Enlarged Benign-Appearing Cervical Lymph Nodes by Ultrasonography Are Associated with Increased Likelihood of Cancer Somewhere Within the Thyroid in Patients Undergoing Thyroid Nodule Evaluation

Abstract

Background:

Benign-appearing cervical lymph nodes (CLN) are easy to assess during an ultrasonography (US) evaluation for a guided fine-needle aspiration biopsy of a suspicious thyroid nodule, but their clinical significance regarding thyroid cancer risk is not known. Non-malignant-appearing nodes may be an indicator of early malignancy in the thyroid. We hypothesize that there is an increased prediction of thyroid cancer when benign-appearing enlarged CLN (ECLN) > 1 cm in any dimension are present during an US evaluation of thyroid nodules.

Method:

A review of 269 consecutive patients' charts sent for thyroid nodule assessment that underwent thyroidectomy was conducted to compare ECLN, with the presence of thyroid cancer during an ultrasound-guided fine-needle aspiration biopsy of the thyroid nodule. Surgical excision pathology confirmed all abnormal cytology reports.

Results:

From the final 265 charts reviewed, 213 had benign thyroid pathology and 52 had thyroid cancer. Sex, number, and size of the biggest thyroid nodule were not different between groups. Patients with cancer were on average 10 years younger and had higher thyroid-stimulating hormone (TSH) values (p < 0.003) as well as a 10-fold increase in enlarged non-malignant-appearing lymph nodes than their peers without cancer. The presence of ECLN had an 82% sensitivity, 90% specificity, and a 68% positive predictive value for thyroid cancer. There was also an 80% negative predictive value when enlarged lymph nodes were not present. In 8 of the 37 (21.6%) patients with malignancy and ECLN, the primary dominant thyroid nodule was negative on cytologic evaluation, but malignancies were confirmed on surgical specimen, in contralateral nodules on the same side as the ECLN. These nodules were mostly subcentimeric, ranging from 0.2 to 1.14 cm and were not biopsied due to their inconspicuous appearance. After multiple logistic regression analysis, enlarged lymph nodes had a 53.8 odds ratio for cancer (20.49–141.33, p < 0.01).

Conclusion:

Discovering the presence of ECLN in routine assessment of thyroid nodules is an easy and fast surveillance technique that increases the predictive value in diagnosing thyroid cancer, especially when the enlarged lymph nodes are on the same side as the thyroid nodule.

Introduction

B enign-appearing cervical lymph nodes (CLN) are easy to assess during an ultrasonography (US) evaluation for a guided fine-needle aspiration (FNA) biopsy of a suspicious thyroid nodule, but their clinical significance regarding thyroid cancer risk is not known. There is an abundance of data on defining malignant lymph nodes with recommendations for FNA with thyroglobulin (Tg) needle wash for diagnosis of metastasis. There are no published studies on the presence of non-malignant-appearing nodes and thyroid cancer malignancy. This raises the question whether non-malignant-appearing nodes may be an indicator of early malignancy in the thyroid.

We specifically examined, in a retrospective review, what the reported lymph node status was before their thyroidectomy to determine the malignancy rate when benign-appearing enlarged CLN (ECLN) > 1 cm in any dimension are present during an US evaluation of thyroid nodules. The same physician performed all thyroid biopsies. All procedures and pathology reviews were performed at the Thyroid Clinic of the Texas Diabetes Institute, University of Texas Health Science Center, between August 2006 and March 2008.

Thyroid nodules are a common finding in the patient population with a prevalence of 3%–8% by palpation and 20%–76% by thyroid US with a 13% risk of thyroid malignancy (1 –3). Thyroid US enables physicians to consider the malignant potential of the nodules based on their observed characteristics (3). The use of guided FNA biopsy is the standard technique in establishing malignancy (1,4,5). The goal of this technique is to determine whether the nodules are benign or malignant, thereby enabling patients with thyroid cancer to receive appropriate treatment at an early stage with reduced morbidity and mortality. In addition, early detection of benign nodules reduces unnecessary tests and procedures. Thyroid nodules are classified based on features that determine which nodules may be malignant and which may be benign (6 –8). The American Association of Clinical Endocrinologists guidelines state that guided FNA biopsy should be performed on all hypoechoic nodules ≥10 mm with one or more of the following characteristics: irregular margins, chaotic intranodular vascular spots, a taller than wider shape, or microcalcifications (2).

