Research On The Nomogram Prediction Model Of Cervical Lymph Node Metastasis In Medullary Thyroid Carcinoma Based On Ultrasound And Clinical Features

Background A quantitative diagnostic nomogram model of MTC lateral lymph node metastasis based on ultrasound and clinical characteristics was investigated to accurately predict the probability of lateral lymph node metastasis and to provide a more effective reference for clinical decision-making.Methods The clinicopathological and ultrasound imaging data of 110 patients with surgically treated MTC in Zhejiang Cancer Hospital from February 2011 to December 2022 were retrospectively analyzed.The patients were divided into training sets(n=77)and testing sets(n=33)with a ratio of 7:3,and divided into lateral lymph node metastasis(LLNM)postive group and LLNM negative group according to postoperative pathologic results of the MTC lymph node metastasis.Detailed records of patient’s gender,age,calcitonin(Ctn),carcinoembryonic antigen(CEA),platelet-to-lymphocyte ratio(PLR),central lymph node metastasis(CLNM)and ultrasound features.The ultrasound features of MTC primary tumor was analyzed,including number,location,size,morphology,boundary,aspect ratio,relationship to the envelope,internal blood flow,calcification,and whether the ultrasound indicates lateral lymph node metastasis(LLNM).The independent risk factors for MTC with lateral lymph node metastasis were screened by logistic regression analysis,and a clinical ultrasound prediction model was constructed,and the ROC curve was used to compare the diagnostic efficacy and accuracy of the ultrasound imaging features and the diagnostic model.Thirty patients with postoperative pathologically confirmed MTC from January 2020-December 2022 were included in the test set for testing,internal validation of the model,and application of ROC curves to assess the actual efficacy of the predictive model.The study has been reviewed and filed by the ethics committee of our hospital.Results After preliminary screening by univariate analysis,multivariate logistic regression analysis showed that capsule involvement,US reported lateral lymph node metastasis and central lymph node metastasis were independent risk factors for lateral lymph node metastasis(all P<0.05).Ctn,as a specific serum marker for the diagnosis and evaluation of lateral lymph node metastasis in MTC,was included in the model construction together with capsule involvement,US reported lateral lymph node metastasis and central lymph node metastasis,and and a diagnostic nomogram model for lateral lymph node metastasis was constructed based on the above four risk predictors.The model showed good predictive ability in both the training and validation sets with AUCs of 0.948(95%CI=0.901-0.994)and 0.930(95%CI=0.836-1.0),respectively.After calibration curves,it was found that the calibration of the model was good.In the decision curve analysis,the nomogram model performed well in terms of clinical efficacy.Conclusions 1.capsule involvement,US reported lateral lymph node metastasis and central lymph node metastasis ware independent predictors of MTC with lateral lymph node metastasis,which indicate the surgeons should make carefully lateral lymph node metastasis dissection according to different factors.2.The quantitative diagnostic nomogram model of MTC with lateral lymph node metastasis constructed based on ultrasound features and clinical predictors has good predictive efficacy and can be used as a simple,non-invasive and effective quantitative tool to help clinicians screen suitable candidates for prophylactic lateral lymph node metastasis dissection.However,the results of this study still need to be verified by more prospective and multi-center studies.