From January 2010 to June 2012,
6118 focal thyroid nodules underwent US-guided FNA.
A total of 411 of 6118 (6.7%) focal thyroid nodules were diagnosed as AUS/FLUS,
and 246 were excluded because of a lack of further evaluation,
such as surgery,
rFNA,
or follow-up US (n=240) or because further FNA showed AUS/FLUS again without additional histological confirmation (n=6).
A total of 162 patients (146 women and 16 men) with 165 thyroid nodules were analyzed in the present study.
The mean age of patients was 49 years (range,
22 – 76 years).
The mean lesion size was 13 mm in longest diameter (range,
3 – 52 mm).
There were nine histological subcategories of AUS/FLUS nodules (1): microfollicles in a sparsely cellular aspirate with scant colloid (group 1,
n=2),
Hurthle cells in a sparsely cellular aspirate with scant collioid (group 2,
n=3),
mild follicular cell atypia with drying/ clotting artifact (group 3,
n=17),
Hurthle cells in a cellular aspirate (group 4,
n=3),
focal features of papillary carcinoma in an otherwise predominantly benign-appearing (group 5,
n=71),
cyst-lining cells with nuclear atypia in an otherwise predominantly benign-appearing (group 6,
n=0),
follicular cells with regenerative atypia (group 7,
n=3),
atypical lymphoid infiltrate (group 8,
n=1),
not otherwise categorized (group 9,
n=65).
We confirmed diagnosis by surgery (n = 100),
rFNA (n = 42),
rFNA with surgery (n = 23).Malignant nodules were confirmed by surgery (n=88) or rFNA(n=3).
Nodules were deemed benign if confirmed as benign using surgery (n=35) or rFNA (n=39).
The interval between initial FNA and rFNA were three to six months in thyroid nodules that suspicious malignant US features and six months in that indeterminate US features (2).
The χ2test was used to compare the risk of malignancy according to US features and BRAF analysis.
Logistic regression analysis was performed to assess the odds ratio for the risk of malignancy according to specific US features (shape,
margin,
echogenecity,
calcification).
Relative 95% confidence intervals (CIs) were also calculated.
The analysis was performed using the STATA software (version 9.0,
StataCorp).
In all analyses,
P < 0.05 was taken to indicate statistical significance.