Therapeutic Role of Synthetic Lethality in ARID1A-Deficient Malignancies

In 1998, it was discovered that mammalian switch/sucrose nonfermenting (SWI/SNF) complexes are associated with cancer.^[1] Twenty percent of all cancer types are associated with mutations in the 29 genes that make up the SWI/SNF complex, suggesting that this complex plays a crucial role in carcinogenesis and that genetic disruption of this complex may lead to tumor development.^[2–4] Of note, molecular alterations to AT-rich interaction domain 1A (ARID1A) are found in approximately 10% of human malignancies, making it the most frequently mutated gene in the complex.^[5]ARID1A is located on chromosome 1p36.11 and is a region of the genome frequently lost in cancer.^[6]

ARID1A mutations are prevalent in various cancers, including gynecologic, gastrointestinal, pancreatic, breast, urothelial, renal cell carcinoma, and non–small cell lung cancer (NSCLC).^[7–10] ARID1A contributes to signaling via key molecular pathways, such as phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA), phosphatase and tensin homolog, tumor protein p53, and protein phosphatase 2 scaffold subunit A alpha.^[11] In hepatocellular carcinoma, ARID1A displays a dual role, with its gain of function triggering tumor initiation and loss promoting tumor progression.^[12,13] ARID1A loss predicts poor overall survival in gastric cancer and suggests potential sensitivity to programmed death-1–programmed death-ligand 1 (PD-1–PD-L1) immune checkpoint therapies.^[14–16] ARID1A is essential for maintaining the integrity of pancreatic acinar cells and is a barrier to transformation and epithelial-mesenchymal transition in the pancreas.^[17] Loss of ARID1A, in the presence of the KRAS mutation and P53 loss, the most common molecular alterations in pancreatic ductal adenocarcinoma, exacerbates the aggressive behavior of pancreatic ductal adenocarcinomas.^[17] In breast cancer, ARID1A mutations are linked to cancer aggressiveness, prognosis, epigenetic regulation, and cell invasion through interactions with the RAB11FIP1 gene.^[9,18–23] In renal cell carcinoma, ARID1A silencing leads to increased cell proliferation and reduced cell death, with lower expression levels correlating with poor prognosis.^[24,25]ARID1A mutations are also identified in cancers of unknown primary origin, melanoma, colorectal cancer, and NSCLC.^[8,10,26,27] While the role of ARID1A in colorectal cancer remains unclear, its loss in NSCLC has been linked to poorer survival rates, indicating its potential as a valuable prognostic marker.^[28–30] Because ARID1A mutations are found in many cancer types and interact with other critical molecular pathways, an understanding of the association between these aberrant signaling pathways could lead to the use of a synthetic lethal strategy in which the simultaneous inactivation or alteration of two genes leads to cell death, while the inactivation or alteration of either gene alone does not.^[31,32]

This article aimed to review the role of ARID1A in cancer biology, its interaction with other signaling pathways, and the potential therapeutic strategies for targeting ARID1A-deficient cells, including epigenetic drugs and immune checkpoint blockade therapies.

ARID1A regulates gene transcription by directly controlling cancer-related gene expression or indirectly by recruiting or activating histone-modifier enzymes.^[33,34] This regulates immunologic response, chromatin arrangement, cell differentiation, and development.^[33–35]ARID1A may behave as a tumor suppressor or oncogene, depending on the cancer type and stage.

ARID1A, also known as BAF250a, is a key component of the SWI/SNF complex, belonging to the BRG1-associated factor (BAF) subclass.^[36–38] Inactivation or silencing of ARID1A has been associated with dysregulated transcriptional programs, cell cycle control, DNA damage response, checkpoint signaling, regulation of p53 targets, and telomerase activity in various types of tumors.^[39,40] ARID1A inhibits carcinogenesis by binding to yes-associated protein and transcriptional co-activator with PDZ-binding motif. This linkage prevents yes-associated protein/transcriptional co-activator with PDZ-binding motif from binding to the TEA-domain transcription factor, which is required to activate downstream target genes responsible for cellular proliferation.^[41] The posttranslational acetylation of Lys120 in the DNA-binding domain of p53 regulates apoptosis without impacting cell cycle control. ^[42] However, ARID1A mutations upregulate HDAC6, which deacetylates Lys120 on P53.^[43] This process inhibits P53’s proapoptotic action and promotes cancer. The upregulation of HDAC6 expression is also one of the consequences of ARID1A inactivation, which leads to deacetylation of Lys120 of P53. This posttranslational modification that is proapoptotic in nature and specifically governs apoptosis without affecting the regulation of the cell cycle. The inhibition of P53K120Ac by ARID1A mutations leads to the suppression of the apoptosis-promoting function of P53, thereby facilitating the advancement of cancer. Cell proliferation was observed to be elevated in cell lines carrying wild-type (WT) ARID1A with shRNA-mediated suppression of ARID1A but had no impact on ARID1A-null cells.^[44] However, restoring ARID1A in cells where it had been removed markedly reduced proliferation, providing more evidence for its function as a tumor-suppressor gene.^[44]

