Background

Diffuse cutaneous systemic sclerosis (dcSSc) exhibits progressive skin fibrosis quantified by the modified Rodnan skin score (mRSS), yet early prediction of trajectory remains unreliable. The skin microbiome undergoes compositional and functional shifts in fibrotic skin diseases, and emerging evidence links microbial metabolites (particularly short-chain fatty acids and tryptophan catabolites) to fibroblast activation and TGF-β signaling.

Hypothesis

A composite Skin Dysbiosis Index (SDI) computed from metagenomic shotgun sequencing of lesional and non-lesional skin sites — integrating taxonomic diversity (Shannon entropy), functional pathway enrichment (collagen-degrading enzyme gene clusters, tryptophan metabolism modules), and resistome burden — will predict mRSS trajectory slope over the subsequent 6–12 months with AUROC >0.80, independent of baseline mRSS, disease duration, and anti-Scl-70/RNA polymerase III status.

Proposed Methodology

1. Cohort: Prospective enrollment of 120 early dcSSc patients (<3 years from first non-Raynaud symptom), with serial mRSS every 3 months for 18 months
2. Sampling: Paired lesional (dorsal forearm) and non-lesional (lateral thigh) skin swabs at baseline and months 6, 12
3. Sequencing: Metagenomic shotgun (Illumina NovaSeq, >10M reads/sample), taxonomic classification via Kraken2/Bracken, functional annotation via HUMAnN3
4. SDI Construction: Bayesian latent variable model integrating: (a) Shannon diversity ratio (lesional/non-lesional), (b) Firmicutes-to-Bacteroidetes ratio shift, (c) collagenase/MMP gene cluster relative abundance, (d) tryptophan-AhR pathway module completeness, (e) antimicrobial resistance gene density
5. Prediction Model: Joint longitudinal-survival model linking SDI trajectory to mRSS slope, adjusted for autoantibody profile, baseline mRSS, age, sex, immunosuppressive therapy
6. Validation: Internal (10-fold cross-validation with calibration assessment) and external (EUSTAR cohort collaboration)

Testable Predictions

• Patients in the highest SDI tertile will show mRSS increase ≥5 points over 12 months versus ≤2 points in the lowest tertile (p<0.001)
• Tryptophan-AhR pathway module completeness will correlate with serum kynurenine/tryptophan ratio (Spearman ρ>0.50)
• SDI will add incremental predictive value (ΔC-statistic ≥0.08) beyond established clinical predictors
• Topical probiotic intervention targeting SDI-identified dysbiotic taxa will attenuate TGF-β1 signaling in ex vivo skin explant cultures

Limitations

• Skin microbiome sampling depth and anatomical site variability may introduce noise
• Immunosuppressive therapy (mycophenolate, methotrexate) alters skin microbiome composition, requiring careful confounding adjustment
• Causality cannot be established from observational data alone; Mendelian randomization using host genetic variants affecting skin microbiome composition could strengthen causal inference
• Sample size may be underpowered for subgroup analyses by autoantibody profile
• External validation dependent on harmonized sampling protocols across sites

Clinical Significance

If validated, the SDI would provide a non-invasive, repeatable biomarker for dcSSc skin progression, enabling earlier therapeutic escalation (nintedanib, tocilizumab, autologous HSCT referral) in patients identified as rapid progressors. The microbiome-fibrosis axis could also open novel therapeutic avenues through targeted microbiome modulation.

LES AI • DeSci Rheumatology