Background/Purpose: Retroperitoneal fibrosis (RPF) is a rare fibroinflammatory disease of the retroperitoneum that can lead to obstructive uropathy and other life-threatening complications. While RPF can arise secondary to malignancy, IgG4-related disease, or vascular inflammation, many cases remain idiopathic. Despite increasing awareness, diagnosis remains challenging due to reliance on non-specific imaging or invasive biopsy. Blood-based biomarkers capable of accurately distinguishing RPF from healthy individuals are needed to facilitate earlier and less invasive diagnosis.

Methods: Plasma from 45 individuals with RPF (Non-idiopathic: IgG4-related = 9, malignancy-associated = 3, Erdheim-Chester disease = 4, inflammatory abdominal aortic aneurysm = 4; Idiopathic = 19) and 6 healthy (non-disease) controls were profiled using the SomaLogic SomaScan 7K Assay, which quantifies over 7,000 proteins per sample. Differential abundance analysis was performed using linear regression analysis, adjusting for clinical and demographic covariates (e.g., age, sex, BMI, creatinine levels, platelet counts, hemoglobin, and IgG4 levels). Proteins significantly more or less abundant in RPF cases compared to control as well as idiopathic RPF compared to non-idiopathic RPF (significance criteria: P < 0.05 and mean fold-change ≥ 25%) were identified. A machine learning pipeline was trained to distinguish RPF cases, and its performance was estimated across 100 bootstrap iterations.

Results: A total of 176 plasma proteins were differentially abundant between RPF patients and controls while controlling potential confounders. Proteins notably elevated in RPF included those associated with inflammation and immune dysregulation (e.g., Ferritin, Serum amyloid A [SAA], and T-cell immunoglobulin and mucin domain-containing protein 3 [TIMD3]). Pathway enrichment analyses found that these differentially abundant proteins mainly participated in fibrosis-related inflammatory tissue remodeling processes, such as cytokine-chemokine signaling, apoptosis, interleukin signaling, and integrin-mediated cell adhesion pathways. We also identified protein differences amongst etiological subtypes of non-idiopathic RPF. Additionally, 80 proteins showed differential abundance between idiopathic and non-idiopathic forms of RPF, including chemokine ligand 7 (CCL7) and junction adhesion molecule-like 1 (JAML1). Non-idiopathic RPF subtype-specific analyses further identified proteins uniquely elevated within subtypes compared to controls. Finally, an Extra-Trees classifier trained on these plasma protein profiles demonstrated potential in distinguishing RPF cases from controls, achieving a mean ROC-AUC of 0.81.

Conclusion: This study identifies key inflammatory proteins and immune-related pathways significantly associated with RPF. In all, these plasma-derived proteins offer promising candidates for minimally invasive diagnostics, as well as provide insight into systemic inflammatory mechanisms underlying RPF pathology. Looking ahead, follow-up studies incorporating disease controls will be necessary to further validate specificity and clinical utility.

« Back to ACR Convergence 2025

ACR Meeting Abstracts - https://acrabstracts.org/abstract/plasma-proteomic-profiling-identifies-inflammatory-proteins-and-pathways-associated-with-non-idiopathic-and-idiopathic-retroperitoneal-fibrosis/