DEVELOPING AND EVALUATING A LABORATORY-BASED FRAILTY INDEX (FI-LAB) FOR THE PREDICTION OF LONG-TERM HEALTH OUTCOMES IN SYSTEMIC LUPUS ERYTHEMATOSUS

O013 / #551 Topic: AS23 - SLE-Diagnosis, Manifestations, & Outcomes ABSTRACT CONCURRENT SESSION 02: SLE METRICS – IMPROVING OUTCOMES & MEASURES 22-05-2025 1:40 PM - 2:40 PM Background/Purpose Frailty is a useful measure of health status in systemic lupus erythematosus (SLE), but it is not routinely captured in existing SLE datasets. In other populations, frailty indices constructed exclusively from laboratory data have been shown to predict adverse health outcomes. We aimed to construct and evaluate the first laboratory-based frailty index (FI-Lab) for people living with SLE. Additionally, we compared the FI-Lab to an existing clinical frailty index with respect to the prediction of future health outcomes. Methods This study used existing data from a single-center prospective cohort of adult SLE patients followed annually with standardized clinical and laboratory assessments. We included the first study visit for each patient occurring between 2010 and 2019, with follow-up data available until June 2024. All participants met the 1997 revised American College of Rheumatology (ACR) classification criteria for SLE. A 30-item FI-Lab was constructed by adapting the list of laboratory variables previously identified by Ellis et al.[1] Baseline FI-Lab scores were calculated for each patient. Using clinical data, a baseline Systemic Lupus International Collaborating Clinics Frailty Index (SLICC-FI) score was calculated for each patient. Organ damage accrual was defined as the change in SLICC/ACR Damage Index (SDI) score from the baseline visit to the last follow-up visit. Mortality was defined as any recorded death within the follow-up period. Cox proportional hazards regression was used to examine the association between baseline FI-Lab scores and all-cause mortality risk, while negative binomial regression was used to evaluate the association of baseline FI-Lab scores with organ damage accrual during follow-up. To compare the performance of models containing the baseline FI-Lab and/or SLICC-FI as predictor variables, we used Akaike information criterion (AIC), Harrell’s C-statistic, and pseudo-R 2 values. Results The 283 included patients (89% female) had a mean (SD) age of 47.7 (15.1) years and a median (IQR) disease duration of 8.3 (2.6-19.8) years at baseline. The 97 patients (34.3%) classified as frail at baseline (based on FI-Lab scores > 0.21) had increased mortality risk [hazard ratio 3.71; 95% CI 1.82-7.54] compared to nonfrail patients (Figure 1). Baseline frailty was also associated with a higher rate of organ damage accrual during follow-up [incidence rate ratio 2.26; 95% CI 1.59-3.22]. A weak correlation existed between baseline FI-Lab and SLICC-FI scores (r s =0.37, p<0.001). In unadjusted analysis, higher baseline FI-Lab and SLICC-FI scores were both associated with increased mortality risk during follow-up. However, after multivariable adjustment, only the FI-Lab maintained a significant association with mortality risk (Table 1). Both the FI-Lab and the SLICC-FI were significant baseline predictors of organ damage accrual during follow-up, and the multivariable model that included both frailty measures was superior to the models containing either the FI-Lab or the SLICC-FI alone (Table 1). Figure 1. Kaplan-Meier survival curves for mortality risk during follow-up among SLE patients who were classified as frail at baseline (in red) versus non-frail SLE patients (in blue) based on laboratory-based frailty index (FI-lab) scores. Table 1. Association of baseline FI-Lab and SLICC-FI scores with mortality risk and organ damage accrual during follow-up (n=274). Conclusions An FI constructed from routinely collected laboratory variables can measure frailty and predict future health outcomes in SLE. The FI-Lab may serve as a convenient screening tool to detect subclinical deficit accumulation and promote early risk mitigation among SLE patients. References: [1.] Ellis HL. CMAJ 2020;192(1):E3-8.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.