Neuroradiology/Head and Neck Imaging
Original Research
Role of Ultrasound Diagnosis in Assessing and Managing Thyroid Nodules With Inadequate Cytology
Audio Available | 
Share
OBJECTIVE. This study aimed to assess the diagnostic accuracy of a specific ultrasound classification system in evaluating thyroid nodules with inadequate cytology, defined as a cytologic result that shows insufficient cellularity and does not provide useful information. Ultrasound diagnoses were made in accordance with a specially devised ultrasound classification system.
SUBJECTS AND METHODS. From January 2008 to December 2009, 1036 patients with 1289 thyroid nodules (largest diameter ≥ 5 mm) diagnosed by ultrasound and subsequent ultrasound-guided fine-needle aspiration (FNA) were enrolled in the study. Each thyroid nodule was prospectively classified on the basis of its ultrasound features by a single radiologist into one of five diagnostic categories: benign, probably benign, borderline, possibly malignant, or malignant. Solid nodules were classified using all five categories, whereas partially cystic thyroid nodules were classified using four (borderline was omitted). Repeated ultrasound-guided FNA was performed on all nodules for which the initial ultrasound-guided FNA revealed inadequate cytology.
RESULTS. Of 96 nodules with inadequate cytology (96/1289, 7.4%), 22 were surgically removed, and the ultrasound diagnoses and cytopathology results of all 96 were compared. A borderline ultrasound diagnosis was assigned to 13 nodules. The sensitivity, specificity, positive and negative predictive values, and accuracy were calculated for all nodules excluding all the borderline nodules (100%, 95.1%, 33.3%, 100%, and 95.2%, respectively) and with four borderline nodules reclassified as benign (100%, 95.3%, 33.3%, 100%, and 95.4%, respectively). The values obtained with these approaches were not significantly different (p > 0.05).
CONCLUSION. The selective use of repeated ultrasound-guided FNA for nodules with initial inadequate cytology may be preferable to its unconditional use depending on the ultrasound diagnoses of thyroid nodules.
Keywords: diagnosis, fine-needle aspiration, inadequate cytology, thyroid nodule, ultrasound
Ultrasound-guided fine-needle aspiration (FNA) is an easy-to-use and accurate tool for evaluating thyroid nodules and is the first method of choice for distinguishing between malignancy and benignancy in nodular thyroid disease. Many investigators have reported that ultrasound-guided FNA has high diagnostic adequacy and efficacy when used to assess thyroid nodules [1–4]. However, the incidence of inadequate cytology—defined as a cytologic result that shows insufficient cellularity and does not provide useful information—varies greatly among studies [1–11]. Inadequate cytology is an important limitation of ultrasound-guided FNA in assessing thyroid nodules. To avoid delays in the diagnosis of thyroid malignancy and unnecessary surgery of benign lesions, the American Thyroid Association recommends repeated FNA if cytology from the initial FNA is inadequate [12]. In general, patients whose ultrasound-guided FNA results for thyroid nodules exhibit inadequate cytology mostly undergo the procedure again regardless of the ultrasound features of the nodule. However, the risk of malignancy in thyroid nodules with inadequate cytology has variably been reported to be approximately 2% to 51% [5–11].
In several studies, investigators have reported the ultrasound features affecting the nondiagnostic cytology of FNA [9, 13–15]. To our knowledge, however, no study has investigated the feasibility of ultrasound diagnosis in predicting malignancy and determining the necessity of repeated FNA for the nodules exhibiting inadequate cytology on the basis of the use of a specific ultrasound classification system. In the current study, we aimed to assess the feasibility and role of ultrasound diagnosis in predicting the malignancy of thyroid nodules showing initial inadequate cytology using a specific ultrasound classification system.
| Subjects and Methods |   |
|---|
From January 2008 to December 2009, a consecutive series of patients underwent thyroid ultrasound and ultrasound-guided FNA at our hospital. Of these, 1036 patients (876 women and 160 men; mean age, 49.0 ± 12.0 [SD] years) who had undergone ultrasound-guided FNA for one or more thyroid nodules that were 5 mm or greater in the largest diameter were enrolled in this study. We obtained informed written consent from all the patients before performing ultrasound-guided FNA. This study was approved by the institutional review board of our hospital.
