Authors: Jae Ho Shin, Yeseul Kim, Min Kyoung Lee, Jung Hwan Baek, So Lyung Jung
Categories: Review Article, Thyroid, Ultrasonography, Biopsy
Source: Ultrasonography
Doi: 10.14366/usg.25201
Authors: Jae Ho Shin, Yeseul Kim, Min Kyoung Lee, Jung Hwan Baek, So Lyung Jung
Ultrasound-guided fine needle aspiration biopsy (FNAB) for thyroid lesions has proven safe and effective; however, FNAB has limitations, including the frequent occurrence of nondiagnostic and inconclusive results and technical challenges in specific clinical scenarios. To address these limitations, core needle biopsy (CNB) was introduced, and substantial advancements have been made in CNB devices and techniques. Evidence from large patient cohorts indicates that CNB is not only safe and reliable but also provides superior diagnostic accuracy compared with FNAB. This review discusses recent advancements in CNB techniques to optimize tissue sampling, the diagnostic performance of CNB, and patient-centered perspectives, highlighting the potential of CNB as a selective first-line biopsy modality.
Advancements in high-resolution ultrasound (US) have considerably improved image quality and overall detection of thyroid nodules, contributing to an increase in the number of thyroid biopsies. Fine needle aspiration biopsy (FNAB) of thyroid nodules has been widely regarded as a reliable and cost-effective diagnostic tool [1–4]. In addition to its high sensitivity for differentiating malignant from benign nodules, FNAB’s straightforward capillary-sampling mechanism, combined with procedural efficiency, makes it well suited for outpatient practice and a preferred first-line biopsy modality [2,5,6]. Nevertheless, a major limitation of FNAB is the frequent occurrence of inconclusive results, including nondiagnostic outcomes and atypia of undetermined significance [7–12].
Core needle biopsy (CNB) has undergone substantial advances in devices and techniques, and its use has expanded beyond that of an alternative biopsy modality because of better diagnostic performance than FNAB [13–16]. For example, several studies reported that CNB outperformed repeat FNAB in the re-evaluation of inconclusive FNAB results and outperformed FNAB in challenging scenarios, such as nodules with rim calcification, macrocalcifications, or cystic components [17–20]. In addition, major progress in CNB practice has been shaped by two guideline documents published in 2017 by the Korean Society of Thyroid Radiology and in 2020 by the Korean Thyroid Association [21,22]. Both guidelines provide evidence-based practical recommendations aimed at improving the efficiency, safety, and diagnostic performance of CNB.
This review highlights technical innovations and improvements addressed in the two major guideline publications from the Korean Society of Thyroid Radiology and the Korean Thyroid Association, along with advancements and findings that may not have been fully covered in those documents [21,22]. The review also provides a comprehensive overview of the diagnostic performance of CNB, management of complications, and patient-centered perspectives, thereby highlighting the emerging role of CNB as a selective first-line biopsy modality.
Before the development of modern thyroid-dedicated CNB, the Vim-Silverman needle (MacGregor Instrument Co., Camas, WA, USA) and large-bore needles (e.g., 14-gauge needles) were used to obtain thyroid nodule specimens through the 1980s [21–24]. The Vim-Silverman needle consisted of a stylet for puncture, a cannula to guide the target tissue, and a bifid needle for specimen retrieval. Biopsies were typically performed under palpation guidance, with solid portions targeted for sampling, whereas cystic regions were aspirated using a syringe with a thinner needle (21–23G) [23]. In essence, early CNB techniques involved puncturing and nicking the tissue.
Concerns have been raised regarding the invasiveness of CNB and the potential for increased hemorrhage risk due to the use of thick needles. Modern CNB devices with thinner calibers (18G–21G) have been developed to address these concerns [21]. Nasrollah et al. [25], for example, demonstrated the feasibility of CNB using a 21-G device (Menghini cutting needle; diameter, 0.514 mm) in 40 consecutive indeterminate nodules previously assessed with FNAB. In that study, 21-G CNB achieved a high specimen adequacy rate (97.5%) without major complications and enabled identification of 35% of benign lesions.
