Background Belimumab is used to treat several systemic lupus erythematosus (SLE) manifestations and predictors of response are advisable in clinical practice.

Objectives To explore the effects of belimumab treatment on skin involvement in SLE patients in a real-life setting.

Methods SLE patients treated with belimumab (10 mg/kg day 0, 14, 28 and then every 28 days), as add-on therapy in 11 Italian cohorts were prospectively followed-up for 24 months. Skin involvement was measured by the CLASIa score (Cutaneous Cutaneous LE Disease Area and Severity Index Activity) at baseline and every six months until month 24. Damage was measured by CLASI-Damage (CLASId). Response was defined as CLASIa <2 at measured timepoints.

The following variables were tested to determine baseline predictors of response at 12, 18 and 24 months: gender, age, disease duration, SLE activity index 2000 (SLEDAI-2K) ≥10, prednisone dose >7.5 mg/day, baseline CLASIa and CLASId, concomitant immunesuppressants (yes/no) and number of previous immunesuppressants. Statistical analysis was performed with SPSS 22.0 software.

Results 188 patients were studied, among whom 48 (25.5%) had skin involvement as the leading feature for belimumab therapy. Mean follow-up period was 17.5±10.96 months. Thirty-eight patients completed a 6-month follow-up; 27 completed a 12-month follow-up; 19 completed a 18-month follow-up and 15 completed a 24-month follow-up. Fourteen patients discontinued due to adverse events (7/14, 50%), lack of efficacy (4/14, 28.5%) or other causes (3/14, 21.4%).

CLASIa, daily prednisone dosage and SLEDAI-2K significantly decreased during the follow-up (p<0.001). CLASIa<2 was achieved by 25/38 patients (65.7%) at 6 months, 19/27 (70%) at 12 months, 11/19 (58%) at 18 months and 10/15 (67%) at 24 months. A lower baseline CLASIa was associated with CLASIa response at 18 and 24 months (responders vs. non-responders: 3.4±2.4 vs. 9.5±5.6; 3.5±2.5 vs. 11.2±6.5; p<0.005 for both), while a lower baseline CLASId was associated with CLASIa response at each timepoint (responders vs. non-responders: at 12 months: 0.6±0.9 vs. 2.4±2.5; at 18 months: 0.5±0.8 vs. 2.5±2.3; at 24 months: 0.6±0.8 vs. 3.25±2.7; p=0.01 for all). Multivariate analysis was only performed at 12 months due to low patient number at 24 months. CLASId was the only independent negative predictor of response (OR 0.52, p=0.05). Notably, CLASId remained stable during the follow-up.

Conclusions Belimumab use in cutaneous involvement is associated with reduced activity in skin lesions and hindrance of skin damage. Early use of belimumab before damage is established is likely to be associated with a better response.

Acknowledgements dr. Margherita Zen, dr. Maddalena Larosa, dr. Francesca Ometto, dr. Mara Felicetti for helping in data collection and statistical analysis

Disclosure of Interest None declared