Core needle biopsy in the management of thyroid nodules with an indeterminate fine-needle aspiration report

Abstract

Ultrasonography (US)-guided fine-needle aspiration biopsy (FNAB) cytology is widely used but is limited due to its pathologically indeterminate results in diagnosing thyroid nodules. Recently, US-guided core-needle biopsy (CNB) was introduced as an effective and safe technique for diagnosing indeterminate thyroid nodules. Using CNB, information about architectural histologic structure such as nodule capsule or more immunochemical staining can be obtained which lead to a more accurate diagnosis. Up to 98% of indeterminate thyroid lesions can be classified as malignant or benign when CNB is used for follow-up analysis. Other evidences revealed the effectiveness of CNB in reducing inconclusive results and improving the diagnostic performance of thyroid nodules initially diagnosed as AUS/FLUS by FNAB. In this review, we investigate how to deal with indeterminate thyroid nodules diagnosed by FNAB and determine how CNB has a role in diagnosing these indeterminate thyroid nodules.

Introduction

Ultrasonography (US)-guided fine-needle aspiration biopsy (FNAB) cytology is widely used as a minimally-invasive tool for evaluating thyroid lesions (1,2). However, FNAB is limited due to pathologically indeterminate results in about 10 to 20% of the cases (3). Even though the American Thyroid Association (ATA) guidelines recommend repeat FNAB for these cases, repeat FNAB shows 1–7% of non-diagnostic results and 3.8–31.0% of indeterminate results (4-7). It is known that the malignancy risk of the AUS/FLUS category is 6–18% based on the 2017 Bethesda system (7). However, there has been a tendency to overuse the diagnosis of AUS/FLUS and the reported malignancy risk varies from the proposed rate by 14% to 38% (8,9).

Due to the variable risk of malignancy of these indeterminate thyroid nodules, other guidelines (2,10) recommend evaluating these nodules by diagnostic surgery. However, the majority (70–80%) of indeterminate thyroid nodules are benign according to surgical histology (11). Other solutions for dealing with indeterminate thyroid nodules, such as the molecular testing or gene expression classifier, or clinical studies together with US findings are investigated, but such parameters are still controversial (12-15). Recently, core-needle biopsy (CNB) was introduced as a safe and effective tool in diagnosing indeterminate thyroid nodules in order to prevent unnecessary surgery (16,17). Some studies reported that as much as 98% of indeterminate thyroid nodules are able to be classified as malignant or benign when CNB is used for follow-up analysis (18-22).

Therefore, in this review, we investigate how to deal with indeterminate thyroid nodules diagnosed by FNAB and using CNB.

Strategies to deal with indeterminate FNAB results

Various additional tests have been suggested for thyroid nodules with previously indeterminate FNAB results, such as repeat FNAB (23), molecular testing (12-15), diagnostic surgery (2,10) or CNB (16,17,24). The recently revised ATA management guidelines proposed that repeat FNAB or molecular testing be used and in case the results are inconclusive, that excision be performed with consideration of the clinical and radiologic features and patient preferences (25).

Repeated FNAB can be easily performed if the primary FNAB result is indeterminate, however, the benefit in patient management is still unclear because there is still a chance to obtain indeterminate results (i.e., non-diagnostic and AUS/FLUS results) (26). Only a small chance to diagnose malignancy (1.35% to 24.77%) is reported (6,9,27-32). Application of immunohistochemical stains such as galectin-3, cytokeratin-19, HBME-1, or BRAF^V600E has been introduced for these nodules (33,34). The most important drawback is that immunohistochemical stains do not guarantee significant improvement in the diagnostic accuracy. Diagnostic surgery can provide a definitive diagnosis, although the procedure itself not only has a risk of complications but also holds a high possibility for benign results (11).

Some recent studies showed that histologic examinations by CNB can accurately diagnose a large percentage of indeterminate thyroid nodules (16,17) and with a very low rate of minor complications (35). Although CNB has a risk of repeated indeterminate results, CNB demonstrated a summary sensitivity of 91% (95% CI, 81–96%) and specificity of 99% (95% CI, 98–100%) by a meta-analysis using data collected from 10 CNB studies with 1,733 patients (36).

