PDF
ePub
Citation
Print
INTRODUCTION
Ultrasonography (US) guided-fine-needle aspiration (FNA) has been used as the primary diagnostic tool for thyroid nodules, owing to its safety, simplicity, cost-effectiveness, and reliability.1 Since early 2010, several articles have demonstrated the safety and tolerability of core-needle biopsy (CNB) for the evaluation of thyroid nodules, as well as its diagnostic efficacy.2345678 Unlike FNA, which has been validated as a standard diagnostic method for over half a century, the value of CNB has not yet been fully evaluated for diagnosing thyroid nodules.91011
The main limitation of FNA is nondiagnostic and indeterminate results, corresponding to categories I, III, and IV of the Bethesda System for Reporting Thyroid Cytopathology (BSRTC).12 To overcome this weakness, some investigators have suggested CNB as alternative or complementary diagnostic tool for thyroid nodules,5813141516 even though CNB does not show superior diagnostic performance to FNA for diagnosing thyroid cancer and poses a higher rate of post-biopsy hematoma than FNA.117 Many favorable results for CNB have been found in studies conducted by experienced radiologists and CNB has even been used to diagnose small nodules,58151618 even micronodules not indicated for FNA in prestigious guidelines.1920 Considering the thickness of the core needle, the potential risks of CNB might be greater than the expected benefits in patients with thyroid nodules, especially in those with small nodules, even though CNB shows better diagnostic accuracy than FNA.
The purpose of our study was to compare the diagnostic performances of FNA and CNB performed simultaneously at outside clinics for thyroid nodules according to nodule size.
MATERIALS AND METHODS
Our Yonsei University Institutional Review Board approved this retrospective study (IRB No. 4-2019-0745), and the requirement for informed consent was waived for review of images and medical records.
Study population
Between July 2012 and May 2019, patients who underwent both FNA and CNB for a thyroid nodule from one of 71 outside clinics were included. Among these patients, 320 patients who underwent surgery in our institution were finally included in our study. The median age of patients was 42 years (range, 19–74 years), and the median size of nodules was 10 mm (range, 3–62 mm). There were 250 female patients and 70 male patients (Table 1).
Staging US and cytopathology reports
Preoperative staging US was performed by one of 33 radiologists with 1–22 years of experience in thyroid imaging. The radiologists prospectively recorded the size and US features of thyroid nodules that had undergone FNA or CNB at outside clinics into our institutional database. The recorded US features were composition, echogenicity, margin, calcifications, and shape.21
In our institution, all cytopathologic slides from outside clinics were reviewed by one of the 15 cytopathologists in our institution according to the review schedule before surgery. Cytology reports were based on the six categories of the BSRTC.12 The CNB pathologic reports were categorized into the same six categories of the Bethesda System according to a recently proposed guideline by a radiologist (K.J.Y.), because the diagnostic categories of CNB have yet to be standardized for thyroid nodules (Fig. 1).22
Data and statistical analysis
Surgical pathologic diagnosis was defined as the standard reference. One nodule was interpreted as “malignancy” on FNA and “suspicious for malignancy (SUSP)” on CNB, although no residual cancer was noted in the surgical specimen. In this study, the nodule was considered as a cancer because central lymph node metastases were seen on the pathologic samples. We compared demographics between benign and malignant nodules, using the chi-square or Fisher's exact test for categorical variables and the Mann-Whitney U test for continuous variables.