Thyroid ultrasound was performed by a radiologist with 8 years of relevant experience using a high-resolution ultrasound instrument (iU 22, Philips Healthcare) equipped with a 12-5–MHz linear probe. We used two different categoric systems for classifying solid thyroid nodules (defined as purely solid or as predominantly solid with a cystic component accounting for < 10% of the total volume) or partially cystic thyroid nodules (defined as thyroid nodules with a cystic component accounting for ≥ 10% of the total volume). On the basis of ultrasound features, solid thyroid nodules were prospectively classified into one of five categories: benign, probably benign, borderline, possibly malignant, or malignant. Partially cystic thyroid nodules were grouped into one of the following four categories according to their ultrasound features: benign, probably benign, possibly malignant, or malignant.
For solid thyroid nodules, the ultrasound features deemed to indicate benignancy included an ovoid or flat shape, isoechogenicity, a smooth margin, and peripheral vascularity. The ultrasound features of thyroid nodules that are still indeterminate for benignancy or malignancy, termed “borderline features” in this study, included hypoechogenicity; centrally predominant vascularity; and macrocalcifications, such as eggshell calcifications and intranodular macrocalcifications. Solid thyroid nodules diagnosed as malignant were characterized by marked hypoechogenicity, a spiculated margin, microcalcifications, taller-than-wide shape, and associated lymphadenopathy with intranodal cystic components or microcalcifications in the perithyroidal region.
For partially cystic thyroid nodules, the ultrasound characteristics of benignancy included a configuration that was either concentric or eccentric with a blunt angle, a smooth free margin, peripheral or no vascularity, a spongiform appearance or daughter cysts in the solid component, an intranodular comet-tail artifact formed by colloidal crystals, and isoechogenicity. The ultrasound features of a malignant partially cystic thyroid nodule included an eccentric configuration with an acute angle, microcalcifications, macrolobulation or irregularity of the free margin, perinodular infiltration, a centripetal vascularity in the pedicle, and associated lymphadenopathy with intranodal cystic components or microcalcifications in the perithyroidal region. The criteria for ultrasound features that we used in this study are based on previous studies [16–22].
The criteria for ultrasound diagnosis of thyroid nodules differed on the basis of the type of nodule (solid thyroid nodule or partially cystic thyroid nodule) (Fig. 1). For solid thyroid nodules, those with three or more ultrasound features of benignancy and no malignant or borderline ultrasound features were considered benign. Solid thyroid nodules with one or two ultrasound features of benignancy and no malignant or borderline ultrasound features were considered to be probably benign. Those with one or more borderline ultrasound features and no ultrasound features of malignancy regardless of benign ultrasound features were considered to be borderline. Those with one ultrasound feature of malignancy regardless of borderline or benign ultrasound features were considered to be possibly malignant. Solid thyroid nodules with two or more ultrasound features of malignancy regardless of borderline or benign ultrasound features were considered malignant.
The criteria underlying the ultrasound diagnosis of partially cystic thyroid nodules were as follows: partially cystic thyroid nodules with three or more ultrasound features of benignancy and no features of malignancy were considered benign. Partially cystic thyroid nodules with one or two ultrasound features of benignancy and no features of malignancy were considered to be probably benign. Partially cystic thyroid nodules with one ultrasound feature of malignancy regardless of other benign features were considered to be possibly malignant. Partially cystic thyroid nodules with two or more ultrasound features of malignancy regardless of other benign features were considered malignant.
Ultrasound-guided FNA was performed immediately after thyroid ultrasound by the same radiologist. All 1036 patients underwent ultrasound-guided FNA; they had 1289 nodules (nodule size range, 0.5–9.8 cm; mean size, 1.5 cm). For each sample, a smear was prepared on 4–6 slides, fixed in 95% ethanol, and sent to the department of pathology for Papanicolaou staining. In case of partially cystic thyroid nodule, the remaining aspirate in the syringe was sent for cell analysis.
