2A inhibitors (16). Thus, MTA inhibition of PRMT5 results in a condition in which MTAP-del cancer cells are susceptible to additional disruption of the PRMT5 pathway, rendering these cells potentially vulnerable to targeted synthetic lethality. In consensus, RNA interference screening in MTAP-del cancer cells identified an increased sensitivity to PRMT5 inhibition for cell viability (4, 5, 17).

Tumor-Nonselective (SAM- and Substrate-Competitive) PRMT5 Inhibitors

Initial attempts to develop anticancer drugs targeting the PRMT5 pathway focused on the inhibition of PRMT5 in a SAM- and/or substrate-competitive manner, regardless of MTAP status (Fig. 1). In preclinical studies, these PRMT5 inhibitors demonstrated encouraging antiproliferative activity (38–40). Subsequently, several PRMT5 inhibitors entered early clinical studies, including JNJ-64619178 (NCT03573310), GSK3326595 (METEOR-1/NCT02783300 and NCT03614728), PF-06939999 (NCT03854227), PRT543 (NCT03886831), and PRT811 (NCT04089449), mainly in patients with solid tumors or hematologic malignancies.

Despite the preclinical potential of these PRMT5 inhibitors, their clinical development has been impeded by safety concerns. Potent inhibition of PRMT5 in the absence of tumor cell selection can result in off-target cytotoxicity, particularly in hematopoietic cells in which it plays an essential role in regulating their proliferation, differentiation, and maintenance (41). Across clinical trials for the tumor-nonselective PRMT5 inhibitors, high rates of grade ≥3 treatment-related adverse events (TRAE) were reported (23%–51%), most commonly anemia (up to 29%) and thrombocytopenia (up to 21%; refs. 42–47). Preliminary efficacy data from five of the six studies showed low objective response rates (ORR), ranging from 0% to 7% (42, 44–47). The NCT03614728 trial for GSK3326595 was ultimately terminated after meeting futility criteria at the second interim analysis (47), and the NCT03854227 trial for PF-06939999 was terminated by the sponsor for strategic reasons (ClinicalTrials.gov: NCT03854227; ref. 46). Efficacy and safety data for the brain-penetrant PRMT5 inhibitor PRT811 seemed more promising, although the patient populations were small: two of 10 patients with splicing gene–mutated metastatic uveal melanoma and three of 16 patients with IDH1/2-mutated high-grade glioma experienced a response, including two complete responses in those with glioma (43). As with the other PRMT5 inhibitors, the most common grade ≥3 TRAEs were hematologic although they were reported at lower rates for PRT811 (thrombocytopenia, 9.8%; anemia, 3.3%; ref. 43).

As these SAM- and substrate-competitive PRMT5 inhibitors target PRMT5 in both tumor tissue and healthy cells, they lack the ability to selectively target and kill cancer cells, resulting in narrow therapeutic windows largely limited by hematologic toxicities (42–47). Furthermore, likely because of the essential role for PRMT5 in T-cell survival and proliferation (48), the tumor-nonselective PRMT5 inhibitor EPZ015666 has been associated with reduced T-cell viability, proliferation, and functionality, compromising the antitumor immune response (49). Collectively, these data highlighted the need for inhibitors that selectively target PRMT5 activity in tumor cells.

MTAP-del Tumor-Selective (MTA-Cooperative) PRMT5 Inhibitors

Concurrent with early studies of tumor-nonselective PRMT5 inhibitors, PRMT5 was identified as a potential synthetic lethal target in MTAP-del cells through RNA interference screening (4, 5, 17), presenting an opportunity to target PRMT5 inhibitors to cancer cells characterized by MTAP deficiency. Because of the high concentration of MTA in MTAP-deficient cells (5), MTAP-del cancer cells are hypothesized to be distinguished from MTAP wild-type cells by a high proportion of PRMT5–MTA complexes. Given the partial inhibition of PRMT5 activity by MTA, it was theorized that inhibitors selectively binding to and stabilizing the PRMT5–MTA complex could further inhibit PRMT5 activity and induce synthetic lethality specifically in MTAP-del cancer cells while sparing healthy MTAP wild-type cells, thereby circumventing the off-target effects of nontumor-selective PRMT5 inhibitors (Fig. 1; refs. 50–55). Several novel compounds evaluated in preclinical and clinical studies exploit this vulnerability, including the MTA-cooperative PRMT5 inhibitors BMS-986504 (formerly MRTX1719), AMG 193, AZD3470, and TNG908 (currently superseded by two newer agents, TNG462 and TNG456).

Preclinical data for MTA-cooperative PRMT5 inhibitors

BMS-986504/MRTX1719, AMG 193, AZD3470, and TNG908 were individually discovered using high-throughput screening techniques for small molecules that bound selectively to PRMT5 in the presence of MTA (50, 52, 55, 56). Molecular modeling and structure-based drug discovery approaches were then used to iteratively optimize potency and selectivity for the PRMT5–MTA complex to develop the final compounds. TNG908 was specifically designed as a brain-penetrant MTA-cooperative PRMT5 inhibitor targeting the high prevalence of MTAP deletions in glioblastomas and cancers with a high incidence of brain metastases (55). To increase its effectiveness, TNG908 precursors further underwent screening to improve permeability and reduce efflux across the blood–brain barrier (55). Subsequently, TNG462 was developed following additional optimization of the TNG908 molecular structure to increase potency and selectivity, followed by TNG456 (57). X-ray crystallography of the PRMT5-binding structures of BMS-986504/MRTX1719, AMG 193, and TNG908 showed that they occupied the substrate binding site of the PRMT5–MTA complex and formed a tight van der Waals interaction with MTA while interfering with the interaction between PRMT5 and SAM, contributing to long dissociation half-lives and high binding selectivity for PRMT5–MTA versus PRMT5–SAM (51, 52, 55).

Cell viability assays of these MTA-cooperative PRMT5 inhibitors indicated a high degree of selectivity for MTAP-del cells compared with MTAP wild-type cancer cell lines (50–52, 55, 57–59). Daily oral administration in xenograft models of MTAP-del lung cancer (for BMS-986504/MRTX1719) and MTAP-del colorectal cancer (for AMG 193) showed dose-dependent tumor growth inhibition, particularly with prolonged drug exposure and without marked toxicity or weight loss (50–52), suggesting potential for a wide therapeutic window. Importantly, significant hematologic toxicity was not observed in mouse models treated with AMG 193 (52), and assessment of human bone marrow showed reduced toxicity from BMS-986504/MRTX1719 or AZD3470 compared with tumor-nonselective PRMT5 inhibitors (51, 56).