To calculate the diagnostic performances of both FNA and CNB, we defined various combinations of test-negatives and test-positives according to the Bethesda categories to assess malignancy and designated these combinations as Criteria 1 to 5 (Table 2).4814162324 In Criteria 1, “benign” was considered as test-negative, and “SUSP” and “malignancy” were considered as test-positive. In Criteria 2, “benign” was considered as test-negative, and “follicular neoplasm or suspicious for follicular neoplasm (FN/SFN),” “SUSP,” and “malignancy” were considered as test-positive. In Criteria 3, “nondiagnostic,” “benign,” and “atypia of undetermined significance or follicular lesion of undetermined significance (AUS/FLUS)” were considered as test-negative, and the remaining categories were considered as test-positive. In Criteria 4, “SUSP” and “malignancy” were considered as test-positive, and the remaining categories were considered as test-negative. In Criteria 5, “malignancy” was considered as test-positive, and the remaining categories were considered as test-negative. We calculated the sensitivites, specificities, accuracies, positive predictive values (PPVs), negative predictive values (NPVs), and areas under the receiver operating characteristic curve (AUCs) along with 95% confidence intervals (CIs) of FNA and CNB to diagnose thyroid malignancy. These variables were compared using the generalized estimated equation and the AUCs using the Delong method. All diagnostic performances of FNA and CNB were calculated and compared separately for all nodules, nodules ≥10 mm, and nodules ≥20 mm, respectively. All analyses were performed with SPSS version 25 (IBM Corp., Armonk, NY, USA) and SAS (version 9.4, SAS Institute Inc., Cary, NC, USA). A p value of less than 0.05 was considered to indicate a significant difference.
RESULTS
Baseline clinicopathological characteristics
There were 279 malignant nodules in 279 patients and 41 benign nodules in 41 patients (Tables 1 and 3). The median age (42 years, range; 19–74 years) of the patients with malignant nodules was not significantly different from that (39 years, range; 22–68 years) of patients with benign nodules. The median size of benign nodules was 23 mm, which was significantly larger than that of malignant nodules (10 mm) (p<0.001). There was no statistical association between malignancy risk and gender. Significant differences were evident between benign and malignant nodules for the FNA and CNB results (p<0.001).
The proportions of the “benign” and “SUSP” categories were higher in FNA than in CNB (p=0.049 and p<0.001, respectively). In contrast, the proportions of the “FN/SFN” and “malignancy” categories were higher in CNB than in FNA (p<0.001).
Diagnostic performances of FNA and CNB
Table 4 demonstrates the diagnostic performances of FNA and CNB according to nodule size using Criteria 1–5.
When applying Criteria 1 and 2, diagnostic performance mostly did not differ between FNA and CNB in all nodules, except for NPV, which was better for FNA FNA (66.7%; 95% CI: 44.9–88.4 for Criteria 1 and 48.8%; 95% CI: 33.5–64.1 for Criteria 2) than for CNB (12.5%; 95% CI: 0–35.4 for Criteria 1 and 11.5%; 95% CI: 0–23.8 for Criteria 2) (p<0.001). When applying Criteria 3, specificity and PPV were higher for FNA (92.7%; 95% CI: 84.7–100 for specificity and 98.5%; 95% CI: 96.9–100 for PPV) than for CNB (63.4%; 95% CI: 48.7–78.2 for specificity and 93.6%; 95% CI: 90.5–96.7 for PPV) (p<0.001 and p=0.003, respectively). Accuracy and NPV were not statistically different between FNA and CNB. The sensitivity (78.9%; 95% CI: 74.1–83.6) of CNB was higher than that (72%; 95% CI: 66.8–77.3) of FNA (p=0.036). When applying Criteria 4, the diagnostic performance of FNA was similar to that of CNB. When applying Criteria 5, sensitivity, accuracy, and NPV were significantly higher for CNB than FNA.
In nodules ≥1 cm, FNA showed superior diagnostic performance over CNB in terms of NPV with Criteria 1 and 2 and specificity and PPV with Criteria 3. The other measures of diagnostic performance were not different between FNA and CNB when Criteria 1–3 were used. When Criteria 4 was applied, none of the diagnostic performance measures of FNA significantly differed from CNB. When applying Criteria 5, sensitivity, accuracy, and NPV were significantly higher for CNB than for FNA.
In nodules ≥2 cm, FNA showed superior diagnostic performance over CNB in terms of NPV with Criteria 1 and 2 and specificity and PPV with Criteria 3. The other measures of diagnostic performance were not different between FNA and CNB when Criteria 1–3 were used. When Criteria 4 and 5 were applied, the diagnostic performance of FNA did not statistically differ from that of CNB.
