单发甲状腺乳头状癌同侧中央组淋巴结转移CT征象分析

Multivariate analysis of CT signs of ipsilateral central lymph node metastasis in single papillary thyroid carcinoma

目的探讨单发PTC同侧中央组淋巴结转移(ipsilateral central lymph node metastasis,ICLNM)的各种CT征象,通过多因素回归分析明确其诊断价值。方法回顾分析经手术及病理证实的302例直径〉1.0 cm单发PTC的CT资料,通过受试者操作特性曲线(receiver operating characteristic,ROC)获得淋巴结转移直径的最佳阈值,并采用多因素回归分析淋巴结大小、强化程度、钙化或囊变、中央区浑浊和侧颈部淋巴结阳性与ICLNM阳性的关系’结果302例PTC中,病理证实ICLNM阳性和阴性比例分别为63.6%(192/302)和36.4%(110/302)。根据ROC曲线,随着淋巴结直径的增大,诊断淋巴结转移的敏感度降低而特异度增髙,当阈值为0.4 cm时,约登指数最大(0.358),此时敏感度和特异度分别为50.5%和80.3%。多因素分析显示直径≥0.4cm、高强化、中央区浑浊和侧颈部淋巴结阳性是1CLNM阳性的独立风险因子,OR值分别为4.189(95% CI 2.037~8.617),3.875 (95% CI 1.561~9.617),4.054(95% CI 2.230~7.371)和8.735 (95% CI 1.093~69.831);钙化或囊变在ICLNM中差异无统计学意义。结论直径≥0.4cm、高强化冲央区浑浊和侧颈部淋巴结阳性是PTC时ICLNM阳性的独立风险因子,钙化或囊变虽不是独立风险因子,对ICLNM阳性却具有高度准确度,准确识别这些征象,对于外科医生采取更彻底的手术治疗方案及降低术后复发具有重要意义。

Objective To investigate the diagnostic value of CT signs of ipsilateral central lymph node metastasis (ICLNM) in single papillary thyroid carcinoma (PTC) by multivariate regression analysis. Methods The CT data of 302 single PTC with diameter >1.0 cm confirmed by operation and pathology were retrospectively analyzed. The optimal thresholds of lymph node metastasis diameter were obtained by receiver operating characteristic (ROC) curve analysis. And multivariate regression analysis was used to analyze the relation between lymph node size, degree of enhancement, calcification or cystic degeneration, central turbidity, positive lateral cervical lymph nodes and the ICLNM positivity. Results In 302 PTC, the proportion of ICLNM positive and negative was 63.6%(192/302) and 36.4%(110/302), respectively. According to the ROC curve, with the increase of lymph node diameter, the sensitivity of diagnosing lymph node metastasis decreased and the specificity increased. When the threshold was 0.4 cm, Youden index was the largest (0.358), and the sensitivity and specificity was 50.5% and 80.3%, respectively. Multivariate analysis showed that the diameter ≥0.4 cm, high enhancement, central turbidity and lateral cervical lymph nodes positivity were the independent risk factors of ICLNM, and the OR values were 4.189(95% CI (2.037-8.617)], 3.875 [(95% Cl (1.561-9.617)], 4.054[(95%6/(2.230-7.371)] and 8.735 [(95% CI (1.093-69.831)], respectively. Calcification or cystic degeneration was not statistically significant in ICLNM. Conclusions The diameter ≥0.4 cm, high enhancement, central turbidity and lateral cervical lymph nodes positivity are the independent risk factors of ICLNM. Although calcification or cystic degeneration is not the independent risk factor, it has high accuracy for ICLNM positivity. The accurate identification of these signs can help surgeons to take a more thorough surgical treatment and has great significance to reduce postoperative recurrence.