The diagnostic value of add-on thyroid cell block in the evaluation of thyroid lesions

INTRODUCTION

The incidence of thyroid neoplasm has been increasing dramatically. It is expected to be the fourth common cancer worldwide by 2030.^[1] At the local level, thyroid nodules appear to be fairly frequent in Saudis, perhaps exceeding the 10% incidence of Western adult population.^[2] The sensitivity and specificity of fine-needle aspiration (FNA) as a presurgical predictive tool for the evaluation of thyroid nodules have been well established, precisely with the application of the Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) characterization.^[3-5] The reported sensitivity and specificity for malignant cases were of 94% and 98.5%, consecutively.^[6] However, false positive and false negative calls resulted from FNA readings can lead to remarkable diagnostic pitfalls.^[7] False negative results are often caused by inadequate aspirated material; therefore, it is essential to consider additional technique within FNA procedure to enhance the sample adequacy. In some institutions, cell-block materials are performed routinely per the College of American Pathologists 2016 Practice Survey, which reported 73.7% laboratories (575 out of 735 respondents) processing cell-blocks as a common cytopreparation.^[8] Nevertheless, limited literatures support their pivotal role in the diagnostic efficacy while others indicate their disadvantages such as labor intensive, skill dependent, and waste of resources.^[9,10] This study focuses on the applicability of cell-block implementation in increasing diagnostic effectiveness of thyroid FNA.

MATERIAL AND METHODS

This study received approval from the Committee of Biomedical Ethics at King AbdulAziz University, Jeddah, KSA (No. 402-21). Two years retrospective review (2020– 2021) of the cytopathology archive yielded a total of 252 thyroid FNA cases. All FNAs were done by radiologists under ultrasound guidance using 22-gauge needles (Terumo Neolus^® Hypodermic needles). However, larger needles diameter was also used precisely for cystic nodules to tackle the sloid component. Adequacy assessment was achieved based on TBSRTC criteria.^[3] Average of two aspirates was done per case, diff-quick stained slides were prepared as well as alcohol-fixed slides for Papanicolaou staining. The needles were rinsed immediately in specimen vials containing SurePath preservative fluid (BD Diagnostics-TriPath, Burlington). For cell-blocks production, needle rinse fluid was centrifuged at 2000 rpm for 5 min and the supernatant was discarded. 10–15 mL of the cell-block preparation mix (200 ml of 100% absolute EMPRATA® ethanol alcohol + 200 mL of 100% absolute Sigma-Aldrich® formalin + 1600 ml of distilled water) was added and the specimen was gently agitated. Then, the specimen was centrifuged at 3400 rpm for 4 min and the supernatant was discarded. To prepare for processing, the pellet was wrapped loosely in a specimen paper with formalin drop, followed by placement in a sealed labeled tissue cassette.

The TBSRTC criteria were applied throughout the diagnosis process.^[3] A presence of six clusters of minimums of ten follicular cells on one slide is required for the adequacy assessment call during the rapid on-site evaluation. Although, conditions such as colloid nodules, lymphocytic thyroiditis, and the presence of atypical cells are excluded from these criteria. Images were taken by Nikon’s light microscope.

RESULTS

A total of 252 thyroid FNA cases were collected over 2 years from the right lobe, left lobe, bi-lobed, or isthmus locations [Table 1]. Filtered patients were outweighed by female gender (80%) with different age groups ranging from 18 to 76-years-old. Of those cases, a total of 920 cytology slides and their matched 150 cell-blocks were classified based on TBSRTC category as follows: 38% were benign (II). Cystic configuration and scant cellularity contributed to the 18.5% non-diagnostic category (I). The distributions of AUS/FLUS (III) and FN/SFN (IV) were 18.5% and 17%, respectively. Category-V constituted 2% of the suspicious papillary carcinoma (PTC) cases. PTC was the most common malignancy in the 6% Category-VI, followed by medullary carcinoma (MTC) [Figure 1]. On reviewing, cell blocks were grouped into three classes based on their potential values to the cytology interpretation: Inadequate cell-block, cytology smear findings look alike cell-block, and the last group which provided valuable information to the final diagnosis [Table 2].

Close

Figure 1:

The distribution of cytology smear and their matched cell block cases in various TBSRTC categories is as follows: I — non-diagnostic 18.5%, II — benign 38%, III — atypia of undetermined significance (AUS)/follicular lesion of undetermined significance (FLUS) 18.5%, IV — follicular neoplasm (FN)/suspicious for follicular neoplasm (SFN) 17%, V — suspicious for malignancy (SM) 2%, and VI — malignant 6%.

Export to PPT

Our data showed that in 63% of the total thyroid FNA cases, cell-blocks interpretation showed no superior value to cytology slides. Of note, aspirates readings and cell-blocks interpretation were interchangeable in 35% of cases. For example, characteristic features of PTC including intranuclear cytoplasmic pseudoinclusion and nuclear grooving were present in both cytology smears and cell-blocks [Figure 2]. On the other hand, the production of cell-blocks added a value to the diagnosis in 2% of FNA cases. For instance, the morphological impression of MTC [Figure 3a and b] was confirmed by immunoreactivity toward calcitonine and CEA [Figure 3c and d].

Close

Figure 2:

Characteristic features of papillary thyroid carcinoma. Classic intranuclear cytoplasmic inclusion and nuclear groove appeared in both cytology smears (a and b) and cell block material (c and d).

Export to PPT

Close

Figure 3:

The value of cell block confirming the impression of medullary thyroid carcinoma. The appearance of loose clusters of plasmacytoid cells was identified in both methods (a and b). Positive immunohistochemistry staining for Calcitonin (c) and CEA (d).

