细胞学未定性的甲状腺结节良恶性鉴别的预估模型建立

A Predictor Model to Select Malignant Thyroid Lesions in Patients with Indeterminate FNA Cytology

目的通过评估术前甲状腺结节细针穿刺细胞学检查结果未定性患者的临床病理资料、甲状腺影像学报告及数据系统(TI-RADS)诊断结果、超声弹性成像(UE)及细胞学检查结果,并对其进行半定量分析,建立一个实用的预估模型用以鉴别良恶性从而指导进一步的治疗。方法回顾性分析167例术前甲状腺结节细针穿刺细胞学检查结果为未定性患者,收集其临床病理资料、超声影像学资料、术前细胞学穿刺结果和术后病理结果,分析得出其相关因素并予不同评分。结果多因素分析发现TI-RADS评分、超声弹性评分及术前细针穿刺细胞学结果在良性结节组与恶性结节组间差异有统计学意义(P=0.000),根据比值比(OR值)赋分并建立模型,发现以评分≥4分为分界值,其敏感度为79.4%,特异性为85.9%。结论该模型对术前甲状腺结节细针穿刺细胞学检查结果未定性的甲状腺结节良恶性鉴别具有一定的预测价值,当评分≥4时需考虑结节恶性可能。

Objective To build and assess the diagnostic accuracy of a helpful preoperative scoring system using the clinical differ-ences, conventional US findings, sonographic elastography （ SE） and cytopathology results in predicting thyroid malignancy with indeter-minate cytology.Methods From January 2012 to March 2014,167 patients with indeterminate cytology diagnosed by ultrasound-guided fine-needle aspiration biopsy were retrospectively analyzed.The clinical differences, US and cytopathology data were collected to deter-mine the relevant factors and different scores.Results There were significant differences（P〈0.05） among FNA cytopathology, thyroid imaging reporting and data system （ TI-RADS） categorization, sonographic elastography （ SE） between the malignant group and the be-nign group.The index point≥4 was the delimitation value,with which the nodules where characterized by a 79.37% sensitivity and a 85.85%specificity.Conclusion Our scoring system for prediction of thyroid malignancy can be another choice for patients with indeter-minate US-FNA cytological results.When the score is equal or larger than 4,the possibility of malignancy should be considered.