| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
|
by the American Institute of Ultrasound in Medicine J Ultrasound Med 25:1531-1537 • 0278-4297 Ultrasonographic Differentiation of Benign From Malignant Neck Lymphadenopathy in Thyroid CancerClinical Department of Nuclear Medicine and Radiation Protection, University Hospital Rebro, Zagreb, Croatia (S.K.K., I.B., G.H.H., D.D.); Department of Epidemiology, Zagreb Public Health Institute, Zagreb, Croatia (V.T.); and University Hospital "Sestre Milosrdnice," Zagreb, Croatia (K.K.). Address correspondence to Sanja Kusacic Kuna, MD, Clinical Department of Nuclear Medicine and Radiation Protection, University Hospital Rebro, Kispaticeva 12, 10 000 Zagreb, Croatia. E-mail: sanja.kusacic-kuna{at}zg.t-com.hr
Objective. The aim of this study was to determine whether ultrasonography itself was able to distinguish benign from malignant lymphadenopathy in patients with thyroid cancer. Methods. We evaluated lymph nodes in a group of patients with thyroid cancer. Nodes were detected and measured by ultrasonography, and their shape, echogenicity, size, and location were noted. Ultrasonographically guided fine-needle aspiration biopsy (FNAB) was performed, and smears were analyzed cytologically. Results. Ultrasonographically guided FNAB was performed in 578 neck nodes in a group of 631 patients with thyroid cancer. In most cases, metastases had a round shape and various echo structures, with a predomination of hypoechoic nodes without a hilum. There were statistical differences in size between metastatic and benign nodes in terms of maximum diameter, minimum diameter, and volume. Among these, minimum diameter and the shape of the nodes seemed to be the most reliable in suggesting malignancy. A round shape with a longitudinal/transverse ratio of less than 2 of hypoechoic nodes indicated the presence of metastases, and we then performed FNAB. The absence of an echogenic hilum and the presence of cystic portions and calcifications were significantly greater in malignancies than in benign lesions (P < .001). In most cases, metastatic nodules were situated in the lower third of the neck. Reactively enlarged nodes occurred more frequently in the upper part of the neck. Conclusions. Ultrasonography itself cannot distinguish benign from malignant lesions, but an echographic appearance suggests malignancy and helps in the selection of the node to aspirate with ultrasonographically guided FNAB, which is crucial for a final diagnosis.
Key Words: lymph nodes • thyroid cancer • ultrasonography Abbreviations: FNAB, fine-needle aspiration biopsy • L/T, longitudinal/transverse
Thyroid cancer, especially papillary carcinoma, metastasizes most often into cervical lymph nodes, and it is not unusual to find thyroid cancer neck metastases at the time of presentation of the disease. Early detection of metastases is of great clinical importance because it enables more successful surgery and radiation therapy treatments. Because of their anatomic position, enlarged cervical lymph nodes are not easily palpable, especially when they are small and located behind sternocleidomastoid muscles, deep in the paratracheal region, or behind a carotid artery or jugular vein. Moreover, because of their biological behavior, radioiodine does not accumulate in these nodes.1 Several studies have shown that ultrasonography has higher sensitivity than palpation for the detection of enlarged lymph nodes in patients with suspected regional lymph node metastases.2–5 Ultrasonography is sensitive in detecting superficial lymphadenopathy; it provides valuable diagnostic information with a high degree of diagnostic accuracy; and it is easily tolerated by patients and is cheaper and faster to perform than other methods. Additionally, it provides the ability to perform ultrasonographically guided fine-needle aspiration biopsy (FNAB), which informs us about the nature of the pathologic process. Because of this, ultrasonography is commonly the first imaging modality that is used. Normal cervical lymph nodes are rarely visualized by ultrasonography; however, hyperplastic lymph nodes appear in many diseases, and they are echographically visualized with high accuracy.6–9 Benign as well as malignant lymph nodes have typical ultrasonographic morphologic characteristics,10,11 and accepted echographic criteria can help in their differentiation. Most reports have been in agreement about the echographic appearance of benign lymph nodes and have described them as oval and hypoechoic, with a smooth border and a hyperechoic hilum, which represents hilar fat and small vessels. Malignant nodes are described as being more round and of a heterogeneous structure (commonly hypoechoic with a loss of the hilum, sometimes with cystic degeneration or calcifications)12–16; however, ultrasonographically guided FNAB still remains the most promising tool for confirming or excluding malignancy.17–19 This study evaluated the ultrasonographic characteristics of cervical lymph nodes in patients with thyroid carcinoma to determine a more reliable method of distinguishing benign from malignant lymphadenopathy.