DISCUSSION
In this study, the diagnostic performance of FNA was mostly the same as that of CNB, regardless of nodule size, except for NPV, which was better for FNA than for CNB when applying Criteria 1 and 2. When applying Criteria 3, the specificity and PPV of FNA were superior to CNB regardless of size. When applying Criteria 4, diagnostic performance did not differ between FNA and CNB, regardless of size, unlike the results of a previous study that found CNB to have higher sensitivity and accuracy than FNA.4 After applying Criteria 5, diagnostic performance did not differ between FNA and CNB in nodules ≥2 cm (not small nodules). However, in nodules ≥1 cm and all nodules, the sensitivity, accuracy, and NPV of CNB were better than those of FNA, findings which were similar to those of a published study.8
Several investigators have suggested that CNB is superior to FNA and that it may be a potential first-line diagnostic tool.481416 In these studies, the “nondiagnostic,” “benign,” “AUS/FLUS,” and “FN/SFN” categories were regarded as test-negative, and the “nondiagnostic,” “benign,” “AUS/FLUS,” and “FN/SFN” categories with and without the “SUSP” category were regarded as test-positive, which were the same designations used for Criteria 4 and 5, respectively, in our study.481416 In other studies, the diagnostic performance of FNA was calculated by regarding the “benign” category with and without the “nondiagnostic” category as test-negative (Criteria 1 and 2 in our study, respectively) or by regarding the combination of the “nondiagnostic,” “benign,” and “AUS/FLUS” categories as test-negative (Criteria 3 in our study).2324 According to BSRTC, the risk of malignancy was 15–30% and 60–75% for thyroid nodules categorized as “FN/SFN” and “SUSP,” respectively.12 Considering that “FN/SFN” and “SUSP” nodules have relatively high risks of malignancy and are recommended for surgery according to the BSRTC, a nodule with the above cytologic categories should not be considered as test-negative when calculating the diagnostic performances of FNA or CNB (Criteria 4 and 5). Based on the results of our study, we need to reconsider previous research that found CNB to be more reliable or accurate than FNA, and CNB cannot be considered as a first-line diagnostic tool for diagnosing thyroid nodules without more supportive evidence.
To obtain tissue samples without damage to the adjacent tissues, it is important to visualize the entire needle. Therefore, the core needle should be positioned parallel to the transducer. Unlike the core needle in CNB, the fine needle in FNA can be inserted in various directions with to-and-fro motions to obtain cells from different sites of the thyroid nodule for representative sampling.25 Differences in the methods of FNA and CNB may explain their different diagnostic performances.
Percutaneous CNB is a useful diagnostic tool for diagnosing soft tissue lesions, especially breast lesions, and surgical excision biopsies can be avoided with its use.2627 CNB has also been reported as a safe tool for diagnosing thyroid nodules with low complication rates.2 However, as the core needle is thicker than the fine needle, it might cause more severe and permanent damage to the adjacent major organs than the thinner fine needle (Supplementary Figs. 1 and 2, only online).
This study has several limitations. First, patient selection bias exists because we included patients who underwent both FNA and CNB before surgery. Second, we could not evaluate the experience levels of the physicians who performed both FNA and CNB as these procedures took place at outside clinics. However, the results of this study can reflect the real clinical value of FNA and CNB. Third, the results of the first procedure may have affected the next procedure. Commonly, FNA is performed before CNB even when they are both performed during the same session/visit. Therefore, the results of CNB can be influenced by the procedures undertaken for FNA. Fourth, although our findings indicate that CNB cannot be a main diagnostic tool for thyroid nodules, it can still play a role in diagnosing thyroid lymphoma, anaplastic carcinoma, and metastasis from other organs.282930 In this study, we were unable to evaluate the selective indications of CNB. Finally, we did not assess diagnostic performance for predicting neoplasm because there were only a small number of FNs in this study.
In conclusion, CNB did not show superior diagnostic performance to FNA for diagnosing thyroid nodules.
XML Download