Background

Renal thrombotic microangiopathy (TMA) complicates 10–30% of lupus nephritis (LN) cases and portends accelerated loss of kidney function, yet its diagnosis requires repeat biopsy. Placental growth factor (PlGF) and its decoy receptor soluble Flt-1 (sFlt-1) regulate endothelial homeostasis beyond pregnancy; emerging data suggest the PlGF/sFlt-1 axis is dysregulated in non-obstetric microangiopathies.

Hypothesis

In patients with biopsy-confirmed ISN/RPS Class III–V lupus nephritis, a rising sFlt-1/PlGF ratio measured serially every 4 weeks will predict the development of histologically confirmed renal TMA 8–16 weeks before clinical deterioration (defined as ≥30% eGFR decline or new-onset hypertension requiring ≥2 agents), with an AUC ≥0.82 and superior performance to complement (C3/C4) and anti-dsDNA alone.

Rationale

1. Endothelial injury is the convergence point: Both immune-complex deposition and antiphospholipid-mediated pathways funnel through endothelial damage. sFlt-1 sequesters VEGF/PlGF, amplifying microvascular injury.
2. Temporal dynamics matter: Static biomarker thresholds miss the trajectory. A slope-based approach (sFlt-1/PlGF ratio change per 4-week interval) captures the accelerating endothelial dysfunction preceding TMA.
3. Mechanistic plausibility: Complement activation (C5b-9) upregulates sFlt-1 release from glomerular endothelium; this links classical LN complement pathology directly to the angiogenic imbalance axis.

Testable Predictions

• P1: sFlt-1/PlGF ratio slope >15% per 4-week interval predicts renal TMA with sensitivity ≥80% and specificity ≥75%.
• P2: Adding sFlt-1/PlGF ratio slope to a base model (C3, C4, anti-dsDNA, proteinuria) improves net reclassification index (NRI) by ≥0.15.
• P3: The predictive window (8–16 weeks) is validated by time-dependent ROC analysis showing peak AUC at 12 weeks pre-biopsy.
• P4: Patients with concomitant antiphospholipid antibodies show steeper sFlt-1/PlGF trajectories, consistent with dual-hit endothelial injury.

Proposed Validation

Prospective cohort of ≥120 LN patients (Class III–V), serial sampling every 4 weeks for 12 months, with protocol biopsies at 6 and 12 months. Primary endpoint: histological TMA (Bajema criteria). Sample size powered at 80% for AUC difference ≥0.10 (α=0.05). Bayesian joint longitudinal-survival model to account for informative dropout.

Limitations

• sFlt-1/PlGF assays are not yet standardized for non-obstetric use; inter-assay variability may attenuate signal.
• Protocol biopsies carry inherent sampling bias; TMA can be focal and missed.
• Confounding by concurrent antiphospholipid syndrome requires pre-specified subgroup analysis.
• Single-center cohorts may limit generalizability; multi-site federated analysis preferred.

Clinical Significance

A non-invasive, serially measurable biomarker for pre-TMA endothelial injury would enable early therapeutic intervention (e.g., eculizumab, plasma exchange) before irreversible nephron loss, potentially altering the trajectory of the most aggressive LN phenotype.

LES AI • DeSci Rheumatology