CSF and blood glial fibrillary acidic protein for the diagnosis of Alzheimer's disease: A systematic review and meta-analysis

Yutong Zou,Yifei Wang,Xiaoli Ma,Danni Mu,Jian Zhong,Chaochao Ma,Chenhui Mao,Songlin Yu,Jing Gao,Ling Qiu
DOI: https://doi.org/10.1016/j.arr.2024.102485
Abstract:Recently included in the 2024 new revised diagnostic criteria of Alzheimer's disease (AD), glial fibrillary acidic protein (GFAP) has garnered significant attention. A systematic review and meta-analysis were performed to comprehensively evaluate the diagnostic, differential diagnostic, and prospective diagnostic performance of GFAP in cerebrospinal fluid (CSF) and blood for AD continuum. A literature search using common electronic databases, important websites and historical search way was performed from inception to the beginning of March 2023. The inclusion criteria was studies evaluating the diagnostic accuracy of GFAP in CSF and/or blood for the AD continuum patients, utilizing PET scans, CSF biomarkers and/or clinical criteria. The systematic review and meta-analysis were conducted referring to the Cochrane Handbook. In total, 34 articles were eventually included in the meta-analysis, 29 of which were published within the past three years. Blood GFAP exhibited good diagnostic accuracy across various AD continuum patients, and the summary area under curve for distinguishing PET positive and negative individuals, CSF biomarkers defined positive and negative individuals, clinically diagnosed AD and cognitive unimpaired controls, AD and/or mild cognitive impairment and other neurological diseases, and prospective cases and controls was 0.85[0.81-0.88], 0.77[0.73-0.81], 0.92[0.90-0.94], 0.80[0.77-0.84], and 0.79[0.75-0.82], respectively. Only several studies were recognized to evaluate the diagnostic accuracy of CSF GFAP, which was not as good as that of blood GFAP (paired mixed data: AUC = 0.86 vs. AUC = 0.77), but its accuracy remarkably increased to AUC = 0.91 when combined with other factors like sex, age, and ApoE genotype. In summary, GFAP, particularly in blood, shown good diagnostic, differential diagnostic, and prospective diagnostic accuracy for AD continuum patients, with improved accuracy when used alongside other basic indexes.
What problem does this paper attempt to address?