Correlation Analysis Between Ultrasound Features Of Multifocal Papillary Thyroid Microcarcinoma(mPTMC)and Lymph Node Metastasis(LNM)

Objectives:To analyze the clinical characteristics、ultrasound characteristics and ACR scores of mPTMC,screen out independent risk factors for LNM in mPTMC.Construct a logistic regression equation to predict the risk of LNM,and establish a risk prediction system with a scoring method which was hoped to evaluate the aggressiveness of mPTMC and provide a basis for its treatment decision.Materials and methods:A total of 450 patients（1183 lesions in total）who underwent surgical treatment for thyroid cancer in our hospital from January 2019 to December 2019,were pathologically diagnosed as mPTMC after surgery,and who underwent ultrasound examination in the ultrasound department of our hospital before surgery),of which348 cases were in the modeling group（including 901 lesions）,and 102 cases were in the verification group（including 282 lesions）.Single-factor Logistic regression analysis was performed on the clinical characteristics、ultrasound characteristics and ACR scores of each case in the modeling group,and meaningful（P<0.05）factors were screened out as independent variables,and multivariate binary logistic regression analysis was performed to obtain independent risk factors for LNM in mPTMC.The inclusion criterion was P<0.05,and the exclusion criterion was P>0.05.According to the standard regression coefficient（βcoefficient,B’）obtained by multivariate Logistic regression analysis,assign independent risk factors to construct a Logistic scoring system for predicting LNM risk.The data of the patients in the verification group were substituted into the scoring system,and the postoperative pathology was used as the gold standard to evaluate the predictive power of the model.Result:The results of binary logistic regression multivariate analysis showed that:X¹male（P=0.001）,X² age≤45 years old（P=0.007）,X⁵ single-case lesion maximum diameter>0.5cm（P<0.000）,X⁷ envelope Invasion（P<0.000）,X⁸ A single lesion with the maximum ACR score of>9 points（P<0.000）andX⁹A single case with multiple lesions ACR score of more than 19 points（P<0.035）are the independent risk factors for LNM in mPTMC.We Obtain the logistic regression equation for predicting the risk of LNM:Y=-2.485+1.315X₁-0.948X_2（2）-0.709X_2（3）+1.556X₅+1.232X₇+0.664X₈+1.644X₉。The score range of the Logistic scoring system is-4 to 17points.Using the Logistic scoring system to predict the risk of LNM in mPTMC,when the score is 4,the sensitivity is 81%,the specificity is 67%,and the area under the curve（AUC）is 0.83.Using the model to predict the risk of LNM in the verification group,the accuracy was 67.6%,the sensitivity was 61.7%,the specificity was 72.7%,the positive predictive value was 65.9%,the negative predictive value was 69.0%,the missed diagnosis rate was 38.3%,and the misdiagnosis rate was27.3.%.Conclusions:1.Male,≤45 years old,maximum diameter≥0.5cm,capsule invasion,the sum of ACR scores>19 and the highest ACR score>9 are independent risk factors for LNM in mPTMC.2.Based on these risk factors,lymph node metastasis in mPTMC patients was established.According to the quantitative model of the model,it is recommended to choose a more thorough treatment for mPTMC patients with high scores.