Core Needle Biopsy Specificity in Suspicious Thyroid Nodules
Core Needle Biopsy Specificity in Suspicious Thyroid Nodules
A total of 52 consecutive patients with preoperative FNA diagnoses, including follicular neoplasm (46%), suspicion of malignancy (48%) or malignancy (6%), that warranted surgical treatment were operated on in Tampere University Hospital (TAUH) between May 2010 and December 2011. This was a prospective study. The patients were operated on according to the management guidelines for thyroid nodules. The extent of the thyroid surgeries ranged from lobectomy (n=29) or total thyroidectomy (n=20) to surgical biopsy of the nodule (n=3). All patients were operated on in TAUH by two experienced surgeons (TH, RS). One patient had a benign preoperative FNA result but was included in the study because clinical evidence suggested it was a malignant tumour.
After resection of the specimen, the suspected thyroid nodule was biopsied under palpation control in the operating room with two punctures each of a 23G FNA needle and a 20G CNB needle (Tru-Core II, Medical Device Technologies Inc, Gainesville, Florida, USA). The FNA samples were fixed in 50% acetone/alcohol, centrifuged and stained with Papanicolaou stain. Two slides were prepared from each sample. The CNB samples were fixed in 10% buffered formalin, routinely processed and embedded in paraffin. Then, 4 μm-thick sections were cut and stained with H&E. A minimum of three sections per sample were analysed. The FNA and CNB samples were analysed blindly without knowledge of either the preoperative FNA or the final histopathological diagnosis. Two cytological and one or two histological research slides in addition to the routine FNA and histology slides were evaluated per patient. Cell blocks from remnant material were not performed in this study. Routinely, cell blocks are beneficial when additional immunohistochemical tests are crucial for diagnosis (ie, in suspected medullary carcinoma and parathyroid lesions). In a large study of VanderLaan and coworkers, cell blocks were performed only in 76.5%. Furthermore, cell blocks are often insufficient (23.5%). Therefore, the cell blocks were not performed in this study.
The statistical analyses were conducted with SPSS V.14.0 software (SPSS Inc., Chicago, Illinois, USA). The Confidence Interval Analysis (University of Southampton, UK) program was used to calculate the sensitivity, specificity and predictive values with their respective CIs (Wilson method). Continuous variables are presented as the mean with the SD. Nominal variables are presented as the number of patients (%).
Materials and Methods
Patients and Tumours
A total of 52 consecutive patients with preoperative FNA diagnoses, including follicular neoplasm (46%), suspicion of malignancy (48%) or malignancy (6%), that warranted surgical treatment were operated on in Tampere University Hospital (TAUH) between May 2010 and December 2011. This was a prospective study. The patients were operated on according to the management guidelines for thyroid nodules. The extent of the thyroid surgeries ranged from lobectomy (n=29) or total thyroidectomy (n=20) to surgical biopsy of the nodule (n=3). All patients were operated on in TAUH by two experienced surgeons (TH, RS). One patient had a benign preoperative FNA result but was included in the study because clinical evidence suggested it was a malignant tumour.
Sample Acquisition and Preparation
After resection of the specimen, the suspected thyroid nodule was biopsied under palpation control in the operating room with two punctures each of a 23G FNA needle and a 20G CNB needle (Tru-Core II, Medical Device Technologies Inc, Gainesville, Florida, USA). The FNA samples were fixed in 50% acetone/alcohol, centrifuged and stained with Papanicolaou stain. Two slides were prepared from each sample. The CNB samples were fixed in 10% buffered formalin, routinely processed and embedded in paraffin. Then, 4 μm-thick sections were cut and stained with H&E. A minimum of three sections per sample were analysed. The FNA and CNB samples were analysed blindly without knowledge of either the preoperative FNA or the final histopathological diagnosis. Two cytological and one or two histological research slides in addition to the routine FNA and histology slides were evaluated per patient. Cell blocks from remnant material were not performed in this study. Routinely, cell blocks are beneficial when additional immunohistochemical tests are crucial for diagnosis (ie, in suspected medullary carcinoma and parathyroid lesions). In a large study of VanderLaan and coworkers, cell blocks were performed only in 76.5%. Furthermore, cell blocks are often insufficient (23.5%). Therefore, the cell blocks were not performed in this study.
Statistical Analysis
The statistical analyses were conducted with SPSS V.14.0 software (SPSS Inc., Chicago, Illinois, USA). The Confidence Interval Analysis (University of Southampton, UK) program was used to calculate the sensitivity, specificity and predictive values with their respective CIs (Wilson method). Continuous variables are presented as the mean with the SD. Nominal variables are presented as the number of patients (%).