The cytologic analysis was categorized as follows: inadequate, benign, indeterminate, follicular neoplasm, suspicious for malignancy, or positive for malignancy. The cytologic results were deemed unsatisfactory when fewer than six clusters of thyroid follicular cells containing no identifiable colloid were observed in a preparation. Benign cytology included nodular goiter, nodular goiter with hyperplastic nodules, colloid nodules, cyst contents with or without benign follicular cells, and lymphocytic thyroiditis. Indeterminate cytology was equivalent to the specimens with atypical cells or follicular cells of undetermined significance. Cellular specimens with abundant follicular cells arranged in a microfollicular pattern with little or no colloid or cellular specimens with a predominant population of Hürthle cells were reported as follicular neoplasms. Specimens were considered suspicious for malignancy if they showed features of a malignant neoplasm that were quantitatively or qualitatively insufficient to make a definite diagnosis of malignancy. Specimens showing abundant cells with malignant cytologic features were assigned positive for malignancy.
Thyroid nodules that were diagnosed as benign or probably benign were classified as negative (benign), whereas those diagnosed as malignant or possibly malignant were classified as positive (malignant). Separate diagnostic indexes were calculated for thyroid nodules that were diagnosed as borderline, which were excluded, included in the negative category, or included in the positive category.
We calculated the sensitivity, specificity, positive and negative predictive values, and accuracy of ultrasound-based diagnoses using pathologic results as the reference standard and evaluated the results using the McNemar test. A p value of less than 0.05 was considered to be statistically significant. Data analyses were performed using statistics software (SPSS, version 17.0.1, SPSS) for Microsoft Windows.
| Results | |
|---|
For the 1289 thyroid nodules in 1036 patients (mean number of nodules used for ultrasound-guided FNA, 1.24), the incidence of adequate and inadequate samplings was 92.6% (1193/1289) and 7.4% (96/1289), respectively. Of the 96 thyroid nodules with inadequate cytology (82 from women and 13 from men; nodule size range, 0.5–8.2 cm; mean size, 1.6 cm), 58 were solid nodules and 38 were partially cystic thyroid nodules. Twenty-two of these nodules (18 from women and four from men; nodule size range, 0.53–8.2 cm; mean size, 1.6 cm) were surgically removed for the following reasons: ultrasound diagnosis of suspected malignancy (n = 1) (Figs. 2A and 2B), pathologic confirmation (n = 2), accompanied thyroid malignancy in the same or opposite lobe (n = 14), coexistence of indeterminate nodule in the same lobe (n = 1) (Figs. 3A and 3B), and request for removal of nodule by the patient (n = 4). Of the 22 surgical nodules, there were two papillary thyroid carcinomas, three follicular adenomas, and 17 nodular hyperplasias. The ultrasound diagnoses and pathologic results of the 22 surgical nodules with initial inadequate cytology are summarized in Table 1.
 View larger version (21K) | Fig. 1 —Flowchart shows classification system used in this study for ultrasound diagnosis of thyroid nodules. |
Surgical confirmation was not performed for 74 nodules because of a benign ultrasound diagnosis despite inadequate cytology in the repeated ultrasound-guided FNA (n = 42), benign cytology with repeated ultrasound-guided FNA (n = 23), indeterminate cytology in the repeated ultrasound-guided FNA (n = 1), sonographic confirmation beyond inadequate cytology in repeat ultrasound-guided FNA (n = 3; fibrotic regression after spontaneous collapse of benign cystic thyroid nodule [n = 2] and pseudonodule related to thyroiditis [n = 1]), nodules with borderline ultrasound diagnosis and inadequate cytology in the repeated ultrasound-guided FNA (n = 4), or patient refusal to undergo thyroid surgery despite the indeterminate cytology in the repeated ultrasound-guided FNA (n = 1).
All 74 nonsurgical nodules with initial inadequate cytology underwent repeated ultrasound-guided FNA; the cytologic results of the 42 nonsurgical solid nodules included benign (n = 21), indeterminate (n = 1), and inadequate (n = 20) and those of the 32 nonsurgical partially cystic thyroid nodules included benign (n = 2) and inadequate (n = 30). The ultrasound diagnoses and cytologic results in the repeated ultrasound-guided FNA of the 74 nonsurgical nodules are listed in Table 2. Of these nonsurgical nodules, 68 did not show any suspicious features or a significant increase in nodular size (≥ 50% increase in the nodule volume) during follow-up ultrasound, and no malignant cytology was noted in the repeated ultrasound-guided FNA.
