Abstract
Objectives
The purpose of this study was to assess the value of the apparent diffusion coefficient (ADC) in the differential diagnosis between benign and malignant thyroid nodules, particularly those found to have indeterminate cytology with fine needle aspiration (FNA).
Methods
Thirty-eight patients with 42 thyroid nodules underwent neck magnetic resonance imaging consisting of T1-, T2-, and diffusion-weighted imaging. The final diagnosis of all nodules was confirmed by surgery, revealing 23 with benign and 19 with malignant lesions. Preoperative FNA cytology was performed in 38 of 42 nodules, including 15 of indeterminate cytology. The mean ADC values in benign and malignant groups were compared.
Results
There was a significant difference between mean ADC values in benign and malignant nodules and between mean ADC in benign and malignant nodules of indeterminate cytology. A cutoff value for malignant nodules of 1.60 × 10 −3 mm 2 /s yielded sensitivity, specificity, and accuracy of 94.73%, 82.60%, and 88.09%, respectively.
Conclusion
The present study revealed that ADC measurements could potentially quantitatively differentiate between benign and malignant thyroid nodules, even those of indeterminate cytology. We propose that diffusion-weighted imaging evaluation should be used for the assessment of thyroid nodules in addition to FNA cytology.
1
Introduction
Thyroid nodules are commonly encountered in routine medical care. Almost 20% of the population have a palpable thyroid nodule, and approximately 70% have a nodule that can be detected by ultrasound . Most nodules are asymptomatic, with only 5% to 10% of nodules being malignant . The main problem after the discovery of a thyroid nodule is discriminating between benign and malignant lesions.
The management of thyroid nodules is guided by numerous factors, including clinical criteria, findings on ultrasound, and the results of fine needle aspiration (FNA) cytology . Fine needle aspiration cytology is the most widely used and cost-effective preoperative test for initial evaluation of thyroid nodules . However, FNA cannot distinguish between benign and malignant nonpapillary follicular and oxyphilic cell lesions . Traditionally, FNA cytology results are divided into 4 categories: nondiagnostic, malignant, indeterminate, and benign . Therefore, the indeterminate category includes follicular neoplasms and oxyphilic cell neoplasms. Although most thyroid nodules of follicular origin are benign, patients with indeterminate cytologic findings are typically referred to surgery for more accurate diagnosis . This presents a clinical dilemma for evaluating clinicians.
Various conventional imaging techniques such as high-resolution ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), and scintigraphy with radioactive iodine or technetium have been investigated as potential methods for overcoming the difficulties in diagnosing follicular adenoma and carcinoma . However, there is currently no alternative algorithm for more conservative management of patients with thyroid nodules of indeterminate cytology. In addition, positron emission tomography with fluorine 18-labeled fluorodeoxyglucose (FDG-PET) and FDG-PET with CT (FDG-PET/CT) have been used in an effort to distinguish benign from malignant nodules in cases of indeterminate thyroid nodules . However, because there is no clear-cut differentiation between malignant and benign thyroid nodules, FDG-PET/CT may have a role in excluding malignancy in thyroid nodules of indeterminate cytology .
Recently, diffusion-weighted magnetic resonance imaging (DWI) has been applied to the head and neck regions . Diffusion-weighted magnetic resonance imaging may be useful for characterizing head and neck tumors, and differentiating benign and malignant cases based on apparent diffusion coefficient (ADC) assessment . We hypothesize that DWI may enable us to differentiate various thyroid nodule components. This distinction, if possible, would be an important development for planning the management of thyroid nodules. Diffusion-weighted magnetic resonance imaging is a noninvasive technique that can highlight subtle abnormalities and improve the characterization of tissue and pathologic processes because it reflects the random motion of water protons in tissue . As such, the thyroid gland is a particularly interesting organ to study with DWI because of its colloid formation and colloid transport functions. To our knowledge, however, there have been few reports of ADC values in the thyroid gland and no studies using DWI to evaluate thyroid nodules found to have indeterminate cytology with FNA. The purposes of the present study were to evaluate the use of ADC in differentiating between benign and malignant thyroid nodules and to examine the relationship between ADC values and histologic architecture, especially in the management of nodules found to have indeterminate cytology with FNA.
2
Materials and methods
2.1
Subjects
The retrospective study was approved by the institutional review board of our hospital. Because of the retrospective nature of the study, informed consent was waived. Fifty-six patients with thyroid nodules who visited our institute from August 2009 to January 2011 had undergone routine neck MRI. All patients had lesions larger than 1 cm in the greatest minimal transverse diameter. Of these patients, 45 underwent thyroidectomy (total thyroidectomy in 14 patients and unilateral lobectomy in 31 patients), which enabled histopathologic examinations of the thyroid nodules. Of the 45 operated patients, 7 were excluded from the study because of previous radiation therapy in 1 patient with the medullary carcinoma, Hashimoto thyroiditis in 2 patients, pure cystic nodules in 2 patients, and bad image quality and motion artifacts in 2 patients. The remaining 38 patients (13 men and 29 women; age range, 23-79 years; mean age, 55.5 years) with 42 thyroid nodules were included in the study group. In all 42 nodules, the histopathologic diagnosis was confirmed by surgery, benign (n = 23) and malignant (n = 19) nodules. The benign nodule group consisted of adenomatous goiter (n = 16) and follicular adenoma (n = 7). The malignant nodule group consisted of papillary thyroid carcinoma (PTC) (n = 12), follicular thyroid carcinoma (FTC) (n = 4, including 1 oxyphilic cell type), medullary carcinoma (n = 2), and metastatic breast cancer (n = 1).
