Etrumadenant (AB928; Arcus Biosciences) is a potent, orally bioavailable, selective dual antagonist of the adenosine A2A receptor (ADORA2A) and A2B receptor (ADORA2B). It competitively displaces adenosine from both receptor subtypes, blocking the Gs-coupled cAMP/PKA second-messenger cascade that mediates adenosine-induced T-cell exhaustion in the tumor microenvironment (TME). **SDH-specific mechanistic rationale — downstream of the HIF→CD73 axis (supported in non-SDH models only):** SDH loss → succinate accumulation → constitutive PHD2/PHD3 inhibition → permanent HIF-1α/2α stabilization (Mechanism 1). HIF-1α transcriptionally activates NT5E (CD73) via a canonical HRE in the NT5E promoter (Synnestvedt et al. J Clin Invest 2002, PMID 12370277); this is constitutive and oxygen-independent in SDH-deficient cells. Elevated tumor-surface CD73 dephosphorylates extracellular AMP → adenosine, which accumulates in the TME and engages ADORA2A and ADORA2B on infiltrating CD8⁺ T cells, CD4⁺ T cells, and NK cells. Adenosine receptor activation raises intracellular cAMP via Gs → adenylyl cyclase → PKA phosphorylation of CREB → suppression of effector gene programs (IFN-γ, perforin, granzyme B); impaired TCR proximal signaling (ZAP-70, LAT phosphorylation); and enhanced FoxP3⁺ Treg immunosuppressive activity. Leone & Emens (J Immunother Cancer 2018, PMID 29914571) reviewed the CD39-CD73-A2A/A2B axis as a major non-checkpoint immune-evasion pathway with a specific combination rationale alongside PD-1/PD-L1 inhibitors. Etrumadenant blocks this final receptor-signaling step, restoring T-cell effector function regardless of extracellular adenosine concentration. This is mechanistically complementary — and potentially combinable — with oleclumab (Mechanism 42), which targets the upstream NT5E/CD73 enzyme that produces adenosine on the tumor cell surface. The two interventions target different proteins, different cell types (tumor cell CD73 vs. T-cell ADORA2A), and different biochemical reactions (AMP hydrolysis blockade vs. receptor antagonism), making combined upstream + downstream adenosine pathway blockade a rational combination hypothesis. **Distinction from oleclumab (existing engine entry):** - Oleclumab: anti-CD73 IgG1 monoclonal antibody; targets tumor-surface NT5E; blocks AMP→adenosine production; requires IV administration; Phase 2 in NSCLC (NCT05061550, NCT03334617). - Etrumadenant: oral dual ADORA2A/ADORA2B small-molecule antagonist; targets T-cell/NK-cell adenosine receptors; blocks adenosine signal transduction; oral dosing; distinct clinical development history. Etrumadenant also antagonizes ADORA2B, which is expressed on myeloid-derived suppressor cells (MDSCs) and dendritic cells (DCs) in the TME; ADORA2B signaling on DCs limits their maturation and antigen-presenting capacity (Hatfield & Sitkovsky, Curr Opin Pharmacol 2016, PMID 27429212), so dual A2A/A2B blockade provides a broader immunostimulatory profile than ADORA2A-selective agents. **Clinical data:** - NCT03629756 (Phase 1; Arcus Biosciences; multiple solid tumors including NSCLC, HNSCC, breast, colorectal, melanoma, bladder, ovarian, endometrial, gastroesophageal, RCC, prostate; n=48; COMPLETED 2021): first-in-human Phase 1 dose escalation and expansion; established oral bioavailability, pharmacokinetics, target engagement (decreased plasma adenosine after dosing), and safety profile. Key toxicities manageable; myelosuppression not a dominant feature (unlike AURKA or PARP inhibitors). - NCT04262856 (Phase 2; Arcus Biosciences; front-line non-small cell lung cancer; domvanalimab + zimberelimab ± etrumadenant; n=151; COMPLETED 2025): Phase 2 combination efficacy data with anti-PD-1 (zimberelimab) ± anti-TIGIT (domvanalimab); establishes Phase 2 dosing, tolerability, and pharmacodynamic markers in a solid tumor IO combination context. Note: NCT03381274 is not an etrumadenant trial. It combined oleclumab with AZD4635, a different A2AR antagonist, in EGFR-mutant NSCLC, so A2AR antagonist + CD73 inhibitor combinations have been studied clinically. **Key limitation:** No published data test etrumadenant or any A2AR antagonist in SDH-deficient GIST, PPGL, or RCC cell lines, animal models, or patient-derived xenografts. ADORA2A expression in SDH-deficient tumor infiltrating lymphocytes specifically has not been quantified. CD73 expression in SDH-deficient tumor tissue (the upstream source of adenosine) has not been systematically confirmed at the protein level (the limiting upstream uncertainty shared with oleclumab). The mechanistic chain from SDH loss through pseudohypoxia, HIF-1α, CD73, adenosine, and ADORA2A on T-cells is well-established in non-SDH-specific hypoxia and cancer models but has not been demonstrated with SDH-null isogenic pairs. Required next step: CD73 protein expression in SDH-null vs. SDH-intact isogenic cell lines; extracellular adenosine quantification in conditioned medium; restoration of co-culture T-cell killing by etrumadenant (or oleclumab) in SDH-deficient cancer cell + T-cell co-culture assays. Evidence_score 22 (theoretical): mechanistic chain is well-supported (PMID 12370277, PMID 29367423, PMID 29914571) and each mechanistic step is supported in non-SDH hypoxia/cancer contexts only; oral clinical-stage drug with Phase 1 and Phase 2 safety and combination data; score held at 22 due to complete absence of SDH-specific preclinical data and unconfirmed CD73 protein expression in SDH-deficient tumors.