Lymphatic drainage is a well-known route for malignant cells to metastasize to other parts of the body. When thyroid cancers metastasize, they typically metastasize to the local CLN. Therefore, it would not be surprising if malignant cells were found in the associated CLN and that these lymph nodes would react to the presence of these cells (9 –11).

Since detection of normal lymph nodes is common on US (12) and thyroid cancer often metastasizes to CLN, one must differentiate between benign, reactive, and malignant CLN. Normal lymph nodes are typically flat and oval shaped. When viewed with Doppler US, they contain an echogenic hilus and a hilar vessel surrounded by a capsule with a hypoechoic cortex (13). Malignant lymph nodes may show heterogeneity, abnormal shape, internal calcifications, cystic degeneration, and erratic vascularity (1). Specifically, CLN in the submandibular region are generally round with a tendency to enlarge with advancing age, believed to be due to fat deposition (14 –16).

Taking into account the recommendations of the American Association of Clinical Endocrinologists and considering the variable sensitivity in US in predicting malignancy (1), discovering a simple factor that would increase the diagnostic predictive value for thyroid cancer would be a beneficial tool. The aim of this study was to investigate benign-appearing CLN at the time of initial evaluation of thyroid nodules, as a tool to increase both the sensitivity and specificity in determining thyroid malignancy risk.

Materials and Methods

This was a retrospective cross-sectional study with charts examined from all of the patients referred to the Thyroid Clinic of the Texas Diabetes Institute for initial thyroid nodule evaluation between August 2006 and March 2008. The University of Texas Health Science Center at San Antonio Institutional Review Board approved the study, and every patient signed an informed consent before nodule evaluation by ultrasound-guided FNA biopsy. Exclusion criteria included patients with a family history of thyroid cancer, history of radiation to the head or neck, and TSH values below the lower limit of normal or insufficient material from the guided FNA biopsy.

All of the patients had real-time thyroid/neck US performed by the same physician (K.E.H.) using a Biosound Esaote MyLab50 with a 7.5–12-MHz linear array transducer with color Doppler capability. The patients were placed in the supine position with a pillow under the upper back for maximal neck extension. Following standard thyroid US procedures, before guided FNA biopsy, the superficial CLN were evaluated. Size, shape, nodal architecture, color, and power Doppler features (to assess the presence and distribution of intranodal vessels) were recorded. The probe was used first in the transverse plane with the probe placed adjacent and parallel to the clavicle and systematically swept superiorly through the submandibular salivary glands. The screen image contained the most lateral portion of the thyroid, the common carotid artery and internal jugular vein, and imaging levels IV, III, II, and I. A second lateral scan with the probe placed with the internal jugular vein in the medial or inside part of the screen allowed a view of the lateral tissue, again sweeping superiorly noting levels Va and b. In addition, a third scan along the longitudinal plane was examined viewing levels II through V.

The US reports were dictated the same day and routinely included the number of thyroid nodules and their features (location, shape, size, echogenicity, heterogeneity, vascularity, margin clarity, microcalcifications, and/or reverberation artifacts). Those thyroid nodules considered suspicious for malignancy had the following features: increased vascularity, irregular or ill-defined borders, presence of microcalcifications, and being taller than their width. Specimens of the suspicious nodules were obtained using a 25-gauge needle attached to a 20-mL disposable syringe. Typically, four samples were collected and immediately placed onto glass slides. One half of the samples were air-dried, and the other half were placed in 95% alcohol for Papanicolaou staining by the cytology lab.

The same cytopathologist (L.J.F.) evaluated all specimens and was unaware of the patient details other than thyroid nodule size and TSH value. Specimens were classified as (i) benign, including lymphocytic thyroiditis (LT), (ii) follicular lesion (FL), (iii) follicular neoplasm, or (iv) suspicious for cancer. Surgical excision pathology confirmed all suspicious reports obtained by cytopathology.

The US reports also routinely comment on all benign or suspicious CLN. CLN were classified as benign if they had S:L <0.5, fatty hilus, polar vessel, and were <1 cm in any dimension, below level II. CLN were considered reactive if they had S:L <0.5, fatty hilus, polar vessel, and were >1 cm in any dimension, below level II. These were noted on reports as “present, benign in appearance, but significance unknown.” Suspicious nodes had S:L >0.5 and contained at least one malignant feature previously described. These suspicious nodes (six patients) were biopsied and underwent Tg wash. Level II nodes were excluded from this evaluation, since it is common for them to exceed 1 cm in size and have a more normal rounded shape (17). All of the patients also provided a routine antecubital venous blood sample that was assayed for TSH.