ARID1A exhibits context-dependent activity in various cancer models.^[39,45] ARID1A has tumor-promoting properties in the early stages of transformation.^[46] Hepatocellular carcinoma tumors overexpressed ARID1A in 83% of cases relative to healthy liver tissue, and higher ARID1A expression levels have been associated with poorer prognosis.^[47,48] WT ARID1A was also found in primary endometrial cancer tumors, but metastatic subclones from the same patient harbored detrimental mutations.^[11] Loss of ARID1A in mice with APC mutations slowed the progression of ovarian cancer and delayed the onset of colon cancer.^[49] Sun et al^[46] suggested the significance of taking stage, dosage, and tissue context into account when assessing ARID1A’s role and formulating treatment plans to modify the epigenetic machinery in cancer.

Understanding the biology of ARID1A has led to the identification of potential targets for therapeutic interventions to treat ARID1A deficient cancers (Fig. 1). Several biomarker-based studies (Table 1) to evaluate cancer therapeutics in ARID1A-deficient malignancies are ongoing.

The application of synthetic lethality as a strategy for developing cancer therapeutics was introduced after the efficacious employment of poly (ADP-ribose) polymerase 1(PARP) inhibitors in patients with ovarian cancer with BRCA mutations.^[50–54] Synthetic lethal strategy is used to target molecular alterations in gene pairs, which, when simultaneously inactivated, produces cell death.^[31,32] This strategy is widely used in cancers with loss-of-function molecular alteration in tumor suppressor genes. As cross-talk between ARID1A and other signaling pathways have been implicated in carcinogenesis, numerous therapeutic targets, such as PARP, enhancer of zeste homolog 2 (EZH2), PIK3CA, the glutathione metabolic pathway, and histone deacetylase 6 (HDAC6), have been investigated for the treatment of ARID1A-deficient cancers.^[55–62]

ARID1A regulates the DNA damage response (DDR) mechanism, which detects DNA damage and coordinates cellular responses.^[63] The ATR/checkpoint kinase 2 pathway, which is triggered by single-strand breaks (SSBs), and the ataxia telangiectasia-mutated/checkpoint kinase 1 (Chk1) pathway, which is activated by double-strand breaks (DSBs), serve as key DNA damage response regulators.^[64] ATR prevents premature mitotic entrance by activating the G2 checkpoint in the presence of DNA damage.^[65] Cell division cycle 25C (CDC25C) is an essential component of the G2/M transition, and its activity is regulated by Aurora kinase A (AURKA) and the ARID1A/ATR/CHK1 pathway.^[65,66] In colorectal cancer cells with ARID1A mutations, interruption of the ARID1A/ATR/CHK1 pathway promotes tumor cell proliferation by increasing CDC25C activity.^[65] In models with ARID1A mutations associated with elevated levels of checkpoint kinase 2, the inhibition of ATM results in the accumulation of cytosolic DNA and the activation of the cGAS/STING signaling pathway.^[67] This process increases the infiltration of T cells into ARID1A-deficient cells, contributing to enhanced immune response.

Subsequently, Williamson et al^[68] investigated the therapeutic potential of ATR inhibitors in ARID1A-deficient malignancies. Researchers discovered that ARID1A-deficient cancer cells were more sensitive to ATR inhibitors than their WT counterparts, possibly because of their inability to recover from replication stress.^[68] Significantly, ATR inhibition prevented the growth of ARID1A-deficient tumor xenografts in mice with minimal toxicity to healthy tissues.^[68] These results demonstrate the potential of ATR inhibitors as a synthetic lethal therapy for tumors with ARID1A deficiency, highlighting the need for further research and development of ATR-targeting molecules for cancer treatment.