The last ultrasound follow-up examination for all nonsurgical nodules was conducted more than 12 months after the repeated ultrasound-guided FNA. One nodule with indeterminate cytology in the repeated ultrasound-guided FNA showed borderline category in ultrasound diagnosis, but this nodule was not surgically confirmed. Three nodules with suspicious ultrasound features exhibited inadequate cytology in the repeated ultrasound-guided FNA, and they were finally diagnosed as benign lesions on the basis of ultrasound follow-up results (e.g., a disappearance of the nodule or a significant decrease in the nodule volume on follow-up ultrasound), including fibrotic regression after spontaneous collapse of a benign cystic thyroid nodule (n = 2) and a pseudonodule related to thyroiditis (n = 1).
 View larger version (195K) | Fig. 2A—42-year-old woman with solid nodule in right lobe classified as malignant on thyroid ultrasound (true-positive). A, Transverse gray-scale (A) and longitudinal color Doppler (B) ultrasound images show marked hypoechogenicity, spiculated margin, taller-than-wide shape, microcalcifications, and some vascular signal for nodule (≈ 1.0 × 1.1 × 1.8 cm). Cytology results from two consecutive ultrasound-guided fine-needle aspiration biopsies were inadequate, but nodule was surgically removed because of suspicious ultrasound diagnosis. Nodule was pathologically confirmed to be papillary thyroid carcinoma. |
 View larger version (138K) | Fig. 2B—42-year-old woman with solid nodule in right lobe classified as malignant on thyroid ultrasound (true-positive). B, Transverse gray-scale (A) and longitudinal color Doppler (B) ultrasound images show marked hypoechogenicity, spiculated margin, taller-than-wide shape, microcalcifications, and some vascular signal for nodule (≈ 1.0 × 1.1 × 1.8 cm). Cytology results from two consecutive ultrasound-guided fine-needle aspiration biopsies were inadequate, but nodule was surgically removed because of suspicious ultrasound diagnosis. Nodule was pathologically confirmed to be papillary thyroid carcinoma. |
 View larger version (173K) | Fig. 3A—46-year-old woman with solid nodule in left upper lobe classified as possibly malignant in ultrasound diagnosis (false-positive). A, Transverse (A) and longitudinal (B) gray-scale images show marked hypoechogenicity, relatively smooth margin, round shape, and intranodular macrocalcification for nodule (arrow, ≈ 0.7 × 0.8 × 0.8 cm). Cytology results from two consecutive ultrasound-guided fine-needle aspiration biopsies were inadequate, but nodule was surgically removed because of coexistence of nodule of indeterminate cytology result for sample from left lobe. Nodule was pathologically confirmed to be nodular hyperplasia. |
 View larger version (179K) | Fig. 3B—46-year-old woman with solid nodule in left upper lobe classified as possibly malignant in ultrasound diagnosis (false-positive). B, Transverse (A) and longitudinal (B) gray-scale images show marked hypoechogenicity, relatively smooth margin, round shape, and intranodular macrocalcification for nodule (arrow, ≈ 0.7 × 0.8 × 0.8 cm). Cytology results from two consecutive ultrasound-guided fine-needle aspiration biopsies were inadequate, but nodule was surgically removed because of coexistence of nodule of indeterminate cytology result for sample from left lobe. Nodule was pathologically confirmed to be nodular hyperplasia. |
TABLE 1: Ultrasound Diagnoses and Pathologic Results in 22 Surgical Nodules With Inadequate Cytology
Of 74 nonsurgical nodules, 68 were classified as benign in the calculation of the diagnostic index of ultrasound diagnosis on the basis of only their benign features in follow-up ultrasound and no malignant cytology in the repeated ultrasound-guided FNA. When 13 nodules that were diagnosed as borderline were excluded from the analysis, the sensitivity, specificity, positive and negative predictive values, and accuracy of ultrasound diagnoses in differentiating between malignancy and benignancy were 100%, 95.1%, 33.3%, 100%, and 95.2%, respectively. When four nonsurgical nodules with the borderline ultrasound diagnosis and benign cytology result in the repeated ultrasound-guided FNA (i.e., four nonsurgical borderline nodules) were reclassified as benign or malignant, the sensitivity, specificity, positive and negative predictive values, and accuracy of ultrasound diagnoses were 100% and 100%, 95.3% and 90.6%, 33.3% and 20%, 100% and 100%, and 95.4% and 90.8%, respectively.