Preoperative FNA cytology was performed in all nodules, except for 4 nodules in the benign nodule group. When nodules were anteriorly located and easy to palpate, FNAs were performed with palpation. Otherwise, FNAs were performed via ultrasonographic guidance. We classified lesions into 4 categories according to FNA cytology: benign, malignant, indeterminate, and nondiagnostic. The results revealed 2 nodules in the malignant nodule group with nondiagnostic cytologic findings. Consequently, preoperative FNA cytology was available in 36 of the 42 nodules (19 benign nodules and 17 malignant nodules) including 15 nodules of indeterminate cytology. A thyroid nodule of PTC was diagnosed as malignant (suggestive of PTC) in 7 of 12 cases that were later confirmed to be PTC, whereas all cases of FTC were categorized as indeterminate cytology except for 1 nodule that was misdiagnosed as malignant (suggestive of PTC). The histopathologic type and preoperative FNA cytology of 42 thyroid nodules are given in Table 1 .
| Histopathologic type | Preoperative FNA cytology | |||
|---|---|---|---|---|
| M | B | I | N/A | |
| Benign (n = 23) | 0 | 11 | 8 | 4 |
| Adenomatous goiter (n = 16) | 0 | 9 | 5 | 2 |
| Follicular adenoma (n = 7) | 0 | 2 | 3 | 2 |
| Malignant (n = 19) | 10 | 0 | 7 | 2 |
| PTC (n = 12) | 7 | 0 | 4 | 1 |
| FTC (n = 4) | 1 | 0 | 3 | 0 |
| Others (n = 3) | 2 | 0 | 0 | 1 |
2.2
MRI study
2.2.1
Data acquisition
The MRI study was performed using a 1.5-T MRI unit (Achieva Nova Dual, Philips Medical Systems, Best, The Netherlands). A neck array coil was used. T2-weighted (T2w) and T1-weighted (T1w) axial images were acquired using a spin-echo sequence (TR/TE = 2500/80 milliseconds for T2w and repetition time (TR)/echo time (TE) = 570/12 milliseconds for T1w) with a field of view of 12 cm and slice thickness of 3 mm. Diffusion-weighted magnetic resonance imagings were acquired using an echo planar imaging sequence (TR/TE = 5000/60 milliseconds) with 2 b values (0 and 1000 s/mm 2 ) with a field of view of 20 cm and a slice thickness of 3.34 mm.
2.3
Data analysis
Apparent diffusion coefficient maps were then generated on the manufacturer’s magnetic resonance (MR) console. For each patient, circular regions of interest (ROIs) for normal thyroid gland and thyroid nodules were drawn by 1 neuroradiologist based on multislice T1w and T2w images. The neuroradiologist was blinded to the histopathologic diagnosis. At this time, ROIs were carefully placed to avoid necrotic, cystic changes and hemorrhagic regions. Consequently, the areas of ROIs placed in the nodules and normal areas of thyroid tissues were between 20 and 50 mm 2 according to the size of the nodules and normal thyroid glands. In thyroid nodules, 1 ROI was used when the focal lesion had a diameter less than 3 cm. Three ROIs were used when the lesion had a diameter of more than 3 cm, and the average of these 3 measurements was used. The ADC value of normal thyroid glands was measured at 1 place in the contralateral or ipsilateral thyroid gland as a reference standard. The ADC value was automatically calculated when the ROI was drawn. The difference between the ADC values of the pathologic lesions and the normal thyroid gland and the difference between the ADC values of the benign and malignant nodules were evaluated. Statistical analysis was performed using Excel and commercially available statistical software (Ekuseru-Toukei 2008; Social Survey Research Information, Co, Ltd, Tokyo, Japan). Mann-Whitney U tests were used to analyze the difference in ADC values between 2 groups. Receiver operating characteristic curves were used to determine the cutoff point with highest accuracy and sensitivity. P < .05 was considered statistically significant.
