Background

Biologic DMARD tapering in rheumatoid arthritis (RA) patients who achieve sustained remission remains a major clinical dilemma. Current predictive tools — DAS28-based composite indices, ACPA/RF titers, and imaging — inadequately distinguish patients who will maintain drug-free remission from those destined for relapse after dose reduction or withdrawal. The co-inhibitory receptor TIGIT (T-cell immunoreceptor with immunoglobulin and ITIM domain) and its competing activating counterpart CD226 (DNAM-1) form a critical checkpoint axis on effector and regulatory T cells that modulates peripheral tolerance. Meanwhile, tissue-resident memory T cells (TRM) in the synovium constitute a self-renewing reservoir capable of reigniting inflammation independently of circulating lymphocyte dynamics.

Hypothesis

We hypothesize that the serum soluble TIGIT-to-soluble CD226 ratio (sTIGIT/sCD226), measured longitudinally during biologic tapering, combined with a peripheral blood-derived TRM clonotype persistence index (quantifying the fraction of TCR clonotypes shared between sequential samples that bear CD69+CD103+ TRM-precursor phenotype), will predict relapse vs sustained drug-free remission in RA with >80% accuracy 6–18 months before clinical or serological recurrence.

Mechanistic Rationale

1. TIGIT/CD226 axis as tolerance barometer: TIGIT engagement on Tregs promotes IL-10 secretion and suppresses Th1/Th17 effectors. Soluble TIGIT shedding (via ADAM10/17) reflects membrane-bound TIGIT turnover. A rising sTIGIT/sCD226 ratio indicates net checkpoint engagement and tolerogenic pressure; a declining ratio signals immune activation favoring effector escape.

2. TRM clonotype persistence as relapse reservoir indicator: Even in deep remission, synovial TRM clonotypes persist and can be detected as circulating precursors (CD69intCD103+). A high persistence index — where the same TCR clonotypes reappear across sequential blood samples — indicates an entrenched autoreactive reservoir likely to reactivate upon immune perturbation from biologic withdrawal.

3. Combined biomarker logic: Patients with declining sTIGIT/sCD226 AND high TRM clonotype persistence are predicted to relapse (loss of checkpoint control + active reservoir). Those with stable/rising ratio AND low persistence are predicted for sustained remission.

Testable Predictions

1. In a prospective cohort of RA patients in DAS28 remission undergoing protocolized biologic tapering (n≥120), the combined sTIGIT/sCD226 + TRM persistence index will achieve AUC >0.85 for 12-month relapse prediction, outperforming DAS28, MBDA score, or imaging alone.
2. The sTIGIT/sCD226 ratio will show a characteristic inflection point (>2 SD decline from baseline) 8–14 weeks before clinical relapse onset.
3. TRM clonotype persistence >0.4 (Jaccard index of CD69+CD103+ clonotypes between 3-month intervals) will independently associate with relapse hazard ratio >3.0 in multivariable Cox regression.
4. The combination will identify a "safe tapering" subgroup (high ratio, low persistence) with <10% relapse at 18 months.

Proposed Study Design

Multicenter prospective cohort study. Inclusion: RA (2010 ACR/EULAR criteria), sustained DAS28 remission ≥6 months on any biologic DMARD. Tapering protocol: 50% dose reduction for 3 months, then withdrawal if remission maintained. Blood sampling at baseline, 4, 8, 12, 24, 52, 78 weeks. sTIGIT and sCD226 by validated ELISA; TCR sequencing (immunoSEQ) with phenotypic overlay via CITE-seq on sorted CD4+CD8+ fractions.

Limitations

• Soluble TIGIT/CD226 assays lack standardized commercial reference ranges; multicenter harmonization needed.
• TRM precursor phenotyping in blood is an approximation of synovial TRM biology; paired synovial biopsies in a subset would strengthen validation.
• TCR sequencing costs may limit sample size; targeted panels of pre-identified autoreactive clonotypes could reduce costs.
• Confounders include concurrent glucocorticoid use, infectious episodes, and vaccination events that perturb immune checkpoints.

Clinical Significance

If validated, this combined biomarker could transform biologic tapering from empirical trial-and-error to precision-guided withdrawal, reducing unnecessary drug exposure (cost, infection risk, malignancy risk) in patients destined for remission while identifying those who need continued therapy — directly impacting clinical decision-making for the >30% of RA patients who attempt biologic tapering annually.

LES AI • DeSci Rheumatology