When 74 nonsurgical nodules and three surgical borderline nodules were excluded, the sensitivity, specificity, positive and negative predictive values, and accuracy of ultrasound diagnoses in differentiating between malignancy and benignancy were 100%, 94.1%, 66.7%, 100%, and 94.7%, respectively (Table 3). The diagnostic efficacies of thyroid ultrasound obtained were considerably high under the following three conditions: when four nonsurgical borderline nodules were excluded, when the same nodules were reclassified as benign, and when all the nonsurgical nodules and three surgical borderline nodules were excluded (p = 0.125, p = 0.125, p = 1, respectively, McNemar test). However, a significantly low diagnostic efficacy of thyroid ultrasound was obtained when four nonsurgical borderline nodules were classified as malignant (p = 0.008, McNemar test).
TABLE 2: Ultrasound Diagnoses and Cytologic Results for Repeated Ultrasound-Guided Fine-Needle Aspiration of 74 Nonsurgical Nodules With Inadequate Cytology
| Discussion | |
|---|
Ultrasound-guided FNA is the best choice for the diagnostic evaluation of nodular thyroid disease because it is a simple, safe, accurate, and cost-effective technique. The use of ultrasound-guided FNA to evaluate thyroid nodules has improved the detection rates for thyroid cancer and decreased the number of benign thyroid surgeries performed [1–4]. However, there are several limitations of ultrasound-guided FNA used to diagnose thyroid nodules. The major limitation of ultrasound-guided FNA is the possibility of inadequate cytology. The incidence of inadequate cytology in the repeated ultrasound-guided FNA has been reported to be relatively high, ranging from 25% to 38% [8–11].
TABLE 3: Diagnostic Indexes of Ultrasound Diagnoses for Thyroid Nodules With Inadequate Cytology When Cytopathologic Results and Long-Term Follow-Up Ultrasound Are Used as Reference Standard
There are several possible causes of inadequate FNA for thyroid nodules including technical error, palpation-guided FNA, cystic changes in the nodule, position of the nodule within the thyroid, and small nodular size [2, 3, 9]. In our study, the inadequate sampling rates in the initial and repeated ultrasound-guided FNAs were 7.4% (96/1289) and 67.6% (50/74), respectively. However, we used a different criterion of nodule size for ultrasound-guided FNA (largest diameter ≥ 5 mm) unlike the American Thyroid Association guidelines (largest diameter ≥ 10 mm), and this criterion is recommended by the Korean Society of Thyroid Radiology [16].
To overcome the limitations of ultrasound-guided FNA of thyroid nodules, some investigators have suggested the use of core needle biopsy [23, 24]. Recently, Lee et al. [5] recommended judicious application of repeated ultrasound-guided FNA for the management of thyroid nodules with initial inadequate FNA. In our study, the incidence rate of inadequate cytology in the repeated ultrasound-guided FNA was 67.6% (50/74), which is much higher than that of inadequate cytology in the initial ultrasound-guided FNA (7.4%, 96/1289). This discrepancy can be explained by the high percentage of partially cystic thyroid nodules (43.2%, 32/74), seven eggshell calcified nodules, and two fibrotic nodules after spontaneous collapse of benign cystic thyroid nodules in the repeated ultrasound-guided FNA. Various studies have reported the risk of malignancy in thyroid nodules with inadequate cytology to be approximately 2% to 51% [5–11], but no possible cause of false-negative FNA of thyroid malignancy has been explained to date. Recently, Jung et al. [15] reported that no histopathologic component was found to correlate with the inadequate results of FNA in papillary thyroid carcinomas; however, two histopathologic characteristics—uneven distribution and fibrosis—were correlated with the ultrasound findings.
The results of our study showed a low incidence of malignancy in thyroid nodules with initial inadequate ultrasound-guided FNA (3.1%, 3/96) and a high incidence of inadequate cytology in the repeated ultrasound-guided FNA (67.6%, 50/74). Therefore, the results of our study indicate that it is not justified to unconditionally perform repeated ultrasound-guided FNA of nodules with inadequate cytology from the initial ultrasound-guided FNA.
