Differential diagnosis of diffuse sclerotic thyroid papillary carcinoma and Hashimoto's thyroiditis using fine‐needle aspiration cytology, BRAFV600E, and ultrasound elastography

Abstract Background The diffuse sclerosing variant of papillary thyroid carcinoma (DSV‐PTC) has ultrasound findings that are similar to Hashimoto's thyroiditis (HT), resulting in under‐diagnosis. DSV‐PTC combined with HT is also common, so early and accurate diagnosis of DSV‐PTC using a variety of diagnostic techniques, including FNAC, BRAFV600E mutation detection, and ultrasound elastography, is critical. Objective To assess the diagnostic value of fine‐needle aspiration cytology (FNAC) and BRAFV600E detection in combination with ultrasound elastography in the diagnosis of DSV‐PTC. Methods We performed a retrospective analysis of 40 patients with pathologically confirmed DSV‐PTC and 43 patients with HT admitted to our hospital's ultrasound department between January 2015 and December 2020. Preoperative FNAC, BRAFV600E mutation detection, and ultrasound elastography imaging were all performed on all patients. For a definitive diagnosis, the results of these tests were compared to postoperative pathological findings. The diagnostic value of FNAC, BRAFV600E mutation detection, ultrasound elasticity imaging, and their combination for DSV‐PTC diagnosis was assessed. Results The mean elastic strain rate ratio (E1/E2) of the 40 DSV‐PTC cases was 5.75 ± 2.14, while that of the 43 HT cases was 2.81 ± 1.20. The receiver operating characteristic (ROC) curve was generated using the average value of E2/E1. The area under the ROC curve was 0.910, and the optimal E2/E1 cut‐off value was 4.500. When FNAC, BRAFV600E mutation detection, and ultrasound elasticity imaging detection were combined, the diagnostic sensitivity, specificity, negative predictive value, positive predictive value, and accuracy of DSV‐PTC diagnosis were 92.5%, 95.3%, 93.2%, 94.9%, and 94.0%, respectively, which were significantly higher than the single technique (p < 0.05). Conclusions The use of FNAC, BRAFV600E mutation detection, and ultrasound elastography in combination is more helpful in establishing an accurate diagnosis of DSV‐PTC than using a single diagnostic technique alone.


| INTRODUCTION
Papillary thyroid carcinoma (PTC) is the most common type of thyroid cancer, accounting for more than 70% of all thyroid cancers. Although the tumor, node, and metastasis staging system is the most commonly used parameter for determining therapeutic plans, recent research has shown that different histopathologic variants of PTC can have different clinical courses and patient prognoses. PTC sonographic criteria include a taller-than-wide shape, an irregular margin, microcalcifications, and marked hypoechogenicity. Sonography's role has expanded to allow for the characterization of PTC variants based on sonographic features. Tall cell and diffuse sclerosing PTC appear to have more aggressive clinical courses and poor prognoses.
Diffuse sclerosing papillary thyroid carcinoma (DSV-PTC) accounts for 0.7%-6.6% of all papillary thyroid carcinomas. 1,2 It is more invasive than traditional PTCs, with higher rates of recurrence and metastasis. Hashimoto's thyroiditis (HT) is chronic lymphocytic thyroiditis. 3,4 Its ultrasound characteristics are similar to those of DSV-PTC, resulting in a missed or incorrect diagnosis.
Cases of DSV-PTC with HT are also fairly common; thus, early detection of DSV-PTC is critical. Furthermore, improving the accuracy of preoperative imaging diagnosis is critical in clinical treatment selection. 5,6 Fine needle aspiration cytology (FNAC) is the preferred method for preoperative assessment of benign and malignant thyroid lesions, 7,8 but it can produce false-negative results due to the operator's technical skill, sampling area, and incomplete sampling. Gene identification is currently a research hotspot in the diagnosis of PTC.
The BRAF V600E mutation is found in approximately 60% of PTC patients. 9,10 In the diagnosis of PTC, BRAF V600E has high specificity.
However, there are few references in diffuse thyroid disease, particularly in DSV-PTC diagnosis.
In recent years, ultrasound elastography has emerged as ultrasonic imaging technology. The hardness of thyroid lesions can be estimated by measuring the degree of deformation after compression, and the histological characteristics of thyroid lesions can be determined by revealing the elastic hardness. 11,12 The strain rate ratio is a newly developed parameter for calculating the hardness of a tissue quantitatively; it avoids the subjectivity of elastic graph grading. The hardness of tissue can be quantified by calculating the ratio of the elastic strain rate of the tissue surrounding the lesion to the elastic strain rate of the lesion (E1/E2), which can be used to differentiate between benign and malignant tissues. Some researchers are currently using real-time ultrasound elastography to diagnose thyroid cancer, 12,13 but there are few reports on the use of conventional ultrasound combined with realtime ultrasound elastography in the analysis of DSV-PTC ultrasound imaging features.
In this study, DSV-PTC was diagnosed using preoperative FNAC, BRAF V600E mutation detection, and ultrasound elastography, and the diagnostic value of combining preoperative FNAC, BRAF V600E mutation detection, and ultrasound elastography were assessed.

