Background/Purpose: Evaluation of lupus flares is inconsistent using glossary-based definitions for grading (1). BILAG 2004 rates flares by new/worse organ scores. The SELENA-SLEDAI flare index (SFI) combines treatment change, new/worse features, and a weighted physician’s global assessment (SFI-PGA) visual analogue scale (VAS). A modified SFI (mSFI) improves accuracy by excluding medication criteria while clinicians differentiate mild from moderate flare (2). The Rapid Evaluation of Activity in Lupus (LFA-REAL^TM) expands the SFI-PGA and other earlier systems (3) for comprehensive VAS-based evaluation of individual symptoms, organs, or global disease. We compared mSFI and BILAG flares during a clinical trial, using SFI-PGA and LFA-REAL^TM to clarify their differences.

Methods: Disease activity (SLEDAI, BILAG 2004, mSFI, SFI-PGA, and LFA-REAL^TM) was rated monthly in a clinical trial. Severity of mSFI and BILAG flares was compared to SFI-PGA and LFA-REAL^TM changes by receiver operating characteristic (ROC) curve analysis.

Results: 50 SLE patients were examined at 430 monthly visit pairs. The mSFI defined 108 flares (73 mild, 34 moderate, 1 severe); BILAG only 77 (63 mild, 8 moderate, 6 severe) (Table 1). The mSFI rated 36 mild BILAG flares as mild and 26 as moderate. These were effectively discriminated by SFI-PGA or LFA-REAL^TM (p<0.001). There were 34 moderate mSFI flares vs only 8 by BILAG (new/worse B in ≥2 organ domains). Also, mSFI flare was scored in 32 of 353 visit pairs with no BILAG flare (27 with mild flare in <3 BILAG domains). Both SFI-PGA and LFA-REAL^TM captured flares by mSFI or BILAG (ROC analysis, p<0.0001 for all). Mild vs moderate/severe mSFI flares were distinguished by SFI-PGA or LFA-REAL^TM. Only LFA-REAL^TM reflected mild vs moderate/severe flares by BILAG (Table 2).

Table 1. Flares by mSFI and BILAG 2004.

Table 2. ROC analysis of changes in SFI-PGA and LFA-REAL^TM in differentiating mild from moderate/severe flares by mSFI and BILAG 2004.

Conclusion: BILAG and mSFI flares are captured by changes in SFI-PGA or LFA-REAL^TM. BILAG flare thresholds are higher (mild flares require increase in 3 organs; moderate flares in a single organ default to mild). SFI compresses these into “mild-moderate”, which hinders clinically significant distinctions. The mSFI clinician’s discrimination of mild vs moderate/severe flare is supported by SFI-PGA or LFA-REAL^TM changes. LFA-REAL^TM can also distinguish mild vs moderate/severe BILAG flares, likely due to its expanded multisystem scale.

References: 1. Isenberg D, et al. Arthritis Care Res. 2018;70:98-103, 2. Thanou A, et al. Rheumatology 2014;53:2175-2181, 3. Petri M, et al. J Rheumatol.1992;19:53-9.

« Back to 2018 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/systemic-lupus-flares-based-on-bilag-and-sledai-rules-are-inconsistent-but-may-be-better-understood-using-visual-analogue-scales/