PO.4.77 Agreement between lldas and expert assessment in identifying SLE patients with LDA: study on a real-world cohort of caucasian patients

Abstract

Purpose

Both lupus low disease activity state (LLDAS) and remission have been proven to be good and achievable targets in the management of SLE. Nevertheless, considerable overlap between LLDAS and remission exists: an average of 80% of patients in LLDAS also meet the definition of remission in different cohorts worldwide, raising the question whether LLDAS definition is too close to definition of remission. Our aim was to evaluate the performance of LLDAS in identifying patients in LDA, defined according to gold standard, which is physician judgement.

Methods

We prospectively collected data of SLE patients attending our outpatient clinic from October 2021 to January 2022. Each patient received a complete clinical evaluation and review of recent laboratory tests by a rheumatologist expert in SLE, who classified patients in the following states: remission, LDA, active disease. Each category was mutually exclusive. The definitions of LLDAS and remission were also applied. LLDAS was defined, according to Franklyn et al., as SLEDAI-2k≤4 without major organ activity (including renal, cardiac and fever), no new disease activity, PGA≤1 (0–3), stable immunosuppressive therapy and prednisone equivalent dose up to 7.5 mg/day. Remission was defined according to the DORIS definition as clinical SLEDAI-2k=0 and PGA <0.5 in patients treated with standard immunosuppressive therapy and a prednisone equivalent dose ≤5 mg/die. In addition, patients fulfilling the definition of LLDAS but not that of remission (LLDAS/no remission) were identified. Cohen’s kappa coefficient was used to assess the agreement between expert definition of LDA and LLDAS.

Results

During the follow-up we enrolled 207 patients with SLE (mean±SD age 46±12.9 years, mean±SD disease duration 9±6 years, 84.5% female). Among them, 154 (74.4%) were in LLDAS, of which 29 (14%) were in LLDAS/no remission, meaning an overlap between LLDAS and remission consisting of 125 (81.2%) patients. According to expert opinion, LDA was observed in 45 (21.7%) and remission in 128 (61.8%) patients. The agreement between expert opinion and LLDAS in discriminating active patients from LDA+remission was overall good (Cohen’s k 0.67). However, definition of LLDAS failed to discriminate patient in LDA from patients in remission as identified by the experts (Cohen’s k -0,02). We also analyzed the agreement after removing patients in remission from the pool of LLDAS (LLDAS/no remission): the agreement between expert definition of LDA and LLDAS/no remission markedly improved (Cohen’s k 0,68). Notably, 9 out of 16 patients were in LDA according to the expert opinion but not to LLDAS due to minimal renal and serosal involvement.

Abstract PO.4.77 Table 1

Conclusions

Our analysis shows that LLDAS is effective in discriminating patients with active diseases from those in LDA/remission. However, the great majority of patients in LLDAS are in remission and some patients with LDA are not identified by LLDAS. Thus, LLDAS should be implemented to capture all patients in LDA and to discriminate patients in LDA from those in remission.