Modern thyroid-dedicated CNB devices typically have shorter needle lengths than liver or lung biopsy devices and are categorized by two main the location of the cutting mechanism and the type of spring-action mechanism. The cutting mechanism may be side-cutting or end-cutting, with side-cutting devices now standard for head, neck, and thyroid biopsies. End-cutting needles use a hollow trocar to extract cylindrical specimens, with the cutting action occurring at the tip [26,27]. Automatic CNB devices use a “double-action” mechanism in which the inner needle is propelled by a spring for penetration, followed by spring-powered advancement of the outer cutting cannula to obtain a tissue sample [21]. In contrast, semi-automatic devices use a “single-action” mechanism in which the inner stylet is advanced manually, followed by spring-powered firing of the cutting cannula. Owing to this single-action mechanism, semi-automatic devices have limitations in penetrating hard tissue, fibrotic components, and calcified nodules. Nevertheless, semi-automatic devices offer distinct advantages, including greater maneuverability for adjusting the needle trajectory and finer control of the cutting cannula, thereby enhancing procedural safety. Park et al. [28] reported comparable rates of nondiagnostic results with semi-automatic CNB devices (0.7%) and automatic CNB devices (1.8%). Furthermore, the overall complication rate remained low, with only 2.4% minor complications. Although that study suggested that both CNB device types are effective and safe, device selection requires careful consideration of factors such as thyroid nodule size, characteristics, location, and operator experience.
Standard CNB techniques are widely practiced in Korea and recommended by guidelines published by the Korean Society of Thyroid Radiology and the Korean Thyroid Association [21,22]. US-guided CNB consists of vessel mapping, injection of 1% lidocaine, puncture, targeting, and specimen sampling [22]. Meticulous vessel mapping is a critical preparatory step before puncture and insertion of the CNB device through the thyroid gland (Fig. 1). A 1% lidocaine solution is injected into the puncture site and perithyroidal region to provide local analgesia. For safe and effective CNB, the needle should be oriented parallel to the US plane to ensure visualization of the entire needle length (Fig. 2). Although a transisthmic approach is preferred because of a shorter distance and lower risk of hematoma formation, the approach may vary depending on factors such as nodule composition (e.g., the solid portion in mixed nodules), nodule location, and vascular distribution (e.g., intranodular or perinodular vessels) [21].
In earlier years, the optimal specimen-sampling method for accurate diagnosis was debated. Marginal targeting was introduced in 2013 by Nasrollah et al. [25]. CNB sampling is performed such that the cutting notch includes three nodular tissue, the nodular margin, and perinodular normal thyroid tissue (Fig. 3). This approach may help differentiate nodular hyperplasia from follicular neoplasm in indeterminate nodules [25,29]. Furthermore, Hahn et al. [30] reported that two marginal-targeting passes are adequate for predicting thyroid neoplasm and malignancy in thyroid nodules with low- and intermediate-suspicion features. In that study, marginal targeting demonstrated high diagnostic performance, with 100% accuracy in differentiating non-tumors from tumors and 88.8% accuracy in differentiating benign from malignant lesions among low- and intermediate-suspicion nodules. However, for malignancies such as classic papillary thyroid carcinoma and thyroid metastasis, a single intranodular sample is typically sufficient rather than multiple biopsies, because pathological assessment primarily depends on nuclear and cytoplasmic characteristics [30–32].
An effective standard CNB technique for nodules in difficult locations is the elevating maneuver. The consensus statement and recommendations of the Korean Society of Thyroid Radiology endorsed its use for small nodules located deep in the posterior thyroid [21]. Although no studies have evaluated its efficacy, this maneuver is frequently used not only for deeply located nodules but also to enable safe and effective biopsy of nodules close to critical anatomical structures, such as the common carotid artery (Fig. 4). After firing, the stylet is lodged in the target nodule. If necessary, the stylet may be carefully advanced through the nodule to facilitate elevation; otherwise, shallow penetration may lead to slip-off during the elevating maneuver. The device is then gently tilted to elevate the nodule away from adjacent critical structures before firing the cutting cannula.