| Subjects
Between January 2015 and December 2020, 121 patients underwent thyroidectomy in our hospital, with all of them undergoing preoperative procedures such as FNAC, BRAF V600E mutation detection, and routine preoperative ultrasound for thyroid calcification. We identi- The pathological diagnosis of DSV-PTC, thyroid lymphocytic infiltration, fibrous hyperplasia, squamous metaplasia, and extensive sand granules were the inclusion criteria for the DSV-PTC group. The sonograms revealed a diffuse thyroid lesion occupying the majority of all of the unilateral or bilateral lobes, as well as stone-like calcification and cervical lymph node metastasis. The Ethics Committee at our hospital reviewed and approved this study, and the patients provided written informed consent.

| Instruments and methods
The diagnostic instruments used were the GE Logiq E9 (GE Medical Systems, American General) and Hitachi Vision Preirus System (Hitachi Medical Corporation, Tokyo, Japan), both of which were equipped with real-time ultrasonic elastography technology.
The frequency of the probe was set to 5-12 MHz. Transverse and longitudinal scanning of the thyroid or the lesion was performed using two-dimensional grayscale imaging during a conventional ultrasound. The lesion and cervical lymph nodes were observed and recorded for their location, size, shape, edge, boundary, echo, calcification, color blood flow distribution, and elastic grade. Submaxillary lymph nodes were used as normal controls, and abnormal lymph nodes were found to be more than 5 mm in diameter, disordered in structure, loss of medulla, high cortical echo, sand body calcification, and abundant and disordered blood flow signals. Sand granular calcification in lymph nodes was classified into four grades: grade 0 (no calcification), grade 1 (sparse distribution), grade 2 (a few more calcification points of varying thickness), and grade 3 (dense and scattered in lymph nodes). The degree of blood flow signal in lymph nodes was classified into four levels: 0 (no flow); 1, sparse spot-like blood flow; 2, rich blood flow distribution without affecting parenchymal echo; and 3, rich blood flow completely filling lymph nodes and covering parenchymal echo.
Following routine ultrasound, real-time ultrasound elastography was initiated. A sampling frame that included the observed lesion, surrounding thyroid tissue and the sternocleidomastoid muscle in front of the thyroid was chosen. The image was viewed, a relatively stable image frame was chosen, and the regions of interest, A and B, were delineated in the sampling frame; A was the lesion area, and B was the reference area, which was the sternocleidomastoid muscle on the ipsilateral side of the thyroid, where the lesion was located.
To calculate the elastic strain rate in the lesion area, the instrument system calculated the average elastic strain variable and the ratio of the two regions of interest.
Patients undergoing FNAC were positioned supine, with a soft pillow on the neck and shoulder, head, back, and anterior neck exposed. Prior to surgery, ultrasound positioning and conventional disinfection were performed. Under the guidance of an ultrasound, a 22G fine needle was used for multi-point puncture and suction of materials in the area with a high concentration of thyroid calcification and suspicious low-echo areas. The suspected malignancy was punctured ≥5 times with multiple points of lift and insert, and 4-6 smears were taken and fixed with 95% ethanol. The FNAC procedure for suspicious lymph nodes is the same as described above.
The FNA cell extract was injected into Eppendorf tubes and stored at À80 C. Following the puncture, the patient was instructed to compress and wrap the neck with sterile gauze for 20 min before being observed for 30 min. The patient was instructed to leave the hospital after a color ultrasound confirmed that there was no neck bleeding. According to the manufacturer's instructions, we performed an immunohistochemical analysis of BRAF V600E expression (all markers were purchased from Dako, Dakopatts, Denmark).