High-resolution thyroid ultrasound is regarded as the most useful diagnostic tool for evaluating nodular thyroid disease. In many studies, investigators have shown that thyroid ultrasound is a feasible method for predicting malignancy in thyroid nodules [17–22], but controversy still exists about the potential role of ultrasound in the differentiation of benign thyroid nodules from malignant ones. In particular, Kim et al. [17] have emphasized that thyroid ultrasound could be helpful in the differentiation of benign nodules from malignant ones. Recently, Kwak et al. [14] suggested that repeated ultrasound-guided FNA should be performed for thyroid nodules with suspicious ultrasound features when discrepancies exist between sonographic and cytologic findings. In the current study, the ultrasound diagnosis for the nodules having inadequate cytology in the initial FNA was excellent, although three benign thyroid nodules showed a false-positive result in the ultrasound diagnosis and 10 borderline nodules were included. In addition, we attempted to use a different ultrasound classification system according to solid thyroid nodules or partially cystic thyroid nodules, and we diagnosed thyroid nodules in accordance with a five- and four-category scheme for solid nodules and partially cystic thyroid nodules, respectively. Therefore, our ultrasound-based selection criteria for repeated ultrasound-guided FNA for thyroid nodules with initial inadequate cytology are as follows: When ultrasound diagnosis for a thyroid nodule is benign or probably benign, repeated ultrasound-guided FNA may be unnecessary; when ultrasound diagnosis for a thyroid nodule is borderline, repeated ultrasound-guided FNA may be considered; and when ultrasound diagnosis for a thyroid nodule is possibly malignant or malignant, repeated ultrasound-guided FNA should be performed.
Ultrasound classification schemes for thyroid nodules are diverse, but most reports include only two or three categories [17–22]. In this study, the ultrasound diagnoses for thyroid nodules with inadequate cytology were highly accurate except for four false-positive cases. We used the present ultrasound classification schemes for solid nodules and partially cystic thyroid nodules because of their accuracy and the relative ease with which nodules can be assigned to one of the categories [25, 26]. Nevertheless, the important limitation of our ultrasound classification for solid nodules is the presence of the borderline category. Several studies investigating eggshell-calcified nodules have reported that interrupted calcification or thick, hypoechoic rims can indicate thyroid malignancy [27, 28]. However, because those studies had small sample size and involved a single center, further large-scale and multicenter studies are required to confirm these reports.
In our study, 13 nodules with borderline ultrasound category were considered to be benign on the basis of negative results obtained from two consecutive ultrasound-guided FNAs and follow-up ultrasound over 12 months. Thus, we think that thyroid nodules showing borderline category on thyroid ultrasound and inadequate cytology at ultrasound-guided FNA are more likely to be benign nodules.
There are several limitations to the current study. First, the diagnosis of 74 thyroid nodules with inadequate cytology in the initial ultrasound-guided FNA was not surgically confirmed (74/96, 77.1%). Second, we did not investigate the efficacy of ultrasound diagnosis separately for solid nodules and for partially cystic thyroid nodules, and we could not subdivide partially cystic thyroid nodules into subclasses according to their cystic percentage because of the relatively small number of cases. Third, three cytopathologists interpreted the FNA slides, which may have resulted in variable cytologic diagnoses for individual cases. Finally, only one radiologist performed thyroid ultrasound and ultrasound-guided FNA for all the cases. The present data do not represent the diagnostic accuracy of ultrasound for inadequate cytology nodules in ultrasound-guided FNA performed by general operators; therefore, further studies are required.
In conclusion, we recommend the selective use of repeated ultrasound-guided FNA for thyroid nodules with inadequate cytology in the initial ultrasound-guided FNA because of the low incidence of malignancy, high incidence of inadequate cytology in the repeated ultrasound-guided FNA, and the high diagnostic accuracy of thyroid ultrasound with a specific ultrasound classification system.
This work was supported by the Busan Paik Hospital Imaging Research Institute.
This work was supported by the Busan Paik Hospital Imaging Research Institute.
| References | |
|---|
