Rheumatology · Medical Literature
Risk Factors for Pulmonary Involvement in Patients with ANCA-associated Vasculitis: A Systematic Review and Meta-analysis.
Abstract
Pulmonary involvement (PI) in ANCA-associated vasculitis (AAV), especially interstitial lung disease (ILD) and diffuse alveolar hemorrhage (DAH), predicts poor outcomes, but a synthesis distinguishing these phenotypes is lacking. We searched seven databases up to April 1, 2026, and included observational studies. Using univariate and multivariate meta-analyses, we calculated pooled odds ratios (OR) or mean differences (MD). Nineteen studies (2,944 patients) were included. For ILD, robust independent risk factors were age > 65 years (OR = 3.60) and male gender (OR = 2.57). MPO-ANCA positivity was more frequent in the ILD group (OR = 2.29). Potential associated factors included smoking, D-dimer, ANA, low C4, and CA199. MPA subtype, cough, and dyspnea were more frequent in the ILD group. Factors associated with lower ILD frequency included GPA subtype, PR3-ANCA positivity, c-ANCA positivity, ENT involvement, gastrointestinal involvement, and a higher BVAS score. For DAH, potential associated factors included proteinuria, high BVAS score, Cr, BUN, PCT, decreased Hb, and low C3. PR3-ANCA positivity and hemoptysis were more frequent in the DAH group. For overall PI, independent risk factors were age (continuous OR = 1.04, > 65 years OR = 4.20), male gender (OR = 3.08), and high BVAS (OR = 3.03). MPO-ANCA positivity (OR = 2.55) was more frequent in the overall PI group, which largely overlapped with those for ILD, suggesting that ILD is the main driver of overall PI. This first meta-analysis combining univariate and multivariate approaches with an evidence-grading framework for ILD demonstrates distinct clinical patterns for ILD versus DAH, offering a clinically actionable tool for risk stratification and personalized management of PI in AAV.
Clinical relevance
Pulmonary involvement (PI) in ANCA-associated vasculitis (AAV) — encompassing interstitial lung disease (ILD) and diffuse alveolar hemorrhage (DAH) — carries significant prognostic weight, yet no prior synthesis had systematically distinguished the risk factor profiles of these two phenotypes. This meta-analysis of 19 observational studies (n = 2,944) addresses that gap by applying both univariate and multivariate pooling with an evidence-grading framework, providing rheumatologists and pulmonologists with phenotype-specific risk stratification data. Given the divergent pathophysiology and management of ILD versus DAH in AAV, this work is immediately clinically actionable for early risk identification and individualized monitoring strategies.
Key findings
1. For AAV-ILD, independent multivariate risk factors were age >65 years (pooled OR = 3.60) and male gender (pooled OR = 2.57), representing the most robust predictors identified. 2. MPO-ANCA positivity was significantly more frequent in the ILD group (pooled OR = 2.29), while PR3-ANCA positivity, c-ANCA positivity, and GPA subtype were each inversely associated with ILD frequency. 3. Additional factors with potential univariate association with ILD included smoking, elevated D-dimer, ANA positivity, low C4, and elevated CA199; MPA subtype, cough, and dyspnea were also more frequent in the ILD group. 4. Factors associated with lower ILD frequency included ENT involvement, gastrointestinal involvement, and higher BVAS score, suggesting a divergent organ-involvement pattern from DAH. 5. For AAV-DAH, potential associated factors included proteinuria, elevated BVAS score, elevated creatinine, BUN, procalcitonin (PCT), decreased hemoglobin, and low C3, reflecting a pattern of systemic inflammation and renal-pulmonary involvement; PR3-ANCA positivity and hemoptysis were more frequent in the DAH group. 6. For overall PI, independent multivariate risk factors were age as a continuous variable (OR = 1.04 per year), age >65 years (OR = 4.20), male gender (OR = 3.08), high BVAS score (OR = 3.03), and MPO-ANCA positivity (OR = 2.55); the overlap with ILD-specific predictors suggests ILD is the primary driver of overall PI in AAV.
Practice implications
Rheumatologists should prospectively risk-stratify AAV patients for pulmonary involvement at diagnosis using the following high-yield profile: age >65 years, male sex, MPO-ANCA/MPA subtype, and elevated BVAS — this constellation warrants early and systematic pulmonary evaluation including HRCT chest, as ILD is the dominant PI phenotype in this group. Conversely, patients with PR3-ANCA positivity, GPA subtype, or prominent ENT/gastrointestinal involvement have a lower ILD burden but should be monitored for DAH, particularly in the setting of hemoptysis, renal impairment (elevated Cr/BUN), decreased hemoglobin, or low C3. Integrating phenotype-specific risk factor profiles (ILD vs. DAH) into multidisciplinary care pathways can facilitate earlier pulmonary subspecialty referral, guide intensity of surveillance imaging, and inform personalized immunosuppressive management in AAV.