PARP1 is best known for detecting DNA SSBs.^[69] Activated PARP1 mediates recruitment of the SSB repair (SSBR) machinery for DNA repair. Subsequently, PARP1 dissociates from the site to allow the repair machinery to access DNA. This allowed PARP1 to be released from DNA to another location to initiate the SSBR process.^[70] PARP1 has also been implicated in the repair of DNA DSBs by nonhomologous end-joining and alternative end-joining. PARP inhibitors are often used to interfere with DNA repair and induce tumor cell death in homologous recombination–deficient cancers, such as those caused by BRCA1/2 or other genes in the homologous recombination pathway.^[71,72] PARP inhibitors (PARPis) prevent the release of PARP1 from DNA.^[73] This PARP trapping prevents the repair machinery from accessing DNA. Thus, PARP trapping suppressed DNA repair and induced cell death.

PARPis are selectively lethal to cells lacking DNA repair components. In BRCA1/2-deficient cells, PARPis increase the number of DNA SSBs, which are converted into irreversible DNA DSBs during replication.^[74] DNA DSBs trigger cell death through apoptosis when the quantity of DNA damage is too high for repair.^[74] It has been suggested that ARID1A-deficient tumors could be vulnerable to PARPis, as ARID1A mutations can impair the DNA damage response.^[56] PARPi may be useful in treating cancers with ARID1A mutations, including breast cancer and cholangiocarcinoma.^[75,76] A case report detailed the favorable response to olaparib therapy for almost 13 months in a patient with pancreatic ductal adenocarcinoma who had a detrimental ARID1A mutation.^[77] Park et al^[78] discovered that ARID1A-deficient cells are more vulnerable to PARPi because ionizing radiation–induced DSBs make them rely on PARP-dependent repair pathways. Preclinical studies have shown promising results supporting this hypothesis, but its clinical relevance has not been confirmed yet.^[79,80] Ongoing clinical trials are investigating the efficacy of PARPis in ARID1A-deficient tumors, but further research is needed to establish this association.^[81,82]

However, ARID1A alterations have been linked to a paradoxical clinical scenario in ovarian cancer. Despite being correlated with PARPI sensitivity in preclinical contexts, ARID1A mutations appear remarkably to induce resistance to platinum-based chemotherapeutic agents in ovarian cancer.^[83–87] To investigate this paradox, one research group examined the complex mechanisms of platinum resistance associated with these mutations.^[88] They discovered that the absence of ARID1A protein resulted in a remarkable transcriptional upregulation of multidrug resistance-associated protein 2 (MRP2).^[88] MRP2, an efflux pump that facilitates the ATP-dependent active transport of platinum compounds across the cellular membrane, may play a crucial role in promoting platinum resistance.^[88,89] Thus, the observed transcriptional shift induced by ARID1A loss may provide a mechanistic explanation for the platinum resistance observed in patients with ovarian cancer.

ARID1A interacts with EZH2 at its carboxyl terminus and suppresses the interferon response mediated by EZH2. Gene expression patterns indicated that EZH2 and ARID1A directly targeted PI3K-interacting protein 1 (PIK3IP1), a negative regulator of PI3K-Akt signaling.^[56] ARID1A activates PIK3IP1 expression, whereas EZH2 suppresses PIK3IP1 expression. The ARID1A protein appears to inhibit EZH2 activity. The loss of ARID1A function consequently results in the suppression of PIK3IP1, which normally inhibits the PI3K pathway. By inhibiting unrestrained EZH2 activity, the PI3K pathway could be rendered ineffective. Tazemetostat, an oral EZH2 inhibitor, demonstrated significant clinical activity in a phase 2 study, with a 69% objective response rate, a median response duration of 10.9 months, and a median progression-free survival (mPFS) of 13.8 months in patients with EZH2-mutant relapsed or refractory follicular lymphoma.^[90] Tazemetostat showed a respectable safety profile.^[90] Bitler et al^[56] demonstrated that inhibiting EZH2 decreased cell proliferation and increased cell death in cancer cells with ARID1A mutations. Nonetheless, the phase 2 clinical trial (ClinicalTrials.gov identifier: NCT03348631) of Tazemetostat for patients with recurrent ovarian or endometrial cancer is still ongoing.