Antitumor activity was also observed across a variety of other MTAP-del xenograft models, with notable activity in mesothelioma, pancreatic, lung, esophageal, cholangiocarcinoma, and gastric cancer models for BMS-986504/MRTX1719 (51) and diffuse large B-cell lymphoma, pancreatic, lung, melanoma, and esophageal cancer models for AMG 193 (52). Similarly, AZD3470 showed efficacy in 59% of 93 MTAP-del xenograft models for NSCLC, esophageal, pancreatic ductal adenocarcinoma (PDAC), bladder, head and neck squamous cell carcinoma, and gastric tumors (60). TNG908 and TNG462 also exhibited dose-dependent antitumor activity in MTAP-del xenograft models across several different tumor types (55, 57, 59).

Administration of MTA-cooperative PRMT5 inhibitors decreased SDMA levels in a dose-dependent manner in MTAP-del xenograft models in individual studies, with maximal tumor growth inhibition found to correlate with at least 95% SDMA reduction, suggesting that near-complete PRMT5 inhibition may be required to achieve synthetic lethality (52). BMS-986504/MRTX1719 showed SDMA inhibition of 95% to 98% at the two highest dose levels tested preclinically, 50 and 100 mg/kg (51); for TNG462, SDMA levels in xenograft models were undetectable up to 72 hours following the last dose (57). Modulation of SDMA levels by MTA-cooperative PRMT5 inhibitors was selective, with AMG 193 administration inhibiting SDMA by 86% to 93% in MTAP-del tumors versus 26% to 76% in MTAP wild-type tumors (52).

MTA-cooperative PRMT5 inhibition was also investigated in MTAP-del cell lines to assess their effects on pathways downstream of PRMT5. Gene set enrichment analyses showed that treatment with BMS-986504/MRTX1719 or AM-9747, a representative analogue of AMG 193, resulted in gene expression changes in RNA splicing, apoptosis, cell cycle, and/or immune-related pathways (51, 52). Furthermore, AM-9747 treatment was associated with multiple indicators of DNA damage with high selectivity for MTAP-del versus MTAP wild-type cell lines (52). In an integrative dynamic multiomics analysis, AZD3470 treatment in MTAP-del NSCLC cell lines induced alterations consistent with the inhibition of PRMT5, including protein posttranslational modification, gene expression, chromatin accessibility, and RNA splicing, in a manner suggestive of changes in the cell cycle and DNA repair pathways (61). Other studies have found that the predominant impact of PRMT5 inhibition on splicing lies in the induction of intron retention by impairing the removal of detained introns (62, 63). Detained introns were observed in genes for which downregulation is associated with cell cycle defects, apoptosis, and/or senescence and are speculated to be a primary driver for the vulnerability of cells to PRMT5 inhibition, with increased sensitivity in MTAP-del tumor cells (62, 63).

Overall, the preclinical data available thus far indicate that MTA-cooperative PRMT5 inhibitors share a number of common biophysical and biological characteristics. These characteristics include oral bioavailability, substrate-competitive PRMT5 interaction, strong selectivity for and stabilization of the PRMT5–MTA complex over PRMT5–SAM, dose-dependent reduction in SDMA, downstream effects on RNA splicing and cell viability pathways, and selectivity for MTAP-del cells compared with MTAP wild-type cancer cell lines in cell viability assays. Importantly, MTA-cooperative PRMT5 inhibitors conferred robust tumor growth inhibition in MTAP-del xenograft models without the hematologic toxicities commonly associated with tumor-nonselective PRMT5 inhibitors.

Clinical data for MTA-cooperative PRMT5 inhibitors

BMS-986504/MRTX1719, AMG 193, AZD3470, TNG908, and TNG462 have been evaluated in clinical trials, with preliminary data available for BMS-986504/MRTX1719 and AMG 193 [Table 1; NCT03435250 results summary (https://clinicaltrials.servier.com/trial/NCT03435250/study-of-ag-270-in-participants-with-advanced-solid-tumors-or-lymphoma-with-mtap-loss); refs. 51, 53, 64–66].

Table 1.

Clinical data summary for MTA-cooperative PRMT5 and MAT2A inhibitors in MTAP-del tumors.

