With the popularization of ultrasound imaging technology in thyroid nodule screening and the application of thermal ablation technology in treating benign thyroid nodules [1,2], a more effective and accurate biopsy method, either for the assessment of the malignancy risk in thyroid nodules or for the requirement of pathological diagnosis before thermal ablation, has consistently been the focus of clinical research on thyroid nodules.

Fine-needle aspiration (FNA) is the most commonly used tool for diagnosing thyroid nodules because of its convenience, safety, and effectiveness [3–5]. However, FNA also has some limitations, such as its relatively high proportion of inaccurate or inconclusive results [6–8] and its inadequacy for the preoperative management of follicular thyroid carcinoma, lymphoma, medullary carcinoma, and other malignant tumours [9–13]. In those cases, repeat FNAs or even surgery is essential. Unfortunately, repeat FNAs also have a high rate of inconclusive results, and some operations are performed unnecessarily [14–16]. In recent years, several studies have shown that core needle biopsy (CNB) has the potential to be an effective alternative first-line diagnostic tool for thyroid nodules by an experienced operator [9,17–20]. Compared to FNA, CNB significantly reduces the proportion of inconclusive results and has better diagnostic efficacy, and CNB is not associated with any obvious complications [21–23]. However, there will still be inconclusive results in varying degrees or insufficient efficacy in diagnosing malignancies when using CNB [24–26].

Previous studies have focused on comparisons between FNA and CNB, but there are differences in methods and techniques between the two tools. Whether the differences between these tools can become complementary conditions is worth exploring. Consequently, this retrospective study was conducted to determine whether the combination of FNA and CNB is more effective than either modality alone.

Our retrospective study demonstrated that the combined FNA/CNB diagnosis provided a lower proportion of inconclusive results than either FNA or CNB alone, and the diagnostic efficiency for malignant nodules also significantly improved. Especially in the assessment of nodules with size ≤1.5 cm and ACR TI-RADS 4–5, the application of combined FNA/CNB yielded a better diagnostic performance for malignancy than that of FNA or CNB alone.

In our study, FNA had a higher proportion of inconclusive results, and this result was similar to many previous studies [33–35]. Even though a more experienced operator may yield a better specimen, the internal composition of the nodule and the sample size may still be obstacles [27,36]. There are several possible reasons why FNA had higher inconclusive results. First, FNA results are closely related to the internal composition of thyroid nodules. Calcification, cystic degeneration, and an abundant vascular supply can affect the quality of the puncture specimens. Second, the quality of the samples obtained from operators with different clinical experiences can also vary greatly. An operator who is more skilled in their puncture technique is more likely to obtain a more representative and accurate tissue sample. Third, the samples that are obtained from FNA are relatively small, so it is impossible to complete immunohistochemical tests with these types of samples. These limitations of FNA sample quality and quantity also directly affected the diagnostic efficacy of FNA for malignant nodules. In this study, the sensitivity of FNA in predicting malignancy was only 58.1%, which was disappointing.

In recent years, it has been suggested that CNB may be used as an alternative first-line diagnostic tool for thyroid nodules, especially for more experienced operators [18,24,29]. Our study showed that compared with FNA, with CNB, the proportion of inconclusive results was reduced by 21%, and the sensitivity and accuracy in predicting malignancy increased by 21.9% and 4.8%, respectively, indicating satisfactory results. For some nodules with obvious calcification or an abundant blood supply, it is difficult to obtain a sufficient number of valid specimens using FNA. When a nodule is covered with calcified tissue, it is difficult for the fine needle to penetrate the nodule to access the tissue for sampling, and when the nodule has an abundant blood supply, a large number of blood cells are aspirated, resulting in an insufficient amount of follicular cells. In those cases, using a core needle can penetrate the target nodule to obtain a larger tissue sample, which not only provides more information about the internal structure and capsule of the nodules, but also provides additional immunohistochemical staining data for the differential diagnosis [37–39].

With reference to previous studies [18,28,29], and combined with the final malignant probability of the combined diagnosis, criteria for the combined diagnosis were formulated. The results showed that the combined diagnosis yielded a lower proportion of inconclusive results and a higher malignancy diagnosis efficiency than either FNA or CNB alone, suggesting that FNA and CNB can complement each other in terms of technology and methodology. This finding may be the result of the different methods and quantities of samples for FNA and CNB. First, the sampling methods were different. FNA was used to aspirate the target lesions directly, while CNB obtained specimens by reserving a certain forward stroke distance. Second, the specimens that were obtained via FNA versus CNB were different sizes. CNB can obtain a larger amount of tissue and more tissue samples than FNA, which could provide more information about the histological characteristics of the nodule, including the nodular capsule and the related cell nuclear features. Therefore, utilizing both methods to make a diagnosis combines the advantages of FNA and CNB, which contributes to the performance of the joint technique. In terms of inconclusive results, although the combined diagnosis was better than CNB alone, the reduced proportion of inconclusive results cannot be proportional to the procedural burden caused by the combined diagnosis, regardless of the nodule size. In this regard, it is suggested that CNB alone may be enough without the addition of FNA for the diagnosis of thyroid nodules, consistent with a recent study [29]. However, in terms of diagnostic performance for malignancy in thyroid nodules, the combined diagnosis shows more obvious advantages than CNB alone.

