Idiopathic retroperitoneal fibrosis: long-term risk and predictors of relapse

Rationale & Objective: Idiopathic retroperitoneal fibrosis (IRF) is a rare disorder of unknown cause. Medical therapy can induce remission, but disease relapses are common. This study sought to characterize long-term outcomes of IRF and the factors associated with disease recurrences. Study Design: Retrospective cohort study. Setting & Participants: Retrospective analysis of 50 patients with IRF prospectively followed up for 8.9 (IQR, 4.7-12.7) years at a tertiary-care referral center. Exposures: Demographic, clinical, treatment, and laboratory parameters, including measures of autoimmunity. Outcome: Disease relapse. Analytical Approach: Proportional hazards analysis for the subdistribution of competing risks. Results: 49 patients received medical treatment and 35 underwent interventional procedures. All patients experienced a clinical response (defined as regression of disease-related symptoms and hydronephrosis, and decrease in the maximal transverse diameter of the retroperitoneal mass on computed tomography of >50%), 44 of whom responded within 1 year. The remaining 6 responded over a median of 2.95 years after starting therapy. 40 patients were alive at last observation, 1 receiving maintenance dialysis and 15 with estimated glomerular filtration rate < 60 mL/min/1.73 m2. Patient survival at 5, 10, and 15 years was 95%, 84%, and 68%, respectively. 19 (38%) patients had at least 1 relapse (occurring a median of 5.19 years after starting therapy), defined as an increase in serum creatinine level of at least 30% or recurrence/development of hydronephrosis and ≥20% increase in the maximal transverse diameter of the retroperitoneal mass on computed tomography. Cumulative incidences of relapse at 5, 10, and 15 years were 21%, 41%, and 48%, respectively. Baseline antinuclear antibody positivity and male sex were associated with relapse (subdistribution hazard ratios [sHRs] of 5.35 [95% CI, 2.15-13.27] and 4.94 [95% CI, 1.32-18.57], respectively), while higher corticosteroid therapy dosage at 1 year (sHR for relapse per 1-mg/d greater dosage, 0.91 [95% CI, 0.84-0.98]) and treatment with prednisone alone or with tamoxifen (sHR for relapse of 0.25 [95% CI, 0.07-0.85] vs other therapies) were associated with lower rate of relapse. Limitations: Small sample size and variable approaches to therapy. Conclusions: IRF relapses were common and were experienced more frequently by male patients. Corticosteroids alone or with tamoxifen were associated with a lower rate of relapse. The strong association of antinuclear antibody positivity with relapse supports the hypothesis of an autoimmune pathogenesis of IRF.