Drug	Study	N	Dosage	Safety summary	Most common TRAEs	ORR		DCR
MTA-cooperative PRMT5 inhibitors
BMS-986504 (MRTX1719; refs. 51, 64)	NCT05245500 Phase I MTAP-del advanced or unresectable solid tumors	146a	Phase I: 50, 100, 200, 400, 600, or 800 mg once daily or 400 mg twice daily Phase Ib dose expansion: 200, 400, or 600 mg once daily	TRAEs: 78% (grade ≥3: 12%) Serious TRAEs: 3% Discontinued because of TRAEs: 3% No treatment-related deaths	Any grade: Nausea (37%) Vomiting (26%) Fatigue (22%) Diarrhea (19%) Decreased appetite (13%) Anemia (10%) Grade ≥3: Anemia (2%) Neutropenia (2%)	All evaluable patients: 26 of 124 (21%), all PR NSCLC: eight of 26 (31%) PDAC: five of 33 (15%) Mesothelioma: four of 11 (36%) CCA: two of 11 (18%)		All evaluable patients: 87 of 124 (70%) NSCLC: 21 of 26 (81%) PDAC: 23 of 33 (70%) Mesothelioma: nine of 11 (82%) CCA: eight of 11 (73%)
AMG 193 (53)	NCT05094336 Phase I/Ib MTAP-del advanced solid tumors	80	40, 120, 240, 480, 800, 1,200, or 1,600 mg once daily or 600 mg twice daily MTD: 1,200 mg once daily	TRAEs: 85% (grade ≥3: 13.8%) Serious TRAEs: 10% Discontinued because of TRAEs: 2.5% No treatment-related deaths	Any grade: Nausea (48.8%) Fatigue (31.3%) Vomiting (30%) Decreased appetite (22.5%) Dizziness (13.8%) Asthenia (13.8%) Diarrhea (10%) Dysgeusia (10%) Grade ≥3: Fatigue (5%) Decreased lymphocytes (3.8%) Decreased neutrophils (2.5%) Decreased appetite (2.5%)	All evaluable patients: nine of 74 (12.2%), all PR Active dosage levels (800/1,200 mg once daily): nine of 42 (21.4%) Nine responses in eight tumor types at active dosage level (NSCLC, PDAC, CCA, melanoma, esophageal cancer, gallbladder cancer, RCC, and ovarian Sertoli–Leydig cell tumor)		All evaluable patients: 35 of 74 (47.3%) Active dosage levels (800/1,200 mg once daily): 23 of 42 (54.8%)
MAT2A inhibitors
IDE397 (65)	NCT04794699 Phase I MTAP-del advanced solid tumors	28 with RP2D	RP2D: 30 mg once daily	TRAEs: 54% (grade ≥3: 18%) No serious TRAEs No discontinuations due to TRAEs	Any grade: Peripheral neuropathy (25%) Decreased appetite (11%) Increased blood creatinine (11%) Nausea (11%) Fatigue (11%) Grade ≥3: Asthenia (4%)	All evaluable patients with RP2D: Nine of 27 (33%) and one CR (UC) Squamous NSCLC: Three of eight (38%) Adenocarcinoma NSCLC: two of nine (22%) UC: four of 10 (40%)	All evaluable patients with RP2D: 25 of 27 (93%) Squamous NSCLC: Eight of eight (100%) Adenocarcinoma NSCLC: nine of nine (100%) UC: eight of 10 (80%)
S95033 (AG-270) [NCT03435250 results summary (https://clinicaltrials.servier.com/trial/NCT03435250/study-of-ag-270-in-participants-with-advanced-solid-tumors-or-lymphoma-with-mtap-loss); ref. 66]	NCT03435250 Phase I MTAP-del advanced solid tumors or lymphoma (Trial terminated)	40b	50, 100, 150, 200, or 400 mg once daily or 200 mg twice daily MTD: 200 mg once daily	TRAEs: 70% (grade ≥3: 17.5%) Serious TRAEs: 10% Discontinued because of TEAEs: 5% No treatment-related deaths	Any grade: Fatigue (25%) Increased blood bilirubin/hyperbilirubinemia (25%) Thrombocytopenia/platelet count decrease (17.5%) Rash (15%) Anemia (10%)	All evaluable patients: Two of 40 (5%), both PR (sex cord stromal cell cancer and NSCLC) Most common BOR was progressive disease (24 of 40 [60%])	All evaluable patients: 17.5%

Abbreviations: BOR, best overall response; CCA, cholangiocarcinoma; CR, complete response; NR, not reported; PR, partial response; RCC, renal cell carcinoma; TEAE, treatment-emergent adverse event; UC, urothelial cancer.

^aOne additional patient included in the safety and efficacy populations was treated but did not have a fully executed informed consent form and was therefore not considered to be enrolled.

^bRepresents patients enrolled in the S95033 monotherapy arm of the dose-escalation phase; additional patients were enrolled in arms evaluating S95033 in combination with docetaxel or in combination with paclitaxel and gemcitabine.

BMS-986504/MRTX1719 is being assessed in the phase I NCT05245500 clinical trial of patients with advanced, unresectable, or metastatic MTAP-del solid tumors (including NSCLC, PDAC, mesothelioma, and cholangiocarcinoma; refs. 51, 64). Most patients (75%) had received ≥2 previous lines of therapy. In the dose-escalation portion, BMS-986504/MRTX1719 was administered orally at dosages ranging from 50 to 800 mg once daily or 400 mg twice daily. Dose-limiting toxicities (DLT) were observed in two of five patients at the 800 mg once daily dosage (grade 2 seizure and grade 2 vomiting/fatigue) and one of 20 patients at the 400 mg twice daily dosage (grade 2 rash), establishing 600 mg once daily as the maximum acceptable dose (64). A subsequent phase Ib dose expansion included patients treated with 200, 400, or 600 mg once daily, for a total of 147 patients treated across all dose levels as of September 2024.

AMG 193 is being investigated in a phase I/Ib clinical trial (NCT05094336) of patients with advanced CDKN2A-deleted and/or MTAP-del solid tumors (53). Patients had received a median of two prior lines of therapy. In the dose-exploration phase, AMG 193 was administered orally to 80 patients at dosages ranging from 40 to 1,600 mg once daily or at 600 mg twice daily. Eight patients had DLTs (all at dosages ≥240 mg), with two reporting vomiting and one each reporting nausea, fatigue, hypersensitivity reaction, hypokalemia, encephalopathy, and palpitations. The MTD was set at 1,200 mg, for which 11.1% of patients had experienced DLTs. An ongoing dose-expansion phase included 87 patients treated with AMG 193 at 1,200 mg once daily as of May 2024, for a total of 167 patients with PDAC, NSCLC, biliary tract cancer, gastric/esophageal cancer, glioblastoma, or other solid tumors (54).

BMS-986504/MRTX1719 and AMG 193 were well tolerated in their respective clinical trials. For the 147 patients enrolled in NCT05245500 and treated with BMS-986504/MRTX1719, TRAEs were observed in 78%, most commonly nausea, vomiting, and fatigue (64). Grade ≥3 TRAEs were reported in 12% of patients, most commonly anemia and neutropenia (each 2%). Serious TRAEs were observed in 3% of patients, and 3% of patients discontinued because of TRAEs (64). In 80 patients treated with AMG 193 in the dose-exploration phase of NCT05094336, 85% experienced TRAEs (53). Similar to NCT05245500, the most common TRAEs were nausea, vomiting, and fatigue. Grade ≥3 TRAEs were reported in 13.8% of patients, most commonly fatigue (5%) and decreased lymphocyte count (3.8%). Serious TRAEs were observed in 10% of patients, and 2.5% discontinued because of TRAEs (53). No treatment-related deaths were reported in either study. Importantly, although treatment-related anemia is a DLT commonly associated with tumor-nonselective PRMT5 inhibitors, anemia was observed in ≤10% of patients treated with BMS-986504/MRTX1719 or AMG 193, with no dose-limiting hematologic events or patients who discontinued because of hematologic toxicities reported for either agent (53, 64).