Recently, a web-based risk stratification system created using combination of Bethesda III thyroid nodules and US features has been introduced (http://www.gap.pe.kr/xe/Estimation_A) (37). This system is based on a previously published web-based prediction model (http://www.gap.pe.kr/xe/Estimation) (38), which is a simple and easily accessible web-based diagnostic scoring system for the malignancy risk stratification of thyroid nodules. Evaluation of the malignancy risk-stratification system showed good predictive accuracy, with approximate AUCs of 0.83.

CNB in diagnosing indeterminate thyroid nodules

Although it is recorded in the literature that CNB was performed for thyroid nodules in the early 1980s, large-needle biopsy, conducted without US guidance was not recognized in clinical use due to the pain and bleeding risk at that time (39,40). With the widespread use of high-resolution US and introduction of advanced CNB devices such as spring-activated single- or double-action core needles, accurate diagnosis became possible using US-guided CNB with a minimal chance of complications (41). Since then CNB has been reported as an effective and safe method in diagnosing thyroid nodules (18,26,42-44).

By obtaining enough tissue of the nodules using CNB, more information regarding the architectural histologic structure such as the nodule capsule or more immunochemical staining can be obtained (45). In addition, CNB has the advantage of assessing nuclear change, general alterations in the follicular structure, and relationships with adjacent tissues by obtaining adequate material from the nodule (43). We can expect the potential of overcoming the limitation of FNAB, such as poor specimen quality of cellularity and appropriate preservation which leads to misdiagnoses (26,33,46-48). Although several studies revealed that CNB demonstrates no additional benefit to that of FNAB (41,49,50), the role of CNB has been suggested in many recent studies (18,26,46). Up to 98% of indeterminate nodules can be classified as malignant or benign when CNB is performed (18-22).

A recent article investigated that subcategory nodules of nuclear atypia had a higher malignancy risk, of being surgical candidates, of having suspicious US features, and of having malignant CNB results than subcategory nodules of architectural atypia (51). Other studies showed that CNB was helpful for diagnosing subcategory nodules of nuclear atypia but was not, or less helpful for subcategory nodules of architectural atypia (49,52,53). Yet at the same time, another study of 153 consecutive patients suggested that CNB might be more useful in decision making than repeat FNAB in both subcategory nodules of nuclear atypia and subcategory nodules of architectural atypia and it has the potential to be a first-line alternative diagnostic tool for initially diagnosed indeterminate thyroid nodules (53).

Even though there is no clear guideline for CNB in diagnosing indeterminate thyroid nodules yet, evidences revealed the effectiveness of CNB for reducing inconclusive results and improving the diagnostic performance of thyroid nodules with initial indeterminate results (54). In indeterminate lesions, the combined use of repeated FNAB and CNB might be considered. Previous studies with large series of nodules showed that CNB has higher accuracy than repeated FNAB, although the combination of two biopsies even improves the rate of diagnosis (26,46).

Several studies have shown the usefulness of CNB for thyroid nodules with AUS/FLUS results. In a retrospective study comparing three management tools i.e., CNB, repeat FNAB, and diagnostic surgery for previous AUS/FLUS in FNAB, the CNB results were preferable, i.e., 77.8% benign, 20.3% cancer, and 1.8% non-diagnostic, than those of repeat FNAB i.e., 35.2% benign, 16.1% cancer, and 48.6% non-diagnostic, and were comparable to those of diagnostic surgery (18). In a prospective study of concurrent CNB and FNAB, the incidence of non-diagnostic or AUS/FLUS results was lower in CNB, i.e., 3.1% non-diagnostic and 23.6% AUS/FLUS results than in repeat FNAB, i.e., 9.3% non-diagnostic and 39.8% AUS/FLUS results (26).

