Diagnostic accuracy of fine needle aspiration biopsy versus postoperative histopathology for diagnosing thyroid malignancy

Abstract Introduction Ultrasound‐guided fine needle aspiration biopsy (FNAB) is currently widely used for the initial screening of patients with thyroid nodules enabling prevention of unnecessary surgery. The purpose of this study was to retrospectively analyse the diagnostic accuracy of thyroid FNAB compared with postoperative histopathology of a large cohort from Azerbaijan. Methods We evaluated the FNAB results of 738 patients who underwent thyroid surgery at the National Centre of Oncology in Azerbaijan. The measures of diagnostic accuracy were calculated for the ultrasound‐guided preoperative FNAB results (based on the six diagnostic categories of the Bethesda classification) compared with postoperative histopathologic results (benign or malignant) for correspondent areas. Results Considering both DC V and DC VI categories (387 cases) as ‘cytologic‐positive’ and DC II category (72 cases) as ‘cytologic‐negative’, we found 14 false‐positive and 10 false‐negative results. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and diagnostic accuracy were 97.4%, 86.1%, 96.4%, 81.6% and 94.8%, respectively. Conversely, when considering only the DC VI category as ‘cytologic‐positive’, the sensitivity, specificity, PPV, NPV and diagnostic accuracy of FNA were 93.2%, 100%, 100%, 81.6% and 97.1%, respectively. Conclusions The results of our cohort demonstrated high levels of diagnostic accuracy, supporting FNAB's role as a reliable diagnostic tool in the preoperative evaluation of thyroid nodules. The sensitivity, specificity, NPV, PPV and accuracy of thyroid FNAB in our institution were comparable with those of other institutions.


Conclusions:
The results of our cohort demonstrated high levels of diagnostic accuracy, supporting FNAB's role as a reliable diagnostic tool in the preoperative evaluation of thyroid nodules. The sensitivity, specificity, NPV, PPV and accuracy of thyroid FNAB in our institution were comparable with those of other institutions.

K E Y W O R D S
Bethesda classification, fine needle aspiration biopsy, thyroid nodules 1 | BACKG ROU N D Thyroid nodules are common in the general population with the   global incidence ranging between 40,000 and 71,000 per 100,000 persons. 1,2 The majority of thyroid nodules are benign and the risk for malignancy in asymptomatic nodules is 0.45%-13%. [1][2][3] Fine-needle aspiration (FNA) cytology (FNAC) is a reliable and cost-effective method for thyroid nodule evaluation. 4 FNA biopsy (FNAB) defines cases when surgery is required to be performed and decreases the overall incidence of thyroidectomy performed in benign cases. Studies have suggested a high sensitivity and specificity for thyroid malignancies prediction, ranging between 65%-98% and 72%-100%, respectively. 5 Unfortunately, FNAB may have some false-negative and false-positive outcomes. Factors implicated for this rate may include specimen adequacy, sampling technique, the skill of the physician who performed an aspiration and experience of a pathologist. 6 The purpose of this study was to evaluate the measures of diagnostic accuracy of FNAB performed at the National Center of Oncology in Azerbaijan by comparing it with postoperative histological diagnoses and define possible causes of errors leading to false-negative and false-positive results in order to better identify the factors that may affect the accuracy.

| ME THODS
The study covered the retrospective analysis of results obtained from January 2016 to December 2019. Data from 959 patients who had thyroid surgery in the National Centre of Oncology were reviewed from electronic medical records. We considered only patients with surgical indication of suspected thyroid malignancy who underwent FNAB in the same institution (n = 738). The patient's age ranged between 6 and 81 (median 46), and the sex ratio was 637/101 (female to male).
All performed FNABs were ultrasound-guided and ultrasound risk stratification was performed according to TI-RADS (thyroid imaging reporting and data system) and collected samples were smeared, stained with haematoxylin-eosin and submitted for cytological interpretation.
All FNABs were classified in accordance with the Bethesda sys-

| RE SULTS
According to the thyroid FNAB results, 88 cases (11.9%) were inter-  Table 1. Tumours, 8th Edition. 9 Only in 2 of 10 cases, tumour cells were detected also in lymph nodes, and they were classified as N1a.

TA B L E 1 Pathological characteristics of the cases
Postoperative histological examination revealed all tumours to be papillary cancer.

| DISCUSS ION
Thyroid nodules can be caused by different benign and malign disorders. 1 Initial assessment of the patients with a thyroid nodule includes a detailed history and physical examination. Thyroid ultrasound is performed to confirm the presence of nodules, to evaluate additional nodes and cervical lymph nodes and to assess for sonographic features. The next step is a FNA biopsy, which is considered a gold standard for thyroid nodule evaluation. 4,5 FNAB is an economical, minimally invasive and accurate method.
It may be performed by palpation or with ultrasound guidance.
Needle biopsy of simple cysts is not recommended. In this study, all FNABs were ultrasound guided. FNAB is recommended for the thyroid nodules larger than 1 cm; it is not usually performed on nodules smaller than that. The interpretation of the FNAB smears may be influenced by different factors such as technique and specimen preparation, the skill of the operator and the cytopathologist's expertise. 6 Our multidisciplinary team consists of two surgeons, one pathologist who specializes in thyroid disorders, one nuclear medicine specialist, two endocrinologists and one radiologist. The study showed that hyperplastic and adenomatous nodules were the most important factors in the false-positive diagnosis, as they can be misdiagnosed as suspicious for papillary thyroid