Preliminary efficacy data for BMS-986504/MRTX1719 showed that among 124 clinically evaluable patients with a median follow-up of 7.6 months, ORR across all dose levels was 21% (all partial responses), the disease control rate (DCR) was 70%, and the median duration of response (DOR) was 6.9 months (64). Clinical activity was observed in multiple tumor types, including NSCLC (n = 26; ORR 31% and DCR 81%), PDAC (n = 33; ORR 15% and DCR 70%), mesothelioma (n = 11; ORR 36% and DCR 82%), and cholangiocarcinoma (n = 11; ORR 18% and DCR 73%). The median DOR in the NSCLC subgroup was 10.5 months and was not reached for the other tumor subgroups. Among all patients, median time to response was 4.5 months, and 88% of patients with a partial response had initial stable disease (64). Analyses of tumor SDMA levels in patients treated with BMS-986504/MRTX1719 showed dose-dependent reductions, with the greatest reductions at doses of 400 mg once daily or greater (67). The median tumor SDMA levels decreased from an H-score of 285 at baseline to 0 by the beginning of treatment cycle 2 in patients with paired samples, with increased intron retention observed in nine of 11 patients (51, 67). Furthermore, RNA sequencing showed that BMS-986504/MRTX1719 led to the downregulation of pathways related to cell cycle progression, including DNA repair, mitotic spindle formation, and MYC and E2F targets (67).

Seventy-four patients from the dose-exploration phase were included in an AMG 193 efficacy analysis (53). Across all dose levels, the ORR was 12.2%, and the DCR was 47.3%. In 42 patients treated at the active dosage levels (800 and 1,200 mg once daily and 600 mg twice daily), nine objective responses were observed across eight different tumor types: NSCLC, PDAC, cholangiocarcinoma, melanoma, esophageal cancer, gallbladder cancer, renal cell carcinoma, and ovarian Sertoli–Leydig cell tumor. The ORR in patients treated at these active dose levels was 21.4%, the DCR was 54.8%, and the median DOR was 8.3 months. The median time to response was 3.6 months at 800 mg once daily and 1.8 months at 1,200 mg once daily. Pharmacokinetic analysis showed that plasma levels of AMG 193 increased proportionally to the dose in patients treated with 40 to 1,200 mg once daily, and pharmacodynamic modeling of SDMA levels up to 1,600 mg once daily showed a significant dose–response relationship. In eight patients with posttreatment tumor samples, near-complete SDMA elimination was observed at dosage levels of 480, 800, and 1,200 mg once daily (53).

TNG908 is being investigated in a phase I/II study (NCT05275478) of patients with MTAP-del advanced solid tumors. However, early data showed no responses in 23 patients with glioblastoma treated at active doses of TNG908 [Tango Therapeutics press release (https://www.businesswire.com/news/home/20241106825271/en/Tango-Therapeutics-Reports-Positive-TNG462-Clinical-Data-and-Provides-Update-on-PRMT5-Development-Program)]. Enrollment was terminated early to focus on clinical trials of TNG462 (NCT05732831) and TNG456, a next-generation brain-penetrant MTA-cooperative PRMT5 inhibitor (NCT06810544). Finally, although data are not yet available, AZD3470 is being explored in the phase I/IIa PRIMROSE trial (NCT06130553) of patients with MTAP-deficient solid tumors and in the phase I/II PRIMAVERA trial (NCT06137144) of patients with relapsed/refractory hematologic malignancies.

Collectively, early clinical data for orally administered BMS-986504/MRTX1719 and AMG 193 indicate efficacy across a variety of tumor types, including mesothelioma and pancreatic cancer, which are refractory to many other classes of therapies. Both compounds share wide therapeutic windows and lower grade ≥3 TRAE rates than tumor-nonselective PRMT5 inhibitors, particularly for hematologic events that were frequently dose limiting with nonselective PRMT5 inhibitors (53, 64). Interestingly, the median times to first response for BMS-986504/MRTX1719 and the 800 mg once daily dosage level of AMG 193 were relatively long, and a high proportion of responders to BMS-986504/MRTX1719 had an initial response of stable disease. Similar results have been reported from early clinical data for TNG462 [Tango Therapeutics press release (https://www.businesswire.com/news/home/20241106825271/en/Tango-Therapeutics-Reports-Positive-TNG462-Clinical-Data-and-Provides-Update-on-PRMT5-Development-Program)]. This suggests that deepening tumor regression may occur over a longer period relative to traditional targeted cancer treatments, such as tyrosine kinase inhibitors, potentially related to the epigenetic modulatory mechanism of action. The initial trajectory of response is reminiscent of immunotherapy for cancer, which also provides a more durable response than traditional treatments (68). Longer follow-up will be needed to fully understand the clinical benefit and durability of MTA-cooperative PRMT5 inhibitors.

MAT2A Inhibitors

Similar to PRMT5, RNA interference screening has shown that MTAP-del cell viability is sensitive to depletion of MAT2A (17). Inhibition of MAT2A reduces the biosynthesis of SAM, a state that is theorized to induce synthetic lethality in MTAP-del tumor cells, in which PRMT5 activity is already partially inhibited by the high concentration of MTA (Fig. 1). Thus far, MTAP-del tumor-selective synthetic lethality has been reported for several small-molecule MAT2A inhibitors, including S95033 (formerly AG-270) and IDE397.

Preclinical data for MAT2A inhibitors

S95033/AG-270 is a first-in-class MAT2A inhibitor that was developed based on a fragment screening approach, with structure-based design used to iteratively optimize the potency for MAT2A inhibition and produce a metabolically stable, highly potent final compound (69). S95033/AG-270 inhibits the release of SAM from MAT2A and was selective for antiproliferative activity in MTAP-del HCT116 cell lines. Oral administration of S95033/AG-270 inhibited tumor growth in a dose-dependent manner in MTAP-del but not MTAP wild-type xenograft models (69, 70). Similarly, the independently developed IDE397 demonstrated antitumor activity in MTAP-del xenograft models for NSCLC, esophageal, gastric, bladder, head and neck, and pancreatic cancers (71).

