Abstract
Cyclophosphamide (CYC), together with rituximab, is the mainstay for induction therapy in ANCA-associated vasculitis (AAV) [1]. The risk factors influencing the likelihood of relapse in individuals undergoing CYC induction are still unclear. This study aims to score the more robust predictors of relapse in these patients, thereby building a clinically useful tool.
Patients' characteristics were pooled from three cohorts from Italy, Ireland, and Spain. Eligibility criteria were a diagnosis of GPA, MPA, or EGPA, age >18 years, induction treatment with CYC [either as intravenous therapy (IV) and/or oral], and at least 12 months of follow-up. Baseline assessment included clinical data, laboratory testing, and assessment of manifestations of GPA, MPA, and EGPA in each organ system according to the Birmingham Vasculitis Activity Score (BVAS). Relapse was defined as the presence of ≥1 new vasculitis manifestation after a remission lasting at least 3 months; remission was defined as no disease activity (BVAS=0) irrespective of the glucocorticoid use. Time-to-remission and relapse-free survival probabilities were estimated using the Kaplan-Meier method. A multivariate Cox analysis, including the statistically significant factors at a previous univariate analysis, was employed to develop a practical prognostic score, defined as “Relapse Evaluation And Cyclophosphamide Treatment (REACT) score”, in which each independent predictor was assigned a number of weighted points proportional to its β regression coefficient.
The score development cohort comprised 505 patients. 223 patients (44.2%) had MPA, 217 (43.0%) GPA, and 65 (12.9%) EGPA. 183 patients (36.2%) experienced a relapse, with a median time to relapse of 33 months (IQR: 13-61). A multivariate Cox regression analysis proved that PR3-ANCA (HR: 1, 30, 95% CI 1, 01-1, 87), IV CYC (HR: 1.78, 95% CI 1.31-2.41), cardiovascular involvement (HR: 1.82, 95% CI 1.01-3.25), arthralgias/arthritis (HR: 1.46 95% CI 1.08-1.98) and the absence of rapidly progressive glomerulonephritis (HR: 1.37, 95% CI 1.02-1.85) were all independent risk factors of relapse. These variables were used to build the REACT score, with each item receiving 1 point except for CV involvement and IV CYC, each receiving 2 points. We then identified three risk categories according to the score of each patient (Fig. 1A). We identified 138 (27%) patients with low risk of relapse (score 0-1), 252 (49%) with an intermediate risk (score 2-3), and 115 patients (22%) with a high risk (score 4-7). These three groups showed different relapse probabilities on Kaplan-Meier analysis (pairwise comparisons between the groups were all statistically significant) (Fig. 1B).
The REACT score can be employed at diagnosis to predict the risk of relapse in patients with AAV treated with cyclophosphamide induction. Its value needs to be confirmed in external cohorts.
Comments