Background/Purpose: ANCA-associated vasculitis (AAV) treatment includes high dose glucocorticoids (GCs), which are associated with increased body-mass index (BMI), a complication abhorred by patients and associated with increased risk of metabolic and cardiovascular complications. A prior study found increased BMI to be independent of GC exposure for AAV (Arthritis Care Res 2008;59:746). The ability to classify BMI change as an adverse event related to GC exposure or a positive outcome reflecting improved disease activity is critical in studies investigating steroid-sparing strategies. We evaluated this further by investigating the complex relationship between BMI, GC exposure, and disease activity.

Methods: We analyzed patients enrolled in the Rituximab in ANCA-Associated Vasculitis (RAVE) trial. First, we replicated the methods used by the prior study by performing a multivariate linear regression to investigate the relationship between GC exposure (prednisone equivalent) and BMI changes while adjusting for age, sex, baseline BMI and creatinine, GC exposure prior to baseline, disease status (new diagnosis versus flaring), ANCA subtype (PR3- or MPO-ANCA+), randomization arm, and flare during the study. Second, we used mixed effects regression models to determine the time-varying relationship between the cumulative moving average of disease activity and BMI as well as cumulative GC exposure during induction treatment (the first 6 months of the trial) when accounting for confounders. Third, we used the same methods to assess the relationship between BMI and quality of life (QOL), as measured by SF36 scores.

Results: In RAVE (N=197), 99 (50%) patients were male, the mean age was 52.8 (±15.5) years, and the mean baseline BVAS/WG was 8.0 (±3.1). The majority were PR3-ANCA+ (67%). The baseline BMI was 28.8 (±6.3) and the largest BMI change occurred during induction (+1.1 (±2.2) kg/m², P<0.0001). During RAVE, the mean cumulative GC dose was 5,038mg (±2,638), most administered in the first 6 months. In fully adjusted multivariate linear regression models, there was no association between GC exposure and change in BMI at 6, 12, and 18 months (P>0.3 for all analyses). However, in fully adjusted mixed-effects models, both cumulative GC exposure and moving averages of BVAS/WG were associated with BMI change in opposing directions [+0.2kg/m² per 1gm GC (±0.04) and -0.1 per BVAS/WG point (±0.02), both P<0.001]; randomization to rituximab was associated with increased BMI [+0.7 (±0.2) P<0.0001]. Increasing BMI was associated with worse QOL in every SF 36 domain (P<0.03 for all analyses).

Conclusion: During AAV induction, increased BMI was associated with increased GC exposure but also improved disease control and randomization to rituximab. These observations raises important questions regarding whether increased BMI should be classified as an adverse event related to GC therapy as BMI increases may actually indicate effective disease control whereas decreased BMI may augur poorly for long-term disease control. The observed association between rituximab and increased BMI deserves additional evaluation.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/disease-activity-glucocorticoid-exposure-and-rituximab-determine-body-composition-changes-during-induction-treatment-of-anca-associated-vasculitis/