Importantly, individual studies showed that both S95033/AG-270 and IDE397 exhibited dose-dependent reduction in SAM levels in tumors regardless of MTAP status but reduced SDMA levels only in MTAP-del models, suggesting selective inhibition of PRMT5 activity in MTAP-del cells (69–71). Furthermore, treatment with AGI-24512, a precursor of S95033/AG-270, was associated with aberrations in cell cycle progression and RNA splicing, DNA damage, and reduced DNA damage repair in MTAP-del but not MTAP wild-type cells in a manner consistent with disruption of PRMT5 activity (70). MAT2A inhibition is also associated with the upregulation of MAT2A protein (16, 70), in contrast to MTA-cooperative PRMT5 inhibitors that have not been observed to affect the expression of either MAT2A or PRMT5 (67).

Clinical data for MAT2A inhibitors

S95033/AG-270 was evaluated in the phase I NCT03435250 clinical trial of patients with MTAP-del advanced solid tumors or lymphoma (Table 1). During dose escalation, 40 patients received S95033/AG-270 at dosages of 50 to 400 mg once daily or 200 mg twice daily, with the MTD determined to be 200 mg once daily (66). DLTs were reported in four patients who received S95033/AG-270 at 100 mg once daily (grade 2 hypersensitivity), 150 mg once daily (grade 3 hyperbilirubinemia and grade 3 maculopapular rash), 200 mg once daily (grade 3 decreased neutrophil count), or 200 mg twice daily (grade 3 drug-induced liver injury and grade 3 rash). TRAEs were observed in 70% of patients, most commonly hyperbilirubinemia, fatigue, and thrombocytopenia. Only two patients (5%) had an objective response, with 60% of patients experiencing progressive disease as the best response (66). Ultimately, NCT03435250 was terminated early because of a strategic decision to discontinue the S95033/AG-270 clinical development program, with a newer agent, S095035, currently in clinical investigation [NCT06188702 and NCT03435250 results summary (https://clinicaltrials.servier.com/trial/NCT03435250/study-of-ag-270-in-participants-with-advanced-solid-tumors-or-lymphoma-with-mtap-loss); ref. 66].

IDE397 dose escalation and expansion are currently being assessed in the phase I NCT04794699 clinical trial in patients with advanced MTAP-del solid tumors (Table 1; ref. 65). Dose escalation indicated a recommended phase II dosage (RP2D) of 30 mg orally administered once daily. To note, safety and efficacy data from this study were reported in patients who received the RP2D, which could hinder insights into the broader benefit/risk profile relative to an evaluation of all dose levels. Among 28 patients with NSCLC or urothelial cancer treated at this dose level in the dose-expansion phase, 54% reported any-grade TRAEs, most commonly peripheral neuropathy, decreased appetite, increased blood creatinine, nausea, and fatigue. There were no grade ≥3 TRAEs reported in >5% of patients, no serious TRAEs, and no discontinuations due to TRAEs. Among 27 patients with preliminary efficacy results, the ORR was 33%, and the DCR was 93%, including one patient with urothelial cancer who had a complete response. The median time to response was ∼2.7 months, and the median duration of treatment was >6.2 months. A molecular response analysis showed that IDE397 treatment reduced levels of circulating tumor DNA in all patients with evaluable samples (65).

Similar to the MTA-cooperative PRMT5 inhibitors, S95033/AG-270 and IDE397 are orally bioavailable, and IDE397, in particular, has yielded promising safety and efficacy results in preliminary clinical reports. Larger clinical trial experiences will be of interest to understand potential differences with respect to both antitumor activity and safety profiles between MTA-cooperative PRMT5 and MAT2A inhibitors.

Future Directions

The novel nature of the MTA-cooperative PRMT5 and MAT2A inhibitors and their preliminary evidence for efficacy across a variety of MTAP-deficient tumors have opened many potential avenues of current and future investigation, including new drug combinations, identification of predictive biomarkers or particularly sensitive tumor types, and optimization of patient identification for treatment. Further studies to investigate the mechanisms of action of MTA-cooperative and MAT2A inhibitors (e.g., the impact of impaired splicing and detained introns on tumor cell viability) and to identify the particular downstream pathways that are ultimately required for tumor cell survival may aid in these efforts.

Potential combinatorial treatments: MTA-cooperative PRMT5 and MAT2A inhibitors

Given the unique mechanisms of action used by MTA-cooperative PRMT5 and MAT2A inhibitors, these agents may have the potential to complement each other in the treatment of tumor tissue. The combination of the MAT2A inhibitor IDE397 and the MTA-cooperative PRMT5 inhibitor AMG 193 at low doses augmented the depth and durability of the antitumor response in MTAP-del cell lines, including complete regressions that were sustained off-treatment (72). Synergistic effects from combining IDE397 with other MTA-cooperative PRMT5 inhibitors have also been observed in MTAP-del lung adenocarcinoma and pancreatic cancer models, and a pharmacodynamic analysis using SDMA showed that the combination regimens induced an earlier onset and greater extent of PRMT5 inhibition in MTAP-del tumors than either agent alone (73). Although the combination of IDE397 and AMG 193 is currently under clinical investigation (NCT05975073), the study has ended patient recruitment (ClinicalTrials.gov: NCT05975073). It remains to be seen whether the combinations are truly synergistic or whether increased dosing of MTA-cooperative PRMT5 or MAT2A inhibitors alone can produce the heightened antitumor responses observed from combination treatment.