| Diagnostic criteria for DSV-PTC
At least one sonographer performed the preoperative ultrasound examination and diagnosis. Any disagreements were settled with the help of a second ultrasound doctor. DSV-PTC ultrasound images were evaluated for gland (unilateral/bilateral) enlargement and abnormally hypoechoic and diffuse microcalcification ("blizzard-like" changes) with or without nodular lesions (nodules are often ill-defined). Lymph node diameter >5 mm, irregular structure, the disappearance of the medulla, increased cortical echogenicity, sand granular calcification, and rich and disordered blood flow signals were all criteria for cervical metastatic lymph nodes. The following histological criteria were used to diagnose DSV-PTC: (1) diffuse gland enlargement with calcification, (2) extensive squamous metaplasia, (3) dense fibrosis, (4) a large number of sand and gravel bodies, and (5) massive lymphocytic infiltration. The sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy were calculated, and the rates were compared using the chi-square test. A p-value of <0.05 was considered statistically significant 3 | RESULTS

| Comparison of clinical data and ultrasound characteristics between patients with DSV-PTC and HT
The study included 83 DSV-PTC and HT patients (age: 46.78 ± 13.18 years; male/female ratio, 35/48). Table 1

| Comparison of ultrasonic characteristics between DSV-PTC and HT cervical metastatic lymph nodes was analyzed
Thirty-five of the 40 DSV-PTC cervical lymph node metastases measured 12-21 mm in diameter. DSV-PTC cervical lymph node metastases were oval or round in shape, with a rough capsule surface. The echo of affected lymph nodes is similar to that of the thyroid, with internal heterogeneity and patchy foci. Grainy calcification was found in lymph nodes, which were scattered or clustered (Figure 1). Cystic changes in lymph nodes were found in four cases. There were 31 HT cases with enlarged T A B L E 1 Comparison of clinical data between patients with DSV-PTC and HT  Table 1).

| Analysis of the diagnosis of DSV-PTC and HT using the elastic strain rate ratio method
In this study, 40 cases of DSV-PTC and 43 cases of HT were pathologically confirmed. According to the pathological examination results, the mean elastic strain rate ratio of the 40 DSV-PTC cases was 5.75 ± 2.14, while that of the 43 HT cases was 2.81 ± 1.20 ( Figure 1). Notably, the mean elastic strain rate ratio of DSV-PTC was significantly higher than that of HT (p = 0.000). The ROC curve was drawn using the average value of E2/E1. The area under the ROC curve was 0.910, and the optimal critical point of E2/E1 was 4.500 ( Figure 2). As shown in Table 2 In comparison to postoperative pathological findings, the sensitivity, specificity, NPV, PPV, and accuracy of the preoperative diagnosis of PTC using BRAF V600E mutation detection were 10%, 100%, 54.4%, 100%, and 56.6%, respectively. 3.5 | Combined FNAC, BRAF V600E mutation detection, and elastic strain rate ratio method and postoperative pathology comparison

| DISCUSSION
DSV-PTC is a subtype of papillary thyroid carcinoma that accounts for about 1.8% of all cases. [14][15][16] In contrast to the majority of cases, which have good differentiation, slow growth, and a good prognosis, DSV-PTC grows quickly and aggressively, is more prone to lymph node and distant metastasis, and has a poor prognosis. 17  fibrosis, and lymphocyte infiltration. As a result, the sand-like calcification of DSV-PTC is more extensive and dense. In this study, 32 cases covered the entire affected gland, with a diameter of 1-3 mm, and 5 cases were clustered in clusters. However, all 43 HT cases were characterized by diffuse enlargement of the thyroid gland, and the internal diffuse echo was weakened, and streak-like and point-like strong echo could be seen, which was primarily manifested as "network," which was also one of the distinguishing features of DSV-PTC and HT.
In this study, 31 HT patients had lymph nodes in neck region VI with decreased aspect ratio, full shape, decreased parenchymal echo, and most medullolymphatic hilum echoes disappeared. This is due to  In this study, the sensitivity of diagnosis was increased from 70.7%, 77.5%, and 10% to 92.5%, respectively. Furthermore, patients with a positive FNAC result, a high ultrasound elastography strain rate, and a negative BRAF V600E mutation test had a postoperative pathological confirmation accuracy of 98%. As a result, combining FNAC, BRAF V600E mutation, and ultrasound elasticity in the differential diagnosis of DSV-PTC improves diagnostic sensitivity and specificity significantly, allowing some DSV-PTC patients who cannot be clearly diagnosed by US-FNAC to obtain a scientific and accurate diagnosis before surgery. It provides a reliable imaging and molecular basis for patient follow-up treatment, as well as the scientific foundation for the formulation of patient follow-up plans and treatment plans. This study has some limitations. First, this was a retrospective study, and there were some selection biases that may have influenced the results. Second, the sample size of this study is small, and it is expected that future clinical studies with a larger sample size will provide more data for the diagnosis of DSV-PTC.