Dual positive-feedback loops and a bidirectional lactate rheostat converging on the cGAS–STING pathway in rheumatoid arthritis

Rheumatoid arthritis (RA), with persistent synovitis and irreversible joint damage, shows one-third primary or secondary non-response to biologic disease-modifying antirheumatic drugs. We synthesize evidence and propose two positive-feedback loops (Loop 1: NETosis–inflammation; Loop 3: mtDNA–mitophagy) and a bidirectional lactate rheostat (Loop 2) of unresolved sign, converging on cyclic GMP-AMP synthase-stimulator of interferon genes (cGAS–STING). Loop 1 initiates innate inflammation via neutrophil extracellular trap-derived DNA. Loop 2, the most speculative, proposes that glycolysis-driven lactate remodels cGAS via AARS1/2-mediated lactylation; because this modification can inhibit cGAS or stabilize it by impeding MARCHF5 ubiquitination, its net RA direction is unresolved and may be anti-inflammatory. Loop 2’s pro-inflammatory output rests on better-supported lactate-induced mtDNA release fueling cGAS ligand supply (Loop 3 crosstalk); cGAS stabilization remains an unvalidated extrapolation from SLE monocytes, tumors, and in vitro data. Loop 3 may amplify inflammation via TNF-α-dependent repression of PINK1-governed mitophagy, causing cytosolic mitochondrial DNA leakage. An adaptive output-and-reinforcement layer (ACPA/RF, B cells, TPH/Th17 cells, immune complexes) links innate sensing to autoimmune perpetuation. If validated, Loop 1–3 crosstalk with this sign-unresolved rheostat would establish self-sustaining inflammatory machinery explaining chronic progression and anti-TNF non-response. We assess druggability—cGAS–STING antagonists, peptidylarginine deiminase 4 (PAD4) inhibitors, monocarboxylate transporter (MCT) blockers, PINK1 agonists, bromodomain and extra-terminal domain (BET) modulators—and propose biomarker-guided stratified combination pharmacology, hypothesizing that single-loop monotherapy may be vulnerable to compensatory activation. All biomarker cutoffs and combination regimens—including PAD4, MCT, and BET inhibitor triple combination—remain untested in animal models and unvalidated in prospective human cohorts, and none is clinically actionable. Systemic STING blockade may disrupt B-cell tolerance and exacerbate humoral autoimmunity; myeloid-selective strategies and human-validated (non-Cys91) antagonists are therefore favored, though untested. An associated FLS epigenetic imprinting module is a proposed loop-agnostic target in refractory disease.