Endocrine Abstracts

Background

Thyroid nodules with indeterminate cytology represent a therapeutic challenge. The probability of malignancy had been established between 5 and 15%, but it varies among the published data. Nowadays we have molecular tests to help in therapeutic decisions, but they are not available at all centers leaving uncertainty regarding appropriate management.

Objectives

– To know the percentage of malignancy in thyroid nodules with Bethesda III (AUS/FLUS) cytology in our center.

– To establish whether there is an association between the ultrasound characteristics of the B. III nodules and malignancy.

Materials and methods

WECe selected Bethesda III nodules from 1034 FNAs performed in our center between May 2012 – October 2016. Age, sex, ultrasound suspicious characteristics (central vascularization, microcalcifications, taller than wide shape, irregular margins and hypoechogenicity), TIRADS score and definitive histology were registered. The association between malignancy and the cited variables was analyzed.

Results

6.96% (102/1034) of cytologies were B.III. 46.2% (47) of these nodules B.III had surgery, while 20.5% (21) are waiting for surgery and 33% (34) are in clinical ultrasound follow-up. Results in operated patients (n = 47): 76.6% were women; Age 54 [10–75] years. 48.9% (23) had 1 single B.III cytology and in the remaining 51.1% (24) the FNA was repeated (5 B.III, 7 unsatisfactory, 6 benign, 3 suspicious of malignancy and 3 Follicular neoplasm). 42.6% (20) of the B.IIIs nodules operated were thyroid carcinomas: 75% (15) papillary carcinoma, 5% (1) Hurthle cell carcinoma, 10% (2) minimally invasive follicular carcinoma, 5% (1) medullary carcinoma and 5% (1) were anaplastic carcinoma. Within TIRADS 2 nodules, the percentage of malignancy detected was 0%, while in TIRADS 4a was 28.6% and it rises to 75% in TIRADS ≥4b (P = 0.008). The only ultrasound characteristic that was associated with malignancy by itself was irregular margins (P = 0.039). The risk of malignancy increased with the addition of suspicious ultrasound characteristics, without detecting statistical differences here (P = 0.09): 3 or more = 72.7% malignancy; 2 = 50%; 1 = 27.8%; 0 = 0%.

Conclusions

The percentage of malignancy in B.III nodules in our center in these preliminary results is between 34-46.2%. Both the TIRADS and the addition of suspicious ultrasound characteristics correlate well with the presence of malignancy and constitute a helpful tool when considering whether surgery or clinical follow-up is the best option (high percentage of malignancy when there are 2 or more ultrasound data of suspicion or score ≥4b on the TIRADS scale).