Potential combinatorial treatments: MTA-cooperative PRMT5 or MAT2A inhibitors with chemotherapy

Because MTA-cooperative PRMT5 and MAT2A inhibitors can suppress DNA damage repair pathways (70), it has been theorized that they may render tumor cells more sensitive to chemotherapy. An in vitro drug combination viability screen by Gerrick and colleagues (74) identified taxanes, platins, and antifolates (including pemetrexed), among other agents, as synergistic with IDE397 in MTAP-del xenograft models. IDE397 is currently being evaluated in combination with taxanes in the NCT04794699 clinical trial. Several other ongoing and upcoming clinical trials are investigating the combination of PRMT5 pathway inhibitors with chemotherapy, including the combination of BMS-986504/MRTX1719 with standard-of-care therapies in selected solid tumors in NCT07076121 and substudies of NCT05245500, and the combination of AMG 193 with carboplatin, gemcitabine, paclitaxel, pemetrexed, or fluorouracil/oxaliplatin/leucovorin/irinotecan in NCT06333951 and NCT06360354. Interestingly, MTAP deficiency is a predictive biomarker for improved pemetrexed efficacy in breast and lung cancers (75, 76). Because MTAP is responsible for salvaging nearly all the adenine in the human body, rendering MTAP-del tumors susceptible to further perturbation of adenine loss, pemetrexed itself has been found to induce synthetic lethality in MTAP-deficient tumors by targeting enzymes involved in de novo purine synthesis and depleting adenine availability (77). Pemetrexed treatment of seven patients with MTAP-deficient urothelial carcinoma showed a response rate of 43% although all the patients experienced treatment-related anemia (grade 3 anemia in 43%; ref. 77). Hence, clinical results from the combination of AMG 193 and pemetrexed in NCT06333951 will be of particular interest although the safety profile will require close assessment.

Potential combinatorial treatments: MTA-cooperative PRMT5 or MAT2A inhibitors with targeted therapy

KRAS alterations can co-occur with MTAP-del although with variable prevalence across different tumor types (78–80). BMS-986504/MRTX1719 has shown additive activity in combination with KRAS inhibitors in MTAP-del PDAC xenografts (58, 81). AMG 193 treatment has shown similar results in combination with the KRAS^G12C mutant–selective inhibitor sotorasib in PDAC and NSCLC xenograft models (52), with the combination currently being investigated in the phase I NCT06333951 clinical trial. Importantly, although KRAS mutations have been implicated in resistance to several targeted cancer therapies (82, 83), early data for BMS-986504/MRTX1719 from NCT05245500 show responses in patients with KRAS-mutated, MTAP-del tumors, suggesting that KRAS mutations are not a mechanism of resistance for MTA-cooperative PRMT5 inhibition (67).

Several other combinations of PRMT5 pathway inhibitors with targeted therapy are under investigation. Along with various chemotherapies, the previously mentioned viability screen by Gerrick and colleagues (74) identified topoisomerase inhibitors and splicing inhibitors as potentially synergistic with IDE397 in MTAP-del cells. The combination of IDE397 with the TROP-2–targeting TOP1 antibody–drug conjugate sacituzumab govitecan is currently being investigated in the NCT04794699 trial. Additionally, the combination of BMS-986504/MRTX1719 and olaparib, an inhibitor of PARP, which plays a crucial role in DNA damage repair, has shown synergistic activity in MTAP-del cholangiocarcinoma cell lines (84).

A number of other potentially targetable alterations have been identified to co-occur in various types of MTAP-del tumors, including alterations in BRAF, BRCA1, BRCA2, EGFR, FGFR3, PIK3CA, PTEN, SMARCA2, SMARCA4, and TP53 (76, 78–80, 85, 86), which could inform future combinatorial drug studies. In general, MTAP-del has not been found to be mutually exclusive with other targetable alterations (86), potentially rendering MTA-cooperative PRMT5 and MAT2A inhibitors suitable for inclusion in a wide variety of treatment regimens.

Potential combinatorial treatments: MTA-cooperative PRMT5 or MAT2A inhibitors with immunotherapy

MTAP deficiency is associated with a cold tumor immune microenvironment because of the immune suppressive effects of accumulated MTA and/or the associated loss of other genes (e.g., IFNs) in 9p21, leading to a tumor-favorable cytokine profile; decreased tumor-infiltrating leukocytes, with low proportions of T lymphocytes and NK cells and high proportions of immunosuppressive cells; T-cell exhaustion; and activation of immunosuppressive signaling, ultimately promoting tumor growth and resistance to immunotherapy (2, 7, 10, 87). A proteomic network analysis found aberrant immune responses mediated by the PI3K/AKT, MEK/ERK, JAK/STAT, and other tumor-intrinsic signaling pathways in MTAP-del cancer cells (10). Biomarkers associated with response to immunotherapy, including high tumor mutational burden and high tumor PD-L1 expression, have also been reported to occur at lower rates in MTAP-del versus MTAP wild-type tumors (2, 76, 78, 79, 88). An analysis of 42 patients with 9p21 deletion who had received anti–PD-1/–PD-L1 monotherapy for melanoma, lung, renal, head and neck, or esophageal cancers found a response rate of just 4% compared with 27% for a matched 9p21 wild-type control group (2). Altogether, these data suggest that 9p21/MTAP deletion in tumors confers resistance to immunotherapy and may partially contribute to the large proportion of patients in whom immunotherapy fails (2). Interestingly, a type I IFN gene cluster is also linked to and often co-deleted with CDKN2A and MTAP. Among patients with cancer captured in The Cancer Genome Atlas, deletion of this gene cluster along with CDKN2A/MTAP deletions seems to confer comparable overall survival versus patients with only CDKN2A/MTAP deletions (2). However, additional studies will be needed to assess the impact of deletion versus retention of these IFN genes on resistance to immunotherapy in the context of CDKN2A/MTAP deletions.

Some evidence suggests that MTA-cooperative PRMT5 inhibitors may improve the effectiveness of immunotherapy in MTAP-del tumors. In a preclinical study, BMS-986504/MRTX1719 reduced the activation of the PI3K pathway in MTAP-del tumor mouse models and sensitized MTAP-del tumor cells to T cell–mediated tumor killing (89). The combination of BMS-986504/MRTX1719 and an anti–PD-1 treatment conferred better antitumor activity in mice with MTAP-del tumors than either agent alone (89). Clinical data are anticipated from the NCT05245500, NCT07063745, and NCT06333951 clinical trials, which are exploring the combinations of BMS-986504/MRTX1719 or AMG 193 with immunotherapy in patients with MTAP-del advanced tumors. Additional analyses will be needed to investigate whether MAT2A inhibitors can similarly revert resistance to immunotherapy in MTAP-del cancers, as well as to assess the utility of these inhibitors in MTAP-del cancers that fail initial immunotherapy, similar to the use of enfortumab vedotin, an antibody–drug conjugate that has been shown to prolong survival in patients with urothelial carcinoma who previously progressed on immunotherapy (90).