CNB has traditionally been considered less versatile than FNAB because of its larger needle size and limited maneuverability. In radiofrequency ablation of thyroid nodules, hydrodissection is an essential technique that involves fluid injection to create a safe distance between the target nodule and adjacent critical anatomical structures [33]. Similarly, hydrodissection can facilitate sampling in previously challenging locations, such as small nodules near critical structures. Ebrahiminik et al. [34] suggested injecting 0.9% saline solution via a 19-gauge spinal needle between the target thyroid nodule and critical structures, such as the common carotid artery, trachea, esophagus, and tracheoesophageal groove. In all cases in that study, specimen acquisition was technically successful, demonstrating the reliability and safety of hydrodissection-assisted CNB. Hydrodissection can also be performed using thinner needles (21–25 gauge). Gradual injection of solution (generally at a rate of 1–2 mL/s) along the thyroid capsule is performed (Figs. 5, 6). During this process, it is essential to maintain a parallel orientation to the thyroid capsule to optimize separation between the true and false capsules. If resistance is encountered and saline cannot be injected smoothly, the needle position and orientation should be adjusted. In the referenced study, the authors reported using an average of 25.9 mL of saline to separate the target nodule from surrounding critical structures by at least 10 mm [34]. In the authors’ experience, hydrodissection was successfully performed in most cases using a thin 25-gauge needle and less than 10–15 mL of normal saline; however, a larger volume may be required for posteriorly located thyroid nodules.
Calcified thyroid nodules, particularly those that are entirely calcified, pose substantial challenges for FNAB and often yield nondiagnostic results. The main reasons include difficulty penetrating dense calcification and limited maneuverability for capillary sampling. One CNB strategy is to identify “weak” points within a calcified nodule (e.g., areas of interrupted calcification) and advance the stylet with a strong thrust. If penetration remains ineffective, a drilling motion of the stylet at a fixed point may be a useful alternative (Fig. 7) [35]. Once a weak point is identified, the stylet is fired from a safe distance. After the stylet is lodged at the weak point, the CNB device is advanced slowly with alternating rotational motion, mimicking drilling (Video clip 1). Another factor complicating biopsy of densely calcified nodules is prominent posterior acoustic shadowing, which can obscure visualization of the needle and cutting cannula. In this setting, it is important to estimate the position of the cutting notch based on its initial placement before making the final cut. Rather than relying solely on a powerful thrust, a drilling motion should be considered because it may reduce the risk of device ricochet or overshooting. Nevertheless, there is a risk of inadvertently sampling adjacent thyroid parenchyma rather than the intended calcified tissue. To address this issue, visual inspection of the CNB specimen is a simple and effective way to assess sample adequacy; whitish calcified material can be seen grossly in an adequate specimen [21,35]. However, when the specimen is bloody or contains mostly whitish-gray soft tissue, visual inspection may be equivocal. In such cases, a case report investigated the potential use of mammography as an adjunct imaging technique to confirm that the CNB device sampled the calcified component and to provide additional assurance regarding specimen retrieval accuracy [36].
The first published diagnostic report on US-guided CNB was a study by Quinn et al. [37] in 1994, in which the authors demonstrated high diagnostic performance in a cohort of 102 patients. Similarly, a 2003 study reported relatively high diagnostic performance, with 92% accuracy for differentiating neoplastic from non-neoplastic lesions and 96% accuracy for identifying malignant neoplasms [38]. CNB outperforms FNAB because it yields larger specimens, preserves architectural histologic detail, and allows immunohistochemical staining [21,39,40]. Specifically, molecular markers such as Gal-3, HBME-1, CD20, and cytokeratin 19 can be applied to CNB specimens to enhance diagnostic accuracy (Figs. 8, 9) [40].
One key study supporting CNB as a first-line biopsy modality is the meta-analysis by Chung et al. [41]. Thirteen studies were included, comprising 13,585 nodules from 9,166 patients who underwent CNB as a first-line biopsy. The overall nondiagnostic rate for first-line CNB was 3.5%, substantially lower than that reported for FNAB (>10% for initial FNAB and up to 50% for repeat FNAB) [41–43]. Similarly, the rate of inconclusive results was 13.8% for first-line CNB, again lower than rates reported for FNAB (>30%) [10,41]. When Bethesda category VI was considered positive, first-line CNB showed a sensitivity of 80.0% and a specificity of 100% for differentiating malignancy [41]. In contrast, a meta-analysis of FNAB in 16,597 patients—using Bethesda categories III–VI as positive—reported comparable sensitivity (85.6%) but substantially lower specificity (71.4%) [2]. As a first-line modality, CNB yielded few or no false-positive results while maintaining consistently low nondiagnostic and inconclusive rates.