In this study, it was observed that the diagnostic specificity for malignancy reached 100%, regardless of nodule size or the diagnostic method used, which may be related to the diagnostic criteria. When the malignancy criteria are defined as diagnostic results indicative of BSRTC categories V and VI, the pathologist should identify typical malignant pathological changes before making a diagnosis. This may be the reason for the lack of false positives in the present research results, and similar findings can be observed in previous studies [27,28]. Therefore, in cases of the same specificity, the diagnostic sensitivity for malignancy is particularly important. In this study, a total of 155 malignant nodules were diagnosed, of which 154 were ACR TI-RADS 4–5. Therefore, it is obviously unnecessary to advocate all nodules be subjected to the combined procedure to provide an efficiency diagnosis of malignancy. In the subclassification of nodule size, the combined diagnosis of nodules with size less than 1.5 cm and ACR TI-RADS 4–5 showed the best performance. The results showed that the sensitivity of the combined diagnosis reached 91.4%, an increase of 10.7% over CNB alone. In this study, 48 patients received a positive malignant diagnosis by using CNB, but the patients were misdiagnosed on FNA. Fourteen patients were confirmed to have malignant nodules by FNA, but those patients were misdiagnosed on CNB. Therefore, we reviewed the ultrasound images of the misdiagnosed nodules to search for possible causes. For the majority of the nodules at risk for malignancy, using CNB can obtain a larger sample size and more nodule information, resulting in more accurate results. However, for some small nodules, it is possible to obtain inaccurate sample information when using CNB due to the partial volume effect of ultrasound. There may be risk factors, such as haemorrhage, especially for small nodules located on the side of the thyroid envelope, near the inferior thyroid artery, or near the common carotid artery. During CNB, in order to keep a safe distance when using the needle tip, it is possible that the puncture needle may hit the target, but the needle slot does not take a valid sample, leading to a misdiagnosis. In contrast, during FNA, the operator can always control the needle tip inside the nodule to obtain the sample cells. In addition, nodules with malignancy risk are mostly lacking in blood supply, and a substantial amount of sample cells can be easily obtained via FNA. In such cases, FNA is superior to CNB. Therefore, a combined diagnosis can make the two technologies complement each other and can enhance the diagnostic efficiency. Unlike the largest malignant nodules, which had already been screened and treated, many malignant nodules based on exclusion findings are generally small. It is very important to improve the diagnostic sensitivity for malignant nodules, minimizing missed diagnoses to the greatest extent and thus avoiding further metastasis and deterioration of malignant nodules. In addition, CNB has been proven to be safe as an alternative first-line tool for thyroid diagnosis by an experienced operator, the incidences of which major complications have been reported to be as low as 0% to 0.09% [18,29,40]. In other words, it seems safe for experienced operators to add FNA to CNB, which has been verified in our research. Consequently, we believe that the combination of FNA and CNB may be a more valuable method to assess relatively small thyroid nodules at risk of malignancy.

Several limitations of this study should be addressed. First, this was a single-centre retrospective study. Some patients were excluded if they did not have a combined diagnosis or did not receive a final diagnosis, which might have led to case-selection bias. Moreover, not all of the final diagnoses of the patients were determined based on the postoperative pathology, which might have caused a small amount of error. Furthermore, the diagnostic criteria for the combined FNA/CNB diagnosis have not been standardized, especially for the combinations of categories II and III, and further research is needed to confirm the reliability of this standard. Finally, the development of major complications was prevented by the high level of experience of both operators in this study. However, the low degree of serious complications may not be applicable to other operators with different levels of experience in routine practice.

In summary, combined FNA/CNB can achieve a lower proportion of inconclusive results and a more effective diagnostic performance than FNA and CNB alone. For an experienced operator, to obtain the best procedural benefits, CNB alone may be enough without the addition of FNA in thyroid nodules with a size >1.5 cm and ACR TI-RADS 4–5 or nodules with ACR TI-RADS 1–3. However, for nodules with a size ≤1.5 cm and at risk of malignancy (ACR TI-RADS 4–5), to screen thyroid malignancies to the greatest extent and provide an accurate diagnostic basis for further clinical treatment, the combined FNA/CNB technique may be a more valuable and recommended diagnostic method.