Export to PPT

As of cases that were categorized as inadequate or AUS, the addition of axillary cell-block procedure had no remarkable value.

DISCUSSION

Thyroid cancer is on significant rise. Diagnostic approaches for the detection of thyroid neoplasms include routine ultrasound scanning, FNA, biochemical investigations, and molecular techniques. The most reliable and cost effective diagnostic tool is direct smears obtained from FNA, though their false diagnosis rate is high.^[11,12] Different studies confirmed that the use of FNA needle wash material for cell-block could impact TBSRTC Category I.^[9-13] Of note, TBSRTC blue book has no clear recommendation on cell-block processing as a routine practice in thyroid FNA cytopreparations. The addition of cell-block ancillary technique was evaluated in our study to test its potential to enhance maximum use of available FNA material. Over 2 years, 150 cell-blocks generated out of 252 FNA thyroid cases were reviewed in pair with their original smears.

Non-diagnostic results can be attributed to multiple factors including aspirator experience and nature of the nodule (cystic vs. solid or fibrotic vs. calcified).^[14,15] Rarely, metaplastic squamous cells can present as cystic thyroid nodules and cause a considerable diagnostic pitfall.^[16-21] Some studies documented that using ultrasound-guided FNA approach has reduced the rate of non-diagnostic category.^[22-24] In our study, all FNAs were performed under ultrasound guidance. Despite that cell-blocks were initially processed to enhance the sample quality, 96% of the unsatisfactory FNA cases have not gained any benefits from the addition of cell-blocks [Table 2].

In parallel, benign thyroid nodules can be a diagnostic dilemma, such as lymphocytic/Hashimoto thyroiditis cases showing lymphoid component that mimics lymphoma.^[25-28] However, the presence of classic cytologic features and flow cytometry testing can support the benign impression.^[3,26,27] In our study, the characteristic cytologic features of lymphocytic/ Hashimoto thyroiditis were clearly identified on cytology smears by the presence of bland appearing nuclei and oncocytes aggregates mixed with tangled lymphocytes [Figure 4]. Based on these findings, it was not surprising that cell-block technology did not augment the initial cytology call [Table 2].

Close

Figure 4:

Criteria for Lymphocytic/Hashimoto thyroiditis (oncocytes aggregates mixed with lymphocytes) are observed on both smears (a and b) and cell block (c).

Export to PPT

On the other hand, AUS zone is the most challenging category which contributes to unnecessary surgical resections.^[28-33] Therefore, implementation of ancillary techniques could have a great impact on the subsequent management. However, our data showed that that the use of cell-block has not eliminated the number of thyroidectomies in 89% of AUS cases [Table 2].

One of the main advantage of the tissue fragments present in cell-block is maintaining the architecture closely resemble those seen on surgical specimens.^[34] Nevertheless, distinguishing between non-invasive follicular thyroid neoplasm and other categories with follicular patterns such as (Follicular Adenoma, Follicular Carcinoma, and PTC-Follicular variant) is impossible by FNA due to the absence of histological criteria.^[35-40] Therefore, it is not surprising that the use of cell-block has not improved the cytology diagnosis in 48% of our cases (TBSRTC category IV). Besides, cell-blocks recapitulate the initial interpretation in 52% of their paired-cytology cases without additional significant information [Figure 5]. Of note, 67% of TBSRTC Category V cases were diagnosed mainly by cytomorpholoic features seen on smears/liquid based-slides such as pseudoinclusions and longitudinal nuclear grooves. Former features were absent in the accompanying cell-block. While in 33% of cases, cell-blocks reiterate cytology smears featuring characteristic observations of papillary thyroid carcinoma [Figure 2]. Importantly, 33% of malignant cases were accurately based on immunohistochemistry stains applied on processed cell-blocks [Table 2]. For example, the presence of intranuclear pseudoinclusions which occur occasionally in medullary thyroid carcinoma can be mistaken for papillary thyroid carcinoma. Therefore, immunoreactivity toward calcitonin and CEA stains on cell-block material [Figure 3] had a remarkable impact on rendering the final interpretation.^[41-45]

Close

Figure 5:

Benign follicular nodules cytomorphological appearance on smears (a and b) and cell block preparation (c).

Export to PPT

In general, our data showed that cell-block auxiliary testing improved final diagnosis in only 2% of total cases and failed to reduce number of TBSRTC Category I or III cases. In parallel to our findings, the limitation of cell-blocks was described by Horton et al. and Sanchez et al.^[9,10] Application of special enhanced cell blocking methods may improve this because of quantitative improvements. In addition, the exposure of needle rinse to SurePath preservative before histology processing could quantitatively and qualitatively compromise sample adequacy.^[46-48] Therefore, cell-block processing methodology can be optimized using modern techniques such as proprietary preformed gel disc with wells (Nano NextGen cell-blocking™ kit), which concentrates a nuclei-rich sediment.^[49]

To sum up, our study confirmed that the routine processing of cell-blocks is time consuming, impractical, and causes unnecessary delay in the turnaround time. We recommend generating cell-blocks in malignant TBSRTC Categories IV and V if needed for immunostains purposes. In general, the cellularity of cell-blocks can be enhanced by the collaboration between interventional radiologists and interpreting pathologists to yield a robust specimen collection protocols.

CONCLUSION

The generated material from cell-block auxiliary technique did not eliminate TBSRTC Category I or III. It improved cytology diagnosis in 2% of total FNA cases where immunostains are needed. Therefore, our study showed that processing cell-blocks routinely in thyroid aspirates has not significantly aided in increasing the diagnostic yield of unsatisfactory or atypical thyroid samples if routine non-enhancement random methods are used. Further evaluation of optimized cell-block processing methodology utilizing modern enhancement techniques is recommended.