2
Materials and methods
2.1
Subjects
The retrospective study was approved by the institutional review board of our hospital. Because of the retrospective nature of the study, informed consent was waived. Fifty-six patients with thyroid nodules who visited our institute from August 2009 to January 2011 had undergone routine neck MRI. All patients had lesions larger than 1 cm in the greatest minimal transverse diameter. Of these patients, 45 underwent thyroidectomy (total thyroidectomy in 14 patients and unilateral lobectomy in 31 patients), which enabled histopathologic examinations of the thyroid nodules. Of the 45 operated patients, 7 were excluded from the study because of previous radiation therapy in 1 patient with the medullary carcinoma, Hashimoto thyroiditis in 2 patients, pure cystic nodules in 2 patients, and bad image quality and motion artifacts in 2 patients. The remaining 38 patients (13 men and 29 women; age range, 23-79 years; mean age, 55.5 years) with 42 thyroid nodules were included in the study group. In all 42 nodules, the histopathologic diagnosis was confirmed by surgery, benign (n = 23) and malignant (n = 19) nodules. The benign nodule group consisted of adenomatous goiter (n = 16) and follicular adenoma (n = 7). The malignant nodule group consisted of papillary thyroid carcinoma (PTC) (n = 12), follicular thyroid carcinoma (FTC) (n = 4, including 1 oxyphilic cell type), medullary carcinoma (n = 2), and metastatic breast cancer (n = 1).
Preoperative FNA cytology was performed in all nodules, except for 4 nodules in the benign nodule group. When nodules were anteriorly located and easy to palpate, FNAs were performed with palpation. Otherwise, FNAs were performed via ultrasonographic guidance. We classified lesions into 4 categories according to FNA cytology: benign, malignant, indeterminate, and nondiagnostic. The results revealed 2 nodules in the malignant nodule group with nondiagnostic cytologic findings. Consequently, preoperative FNA cytology was available in 36 of the 42 nodules (19 benign nodules and 17 malignant nodules) including 15 nodules of indeterminate cytology. A thyroid nodule of PTC was diagnosed as malignant (suggestive of PTC) in 7 of 12 cases that were later confirmed to be PTC, whereas all cases of FTC were categorized as indeterminate cytology except for 1 nodule that was misdiagnosed as malignant (suggestive of PTC). The histopathologic type and preoperative FNA cytology of 42 thyroid nodules are given in Table 1 .
| Histopathologic type | Preoperative FNA cytology | |||
|---|---|---|---|---|
| M | B | I | N/A | |
| Benign (n = 23) | 0 | 11 | 8 | 4 |
| Adenomatous goiter (n = 16) | 0 | 9 | 5 | 2 |
| Follicular adenoma (n = 7) | 0 | 2 | 3 | 2 |
| Malignant (n = 19) | 10 | 0 | 7 | 2 |
| PTC (n = 12) | 7 | 0 | 4 | 1 |
| FTC (n = 4) | 1 | 0 | 3 | 0 |
| Others (n = 3) | 2 | 0 | 0 | 1 |
2.2
MRI study
2.2.1
Data acquisition
The MRI study was performed using a 1.5-T MRI unit (Achieva Nova Dual, Philips Medical Systems, Best, The Netherlands). A neck array coil was used. T2-weighted (T2w) and T1-weighted (T1w) axial images were acquired using a spin-echo sequence (TR/TE = 2500/80 milliseconds for T2w and repetition time (TR)/echo time (TE) = 570/12 milliseconds for T1w) with a field of view of 12 cm and slice thickness of 3 mm. Diffusion-weighted magnetic resonance imagings were acquired using an echo planar imaging sequence (TR/TE = 5000/60 milliseconds) with 2 b values (0 and 1000 s/mm 2 ) with a field of view of 20 cm and a slice thickness of 3.34 mm.
2.3
Data analysis
Apparent diffusion coefficient maps were then generated on the manufacturer’s magnetic resonance (MR) console. For each patient, circular regions of interest (ROIs) for normal thyroid gland and thyroid nodules were drawn by 1 neuroradiologist based on multislice T1w and T2w images. The neuroradiologist was blinded to the histopathologic diagnosis. At this time, ROIs were carefully placed to avoid necrotic, cystic changes and hemorrhagic regions. Consequently, the areas of ROIs placed in the nodules and normal areas of thyroid tissues were between 20 and 50 mm 2 according to the size of the nodules and normal thyroid glands. In thyroid nodules, 1 ROI was used when the focal lesion had a diameter less than 3 cm. Three ROIs were used when the lesion had a diameter of more than 3 cm, and the average of these 3 measurements was used. The ADC value of normal thyroid glands was measured at 1 place in the contralateral or ipsilateral thyroid gland as a reference standard. The ADC value was automatically calculated when the ROI was drawn. The difference between the ADC values of the pathologic lesions and the normal thyroid gland and the difference between the ADC values of the benign and malignant nodules were evaluated. Statistical analysis was performed using Excel and commercially available statistical software (Ekuseru-Toukei 2008; Social Survey Research Information, Co, Ltd, Tokyo, Japan). Mann-Whitney U tests were used to analyze the difference in ADC values between 2 groups. Receiver operating characteristic curves were used to determine the cutoff point with highest accuracy and sensitivity. P < .05 was considered statistically significant.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