Targeting EZH2 in ARID1A-mutated tumors has been associated with the suppression of PI3K/AKT signaling, and studies have shown a linkage between the ARID1A and PI3K pathways in clear cell ovarian cancer.^[56,91] Class I PI3K is the most abundant of the four classes of PI3K. It plays a central role in cell survival, growth, proliferation, autophagy, differentiation, and metabolism by converting phospho-inositide 4,5-bisphosphate to phospho-inositide 3,4,5-triphosphate in various cellular membranes.^[92–96] Class I PI3K enzymes have a catalytic and regulatory subunit.^[97–99] The binding of regulatory subunits to catalytic subunits stabilizes the catalytic subunit proteins and permits the precise regulation of their enzymatic activity.^[95] Five regulatory subunit proteins are expressed from the following three genes: PIK3R1/p85/p55/p50, PIK3R2/p85, and PIK3R3/p55.^[97–99] Two catalytic subunit proteins are expressed from two genes, PIK3CA/p110 and PIK3CB/p110.^[97–99] Phosphatase and tensin homolog mutations or deletions are frequently observed in various malignancies, activating the PI3K/AKT pathway and facilitating the growth and progression of tumors.^[100–102]

Abnormal activation of this pathway, which serves as a downstream signal transducer for several cell surface receptors, is often associated with the development of cancer, such as somatic changes that activate the PI3K/AKT/mTOR pathway are seen in about 30% of patients with bladder cancer.^[103] In bladder cancer, the absence of ARID1A activates the PI3K signaling pathway, promoting cell proliferation and survival.^[104] Inhibitors of EZH2 and PI3K have proven to be particularly effective against ARID1A-deficient bladder cancer cells, both in vitro and in vivo.^[104] Additionally, breast cancer and gastric cancer cells with ARID1A deficiency exhibited enhanced sensitivity to therapy using small molecule inhibitors targeting the PI3K/AKT signaling pathway.^[59,105] When the PI3K/AKT pathway was inhibited, pancreatic cancer cells deficient in ARID1A were more sensitive to radiation in vitro due to increased apoptosis and a weakened DNA damage response.^[106] In addition, Yang et al^[107] discovered that a combination of PARP and PI3K inhibitors may be used to treat gastric cancer. Based on these findings, therapy of ARID1A-deficient cancers with inhibitors of the PI3K signaling pathway shows promise.

Gamma-glutamate cysteine ligase (GCL) is a rate-limiting enzyme synthesizing the antioxidant glutathione (GSH).^[108] ARID1A collaborates with the BAF complex, BRG1, the catalytic subunit of the SWI/SNF chromatin remodeling complex, RNA polymerase II, and the antioxidant transcription factor NRF2 to create a complex that binds to the transcription start site of SLC7A11.^[62] SLC7A11 encodes a subunit of the transporter XCT that imports cystine into the cell in exchange for glutamate.^[62,109] Thioredoxin reductase then converts the imported cystine to cysteine to produce thioredoxin.^[62] GCL uses both cysteine and glutamate to produce reduced GSH.^[62] Thioredoxin and GSH collaborate to control reactive oxygen species (ROS) levels and avert cell death. In ARID1A-deficient cells, however, SLC7A11 is weakly expressed, and XCT levels are low, resulting in a decrease in intracellular cystine, cysteine, and GSH ^[62] These cells are susceptible to inhibition of thioredoxin reductase and GSH by compounds such as auranofin, APR-246, and buthionine sulphoximine (BSO), which further depletes the antioxidant capacity of ARID1A-deficient cells.^[62] Consequently, ROS levels increase indiscriminately, causing cell mortality. ROS can damage cells and induce cell death At high concentrations, whereas cancer cells rely on the antioxidant GSH to combat excessive ROS, making GSH a prospective cancer treatment target.^[109–111] Under such conditions, inhibition of GGLC (glutamate-cysteine ligase synthetase catalytic subunit)results in apoptotic cell death in ARID1A-deficient gastric cancer cells.^[62,112] Similarly, the GSH inhibitor APR-246 and GGLC inhibitor buthionine sulfoximine (BSO) are effective against ARID1A-deficient ovarian clear cell carcinomas (OCCC) cells.^[113] BSO works by depleting cells of cysteine, a key component in GSH synthesis, which can lead to decreased levels of intracellular GSH.^[114] This depletion of GSH can sensitize cells to oxidative stress and induce cell death.^[115] In preclinical studies, BSO has been shown to have antitumor effects, particularly in combination with other chemotherapeutic agents.^[116] Therefore, inhibition of GSH by synthetic lethal targeting of GCL is a promising therapeutic approach for malignancies lacking ARID1A.

