Pos0195 Does Serial Anca Testing Help Predict the Risk of Relapse in Patients With Granulomatosis With Polyangiitis? A Single-Centre Cohort Study

Background: ANCA-associated vasculitis (AAV) is a frequently relapsing disease for which a meta-analysis concluded that an increase or persistence of ANCA during remission is only moderately predictive of future disease relapse and that serial ANCA measurements during remission are of limited use in guiding treatment decisions for individual patients. However, this analysis used heterogeneous ANCA assays and combined different subgroups of AAV patients or PR3 and MPO-AAV patients, whereas granulomatosis with polyangiitis (GPA) and PR3-AAV have a higher risk of relapse. Objectives: We investigated whether serial ANCA testing with a uniform ELISA assay could be of interest in predicting the relapse risk in a cohort of GPA patients. Methods: To be included, patients had to have GPA follow-up at the Cochin National Referral Centre, at least two measurements of ANCA titer by ELISA, one less than 12 months apart during inactive disease, with a follow-up visit and assessment of disease activity within 12 months of the second ANCA titer measurement. The primary objective was to evaluate the association of an increase in ANCA titer at the time of inactive disease with any relapse within the following 12 months of follow-up. The secondary objective was to evaluate the association of ANCA titer increase with major relapses. The primary endpoint was the cumulative incidence of relapse within the following 12 months. The secondary endpoint was the cumulative incidence of major relapse. We estimated 12-month adjusted restricted mean survival times (RMST) for patients with an increase in ANCA levels compared with those without an increase. To obtain confounder-adjusted relapse-free survival curves for each group, we used direct standardisation (G-computation) using a Cox regression model with potential confounders. The difference in RMST between the two groups and its 95% confidence interval were calculated. Results were analysed globally in patients with an increase in ANCA titer (Increase group) versus those without an increase (No increase group) and in a subgroup focused on ANCA PR3 positive patients. In a sensitivity analysis, we compared patients with doubling of ANCA versus no doubling. Results: 121 GPA patients were analysed in the primary analysis (PR3-ANCA+: 101; MPO-ANCA+: 20) for a total of 357 countable periods. We observed 52 relapses, 5 in the Increase group (unadjusted relapse rate: 18.8%) vs. 47 in the No increase group (unadjusted relapse rate: 13.0%). The adjusted relapse rate was 13.3% (3.2 to 32.0) in 23 patient-periods vs. 13.0% (8.9 to 17.5) in 334 patient-periods. Of the patients with increasing ANCA levels (23 periods), 18 did not relapse during the one-year observation period and 5 relapsed. In PR3-ANCA positive patients (101 patients, 297 periods), the corresponding relapse rate was 15.6% (3.8 to 36.7) vs. 14.5% (9.7 to 19.8) (Figure 1). When restricted to patients with a doubling of the ANCA titer, the relapse rate was 18.0% (4.4 to 41.2) in 18 patient-periods vs. 12.7 (8.7 to 17.1) in 339 patient-periods (Figure 2). Similarly, the RMST in the adjusted analysis was 342.4 (307.8 to 359.9) vs. 343.0 (334.3 to 350.8) days in the two groups for a difference of -0.6 (-35.8 to 18.9) days. The difference in RMST was -1.8 (-39.0 to 19.8) and -9.6 (-53.8 to 13.9) days in patients with PR3-ANCA positivity and a doubling of the ANCA titer, respectively, compared to those without. In the secondary analysis, the adjusted risk of major relapse was 10.6% (1.2 to 29.4) vs. 8.1% (5.2 to 11.7), with a difference in RMST of -4.3 days (-36.3 to 13.5) in the Increase vs. No increase group. Conclusion: Despite using the same ELISA test consistently over time in a homogeneous population of GPA patients followed at a reference centre, the predictive value of an increase in ANCA, a doubling of ANCA or an increase in PR3-ANCA alone is not sufficient to predict relapse at one year to consider escalating treatment on an individual basis. REFERENCES: NIL. Acknowledgements: NIL. Disclosure of Interests: Xavier Puéchal GSK, investigator in academic studies of ANCA-associated vasculitis for which rituximab was provided by Roche Pharma., Michele Iudici: None declared, Elodie Perrodeau: None declared, Claire Goulvestre: None declared, Pascal Cohen investigator in academic studies of ANCA-associated vasculitis for which rituximab was provided by Roche Pharma, Alexis Régent investigator in academic studies of ANCA-associated vasculitis for which rituximab was provided by Roche Pharma., Luc Mouthon investigator in academic studies of ANCA-associated vasculitis for which rituximab was provided by Roche Pharma., Loïc Guillevin investigator in academic studies of ANCA-associated vasculitis for which rituximab was provided by Roche Pharma., Raphaël Porcher: None declared, Benjamin Terrier investigator in academic studies of ANCA-associated vasculitis for which rituximab was provided by Roche Pharma.