During a period of 10 years, we performed ultrasonographic examinations of the neck region in 631 patients (510 women; mean age, 44.1 years; range, 6–81 years) with thyroid cancer, looking for potential metastases before surgery as well as in postoperative follow-up. This was a retrospective study of nodes detected and verified cytologically in this group of patients. Ultrasonographically guided FNAB was performed in 578 neck nodes. Nodes were measured on the screen, and their shape, echogenicity, size, and location were reported. Linear 7.5- and 10-MHz probes were used for real-time B-scans of the neck. Both sides of the neck were checked for the presence of enlarged lymph nodes. Nodes larger than 5 mm in maximum diameter were considered enlarged. The lymph nodes were divided into 3 groups according to their location: the superior third of the neck (including submandibular, submental, and high jugular regions), the mid third of the neck (mid jugular level), and the inferior third of the neck (including low jugular, midline low neck, and supraclavicular regions). We performed ultrasonographically guided FNAB of the most suggestive nodes from the tumor drainage region in one or both sides of the neck. In the case of a solitary node, we punctured it; in the case of multiple nodes, we chose 1, 2, or more nodes, depending on their echographic appearance and position in the neck. Fine-needle aspiration biopsy was performed by a free-hand technique. The transducer was positioned in the craniocaudal direction. Target nodes were reached by 0.6 x 25-mm needles for cell aspiration. Smears were stained by means of the May-Grünwald-Giemsa method. In the case of smears with insufficient material for reliable cytologic diagnosis (5.2%), we repeated the FNAB to obtain sufficient material. All cytologically verified malignant nodes were confirmed histologically after surgical removal. The size of the nodes was measured in 3 dimensions (longitudinal, anteroposterior, and transverse), and the volume was computed. All lymph nodes were scanned in the longitudinal plane, and the largest diameter was measured (maximum diameter). The diameter perpendicular to the longitudinal plane was then measured and defined as the short axis diameter (minimum diameter). The nodes were divided into 2 groups according to their shape: oval and round. All nodes were documented with a multiformat laser camera in a longitudinal nodal plane, that is, the plane that showed the maximum and minimum diameters of the node. Lymph node shape was assessed by measuring the largest and smallest diameters on the same scan and by calculating the longitudinal/transverse (L/T) ratio. Lymph nodes were separated into 2 groups according to their L/T ratio: oval (L/T ratio >2) and round (L/T ratio <2). Echographic characteristics of the nodes, such as their cystic nature, calcifications, and presence of a hilum, were also investigated.
Statistical differences were evaluated with the Mann-Whitney U test, the Kruskal-Wallis test, or the Pearson
Among the 631 patients with thyroid cancer, most had papillary carcinoma (430); others had follicular carcinoma (143), medullary carcinoma (32), a mixed type of carcinoma (20), Hurthle cell carcinoma (15), anaplastic carcinoma (5), and lymphoma (1). In our group, 578 neck nodes were aspirated by ultrasonographically guided FNAB and analyzed cytologically: hyperplastic benign lymph nodes were recorded in 296 (51.2%) (Figures 1  and 2); metastases were recorded in 221 (38.2%) (Figure 3); connective scar tissue was recorded in 21 (3.6%); and other findings such as parathyroid adenoma, seroma, salivary gland cysts, granulomatous inflammation, thymic tissue, and lymphoma were recorded in 40 (6.9%).
The maximum diameter of nodes (mean ± SD) was 1.68 ± 0.96 cm (range, 0.5–13 cm). There was a significant difference in the maximum diameter between metastatic and benign lymph nodes (Mann-Whitney U test, P < .001). Metastases had a significantly greater maximum diameter (2.04 ± 1.34 cm; range, 0.5–13 cm) than benign lymph nodes (1.51 ± 0.51 cm; range, 0.6–3.6 cm). There were significantly more malignant than benign lymph nodes with a maximum diameter of larger than 3 cm and between 2 and 3 cm (14.03% and 21.72% versus 1.35% and 11.82%; P < .001). On the contrary, the frequency of benign nodes was significantly greater in the group with a maximum diameter of 1 to 2 cm (69.93% versus 45.7%; P < .001). There was no significant difference in frequency between metastatic and benign lymph nodes with a maximum diameter of less than 1 cm (18.55% versus 16.89%; P = .62). The mean minimum diameter of observed nodes was 0.77 ± 0.76 cm (range, 0.2–12 cm). Metastatic lymph nodes had statistically greater minimum diameters than benign nodes (1.2 ± 1.2 cm; range, 0.4–12 cm; versus 0.51 ± 0.26 cm; range, 0.2–2.9 cm; P < .001). Almost all (96.96%) benign nodes and half (52.49%) of the metastatic nodes had a minimum diameter of less than 1 cm (P < .001). There were significantly more malignant than benign lymph nodes with a minimum diameter of 1 to 2 cm (37.1% versus 2.77%; P < .001). We also found that malignant lymph nodes were significantly bigger than benign lymph nodes, according to their volume (3.02 ± 6.24 cm3; range, 0.04–59.69 cm3; versus 0.49 ± 1.22 cm3; range, 0.01–19.13 cm3; P < .001).