A meta-analysis by Suh et al. (36) demonstrated that CNB showed higher sensitivity (91%) in diagnosing malignancy than FNA (74%) and with no significant difference in specificity i.e., 99% vs. 100% respectively, and a lower pooled proportion of non-diagnostic results compared with FNAB (5.5% vs. 22.6%). CNB showed much fewer inconclusive results than FNAB and with a pooled proportion of 8.0% (95% CI, 4.4–11.5%) vs. a pooled proportion of 40.2% (95% CI, 25.1–55.3%). Therefore, the authors argue that CNB may be a complementary diagnostic tool in nodules with initially non-diagnostic and AUS/FLUS results on previous FNAB. Also, the National Cancer Institute (40), American Association of Clinical Endocrinologists/American College of Endocrinology/Associazione Medici Endocrinologi (AACE/ACE/AME) (1), and the Korean Society of Thyroid Radiology (KSThR) (55) proposed CNB as an additional diagnostic strategy for thyroid nodules with previous non-diagnostic FNAB results.

CNB guidelines for Bethesda III and IV according to 2016 consensus statement and recommendations from Korean Society of Thyroid Radiology (KSThR)

CNB for atypia (follicular lesion) of undetermined significance in previous FNAB

Even though the malignancy risk of the AUS/FLUS category varies between 15–25% based on the Bethesda system (10), there has been a tendency to overuse the diagnosis of indeterminate and the reported malignancy risk varies from the proposed rate, to reach 14% to 38% (8,9). To improve the accuracy in detection of malignancy and to make better management decision for the AUS/FLUS category, various trials have been suggested such as application of immunohistochemical stains or repeat FNAB or CNB or even diagnostic surgery (33,34). A study reported that CNB results were better in diagnosing thyroid nodules than those of repeat FNAB and equivalent with those of surgical excision (18). Also, a prospective study revealed that the incidence of non-diagnostic or AUS/FLUS was lower in CNB than in repeat FNAB (26). Thus, based on these references, recommendation of the KSThR is to use CNB as good alternative to FNAB for indeterminate thyroid nodules.

CNB for follicular neoplasms

Since follicular neoplasm is challenging to diagnose with FNAB, CNB has been introduced as a complementary method for thyroid nodules because the enough specimen obtained can make detailed histologic evaluation and ancillary immunohistochemical staining possible (26,47,56). Also, ever since new sampling techniques including the capsule of the nodule and the surrounding extranodular parenchyma as well as nodular tissue were introduced, follicular neoplasm and unencapsulated non-neoplastic nodules could be distinguished by identifying the presence of a fibrous capsule on histologic evaluation (57,58). Thus, KSThR recommended CNB for follicular neoplasm as follows (55): (I) CNB can differentiate non-neoplastic nodules from encapsulated follicular neoplasms, (II) CNB cannot differentiate follicular thyroid carcinoma from follicular adenoma.

Limitations of CNB

Although CNB shows high efficacy and safety with competent accuracy, one study demonstrated that up to 36% of the results remained indeterminate after CNB due to insufficient cytology to differentiate nodular hyperplasia from follicular neoplasm (52). Another concern regarding CNB is the possibility of false-negative result. As the tissue is only collected by the side hole of the needle, there is a possibility of capturing the normal thyroid tissue, not the target lesion (18).

Furthermore, no clear guidelines regarding the management of AUS/FLUS results are also limitation of CNB. AACE/ACE/AME guidelines do not recommend the use of CNB in indeterminate nodules due to the limited evidence and the lack of validated reporting systems (59). However, the recently suggested pathology reporting system by the Korean endocrine pathology thyroid CNB study group is expected to be useful in diagnosing CNB specimens (60,61).

Safety of CNB

A retrospective study evaluated 6,169 consecutive patients with 6,687 thyroid nodules and showed overall 53 complications in 50 patients (0.81%), including four major complications including massive hematoma, pseudoaneurysm or voice problems (62).

According to a systematic review and meta-analysis evaluated complications following CNB (36), there was only one major complication among 3,163 patients (0.03%), and which was overnight hospitalization for observation of bleeding. There were 15 minor complications after CNB among 2,608 patients (0.58%), including 12 patients with hematomas, two patients with transient hoarseness, and one patient with hemoptysis. On the other hand, there were no major complications after FNAB in 2,572 patients and the three, minor complications in 2,017 patients include two with hematomas and one patient with transient hoarseness. Other studies controlled post-CNB perinodular hemorrhage with simple manual compression (58,63). Pain during biopsies was controllable with local anesthesia (63).