In preclinical mouse models, ARID1A mutations rendered OCCCs sensitive to treatment with pan–histone deacetylase (HDAC) inhibitors, such as suberoylanilide hydroxamic acid.^[91] This sensitivity is attributed to the recruitment of HDAC2 to ARID1A/EZH2 target genes such as PIK3IP1 in ARID1A-altered cells, not in ARID1A WT cells. As HDAC2, a co-repressor of EZH2, suppresses PIK3IP1 expression in an ARID1A status-dependent manner, inhibition of HDAC2 restores PIK3IP1 expression in ARID1A-deficient cells. As such, suberoylanilide hydroxamine (or vorinostat), a pan-HDAC inhibitor, suppressed the growth of oncogenic ARID1A-mutated OCCCs in orthotopic and genetic mouse models.^[91] Similarly, inhibition of HDAC6 with ACY1215 was associated with a reduction in the growth of ARID1A-mutated but not WT tumors, indicating that HDAC6 inhibitors selectively promoted apoptosis of ARID1A-mutated cells. The above findings provide the rationale for evaluating HDAC6 inhibitors in treating patients with ARID1A-deficient OCCC.

In 2018, Berns et al^[117] reported an in-depth study investigating the sensitivity of OCCC with ARID1A mutations to bromodomain and extra terminal domain (BET) inhibitors. The study identified BRD2, a member of the BET (bromodomain and extra terminal domain) family, as a crucial factor in ARID1A mutant cell line sensitivity to BET inhibition.^[117] Analyses of numerous OCCC cell lines suggest that cell lines containing ARID1A mutations, particularly JQ1 and iBET-762, are more sensitive to BET inhibitors.^[117] To validate these in vitro drug sensitivity findings, the researchers used both OCCC cell line xenografts and patient-derived xenograft models derived from OCCC patients.^[117] The results of these in vivo models corroborated the in vitro findings, highlighting the therapeutic potential of BET inhibitors for patients with ARID1A-mutated malignancies.^[117] Swisher et al^[118] also investigated the efficacy of the oral BET inhibitor PLX2853 as a monotherapy for ARID1A-mutated gynecologic malignancies and in combination with carboplatin for platinum-resistant ovarian cancer in a recent Phase 1b/2a clinical trial. Initial results from the study indicate that both PLX2853 monotherapy and its combination with carboplatin exhibit encouraging antitumor activity and tolerable toxicity profiles.^[118] These results emphasize the potential of BET inhibitors, such as PLX2853, as a novel therapeutic strategy for patients with ARID1A-mutated gynecologic malignancies and provide evidence for future clinical studies in this patient population.

ARID1A represses transcription of AURKA by occupying the AURKA promoter.^[65] Therefore, ARID1A deficiency enhances AURKA transcription, which activates the nuclear localization of cell division cycle 25C to promote the G2/M transition and mitotic entry.^[119] Thus, AURKA inhibition in ARID1A-deficient cells induces G2/M arrest and apoptosis.^[65] As ARID1A has a synthetic lethal interaction with AURKA in colorectal cancer cells,^[65] ARID1A-deficient cells are vulnerable to the action of AURKA inhibitors, warranting further clinical evaluation.

In preclinical OCCC models, ARID1A deficiency rendered cells sensitive to dasatinib, a proto-oncogene tyrosine-protein kinase, Abelson murine leukemia viral oncogene homolog 1, and KIT proto-oncogene receptor tyrosine kinase inhibitor.^[79] Dasatinib showed the highest specific inhibitory effect on ARID1A-mutant OCCC cells compared with the ARID1A WT cell lines.^[79] The sensitivity of ARID1A-mutant OCCC cells to dasatinib is attributed to G1–S cell cycle arrest and addiction of ARID1A-mutant OCCC cells to YES1, a dasatinib target that is highly expressed in OCCCs.^[79] Although these findings are promising for using dasatinib in treating OCCC, limited activity has been observed in clinical trials.^[120,121] However, it should be noted that the sample size was limited, and the study did not assess the ARID1A status in these patients.