Additional evidence supporting CNB as a first-line biopsy modality has been provided by several studies. Hong et al. [44] conducted a prospective study evaluating the potential role of CNB as a first-line biopsy modality and provided practical insights into nodule-specific use. The study used visual assessment of CNB specimens to determine the need for additional sampling, particularly for small nodules prone to mistargeting; nodules >3 cm or those with heterogeneous internal components; and nodules with macrocalcification or fibrosis. Using this approach, CNB demonstrated superior diagnostic performance across nodule sizes. Chen et al. [45] performed FNAB and CNB simultaneously and reported that CNB was a promising first-line biopsy modality, increasing sensitivity and accuracy for predicting malignancy. Notably, CNB showed higher sensitivity for malignancy even in small nodules ≤1.0 cm. Finally, a large cohort-based propensity score analysis by Suh et al. [9] reported higher sensitivity for diagnosing malignancy and a lower nondiagnostic rate for first-line CNB than for FNAB, with Bethesda category VI used to define malignancy.
Diffuse infiltrative lesions are currently categorized as Korean Thyroid Imaging Reporting and Data System (K‑TIRADS) 4 (intermediate suspicion). Overlapping US features, such as hypoechoic lesions with ill-defined margins, can pose diagnostic challenges, particularly in patients with malignancy [46]. In a study by Choi et al. [47], CNB demonstrated 100% sensitivity and 100% specificity for diagnosing thyroid metastasis. Compared with FNAB, CNB showed superior sensitivity (100% vs. 58.6%), making it the preferred biopsy modality in patients with cancer—particularly those with renal cell cancer, lung cancer, or breast cancer—who present with diffusely infiltrative thyroid lesions. Furthermore, patients with cancer who receive adjuvant radiation therapy to the supraclavicular region are at increased risk of radiation-induced hypothyroidism and thyroiditis because of incidental irradiation of the thyroid gland [48]. If radiation-induced thyroiditis mimics metastasis on US, CNB is recommended, particularly when FNAB yields inconclusive results.
Thyroid nodules containing macrocalcifications (punctate echogenic foci >1 mm) may have a malignancy rate as high as 35.3%, necessitating biopsy when the nodule meets biopsy criteria [49,50]. In addition, entirely calcified nodules are associated with a low to intermediate malignancy risk (11.4%–16.1%) [35]. According to the updated K-TIRADS, entirely calcified nodules are classified as intermediate suspicion, and biopsy is recommended when the nodule size meets the criteria [51]. In such cases, a spring-activated CNB device offers an advantage because it provides greater thrust than FNAB or semi-automatic CNB devices, facilitating penetration through dense calcification. In a study comparing the diagnostic efficacy of FNAB and CNB in entirely calcified nodules, CNB outperformed FNAB, yielding fewer nondiagnostic results (7.7% vs. 53.8%) and fewer inconclusive results (15.4% vs. 57.7%) [35]. Similarly, another study reported superior performance of CNB over FNAB in nodules with macrocalcification, with a lower nondiagnostic rate (3.7% vs. 33.3%) [50].
Sudden neck bulging and swelling are an uncommon clinical presentation, for which underlying causes range from de Quervain thyroiditis to rare but aggressive entities such as thyroid lymphoma and anaplastic carcinoma. A meta-analysis of 17 studies reported that CNB is recommended over FNAB in patients with suspected lymphoma or anaplastic carcinoma [52]. Pooled estimates showed that CNB achieved a sensitivity of 94.3% and a specificity of 100% for diagnosing thyroid lymphoma and a sensitivity of 80.1% and a specificity of 100% for diagnosing anaplastic cancer, thereby reducing the need for unnecessary excisional biopsy. Similarly, Ha et al. [53] supported CNB as a first-line modality, reporting that it outperformed FNAB for diagnosing thyroid lymphoma and anaplastic cancer, with higher sensitivity (87.5% vs. 50.6%).
