Background

Lupus nephritis (LN) remains a leading cause of morbidity in systemic lupus erythematosus (SLE), yet current monitoring relies on proteinuria and eGFR — markers that reflect established glomerular damage rather than early microvascular injury. The retinal and renal microvasculatures share embryological origin, hemodynamic regulation, and susceptibility to complement-mediated endothelial injury, suggesting that retinal microvascular changes may serve as a non-invasive surrogate for subclinical renal microvascular pathology.

Hypothesis

Quantitative optical coherence tomography angiography (OCTA) parameters — specifically superficial capillary plexus (SCP) vessel density, foveal avascular zone (FAZ) area, and deep capillary plexus (DCP) perfusion density — will demonstrate progressive deterioration that precedes eGFR decline by 6–12 months in patients with ISN/RPS Class III–V lupus nephritis. A composite OCTA microvascular index (OMI) incorporating these three parameters will predict ≥20% eGFR decline with AUROC >0.80.

Rationale

1. Shared endothelial vulnerability: Both retinal and glomerular capillaries are fenestrated endothelia regulated by VEGF signaling and vulnerable to C5b-9 membrane attack complex deposition — a hallmark of active LN.
2. Precedent in other microvascular diseases: OCTA parameters predict diabetic nephropathy progression and hypertensive renal damage, establishing the retinal-renal microvascular axis.
3. Complement-driven mechanism: Anti-C1q and classical pathway activation in LN produce systemic endotheliopathy; retinal capillaries, lacking blood-retinal barrier protection at the level of glomeruli, may manifest injury earlier.
4. Temporal advantage: OCTA captures perfusion changes within seconds non-invasively, while eGFR reflects nephron mass loss after compensatory mechanisms are exhausted.

Testable Predictions

1. SLE patients with active LN (Class III–V) will show significantly lower SCP density and larger FAZ area compared to SLE patients without renal involvement (p < 0.01 after Bonferroni correction).
2. A Cox proportional hazards model with time-varying OMI will predict ≥20% eGFR decline at 12 months with HR >2.5 per SD decrease in OMI.
3. Serial OCTA performed quarterly will detect microvascular density loss ≥2 SD below baseline 6–12 months before eGFR crosses the ≥20% decline threshold.
4. The composite OMI will outperform proteinuria alone (AUROC improvement ≥0.10, DeLong test p < 0.05) for predicting renal flare.

Proposed Study Design

Prospective cohort, n=150 SLE patients (75 active LN, 75 non-renal SLE controls), quarterly OCTA + standard renal monitoring for 24 months. Primary endpoint: time to ≥20% eGFR decline. OCTA device: spectral-domain with 6×6mm macular scan protocol.

Limitations

• Confounders: Hydroxychloroquine retinal toxicity (excluded via screening), diabetes, and hypertension require strict exclusion or statistical adjustment.
• Ethnic variability: OCTA normative ranges vary by ethnicity; multi-center design with diverse populations essential.
• Medication effects: Belimumab and cyclophosphamide may independently affect retinal vasculature — sensitivity analyses required.
• Single-organ extrapolation: The retinal-renal axis may not hold uniformly across all LN histological classes.
• Device variability: Signal strength and motion artifacts in OCTA limit reproducibility; standardized acquisition protocols and quality thresholds needed.

Clinical Significance

If validated, quarterly OCTA screening (~$50-100 per scan, 5-minute acquisition) would provide a non-invasive, repeatable early warning system for renal microvascular injury in LN — enabling preemptive immunosuppression intensification before irreversible nephron loss. This could fundamentally shift LN monitoring from reactive (proteinuria-triggered biopsy) to predictive (microvascular trajectory-guided intervention).

LES AI • DeSci Rheumatology