ARID1A deficiency is associated with increased PD-L1 expression, mismatch repair (MMR) deficiency, microsatellite instability (MSI), a high mutation load, DDR, tumor microenvironment, and tumor-infiltrating lymphocytes.^[122] As all the above factors are considered predictors of response to the immune checkpoint blockade, it has been hypothesized that ARID1A deficiency may serve as a biomarker of response to immune checkpoint inhibitors (ICIs)^[14,123,124]

Increased PD-L1 expression has been observed in ARID1A-deficient ovarian and gastric cancers.^[122,125] Increased PD-L1 expression in ARID1A deficiency is attributed to upregulation of the PI3K pathway and DSBs induced by impaired ATR activation in ARID1A-deficient cells.^{[34,44,126–128]} In mouse models, mice bearing ARID1A-deficient ovarian tumors treated with the PD-L1 antibody had reduced tumor burden and significantly prolonged survival compared with the control group.^[122] Focusing on the prevalence and impact of ARID1A, ARID1B, and ARID2 mutations, Zhu et al^[129] analyzed genomic and clinical data of a cohort of patients with NSCLC from The Cancer Genome Atlas and five cohorts of patients from the Memorial Sloan Kettering Cancer Center who underwent ICB treatment. Among patients with an available PD-L1 score, higher PD-L1 scores were reported in patients with an ARIDA1B mutation, while lower PD-L1 scores were reported in patients with SMARCA4 mutation. In patients with low PD-L1 scores (< 50), the mPFS was significantly longer in patients with any SWI/SNF complex mutation (8.3 months) compared with WT patients (3.7 months; p = 0.001). They found that these mutations were associated with a higher mutational burden in the tumor and a better response to ICB therapy.^[129] In addition, patients with ARID1A, ARID1B, or ARID2 mutations exhibited superior progression-free survival and overall survival compared with patients without such mutations.^[129] In 2020, Okamura et al^[130] demonstrated the clinical significance of ARID1A alterations as a biomarker for predicting positive outcomes in anti-PD-1/PD-L immunotherapy patients. The researchers conducted a comprehensive investigation on a cohort of patients with cancer, including those with gastric, colorectal, and pancreatic cancers.^[130] Patients with ARID1A mutations (as determined by comprehensive genomic profiling) exhibited substantially prolonged progression-free survival after treatment withICIs targeting the PD-1/PD-L axis.^[130]ARID1A-inactivating mutations can result in considerable increases in PD-L1 expression, which makes tumors with ARID1A deficiency more sensitive to PD-L1 antibodies. Although these findings are compelling, additional research with suitable controls is required to determine the prognostic versus predictive role of ARID1A mutations in immune checkpoint blockade.

MMR is a DNA repair process that primarily corrects mismatched bases to preserve genomic stability.^[131] ARID1A activates MMR by enlisting MSH2.^[122] Thus, ARID1A deletion can lead to MMR deficiency and MSI in several types of cancer, including endometrioid carcinoma of the uterus, gastric cancer, and colorectal cancer.^[132–135] It is unclear if ARID1A loss results in a functional deficiency of ssDNA repair by MSH2 or if alterations in ARID1A result from MSI. To this end, a study was conducted in patients with sporadic MSI endometrial cancer (tumors MLH1 expression because of promoter hypermethylation of the MLH1 gene) and those with germline tumors (Lynch syndrome).^[127] Loss of ARID1A expression was reported in 75% of patients with sporadic MSI tumors as against 14% of those with germline tumors, suggesting that ARID1A could be a causal gene rather than a target gene of MSI. Given that MSI can lead to the accumulation of mutations, producing neoantigens, ARID1A-deficient tumors may be sensitive to immune checkpoint blockade.^[136,137]

High tumor mutation burden (TMB) is more typical in cancers with mutated ARID1A than tumors with WT ARID1A.^[138] DDR gene alterations are emerging as promising predictive biomarkers for immunotherapy response.^[139,140] Patients with high TMB had better immunotherapy responses than those than those who had low TMB, suggesting a potential predictive role for TMB in immunotherapy outcomes.^[130,141] Given the correlation between elevated TMB and enhanced immunotherapy response, it is plausible that ARID1A-mutated tumors would also exhibit enhanced immunotherapy responses. However, additional research is required to validate this hypothesis and establish a direct link between ARID1A mutations, TMB, and immunotherapy outcomes.

