最佳化Myositis Line Blot Immunoassay的臨床應用：整合檢測前臨床評估與更高抗體效價切點的臺灣單一醫學中心世代研究

Optimizing the Clinical Utility of a Myositis Line Blot Immunoassay: An Integrated Approach Combining Pre-test Clinical Assessment and Higher Titer Cut-offs in a Taiwanese Tertiary Center Cohort

目的：本研究旨在評估一商業化的肌炎特異性（MSA）及肌炎相關性（MAA）自體抗體line blot immunoassay（LIA）在臺灣單一醫學中心的診斷效能。探討病人於檢測前的臨床表現與特發性發炎性肌病（IIM）診斷之間的關聯性，並評估使用更高之抗體效價切點對於診斷效能的影響。
方法：本回溯性單一中心研究納入了486位於2024年間在國立臺灣大學醫學院附設醫院接受MSA/MAA LIA檢測的病人。依據臨床診斷，病人被分為IIM組、IIM以外之自體免疫疾病組，以及對照組。我們採用多變項邏輯迴歸分析，以評估檢測的適應症，即病人檢驗前之臨床症狀和症狀數量與IIM診斷之間的關聯性。針對個別與統合的MSA之診斷效能，則透過比較分析製造商建議的抗體陽性切點與更高的切點（抗體效價≥2+）進行評估。
結果：在486位病人中，78位（16.0%）診斷為IIM，223位（45.9%）診斷為其他自體免疫疾病，185位（38.1%）為對照組。診斷為IIM的可能性隨著病人所表現出的臨床症狀數量增加而上升，與自體免疫篩檢組相比，校正後勝算比為1個臨床症狀= 8.20；2個臨床症狀= 11.75；≥3個臨床症狀= 20.06，p值皆<0.05）。採用較高的抗體效價切點（≥2+）增加了統合後的特異度（從0.826提升至0.931）與陽性預測值（從0.441提升至0.632），並伴隨著敏感度的下降。
結論：針對MSA/MAA LIA的應用，本研究結果支持應基於臨床懷疑來謹慎選擇病人進行檢驗，並對抗體效價進行半定量判讀，因為較高的抗體效價切點增加了特定的肌炎特異性自體抗體與IIM診斷的關聯性。

Objectives: To evaluate the real-world diagnostic performance of a commercial myositis-specific (MSA) and myositis-associated (MAA) autoantibody line blot immunoassay (LIA) in a large Taiwanese cohort. We aimed to assess the association between the number of pre-test clinical domains prompting testing and the likelihood of an idiopathic inflammatory myopathies (IIM) diagnosis, and to determine the impact of applying higher titer cut-offs on the assay's diagnostic performance.
Methods: This retrospective, single-center study included 486 patients who underwent MSA/MAA LIA testing at National Taiwan University Hospital in 2024. Patients were categorized as IIM, autoimmune diseases other than IIM, or controls based on clinical diagnosis. We performed a multivariable logistic regression analysis to assess the association between the number of testing indications (clinical domains) and an IIM diagnosis. The diagnostic performance of individual and pooled MSAs was evaluated using both the manufacturer's standard cut-off and a higher cut-off (titer≥2+).
Results: Of the 486 patients, 78 (16.0%) were diagnosed with IIM, 223 (45.9%) had other autoimmune diseases, and 185 (38.1%) were controls. After adjusting for age and sex, an increasing number of clinical domains was associated with a progressively higher likelihood of an IIM diagnosis (adjusted OR: 1 domain = 8.20; 2 domains = 11.75;≥3 domains = 20.06, all p < 0.05) compared to an autoimmune screening reference group. Applying a higher titer cut-off (titer≥2+) was associated with an increase in pooled specificity (from 0.826 to 0.931) and positive predictive value (from 0.441 to 0.632), alongside a decrease in sensitivity.
Conclusions: For optimal use of the MSA/MAA LIA in our cohort, these results support careful patient selection guided by clinical suspicion and a semi-quantitative interpretation of titers, as higher titer cut-offs strengthened the diagnostic association for myositis specific autoantibodies.