Ruibing Li,1,* Jinyang Wang,1,2,* Jianan Wang,1,* Wei Xie,3 Pengfei Song,4 Jie Zhang,4 Yun Xu,5 Decai Tian,5 Lei Wu,3,* Chengbin Wang1,* 

1Department of Laboratory Medicine, the First Medical Centre of Chinese PLA General Hospital, Beijing, 100853, People’s Republic of China; 2School of Laboratory Medicine, Weifang Medical College, Weifang, Shandong, 261053, People’s Republic of China; 3Department of Neurology, the First Medical Centre of Chinese PLA General Hospital, Beijing, 100853, People’s Republic of China; 4School of Advanced Technology, Xi’an Jiaotong - Liverpool University, Suzhou, 215000, People’s Republic of China; 5Center for Neurology, Beijing Tiantan Hospital, Capital Medical University, Beijing, 100050, People’s Republic of China

*These authors contributed equally to this work

Correspondence: Lei Wu; Chengbin Wang, The First Medical Centre of Chinese PLA General Hospital, Beijing, 100853, People’s Republic of China, Email wlyingsh@163.com; wangcb301@126.com

Background: Neuromyelitis optica spectrum disorder (NMOSD) is a group of immune-mediated disorders that often lead to severe disability. The diagnosis and monitoring of NMOSD can be challenging, particularly in seronegative cases, highlighting the need for reliable biomarkers to enhance clinical management. This study aimed to identify serum lipid biomarkers for the diagnosis and monitoring of NMOSD and to assess their potential to improve clinical decision-making.
Methods: We conducted a comprehensive serum proteomic analysis in a discovery cohort of NMOSD patients and controls to identify lipid-related proteins associated with NMOSD. Subsequently, we validated the candidate biomarkers in the retrospective cohort and developed diagnostic models using a random forest algorithm. The association between these lipid biomarkers and disease activity was further evaluated in longitudinal analysis.
Results: Our analysis identified a panel of serum lipid-related biomarkers that demonstrated significant differences between NMOSD patients and controls. The diagnostic models achieved the impressive accuracy of 72% for the full NMOSD spectrum, 72% for AQP4-IgG+ NMOSD, and 68% for double seronegative NMOSD. Importantly, these biomarkers showed a correlation with disease activity, with levels changing from relapse to remission. Additionally, a combination of these lipid biomarkers was found to predict relapse with the AUC of 0.861. A user-friendly smartphone application was developed to facilitate the straightforward “input-index, output-answer” screening process, enhancing both clinical decision-making and patient care.
Conclusion: The diagnostic model based on the serum lipid-related indexes (TC, TG, LDL, HDL, ApoA1, and ApoB) may be the useful tool for NMOSD in diagnosis and monitoring of disease stage, thereby improving the treatment outcome for patients. Future studies should focus on integrating these biomarkers into routine clinical practice to realize their full potential in enhancing NMOSD management.

Keywords: Neuromyelitis optica spectrum disorder, serum proteomic profiles, serum lipid-related indexes, diagnosis biomarker, disease monitoring