Of the 296 benign nodes, 88.2% showed an L/T ratio of 2 or greater (oval shape), and 11.8% showed an L/T ratio of less than 2. Of the 221 malignant nodes, 65.6% showed an L/T ratio of less than 2 (round shape), whereas only 34.4% showed an L/T ratio of 2 or greater. The differences in the distribution of malignant and benign nodes in the two L/T ratio classes were statistically significant (P < .001, Pearson
The nodes were positioned approximately equally in 3 neck areas: 165 (31.91%) in the superior third, 162 (31.34%) in the mid third, and 190 (36.75%) in the inferior third; however, there was a significant difference in the distribution of malignant and benign lymph nodes in different neck areas. The incidence of malignant lymph nodes was significantly higher in the inferior third than in the middle and superior thirds (147 [66.52%] versus 45 [20.36%] and 29 [13.12%]; P < .001). Conversely, benign nodes were found more often in the superior third of the neck (136 [45.94%]) than in the mid third (117 [39.53%]) and inferior third (43 [14.53%]; P < .001) (Figure 4
The incidence of various ultrasonographic findings in benign and malignant nodes was different. A cystic portion was seen in 47 (21.3 %) of the metastatic lymph nodes (Figure 5 ). Calcifications were found in 17 (2.24%) nodes, of which 12 (70.59%) were metastases (Figure 6). A central echogenic hilum was visible in 110 (37.16%) of 296 benign nodes; however, in malignant nodes, a hilum was seen in only 1 (0.45%). All benign nodes were hypoechoic. Most of the metastases were hypoechoic (92.31%); hyperechogenicity in relation to the adjacent muscles was detected in 17 (7.69%) metastatic nodes. A comparison of the ultrasonographic appearances of benign and malignant lymph nodes in patients with thyroid cancer is shown in Table 1.
Normal lymph nodes are difficult to detect because of their high echogenicity, which is similar to that of the surrounding fat tissue. Both inflammatory and malignant diseases lead to enlargement of lymph nodes and to reduction of their echogenicity, whether as a consequence of an increase of intranodal fluid portion or hyperplasia of lymph follicles or tumor cell invasion.10,20 The typical ultrasonographic appearance of a reactively enlarged benign lymph node is a longitudinal or oval shape, with low echogenicity with a smooth border and a hyperechoic hilum.10,21–23 Conversely, the typical ultrasonographic findings of metastatic lymph nodes are enlargement with a round shape and a heterogeneous echogenic structure (commonly hypoechoic but with intra-nodal cystic degeneration and calcifications) with loss of the hilum.12,24 Some authors advocate using nodal size to distinguish benign from malignant lymph nodes.14,15,18,19 Our results showed statistically significant differences between benign and malignant nodes in terms of their maximum diameter, as well as in terms of minimum diameter and volume; however, absolute measurements of the size of the lymph node, such as the maximum diameter, are not sufficient for differentiation of inflammatory from metastatic lymph nodes. The smallest node with a positive FNAB result in our material was 0.8 x 0.4 x 0.7 cm (longitudinal, anteroposterior, and transverse); the largest benign node measured 3.6 x 1.3 x 2.7 cm. Although there were significantly more benign lymph nodes (69.93%) with maximum diameters of 1 to 2 cm, 45.7% of metastases were also in the same group. There were significantly more malignant than benign lymph nodes with maximum diameters of 2 to 3 and up to 3 cm. The size ranges for benign and malignant nodes overlapped; thus, differentiation of those lesions could not be based on size alone, which is concordant with the literature.3 Minimum diameter seems to be a more reliable criterion for differentiation of benign from malignant nodes.18,19 In our study, almost all (96.96%) benign nodes and only 52.49% of the metastases had a minimum diameter of less than 1 cm. In malignant cervical lymphadenopathy, there was an increase in minimum diameter. These findings were noted when the average minimum diameter was compared for the malignant and benign lymph nodes in our series (1.2 and 0.5 cm, respectively). According to other reports,3,14–16 the minimum diameter of metastatic nodes was larger than that of nonmetastatic nodes.