Characteristics and biomarkers associated with response or resistance to MTA-cooperative PRMT5 or MAT2A inhibition

Although early data suggest that MTA-cooperative PRMT5 and MAT2A inhibitors may be efficacious across many different tumor types, thus far, particularly high response rates have been observed for BMS-986504/MRTX1719 in patients with NSCLC and mesothelioma and for IDE397 in patients with squamous NSCLC and urothelial carcinoma (Table 1; refs. 64, 65). BMS-986504/MRTX1719 and AMG 193 are currently being investigated in patients with MTAP-del advanced NSCLC in the NCT06855771 and NCT06593522 clinical trials, respectively. Lower response rates were observed for cholangiocarcinoma and PDAC for BMS-986504/MRTX1719 (64) although the ORR reported for these difficult-to-treat cancers still exceeded historical rates for second-line chemotherapy (91, 92). Larger studies will be needed to understand the mechanisms underlying differential response across tumor types and to assess efficacy in additional tumor types with a relatively high prevalence of MTAP-del.

The discovery of biomarkers predictive for response to MTAP-deficient tumor-selective inhibitors would also be of key importance in identifying patient subgroups with optimal response to MTA-cooperative PRMT5 and MAT2A inhibitors. Data available thus far align with the importance of altered splicing activity associated with PRMT5 inhibition in tumor cells. A high expression ratio of two mutually exclusive PRMT5 co-factors, CLNS1A and RIOK1, has been found to impair PRMT5-mediated spliceosome activity and sensitize tumor cell lines to tumor-nonselective PRMT5 inhibitors (62). Additional studies will be needed to confirm whether this can also act as a biomarker for MTA-cooperative PRMT5 inhibitor efficacy. Genome-wide CRISPR knockout screens have also identified loss of CAAP1 and AKAP17A, which form a complex that may mitigate alternative splicing events such as those induced by PRMT5 inhibition, as sensitizing for MTA-cooperative PRMT5 inhibition in MTAP-deleted cancer cell lines across different histologies (93). Interestingly, the CAAP1 gene is linked to MTAP and CDKN2A on chromosome 9p21 and is co-deleted in 20% of MTAP-del cancers (93). Future research will be needed to explore the role of alternative splicing events related to PRMT5 inhibition in tumor cells and to identify additional biomarkers potentially associated with sensitivity to targeting MTAP-del tumors.

Alternative modalities of MTAP deficiency and identification of tumors sensitive to MTA-cooperative PRMT5 or MAT2A inhibition

Although most clinical studies of PRMT5 pathway inhibitors in MTAP-deficient tumors have been conducted in patients with homozygous MTAP-del tumors, MTAP deficiency can occur from several other modalities, with potential sensitivity to tumor-selective PRMT5 inhibition ultimately determined by the MTA concentration. It is unclear whether other types of MTAP alterations can generate sufficient MTA accumulation in tumor cells, including heterozygous deletion, single-nucleotide variants, MTAP fusions, exon deletions leading to partial MTAP deletion, alternative MTAP splicing, epigenetic silencing, or other methods of MTAP protein loss. As such, it is unknown whether testing for genomic deletion [e.g., through next-generation sequencing (NGS)] or for protein loss (e.g., through IHC) best associates with treatment efficacy and whether this differs by inhibitor class. In a study comparing different MTAP testing modalities in patients with diffuse pleural mesothelioma, MTAP IHC was found to have 96% sensitivity, 86% specificity, and 93% accuracy for 9p21 homozygous deletion and had high agreement with 9p21 copy number determined by combined Fraction and Allele-Specific Copy Number Estimates from Tumor Sequencing/FISH (94). Interestingly, however, a tissue microarray study found that neuroendocrine tumors and Hodgkin lymphomas carried high rates of MTAP deficiency without MTAP deletion (95), suggesting that tumor type should be considered in assays for MTAP deficiency. Of interest will be data from the PRIMAVERA trial (NCT06137144) of AZD3470 in patients with hematologic malignancies, including Hodgkin lymphoma. Although MTAP deficiency is not required for enrollment in PRIMAVERA, MTAP protein loss has been observed in 84% of classical Hodgkin lymphoma samples even in the absence of MTAP deletion, potentially because of epigenetic silencing (96), and baseline tumor tissue will be assessed for MTAP expression for each patient enrolled in the study. The relevance of MTAP epigenetic silencing in solid tumors also remains to be established. To add to the complexities of testing for MTAP deficiency, some NGS assays solely test for CDKN2A deletion rather than MTAP itself. Although MTAP deletion in the absence of CDKN2A deletion is extremely rare, CDKN2A deletions can occur in the absence of MTAP deletions (2, 97), highlighting the importance of including MTAP in NGS assays or confirming MTAP deficiency when CDKN2A is deleted. Additional studies will be needed to assess modalities of MTAP deficiency that confer sensitivity to MTA-cooperative inhibitors and MAT2A inhibitors and to optimize testing for these MTAP alterations across different tumor types.

Conclusions

Overall, the early clinical data from studies of MTA-cooperative PRMT5 and MAT2A inhibitors indicate that MTAP deficiency is an actionable alteration for synthetic lethality and suggest the emergence of a promising new field of anticancer precision medicine. The ability to exploit the vulnerabilities of MTAP-del cells in target tumors represents a significant step in the evolution of PRMT5 pathway inhibition. Data for these tumor-selective inhibitors show notably greater tolerability and a wider therapeutic window than tumor-nonselective PRMT5 inhibitors, with fewer hematologic toxicities, higher response rates, and higher DCRs.

Although many synthetic lethal strategies have been attempted in clinical practice (98), the only anticancer drugs that have received US FDA approval to date are PARP inhibitors targeting BRCA1/2-mutated cancers, which have an increased dependence on PARP for DNA repair (99, 100). The first PARP inhibitor, olaparib, was approved in 2014, followed by rucaparib in 2016, niraparib in 2017, and talazoparib in 2018 (https://www.accessdata.fda.gov/drugsatfda_docs/label/2023/208558s025lbl.pdf, https://www.accessdata.fda.gov/drugsatfda_docs/label/2022/209115s011lbl.pdf, https://www.accessdata.fda.gov/drugsatfda_docs/label/2018/211651s000lbl.pdf, and https://www.accessdata.fda.gov/drugsatfda_docs/label/2020/208447s015s017lbledt.pdf). Preliminary data for drugs targeting the PRMT5 pathway in MTAP-deficient tumors show promise that these inhibitors may eventually comprise a new class of anticancer drugs sharing rare success as a synthetic lethal strategy.