Presented with US features such as scattered parenchymal punctate echogenic foci, the diffuse sclerosing variant of papillary thyroid carcinoma is a rare, aggressive subtype compared with classic papillary thyroid carcinoma [54]. In a study by Yim et al. [54], CNB demonstrated higher sensitivity (96.9% vs. 81.4%) and negative predictive value (90.0% vs. 57.9%) than FNAB, while both modalities showed statistically comparable diagnostic accuracy. Given the sensitivity of 96.9% for differentiating malignancy in the setting of diffuse parenchymal punctate echogenic foci, CNB may serve as a first-line modality in this clinical context to reduce false-negative results.
When CNB is discussed with patients during the consent process, three concerns are commonly cost-effectiveness, pain, and potential complications. CNB is sometimes perceived as unnecessary, with greater discomfort and higher complication risk than FNAB, particularly among patients who previously received nondiagnostic FNAB results or had painful FNAB experiences. Nevertheless, understanding the diagnostic accuracy of CNB—particularly when performed by an experienced operator—is important, as this benefit outweighs the low risk of major complications. This section addresses three key patient perspectives on CNB: cost-effectiveness, tolerability, and complications.
Trimboli et al. [55] reported in 2015 that CNB of nodules previously classified as indeterminate on FNAB prevented unnecessary diagnostic thyroidectomies and substantially reduced hospital costs. In that study, a theoretical cohort of patients with indeterminate nodules was analyzed for cost-effectiveness. If all patients had undergone diagnostic thyroidectomy, the estimated cost would have been approximately 636,400 euros. By reducing unnecessary surgery in patients with benign pathology, CNB reduced expenses by 215,000 euros, representing cost savings of nearly one-third. Furthermore, repeated FNAB may increase overall patient costs, including transportation expenses and income loss due to hospital visits [56,57].
There is no strong recent evidence or comparative study evaluating the cost-effectiveness of CNB by thyroid nodule characteristics or patient demographics, or directly comparing CNB and FNAB cost-effectiveness. Nevertheless, repeated FNAB has up to 31% of patients still receive Bethesda category III results after prior nondiagnostic or Bethesda III outcomes, and a third FNAB is often less diagnostic for nodules with repeated nondiagnostic results [20,58]. Another study reported that a first repeat FNAB after an initially nondiagnostic result yielded a diagnosis in 58% of cases, whereas a second repeat FNAB achieved a 50% diagnostic rate [43]. Considering the diagnostic performance of CNB as a complementary modality, CNB may offer more favorable cost-effectiveness than FNAB because it reduces the likelihood of repeated FNAB or diagnostic lobectomy [59]. Moreover, implementing CNB as a selective first-line biopsy modality may lower overall health care costs and improve procedural efficiency, particularly for nodules prone to nondiagnostic results.
Concerns about discomfort and pain may be barriers to accepting CNB; however, when performed appropriately, CNB has not been associated with greater pain or discomfort than FNAB. Jeong et al. [60] compared tolerability and pain in 100 consecutive patients undergoing FNAB with a 23-G needle and 100 consecutive patients undergoing CNB with an 18-G needle using a 10-cm visual analog scale. The study reported comparable median pain scores for both procedures, with high tolerability (100% for FNAB and 97% for CNB). Notably, CNB performed without lidocaine injection did not increase procedure-related pain; 70% of patients underwent CNB without local lidocaine injection.
When performed by an experienced operator, FNAB and CNB can be completed quickly and accurately. A study of 212 consecutive CNB procedures reported a median procedure time—defined as the interval between skin puncture and device withdrawal—of less than 2 minutes per nodule (approximately 102 seconds), with a 100% technical success rate [44]. In that study, CNB showed consistent time efficiency regardless of nodule size, with nondiagnostic results in <1% of patients and no major complications [44]. Therefore, high tolerability and short procedure time support CNB as an ideal selective first-line biopsy modality in outpatient settings.
