Applicational Value Of 2017 ACR TI-RADS Stratification In Differential Diagnosis Of Benign And Malignant Thyroid Nodules

Objective:To investigate the clinical value of thyroid imaging reporting and data system(TI-RADS)released by the American College of Radiology(ACR)in 2017 for differentiating diagnosis of benign and malignant thyroid nodules.Method:A total of 236 patients with 255 thyroid nodules were retrospectively collected in this study at the First Affiliated Hospital of Hunan Normal University(Hunan Provincial People's Hospital)from January 2017 to August 2019,and were confirmed by ultrasound guided fine needle aspiration cytology(US-FNAC)and surgery.Each nodule was assigned and classified according to the ACR TI-RADS by two senior physicians with more than 5 years of experience in thyroid ultrasound diagnosis.When they disagree,ask another senior physician with the title of deputy director or above to participate in the diagnosis,and reach the final conclusion through joint consultation.1.The stratification of ACR TI-RADS and the total score of nodules were compared with the postoperative histopathological results to draw the receiver operating characteristic(ROC)curve,determine the optimalthreshold to differentiate benign from malignant thyroid nodules according to the Yonden index,and compare the diagnostic value of the two methods.2.According to the management recommendations of ACR TI-RADS and the size of the nodules,thyroid nodules were divided into three group(undealt,follow-up,and FNA respectively),and to analyze the postoperative pathological diagnosis results and cervical lymph node metastasis of thyroid nodule with different management methods.3.According to the Bethesda system for reporting thyroid cytopathology(TBSRTC),Bethesda III,IV and V nodules were defined as cytological indeterminate nodules,and to analyze the diagnostic value of ACR TI-RADS for cytological indeterminate nodules.Result:1.There were 71 benign and 184 malignant nodules in 255 thyroid nodules.They were classified into four categories according to the scores2-14.The malignancy rates of TR2-TR5 were 8.3%(1/12),6.3%(1/16),41.2%(21/51)and 91.5%(161/176),respectively.There was a statistically significant difference among the malignancy rates(P < 0.05).2.ROC curve showed that the accuracy,sensitivity,specificity,positive predictive value,negative predictive value and AUC of ACR TI-RADS stratification were 85.1%,87.5%,78.9%,91.5%,70.9% and0.850,respectively;the corresponding value of the nodal total score were87.1%,91.3%,76.1%,90.8%,77.1% and 0.895,respectively,with the optimal diagnostic threshold of >TR4 and 5.5 points respectively.There was a statistically significant difference of diagnostic value between ACR TI-RADS stratification and the nodal total score(Z=2.907,P < 0.01).3.TR 2 nodules belonged to the undealt group,the rates of malignancy was 8.3%(1/12).The rates of malignancy in TR 3 nodules were 0,0 and 8.3%(1/12)in the undealt group,follow-up group,and FNA group,respectively;the corresponding rates in TR 4 and 5 nodules were66.7%(8/12),41.7%(5/12)and 29.6%(8/27),and 100.0%(9/9),91.3%(63/69)and 90.8%(89/98),respectively.One node of the TR 2 nodules in the undealt group had cervical lymph node metastasis.Two nodes of the TR 4 nodules in the undealt group and one node in the follow-up group had cervical lymph node metastasis,the corresponding value of TR 5were 3 and 23 nodes,respectively.4.Among the 255 thyroid nodules,there were 142 indeterminate nodules,accounting for 55.7%(142/255)of all nodules.In indeterminate nodules,the malignancy rates of Bethesda categories III,IV and V nodules were 39.5%(15/38),80.7%(46/57)and 97.9%(46/47),respectively,and the total malignancy rate was 75.4%(107/142).The malignancy rates of Bethesda categories III in TR 2/3/4,Bethesda categories III in TR5,Bethesda categories IV in TR 2/3/4,Bethesda categories IV in TR5,Bethesda categories V in TR 2/3/4,and Bethesda categories V in TR 5were 17.4%(4/23),73.3%(11/15),35.7%(5/14),95.3%(41/43),100.0%(4/4)and 97.9%(46/47),respectively.5.For cytological indeterminate nodules,the AUC of ACR TI-RADS,TBSRTC and the combined diagnosis of thyroid nodules were0.847,0.820 and 0.913,respectively.The difference of AUC between the diagnoses of the combined method and that of ACR TI-RADS alone and TBSRTC alone were statistically significant(P < 0.05).The accuracy,sensitivity,specificity,positive predictive value and negative predictive value of ACR TI-RADS stratification were 85.9%,87.9%,80.0%,93.1%and 68.3%,respectively;the corresponding value of TBSRTC categories were 81.0%,86.0%,65.7%,88.5% and 60.5%,respectively,after combined,were 83.8%,81.3%,91.4%,82.9% and 86.5%,respectively.Conclusion:1.The stratification of ACR TI-RADS and the total score of nodules have high applicational value in diagnosing thyroid nodules,with the optimal diagnostic threshold of >TR4 and 5.5 points respectively,and the total score of nodules has higher diagnostic value in differentiating benign and malignant nodules than the stratification of ACR TI-RADS.2.If the risk stratification systems of ACR TI-RADS to guide FNA only by the size of nodules,some smaller thyroid malignant nodules with cervical lymph node metastasis will be missed.3.ACR TI-RADS combined with TBSRTC further improved the diagnostic value of thyroid nodules with indeterminate cytology.It can be used as a simple tool to better stratify thyroid nodule malignant risk and may help clinicians further diagnosis and treatment.