Background. The classification of nodules by Thyroid Imaging Reporting and Data Systems (TIRADS) is important in guiding management. Whether sensitivity in identifying thyroid cancers varies with thyroid cancer phenotype remains unclarified. Methods. The ultrasound (US) characteristics of nodules of 26,908 nodular goiter patients were recorded. Fine-needle aspiration cytology (FNA) was performed in all nodules >1 cm irrespective of US findings (n = 25,025) and in nodules between 5 mm and 10 mm with suspicious US characteristics (n = 1,883). Of the 3281 operated cases, 221, 30, and 23 were papillary (PTC), follicular (FTC), and medullary (MTC) cancers, respectively. The US-based indication of FNA, as defined by EU-TIRADS scores, combined with lesion size, was calculated. This study design is unique in avoiding the common selection bias when TIRADS’ sensitivity is tested in a cohort selected for FNA and surgery based on the same US characteristics on which TIRADS is based. Results. The EU-TIRADS score influences decision of FNA in the 10–20 mm range. In such nodules (n = 118), the number of suspicious features (marked hypoechogenicity, microcalcifications, irregular shape, and irregular border) per lesion was lower in FTC (0.7 ± 0.6) than in PTC (1.7 ± 1.0) or MTC (1.8 ± 0.7; ), resulting in EU-TIRADS scores of 4.1 ± 0.6, 4.8 ± 0.3, and 4.9 ± 0.2, respectively (). The EU-TIRADS-based FNA indication rate was lower in FTC (55.5%) compared to PTC (85.0%) and MTC (88.9%) (). Conclusions. EU-TIRADS-defined suspicious US features are less common in FTC than in PTC and MTC. Therefore, a substantial number of FTCs in the 10–20 mm range escape surgery.

中文翻译：

---

基于 EU-TIRADS 的甲状腺结节细针抽吸和细胞学检查忽略了大量滤泡性甲状腺癌

背景。甲状腺成像报告和数据系统 (TIRADS) 对结节的分类对于指导管理很重要。识别甲状腺癌的敏感性是否因甲状腺癌表型而异仍不清楚。方法。记录了 26,908 名结节性甲状腺肿患者的结节超声 (US) 特征。对所有 > 1 cm 的结节进行细针抽吸细胞学检查 (FNA)，无论超声检查结果如何 ( n  = 25,025) 以及具有可疑超声特征的 5 mm 至 10 mm 结节 ( n = 1,883)。在 3281 例手术病例中，221 例、30 例和 23 例分别为乳头状 (PTC)、滤泡状 (FTC) 和髓质 (MTC) 癌。计算了由 EU-TIRADS 评分定义的基于美国的 FNA 适应症，并结合病变大小。当 TIRADS 的敏感性在基于 TIRADS 所基于的相同美国特征选择的 FNA 和手术队列中进行测试时，该研究设计在避免常见选择偏倚方面是独一无二的。结果。EU-TIRADS 分数影响 FNA 在 10-20 mm 范围内的决定。在此类结节 ( n  = 118) 中，FTC (0.7 ± 0.6) 中每个病灶的可疑特征（显着低回声、微钙化、不规则形状和不规则边界）的数量低于 PTC (1.7 ± 1.0) 或 MTC (1.8) ± 0.7；)，导致 EU-TIRADS 得分分别为 4.1 ± 0.6、4.8 ± 0.3 和 4.9 ± 0.2 (）。与 PTC (85.0%) 和 MTC (88.9%) 相比，FTC (55.5%) 中基于 EU-TIRADS 的 FNA 指示率较低 (）。 结论。EU-TIRADS 定义的可疑美国特征在 FTC 中比在 PTC 和 MTC 中少见。因此，10-20 毫米范围内的大量 FTC 逃脱了手术。