Background/Purpose: Neuropsychiatric manifestations of systemic lupus erythematosus (SLE) are common and frequently represent a diagnostic challenge due to its heterogeneous manifestations. Neurofilament light chain (NfL) is a marker of axonal neuronal damage that has been investigated in neurological conditions such as multiple sclerosis, stroke, and dementia. This study aims to evaluate the serum levels of NfL in SLE patients and their association with neuropsychiatric manifestations. Secondary objectives included assessing the association of NfL levels with cognitive dysfunction (CD), and cumulative damage over a longitudinal follow-up.

Methods: A longitudinal study was conducted including 66 patients that met the SLICC/ACR 2019 criteria for SLE. NPSLE was defined according to the SLICC A and B models. Demographic and clinical data including laboratory tests, SLICC/ACR Damage Index (SDI), treatment, and presence of comorbidities were collected. Baseline serum NfL levels were measured using the Single Molecule Array (SiMoA) technique. Cognitive function was assessed through a comprehensive neuropsychological battery. After a mean follow-up of 7.46 ± 1.34 years, 48 patients were re-evaluated for disease activity, damage, and cognitive status. The best cutoff of NfL levels for the detection of NPSLE and CD was evaluated using ROC curve analysis. Univariate and multivariate analysis was used to identify risk factors associated with NPSLE and CD.

Results: The mean age of the 66 SLE patients was 38.10 ± 10.55 years. According to SLICC A model 21 patients had NPSLE and 45 had not; according to SLICC B model 33 patients had NPSLE and 33 had not. Eighteen (27,3%) patients had moderate to severe CD. Patients with NPSLE according to SLICC A had significantly higher NfL levels than non-NPSLE patients (12.90 ± 6.32 vs. 8.08 ± 3.66 pg/mL; p = 0.001). Patients with focal central nervous system (CNS) involvement exhibited higher NfL levels compared to those with diffuse CNS involvement and non-NPSLE patients (Figure 1). NfL showed good accuracy in identifying NPSLE (SLICC A) (AUC = 0.766; 95% CI: 0.631–0.900; p = 0.001) (Table 1). By multivariate analysis, NPSLE per SLICC A and SLICC B were significantly associated with NfL levels >10.54 pg/mL (OR = 22.94, 95% CI: 2.91-180.45; OR = 11.28, 95% CI: 2.52–50.48, respectively) and treatment with immunoglobulin (OR = 10.46, 95% CI: 2.95-37.09; OR = 11.64, 95% CI: 2.42-56.11, respectively). Moreover, NPSLE (SLICC A) was significantly associated with lower age (OR = 0.90, 95% CI 0.82-0.99). CD was independently associated with NfL >10.54 pg/mL (OR = 5.17; 95% CI: 1.06-25,10), diabetes mellitus (OR = 4.80; 95% CI: 1.10-20.99) and lower education level (OR=0.77; IC 95% 0.59-0.99) (Table 2). At follow-up, patients remained cognitively stable and baseline NfL levels were not predictive of worsening of CD and cumulative damage over time. The presence of anti-RNP antibodies was independently associated with mild CD at follow-up (OR = 4.46, 95% CI 1.09–18.30, p = 0.038).

Conclusion: Serum NfL is a promising biomarker for identifying NPSLE and CD in patients with SLE. Although baseline NfL did not predict cognitive decline over time, its association with NPSLE and CD supports its role in disease stratification.

Figure 1. Evaluation of NfL levels in SLE patients with and without neuropsychiatric involvement according to SLICC A and SLICC B model, according to the type of neuropsychiatric manifestation and moderate-to-severe-cognitive dysfunction

Table 1. Accuracy of NfL for the identification of NPSLE according to the SLICC A and SLICC B classification models and for the detection of moderate to severe cognitive dysfunction and sensitivity and

specificity for a cutoff value of NfL >10.54

pg/mL

Table 2. Predictors of NPSLE and Cognitive Dysfunction in SLE Patients

ACR Meeting Abstracts - https://acrabstracts.org/abstract/neurofilament-light-chain-serum-levels-is-associated-with-neuropsychiatric-manifestations-and-cognitive-dysfunction-in-systemic-lupus-erythematosus-patients-a-longitudinal-study/