Abstract
BACKGROUND: Systemic lupus erythematosus (SLE) patients, with or without neuropsychiatric SLE (NPSLE), exhibit greater neuronal impairment compared to healthy individuals in terms of neuronal damage, magnet resonance imaging (MRI) changes and cognitive dysfunction. Interferon (IFN)-α is a key immunopathogenic driver of SLE, being persistently overexpressed in the majority of patients. This longitudinal study aimed to investigate whether disease activity and serum IFN-α levels over time were associated with objective findings of neuronal impairment regarding (i) higher plasma neurofilament light (NfL) concentrations, (ii) structural alterations on MRI, and (iii) cognitive dysfunction upon testing. METHODS: Sixty-six consecutive female SLE outpatients were enrolled in a cross-sectional study. Retrospectively, prior visits with concomitant blood samples (n = 199) were selected from the Lund Lupus Cohort database and biobank. Serum IFN-α concentrations were measured using an electrochemiluminescence immunoassay. IFN-α lupus phenotypes were defined as high (n = 24) or low (n = 33) by considering persistent elevations in serum IFN-α concentrations despite low SLE Disease Activity Index-2000 (SLEDAI-2 K) scores. SLEDAI-2 K lupus phenotypes were defined as moderate-high (n = 31) or low (n = 35) based on SLEDAI-2 K scores from all 576 available visits prior to the study. Ongoing neuronal damage was assessed by plasma NfL concentration measurements using Simoa at the 199 visits. Structural MRI alterations and cognitive dysfunction according to the CNS-Vital Signs test battery were the additional outcomes. Multivariate linear mixed-effect, linear regression, and logistic regression models were used for the statistical analyses. RESULTS: Visits with higher disease activity were associated with higher plasma NfL concentrations (e.g. SLEDAI-2 K total: p = 1.5*10(- 6)). High compared with low IFN-α lupus phenotype patients displayed more cognitive dysfunction (odds ratio 11.0, p = 0.004), and smaller volumes of total grey matter, caudate nucleus, and thalamus (p = 0.036; p = 0.038; p = 0.023). Moderate-high compared with low SLEDAI-2 K lupus phenotype patients displayed larger white matter lesion volumes and smaller total grey matter and thalamus volumes (p = 0.011; p = 0.041; p = 0.005). CONCLUSIONS: The study suggests that disease activity and IFN-α may drive neuronal affliction in SLE, also in the absence of overt neuropsychiatric symptoms, and that controlling disease activity could improve the cerebral outcome.