Several studies compared the complications of FNAB and CNB. A recent study showed no significant differences when comparing the two procedures in terms of pain, tolerability, or complications (64). Two previous studies have also shown similar results regarding the tolerability and pain between the two procedures (35,65).

Cost-effectiveness of CNB

A recent study demonstrated the cost-effectiveness of CNB regarding avoiding unnecessary diagnostic surgery for thyroid nodules with indeterminate pathologic results on FNAB (66). In this study, the authors classified 42.4% of the indeterminate FNAB resulting thyroid nodules as benign according to CNB. This study demonstrated that the cost of a single CNB is less expensive than diagnostic surgery at about 1/6. They also insisted that about one-third of the expense can be saved with CNB compared to undergoing diagnostic thyroidectomy for all nodules with indeterminate results on FNAB by avoiding unnecessary diagnostic surgery.

CNB technical perspective

CNB needs to be performed by experienced operators with US guidance. Operators should decide the suitable CNB needles and the approach route via pre-procedural US to improve the safety and diagnostic accuracy (55). The trans-isthmic approach is usually considered suitable, but operators should completely understand and plan the most appropriate access route. It is also recommended to choose a needle with a similar length with the targeted nodules (26,46) (Figure 1). For the optimal result, the entire needle should be scanned during the procedure and the needles have to be remained parallel to the US probe axis during the procedure. Before firing the core needle, operators should envision the anticipated needle route and make sure that the needle tip would be in the safe place. The location of the specimen notch is adjustable after firing the stylet select most proper sampling target.

Figure 1.

Core needle device. (A) Stylet and specimen notch (arrows); (B) cutting cannula.

There are two strategies for obtaining a specimen with CNB. One is to locate the specimen notch of the biopsy needle confined to the internal portion of the thyroid nodule (58). The other is to biopsy the capsule of the thyroid nodule and the surrounding parenchyma (57) in order to adequately diagnose follicular-pattern lesions (63). Using this modified CNB technique, three components should be obtained for the pathological interpretation, i.e., nodule tissue, nodule-parenchyma border including capsule, and normal thyroid parenchyma (57,67) so as to differentiate nodular hyperplasia from follicular neoplasm (58) (Figure 2). For sampling hard nodules such as severe fibrosis or dense calcification, it is necessary to stab the nodule in order to first identify a weak point. To prevent surrounding structure damage due to needle deflection during the hard nodule CNB, careful evaluation of surrounding structures is vital (68-70).

Figure 2.

US-guided CNB procedure. (A) A 61-year-old female patient with well-defined hypoechoic nodule in the mid-pole of right thyroid gland; (B) planning of trans-isthmic approach with measurement of distance of fire (arrows); (C) firing of stylet. Specimen notch includes nodule, nodule capsule, and small amount of normal thyroid parenchyma; (D) firing of cutting cannula. CNB, core needle biopsy; US, ultrasound.

Visual assessment should be performed in order to confirm whether additional CNB is required (26,46,69,71) after the first CNB. After the visual inspection, the tissue has to be immediately fixed in formalin. In most cases, one or two biopsy sampling is sufficient for the adequate histology diagnosis. A study suggested to obtain at least two core specimens including an intranodular and capsule portion when using a 1.1-cm core device (72), while other researchers argued that one specimen would be enough when using a longer core device, i.e., 1.6- or 2-cm (73). When a nodule has heterogeneous components in US, it is better to harvest tissue in multiple sites of the nodule in order to represent all areas of the nodule. Just in case complication such as bleeding happens, additional sampling could be postponed. Manual compression should be carried out for 20 to 30 minutes immediately after the biopsy.

Conclusions

CNB is safe and effective in diagnosing thyroid nodules. The wide use of US-guided CNB for previous indeterminate FNAB results may be the next successful diagnostic tool in order to reduce unnecessary surgeries.