A Comparative Study of 141 Glial Fibrillary Acidic Protein Immunoglobulin G Positive Cases
Shifeng Zhang, Huilu Li, Peihao Lin, Ding Liu, Zhanhang Wang, Jie Yang, S. N. Ruan, Yinan Zhao, Zhike Lan, Xiao Yang, Y. Lynn Wang, Yong You, Xiuling Wu, Haiyang Wang, Hongli Liu, Huan Yang, Huiyu Feng, Lu Zhang, Houshi Zhou, Qianhui Xu, Teng-Fei Ou, Yuhua Lu, Cong Gao, Wei Qiu, Junwei Hao, Youming Long
Abstract
BACKGROUND: Glial fibrillary acidic protein-immunoglobulin G (GFAP-IgG) positivity is associated with autoimmune GFAP astrocytopathy (GFAP-A), but also with other autoimmune encephalitides and viral infections. We attempted to elucidate the characteristics of GFAP-A in relation to other GFAP-IgG-positive encephalitides and constructed a differential diagnosis model. METHODS: 141 GFAP-IgG-positive cases were identified, including 52 astrocytopathy (GFAP-A group), 48 autoimmune encephalitis (AE-G), and 41 viral encephalitis (VE-G). Multivariate logistic regression was employed to create a diagnostic model, with validation using an external cohort. RESULT: Compared to the AE-G group, the GFAP-A patients showed more onset age ≥ 50 years, headache, fever, consciousness disturbance, MRI radial vascular enhancement, cerebrospinal fluid (CSF) antibody titer grade ≥ 4, and CSF proteins ≥ 700 mg/L, but less female sex, limb numbness, visual disturbances, and CSF chloride ≤ 120 mmol/L. Among these, CSF antibody titer grade ≥ 4, CSF protein ≥ 700 mg/L, and absence of visual disturbances were independent risk factors for GFAP-A diagnosis. Compared to the VE-G group, the GFAP-A patients showed more course ≥ 14 days, onset age ≥ 50 years, limb weakness, serum potassium ≤ 3.9 mmol/L, CSF antibody titer grade ≥ 4, CSF leukocytes ≤ 46*10, MRI radial vascular enhancement, MRI involvement of brainstem, and MRI involvement of spinal cord, but less headache, fever, nausea, and vomiting. Among these, serum potassium ≤ 3.9 mmol/L, MRI spinal cord involvement, and absence of nausea and vomiting were independent risk factors for GFAP-A diagnosis. CONCLUSIONS: Based on critical clinical indicators identified, we constructed a differential diagnosis model for GFAP-A.