Bahattin Erdogan, Emine Dündar, Çivi Çetin Kısmet, Fatih Yılmaz
Department of Pathology, Eskisehir City Hospital, Eskisehir, Turkey.
Department of Pathology, Eskisehir Osmangazi University Faculty of Medicine, Eskisehir, Turkey.
J Cytol. 2024 Oct-Dec;41(4):221-228. doi: 10.4103/joc.joc_26_24. Epub 2024 Nov 8.
The Bethesda System for Reporting Thyroid Cytology (TBSRTC) recommended for the interpretation of needle aspiration cytology of the thyroid, is the most widely used worldwide. Studies have shown that the disagreement between observers, especially in the Bethesda III and IV diagnostic categories, is not insignificant at 10%-40%. In the TBSRTC 2023 version, some definitions were removed and simplified, and molecular pathology was proposed as a complement to cytopathology. The current availability of molecular tests is limited because they can be performed in a few centers and are expensive. Therefore, our study investigated intra- and inter-observer agreement according to TBSRTC 2023 using only immunohistochemically BRAFV600E antibodies.
The study included 173 cases with aspiration cytology evaluated between 2019 and 2022. The immunohistochemical procedure applied BRAFV600E (RM8) monoclonal antibody to cell block sections. All slides were assessed and categorized by three different observers. Data were interpreted using Cohen's kappa and Fleiss's kappa test in the Statistical Package for Social Sciences Windows 2021 program.
For the applied RM8 antibody, sensitivity was 64.71% and specificity was 87.27%. In terms of diagnostic categories, inter-observer agreement was good for Bethesda II ( = 0.606) and moderate for Bethesda III ( = 0.429), Bethesda IV ( = 0.523), Bethesda V ( = 0,464), and Bethesda VI ( = 0.544), respectively.
In conclusion, the study reveals that the 2023 version of TBSRTC provides improvement, especially in the categories of uncertain diagnosis, but is still insufficient to improve cytological diagnostic accuracy, at which point molecular analyses become even more important.
用于甲状腺细针穿刺细胞学检查结果判读的《贝塞斯达甲状腺细胞病理学报告系统》(TBSRTC)是全球使用最广泛的系统。研究表明,观察者之间的分歧,尤其是在贝塞斯达III类和IV类诊断类别中,差异显著,为10%-40%。在TBSRTC 2023版中,一些定义被删除并简化,同时提出分子病理学作为细胞病理学的补充。目前分子检测的可用性有限,因为它们只能在少数几个中心进行,而且费用昂贵。因此,我们的研究仅使用免疫组化BRAFV600E抗体,根据TBSRTC 2023调查了观察者内部和观察者之间的一致性。
本研究纳入了2019年至2022年间进行细针穿刺细胞学检查评估的173例病例。免疫组化程序将BRAFV600E(RM8)单克隆抗体应用于细胞块切片。所有玻片由三名不同的观察者进行评估和分类。数据在社会科学统计软件包Windows 2021程序中使用Cohen's kappa和Fleiss's kappa检验进行解释。
对于应用的RM8抗体,敏感性为64.71%,特异性为87.27%。在诊断类别方面,观察者间一致性在贝塞斯达II类中良好(κ = 0.606),在贝塞斯达III类(κ = 0.429)、IV类(κ = 0.523)、V类(κ = 0.464)和VI类(κ = 0.544)中为中等。
总之,该研究表明,TBSRTC 2023版有改进,尤其是在不确定诊断类别中,但仍不足以提高细胞学诊断准确性,此时分子分析变得更加重要。