ARID1A expression is significantly decreased in tumors with increased levels of tumor-infiltrating lymphocytes (TILs), and biliary tract tumors can be divided into immunologically “hot” and “cold” subgroups.^[142] According to accumulating evidence ARID1A alterations may serve as prospective biomarkers for predicting an immunotherapy response. After anti–PD-1/PD-L1 immunotherapy, patients with ARID1A mutations experienced prolonged progression-free survival, and this association has been observed across several cancer types.^[5,130] In addition, ARID1A mutations have been linked to elevated immune activity in gastrointestinal cancer.^[143] These results suggest a link between ARID1A and the immune response in the tumor microenvironment, which could have significant implications for cancer treatment strategies. Additional research is required to elucidate the precise function of ARID1A alterations in shaping the immune response and predicting immunotherapy outcomes in various types of cancer. By comprehending these predictors, researchers and clinicians can better customize immunotherapy approaches for patients with ARID1A-deficiency malignancies, thereby improving treatment outcomes.

In addition to ARID1A, mutations in other SWI/SNF components, such as ARID1B, SMARCA4, and SMARCB1, have been found in several human cancers.^[144–148] Genes for components of the SWI/SNF chromatin-remodeling complex, ARID1A and its homolog ARID1B, have similar functions.^[149] However, they are mutually exclusive and differ in kinetics.^[150] Their expression profiles during the cell cycle were significantly different. While ARID1A accumulates during the G0 cell cycle phase, its levels continue to decline throughout the remaining phases of the cell cycle; however, ARID1B levels continue to rise even during mitosis.^[55,151] Although mutations in ARID1A or ARID1B in NSCLC are both associated with higher TMB, increased PD-L1 expression, and improved response to ICB, loss of ARID1B destabilizes the SWI/SNF complex and inhibits cell proliferation in ARID1A-mutant tumor cell lines.^[128,152]ARID1B is a potentially attractive therapeutic target for synthetic lethality in tumors with ARID1A mutations.

The frequent co-occurrence of ARID1A and BRG1 mutations in various cancers presents a potential therapeutic opportunity.^[153] When ARID1A function is lost, cancer cells may rely more on other SWI/SNF complex functional components, such as BRG1.^[149] Inhibiting BRG1 can induce synthetic lethality in ARID1A-deficient cancer cells, selectively eliminating them while sparing normal cells with functional ARID1A.^[56,150,154] Developing targeted therapies, like small-molecule inhibitors, to selectively inhibit BRG1 in ARID1A-deficient tumors could offer a promising treatment approach for patients with these cancers.^[155]

SMARCB1 is another component of SWI/SNF chromatin remodeling complex that promotes stability of ARID1A through enhancer formation and function.^[144] Loss of expression of SMARCB1, the hallmark feature of renal medullary carcinomas and rhabdoid tumors, contributes to poor differentiation and aggressive behavior of tumors.^[156,157] Re-expression of SMARCB1 significantly increases protein levels for numerous SWI/SNF subunits, particularly the tumor suppressor subunits ARID1A and ARID1B, represses the oncogenic and ferroptosis resistance programs, and promotes epithelial programs.

ARID1A mutations observed across tumor types result in the loss of ARID1A expression. Loss of function ARID1A alterations negatively impact cellular differentiation, cell cycle, and DNA damage repair. The crosstalk between ARID1A and other signaling pathways allows for the synthetic lethal targeting of ARID1A-deficient cells with inhibitors of the PI3K pathway, PARP, EZH2, and HDACs. Furthermore, the association between ARID1A deficiency and MSI, high TMB, increased PD-L1 expression, and tumor-infiltrating lymphocytes renders them vulnerable to ICI blockade. Large-scale clinical trials are needed to evaluate the agents that target ARID1A.