Correlation between radiological and cytological preoperative risk stratification in thyroid nodules : a retrospective single center analysis

Rani Jaspers, Marian Vanhoeij, Guy Verfaillie, Ine Luyten, Steven Raeymaeckers, Corina Andreescu and Daniel Jacobs-Tulleneers-Thevissen CORRELATION BETWEEN RADIOLOGICAL AND CYTOLOGICAL PREOPERATIVE RISK STRATIFICATION IN THYROID NODULES: A RETROSPECTIVE SINGLE CENTER ANALYSIS PRESENTER: Rani Jaspers ABSTRACT. Objective Preoperative risk stratification for malignancy in thyroid nodules avoids undertreatment and unnecessary surgery. The effectiveness of different preoperative risk stratification classifications based on ultrasonography (TIRADS classification) and cytology (Bethesda classification) have been previously reported. However, there are little data on how both classifications correlate. This study focuses on the correlation between TIRADS and Bethesda classification for preoperative risk stratification of thyroid nodules in a single center. Methods Data of 241 adults, diagnosed during a 1-year period (1st of April 2017 to 31st of March 2018), with a total of 276 thyroid nodules that were classified both according to TIRADS and Bethesda, were retrospectively reviewed. Correlation between radiological and cytological classification was analyzed based on their relative malignancy risk. Results Overall, a significant correlation between both classifications was found (R = 0.362, p<0.001). Categories with the lowest risk of malignancy (being < 5 %; TIRADS 2-3-4a versus Bethesda II; n=134) correlated in 96% of cases. Lesions that were classified by ultrasound in the higher malignancy risk category TIRADS 4b (n=69; 9.2 % malignancy risk) or TIRADS 4c-5 (n=42; 44.4 - 87.5 % malignancy risk) correlated with their respective cytological categories (Bethesda III and IV-V) in 14% and 21% of cases, probably (at least partially) related to an ultrasonographic risk overestimation in these nodules. When looking only at lesions classified by a dedicated radiologist (performing 62% of ultrasounds) this correlation increased to 19% and 27% for TIRADS 4b and TIRADS 4c-5 nodules respectively. Conclusions TIRADS classification is an appropriate risk stratification system for thyroid nodules, especially in benign nodules and when the ultrasound is performed by an experienced radiologist. However, since the good overall correlation is not absolute, both risk stratification systems are complementary in the evaluation of thyroid nodules, mainly for nodules categorized as suspicious for malignancy on radiological evaluation.