Background
Follicular neoplasia of the thyroid (Bethesda IV) is the most common indication for thyroid surgery. As histological verification is the only 100% certain method of confirming malignancy, surgery remains the gold standard for FN treatment. Genetic testing of FN is quite expensive, and it is unavailable in many countries. Meanwhile, its accuracy is still under question due to the limitations of validation and post validation cohorts.
According to various studies, the risk of malignancy of FN is 25-40%. Most cases of FN malignancy are variants of papillary thyroid cancer that require active surveillance alone in nodules sized less than 2 cm.
Materials
A continuous cohort of 4399 patients who underwent surgery at SPbU Hospital in the years 2020-2021 with Bethesda IV cytology obtained by FNAB with MG staining. 3654 patients were female (83,1%) with a mean age of 49,40±15,11. The mean age for male patients was 49,10±14,62.
Results
In (29,5%) patients malignant thyroid nodules were identified by final histological examination. However, in 406 of those patients (31,3%) malignancy was found in non-punctured minor nodules.
Of 980 cases of FN found to be malignant nearly half (46%, N=446) were NIFTP or minimally invasive follicular thyroid cancer and only in 185 (8,8%) cases invasion of at least the nodules’ own capsule was reported. 39 cases involved an invasion of the gland’s capsule (4%). 29 (3,6%) cases had vascular invasion or were suspicious for vascular invasion.
3066 patients had nodules of 2 cm or less. Of those patients 578 had malignant nodules (18,9%). In 262 patients (45,3%) the nodules were identified as NIFTP or minimally invasive follicular thyroid cancer. However, the risk of finding cancer with aggressive features in nodules of 2 cm or less was only 0,9%, as compared to 6,4% in nodules larger than 2 cm. OR=7,4 The risk of finding aggressive cancer in any nodule sized 2 cm or less was 0,16%, while that in larger nodules was 1,5%. Most of the malignant nodules were found to be variants of thyroid cancer meeting the criteria for active surveillance.
We performed ROC analysis to establish cut off size of FN and the risk of missing an aggressive thyroid cancer (AUC=0.84).
Cut off nodule diameter of 3 cm showed a sensitivity of 0.8 (0.64 - 0.92), and a specificity of 0.79 (0.78 - 0.81) for the risk of aggressive cancer. For nodule diameter of 2 cm sensitivity was 0.84 (0.68 - 0.96), while specificity was 0.6 (0.59 - 0.62). NPV was 0.9984 (0.9971 - 0.9994) for the 3 cm diameter and 0.9983 (0.9966 - 0.9996) for the 2 cm diameter.
Conclusion
It seems that patients with nodules of ≤2 cm can safely avoid surgery until the nodules grow in size even if no genetic test is performed. This may also be true for nodules of ≤3 cm, should the surveillance criteria for papillary thyroid cancer be expanded.