Background/Purpose: Factors associated with cumulative glucocorticoid exposure in granulomatosis with polyangiitis (GPA) or microscopic polyangiitis (MPA) have not previously been described. We examined the association between baseline characteristics and cumulative glucocorticoid dose in the Rituximab in ANCA-Associated Vasculitis (RAVE) trial comparing rituximab (RTX) with cyclophosphamide (CYC) followed by azathioprine for severe ANCA-associated vasculitis.

Methods: All patients received at least one 1000 mg dose of intravenous methylprednisolone prior to the initiation of prednisone at 1 mg/kg. Prednisone was reduced to 40 mg/day by 1 month and tapered off by month 6 but could be restarted or increased in response to persistent or increased disease activity. For this analysis, cumulative glucocorticoid doses were calculated monthly during induction (months 0-6) and maintenance (months 7-18). Prednisone and intravenous methylprednisolone were considered separately and when combined. The associations between baseline characteristics and cumulative dose were examined using generalized estimating equations to estimate mean differences and 95% confidence intervals (CI) over time after adjusting for potential confounders. Logistic regression was used to estimate odds ratios for achievement of glucocorticoid-free maintenance, defined as the absence of any glucocorticoid received from month 6 until study discontinuation.

Results: Mean ± SD cumulative prednisone dose was 3458 ± 806 mg at month 6 and 3771 ± 1044 mg at month 18. 89 patients (45%) achieved glucocorticoid-free maintenance. During the maintenance phase, relapsing disease (adjusted mean difference +324 mg over 12 months versus new diagnosis, 95% CI 56 to 592) and overweight or obese body mass index (BMI) (+397 mg, 95% CI 137 to 659) were associated with increased cumulative prednisone dose. In multivariate logistic regression, anti-proteinase 3 (PR3) antibody positivity (OR 0.50, 95% CI 0.26 to 0.97) was associated with decreased odds for achievement of glucocorticoid-free maintenance. Treatment with RTX was associated with decreased cumulative prednisone dose in comparison with CYC during induction (adjusted mean difference -251 mg over 6 months, 95% CI -411 to -91) but not during maintenance (+107 mg over 12 months, 95% CI -189 to +402).

Conclusion: In this clinical trial of GPA or MPA with protocol-defined glucocorticoid tapering, relapsing disease and increased BMI were associated with increased glucocorticoid exposure following induction of remission. PR3 positivity was associated with decreased likelihood of achieving glucocorticoid-free remission. Induction treatment with RTX was associated with a minimal mean decrease in prednisone exposure during induction only. These findings may allow for the prospective identification of patients with GPA or MPA at increased risk for chronic glucocorticoid exposure and toxicity.