The frequency of MTAP deletions across tumor types is similar to or exceeds that of many commonly targeted alterations, including KRAS (∼12%; ref. 101), BRAF (∼8%; ref. 102), EGFR (∼3%; ref. 103), and HER2 (∼3%; ref. 104). Given the fundamental roles of MTAP and PRMT5 in cell viability, it is possible that targeting MTAP-del cancer cells may provide tumor-agnostic benefit, differentiating this strategy from targets such as BRAF^V600E, in which efficacy is limited to certain tumor types (105), although early clinical trial experience has shown that antitumor activity can vary across tumor types. Along with their unique mechanisms of action and favorable safety profile, MTAP-deficient tumor-selective inhibitors could have the potential to act as a backbone for combinations with other therapies, including chemotherapy, targeted therapy, and immunotherapy. Outcomes from the many ongoing clinical trials for these inhibitors are eagerly anticipated, along with future research to identify effective combinatorial regimens and optimize the selection of patient populations for therapies targeting MTAP-deficient tumors. Overall, the pan-tumor applicability, versatility, and safety of MTA-cooperative PRMT5 and MAT2A inhibitors render them among the most promising novel classes of agents introduced in recent years.

Authors’ Disclosures

J. Rodon reports personal fees from Ellipses Pharma, iOnctura, Sardona Therapeutics, MEKanistic Therapeutics, Amgen, Merus, Monte Rosa Therapeutics, AADI Bioscience, BridgeBio, Vall d’Hebron Institute of Oncology, Chinese University of Hong Kong, Boxer Capital, LLC, Tang Capital Advisors, LLC, Guidepoint, and Axiom Pharma, other support from Blueprint Medicines, Merck Sharp & Dohme, Hummingbird Bioscience, AstraZeneca, 280 Bio, Vall d’Hebron Institute of Oncology/Cancer Core Europe, Servier Symphogen, BioAtla, Pfizer, Kelun-Biotech, GlaxoSmithKline, Taiho Pharmaceutical, Roche Pharmaceuticals, Yingli Pharma, Bicycle Therapeutics, Merus, AADI Bioscience, Fore Biotherapeutics, Loxo Oncology, Hutchinson MediPharma, IDEAYA Biosciences, Amgen, Tango Therapeutics, Mirati Therapeutics, Linnaeus Therapeutics, Monte Rosa Therapeutics, Kinnate Biopharma, Debiopharm, Biotheryx, Storm Therapeutics, BeiGene, MapKure, Relay Therapeutics, Novartis, Fusion Pharma, C4 Therapeutics, Scorpion Therapeutics, Incyte, Fog Pharmaceuticals, Tyra Biosciences, Nuvectis Pharma, HotSpot Therapeutics, Adcentrx Therapeutics, Vividion Therapeutics, AstraZeneca, Alnylam Pharmaceuticals, Immuneering Corporation, Alterome, and Exelixis, and nonfinancial support from European Society for Medical Oncology, American Society of Medical Oncology, National Taiwan University Cancer Center, 280 Biot, Dava Oncology, and STOP Cancer during the conduct of the study. M.L. Johnson reports nonfinancial support from Bristol Myers Squibb during the conduct of the study, as well as grants from Adaptimmune, ArriVent BioPharma, Bayer, Black Diamond Therapeutics, BlossomHill Therapeutics, Boehringer Ingelheim, Boundless Bio, Carisma Therapeutics, Conjupro Biotherapeutics, Corvus Pharmaceuticals, D3 Bio, Eli Lilly and Company, Erasca, Exelixis, Genmab, Genocea Biosciences, Harpoon Therapeutics, Helsinn Healthcare SA, IDEAYA Biosciences, IGM Biosciences, Immuneering Corporation, Immunitas Therapeutics, Immunocore Holdings, Incyte Corporation, LockBody Therapeutics, Loxo Oncology, Merus, Mirati Therapeutics, ModeX Therapeutics, Mythic Therapeutics, NeoImmune Tech, Neovia Oncology, NextPoint Therapeutics, OncoC4, Palleon Pharmaceuticals, Pierre Fabre, PMV Pharmaceuticals, Prelude Therapeutics, Puma Biotechnology, RasCal Therapeutics, Ribon Therapeutics, Shattuck Labs, Shanghai Junshi Biosciences, Silicon Therapeutics, Summit Therapeutics, SystImmune, Taiho Oncology, TCR2 Therapeutics, Tempest Therapeutics, TheRas, Tizona Therapeutics, Vividion Therapeutics, and Zymeworks, grants and other support from Amgen, AstraZeneca, Bristol Myers Squibb, Daiichi Sankyo, Genentech/Roche, Janssen Pharmaceuticals, Merck, Novartis, Nuvalent, Pfizer, Regeneron Pharmaceuticals, and Revolution Medicines, and other support from AbbVie, Alentis Therapeutics, BeOne Medicines, Biohaven Pharmaceuticals, Ellipses Pharma, Gilead Sciences, GlaxoSmithKline, Jazz Pharmaceuticals, Ottimo Pharma, and Zai Laboratory outside the submitted work. B. George reports personal fees from Ipsen, Foundation Medicine, Boston Scientific, Astellas Pharma, and Amgen, grants and personal fees from Taiho Oncology, Roche/Genentech, AstraZeneca, and Pfizer, and grants from Hoffman La-Roche, Toray, Hutchison MediPharma, Mirati Therapeutics, CARsgen Therapeutics, GlycoNex, Helix Biopharma, Tvardi Therapeutics, Transcenta, BioNTech, Legend Biotech, Elicio Therapeutics, and Obsidian outside the submitted work. K.C. Arbour reports nonfinancial support from Bristol Myers Squibb during the conduct of the study, as well as personal fees and other support from Bristol Myers Squibb, Revolution Medicines, and Eli Lilly and Company, personal fees from G1 Therapeutics, Amgen, Nuvalent, Merck, and Regeneron Pharmaceuticals, and other support from Verastem Oncology and Genentech outside the submitted work. No disclosures were reported by the other authors.