Complications associated with CNB can be classified as minor or major, depending on severity and the need for medical intervention. Major complications include active bleeding caused by puncture of major vessels that necessitates unexpected hospitalization or surgical intervention. Minor complications are transient, self-limiting adverse events (e.g., hematoma) that require only conservative management [61]. Contrary to the conventional perception that CNB is difficult to manage and associated with more severe complications than FNAB, the overall complication rate of CNB is low [62]. Retrospective studies have reported extremely low major complication rates, whereas the most frequent minor complications are vascular injury–related hematomas and hemorrhages that typically resolve with manual compression without additional hemostatic procedures (Fig. 10) [62,63]. Specifically, large retrospective series reported low minor complication rates (0.79%–2.2%) and extremely low major complication rates (0%–0.09%) [9,62,63]. In addition, no patients with minor or major complications experienced serious morbidity or irreversible disability among 6,687 CNB procedures [62]. Patient-related factors (age and sex), nodule-related factors (location and size), and procedure-related factors (number of attempts and needle cutting length) were not associated with CNB-related complications [62]. A 2018 meta-analysis also reported that CNB is safe regardless of nodule size, with a pooled major complication proportion of <0.1% and a pooled minor complication proportion of 1.08% [64].
For safe procedures, adherence to guidelines and operator experience are important [21,22]. Prevention of hemorrhage is also a critical component of safe CNB. In patients receiving antithrombotic therapy, discontinuation is generally recommended, and decisions should be made after a careful benefit–risk assessment in consultation with the treating physician. For example, discontinuation of aspirin and clopidogrel bisulfate is typically recommended 7–10 days before biopsy [21]. Preprocedure vessel mapping should be performed meticulously to evaluate the location of critical vessels, including the common carotid artery and the superior and inferior thyroid arteries, as well as any anatomic variants that may compromise CNB safety [21]. After the procedure, manual compression for at least 20 minutes should be performed routinely, even if there are no definite signs of hemorrhage. When complications occur, most are minor and can be managed conservatively.
Rare but serious complications, such as iatrogenic pseudoaneurysm caused by CNB, are classified as major because of the potential to compress the upper airway and exacerbate hematoma formation. Noninvasive approaches such as manual compression can induce thrombosis; however, manual compression may be ineffective, particularly for larger or deeply located pseudoaneurysms. Minimally invasive approaches have emerged as effective options for managing iatrogenic pseudoaneurysms. Ham et al. [65] reported that pseudoaneurysms can be treated safely and effectively with US-guided bovine-derived thrombin injection, with a high technical success rate. Jun et al. [66] reported a promising effect of radiofrequency ablation for treating iatrogenic pseudoaneurysm. In that study, a short ablation time (mean, 13.5 seconds) was sufficient to induce thrombosis and facilitate closure of the pseudoaneurysm neck.
Although CNB has numerous advantages over FNAB, several important challenges and limitations remain unresolved. The first limitation is the absence of global standardization of pathologic diagnostic criteria specific to thyroid CNB, despite publication of pathology reporting guidelines by the Korean Endocrine Pathology CNB Study Group [22]. Interobserver variability—particularly in the diagnosis of follicular-pattern lesions, interpretation of architectural or nuclear atypia, and determination of the presence or absence of a tumor capsule—continues to present diagnostic challenges [39,67]. The second limitation is the lack of global standardization of CNB techniques. The third limitation is the absence of clearly established essential and appropriate indications for using CNB as a first-line modality, although the 2016 Consensus Statement and Recommendations from the Korean Society of Thyroid Radiology recommend first-line CNB for patients with suspected rare malignancies such as thyroid lymphoma and anaplastic thyroid carcinoma [21]. In addition, because technical success depends on operator experience, the quality of retrieved CNB specimens and institutional indications for first-line CNB may vary across institutions and clinical settings [14,45].
FNAB has long been recognized as a reliable first-line modality. Initially introduced as a complementary tool, CNB now plays a key diagnostic role because of its diagnostic performance, safety, and efficiency. Future work should include large-scale studies and standardization of advanced CNB techniques and complication management. Future CNB guidelines should also define essential and appropriate CNB indications. Finally, international efforts to standardize pathologic diagnostic criteria for CNB are necessary.