Objective: To evaluate the consistency of measurement results and diagnostic agreement of core cerebrospinal fluid (CSF) biomarkers for Alzheimer disease (AD) in a Chinese population, which were measured across a domestically developed electrochemiluminescence platform (Lifotronic eCL8000), a Japanese chemiluminescence immunoassay platform (LUMIPULSE), and an American electrochemiluminescence platform (Meso Scale Discovery, MSD). Methods: CSF samples were retrospectively collected from 253 subjects recruited at Peking Union Medical College Hospital, Xuanwu Hospital Capital Medical University, and Beijing Tiantan Hospital between July 2013 and September 2019. The cohort was comprised of 132 males and 121 females, with the age of (62.4±13.3) years. All CSF samples were collected via standardized protocols and cryopreserved at -80 ℃. Inter-assay repeatability (coefficient of variation, CV) of the Lifotronic platform on different days was validated. Spearman's rank correlation analysis was used to evaluate the degree of correlation between the results of the Pumen platform and those of the LUMIPULSE and MSD platforms, with the correlation coefficient expressed as ρ. Subsequently, 4 core CSF biomarkers [β-amyloid 40 (Aβ40), β-amyloid 42 (Aβ42), phosphorylated tau (p-tau), and total tau (t-tau)] were quantitatively measured across the three platforms. The consistency of test results between platforms was evaluated using the Passing-Bablok regression model and Bland-Altman analysis. The 95% limits of agreement (LoA) were calculated by Bland-Altman analysis, with good agreement indicated when <5% of points fell outside the limits. Proportional and constant errors between the measurement systems were quantified by the Passing-Bablok regression model, utilizing a robust non-parametric method. 18F-florbetapir positron emission tomography/positron emission tomography/computed tomography (PET/CT) imaging has been performed on 40 of the 253 subjects. With the PET/CT diagnostic results utilized as the gold standard, the diagnostic agreement between the CSF results from the 3 platforms and the PET/CT findings was compared by calculating Cohen's Kappa coefficient. Furthermore, AD diagnostic reference cut-off values for each platform were determined by employing receiver operating characteristic (ROC) curves. Results: The Lifotronic platform demonstrated good repeatability in CSF testing (CV<10%). Spearman's rank correlation analysis was used to evaluate the correlation between the Lifotronic platform and the LUMIPULSE and MSD platforms, with correlation coefficients denoted by ρ. In comparison with the LUMIPULSE platform, all correlation coefficients (ρ) were>0.85. The correlation coefficient (ρ) between the MSD and Lifotronic platforms was>0.80, with a Kappa value of 0.698 (P<0.05). Bland-Altman analysis and Passing-Bablok regression models showed that for t-tau, the mean bias between Lifotronic and LUMIPULSE was 214.95 (95%LoA:-1 490.59-1 920.50), with 99.60% (252/253) of the data points falling within the 95%LoA. In the detection of t-tau, the regression slope between the Lifotronic and LUMIPULSE platforms was 2.041 (Equation: Y=2.041X-75.184), and the regression slope with the MSD platform was 2.360 (Equation: Y=2.360X-87.503). Based on PET/CT imaging validation, the CSF diagnostic results of the Lifotronic platform were consistent with the PET/CT results (Kappa=0.885, P<0.05), as were those of the MSD platform (Kappa=0.714, P<0.05). The Lifotronic platform established AD diagnostic reference thresholds based on the t-tau/Aβ42 and p-tau-181/Aβ42 ratios (the cut-off value for t-tau/Aβ42 was 0.32, and for p-tau-181/Aβ42 was 0.035). The Kappa value between the Lifotronic and LUMIPULSE platforms was 0.928 (P<0.05), and the Kappa value between the Lifotronic and MSD platforms was 0.698 (P<0.05). Conclusion: This study systematically validated the reliability of domestically developed Lifotronic electrochemiluminescence platform and its corresponding assay kits in detecting AD CSF biomarkers, as well as their comparability with international platforms, demonstrating comparable capabilities in terms of accuracy, consistency, and cost control. 目的： 比较我国自主研发的电化学发光平台普门（Lifotronic eCL8000）、日本LUMIPULSE化学发光免疫测定平台以及美国电化学发光平台Meso Scale Discovery（MSD）在检测我国人群脑脊液（CSF）阿尔茨海默病（AD）核心生物标志物检测结果和诊断的一致性。 方法： 回顾性收集2013年7月至2019年9月期间于北京协和医院、首都医科大学宣武医院和北京天坛医院招募的253例受试者的CSF样本，男132例，女121例，年龄（62.4±13.3）岁，所有CSF样本均标准化采集并于-80 ℃冻存。对普门平台进行非同日重复测量一致性（CV）验证；采用Spearman秩相关分析评估普门平台与LUMIPULSE、MSD平台间结果的相关程度，相关系数以ρ表示；随后分别在3个平台上进行CSF样本的4种核心生物标志物［β-淀粉样蛋白40（Aβ40）、β-淀粉样蛋白42（Aβ42）、磷酸化tau（p-tau）和总tau（t-tau）］定量检测；采用Passing-Bablok回归模型与Bland-Altman分析［运用Bland-Altman分析两种测量方法一致性，计算95%一致性界限（LoA），超出该界限的点数<5%，认为具有较好的一致性；Passing-Bablok回归模型利用稳健的非参数方法量化测量系统间的比例与恒定误差］。253例受试者中40例已接受18F-氟贝他苯（18F Florbetapir）正电子发射断层显像/X线计算机体层成像仪（PET/CT）的受试者，以PET/CT诊断结果作为金标准，比较3大平台的CSF生物标志物诊断结果与PET/CT结果的一致性，计算Kappa系数，并采用受试者工作特征（ROC）曲线计算各平台的AD判别cut-off值。 结果： 普门平台在CSF检测中表现出良好的重复性（CV<10%）；与LUMIPULSE平台比较，各个指标相关系数ρ均>0.85。MSD平台与普门平台相关系数ρ>0.80，Kappa值为0.698（P<0.05）。Bland-Altman分析与Passing-Bablok回归模型：t-tau在普门与LUMIPULSE比较时的平均偏差为214.95（95%LoA：-1 490.59~1 920.50），99.60%（252/253）的点位于95%LoA内；在检测t-tau时，普门平台与LUMIPULSE平台回归斜率为2.041（方程：Y=2.041X-75.184），MSD平台回归斜率为2.360（方程：Y=2.360X-87.503）。基于PET/CT影像验证，普门平台的CSF诊断结果与PET/CT结果Kappa值为0.885（P<0.05），MSD平台与PET/CT结果的Kappa值为0.714（P<0.05）。普门平台t-tau/Aβ42 cut-off值为0.32，p-tau-181/Aβ42 cut-off值为0.035；普门与LUMIPULSE平台之间Kappa值为0.928（P<0.05），普门与MSD平台Kappa值为0.698（P<0.05）。 结论： 该研究系统验证了我国自主研发的普门电化学发光平台及其配套试剂盒在AD CSF生物标志物检测中的可靠性与国际平台的可比性，显示出在准确性、一致性和成本控制等方面具有一定的对比能力。.
使用 AI 将内容摘要翻译为中文，便于快速阅读
使用 AI 分析这篇文章的核心发现、关键要点和深度见解
由 DeepSeek AI 提供分析 · 首次使用需配置 API Key
PubMed · 2026-07-07
PubMed · 2026-07-02
PubMed · 2026-07-02