Other work has also shown that the L/T ratio has a high degree of accuracy in the differentiation between benign and metastatic lymph nodes.3,12,13 As mentioned previously, normal cervical lymph nodes typically have an oval shape. Even when cervical lymph nodes are enlarged because of a benign inflammatory process, they usually retain this shape, whereas malignant lymph nodes change their shape to rounded (L/T <2). This was also confirmed in our study: 66% of metastases followed this pattern, whereas 88.2% of benign lymph nodes retained an oval shape (L/T In addition to these morphologic criteria, the position of nodes in the neck also seems to be important in predicting possible malignancy. Some authors have stressed the importance of selection of the node for aspiration19 and have recommended that nodes situated in mid jugular, low jugular, and supraclavicular areas should be considered highly suggestive of metastases. Reactively enlarged nodes occur more frequently in submental and submandibular as well as in subdigastric and high jugular regions. In our series, 66.52% of the metastatic nodules in patients with thyroid carcinoma were situated in the lower third of the neck (including low jugular, midline low neck, and supraclavicular neck regions); 20.36% were found in the mid third, and only 13.12% were found in the superior third of the neck. Benign lymph nodes were found more often in the superior (45.94%) and middle (39.53%) third of the neck (including high jugular, mid jugular, submandibular, and submental regions). Internal architecture can also be useful in the differential diagnosis of cervical lymph nodes. It is thought that a hyperechoic hilum is a good indicator of a benign lymph node3; however, the absence of a hilum is more frequently a sign of malignancy.3,15 In our series, an echogenic hilum was observed in only 1 malignant node but in 37.16% of benign nodes. The same results have been shown by others.3,13,15,16,23,24 In several studies, cystic portions within a cervical lymph node were considered highly suggestive of metastatic papillary thyroid carcinoma.15,16,26–30 We detected cystic portions in 21.3% of metastatic nodes, and all of them were metastases of papillary carcinoma. There was no cystic degeneration among benign lymph nodes in our study; a cysticlike pattern was seen only in seromas. The presence of calcifications can also be useful in predicting the benign or malignant nature of lymph nodes. Most authors have reported calcifications only in malignant nodes.16,24,28 In our material, calcifications were seen mostly in malignant lymph nodes (12 of 17 nodes with calcifications were metastases, and the remaining 5 were detected in postoperative scar tissue). We did not observe any calcifications in reactively enlarged benign lymph nodes. We conclude that detailed ultrasonographic examination of the neck before or after surgery to treat thyroid cancer is compulsory because it gives information about the extent of the disease and enables adequate planning for a therapeutic approach. Ultrasonography itself cannot exactly distinguish benign from malignant lesions, but an echographic appearance can suggest malignancy and help in the selection of the correct nodes to aspirate with ultrasonographically guided FNAB, which is crucial for a final diagnosis. In sampling, nodes with a rounded appearance should be chosen primarily; a round shape with an L/T ratio of less than 2 in hypoechoic nodes without a hyperechoic hilum indicates the presence of metastases. In this study, the absence of an echogenic hilum and the presence of cystic portions and calcifications were significantly greater in malignancies than in benign lesions. The decision to use ultrasonographically guided FNAB as well as the choice of nodes to be punctured should be in concordance with these morphologic criteria of size, shape, location, and internal architecture.
Received May 23, 2006, from the Clinical Department of Nuclear Medicine and Radiation Protection, University Hospital Rebro, Zagreb, Croatia (S.K.K., I.B., G.H.H., D.D.); Department of Epidemiology, Zagreb Public Health Institute, Zagreb, Croatia (V.T.); and University Hospital "Sestre Milosrdnice," Zagreb, Croatia (K.K.). Revision requested June 19, 2006. Revised manuscript accepted for publication July 12, 2006.
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Top
Abstract
Introduction
2 exact test, and a P < .05 was considered to indicate statistical significance. 
2). The tendency of benign nodes to be oval and malignant nodes to be round has also been reported by others.