For statistical analyses, two groups were evaluated: those patients with nodule malignancy determined by cytopathology with surgical excision and those patients without malignancy. The benign (nonmalignant) and malignant groups' age and the size of the largest thyroid nodule found by US were compared using an unpaired t-test. Variables reported as proportions (sex, thyroid nodules found by US, and CLN) were compared using a Pearson's χ ². The TSH values had a symmetric non-normal distribution, so they were analyzed using the Kolmogorov–Smirnov test. Initially, simple logistic regression was performed using enlarged benign-appearing lymph nodes as a predictive variable and the cancer diagnosis as an outcome variable. Following this test, a multiple logistic regression was used to adjust for age, sex, TSH values, number of thyroid nodules, and size of the largest thyroid nodule.

The sensitivity, specificity, and predictive values for cancer diagnosis were calculated using the data obtained from each of the patient's reports.

Results

During the study period, 269 consecutive thyroidectomized patients' charts were reviewed for CLN features and were compared with the biopsy results obtained with guided FNA biopsy. Four charts were excluded from the final analysis due to suppressed TSH (one), positive family history of thyroid cancer (one), and for insufficient material obtained by biopsy (two). Of the final 265 charts reviewed, 213 reports had a benign diagnosis, whereas 52 had a malignancy confirmed by surgical pathology (∼20% of the total). Patients found to have cancer were on average 10 years younger and had two times the TSH levels (p < 0.003), consistent with prior reports of higher TSH values in malignancies (18).

Enlarged benign-appearing lymph nodes were observed about 10 times more frequently in the cancer group than in the benign group. Both groups had similar sex proportions, with women composing 89% and 90% of the benign and cancer patients, respectively. The number of total tumors detected by US and the size of the largest detectable tumor were not different between groups. Demographic data and principal variables are summarized in Table 1.

Table 1.

Demographic Data, Thyroid-Stimulating Hormone Levels, and Thyroid Nodule Characteristics in Patients with Benign Thyroid Nodules or Malignant Nodules

Demographics (264)	Benign (213)	Cancer (51)	p-Value
Age in years (range)	53 (17–83)	45 (21–78)	<0.001
M (%)/F (%)	25 (11)/188 (89)	5 (9)/46 (91)	0.69
TSH mIU/L (range)	1.76 (0.01–20.67)	3.66 (0.03–58)	0.003
Enlarged cervical lymph nodes			<0.001
Present (%)	20 (9)	42 (82)
Not present (%)	193 (91)	9 (18)
Number of thyroid nodules per patient
1 (%)	113 (53)	31 (60)	0.2
2 (%)	66 (30)	13 (25)
3 (%)	18 (8)	6 (11)
4 (%)	10 (4)	0 (0)
5 (%)	6 (2)	1 (1)
Biggest thyroid nodule by US in cm (range)	2.32 (2.15–2.49)	2.54 (2.09–2.99)	0.28

TSH, thyroid-stimulating hormone; US, ultrasonography.

Of the 213 benign diagnoses, 20 had enlarged, but benign by criteria, lymph nodes observed during the US evaluation (false-positives) (9.4%). From these 20 patients, nodule cytopathology reports were benign in 6, with 8 having a diagnosis of LT, 5 having a diagnosis of FL, and 1 having a diagnosis of papillary thyroid cancer (PTC) (found to be benign on surgical excision). Following surgery, three of the five patients with FL (diagnosed on cytopathology) were actually LT on surgical pathology (hemithyroidectomy), and the remaining two patients were benign.

In the 52 malignant cases, only 6 had suspiciously palpable CLN and malignant criteria on lymph node surveillance. The final pathology in these six patients included one with lymphoma, three with invasive stage IVc PTC, and two with stage III PTC (Table 2). These patient lymph nodes were biopsied, underwent Tg wash, and were clearly positive on pathology (6/6) and Tg wash (5/6, one lymphoma).

Table 2.

Distribution of Diagnoses in the Benign and Thyroid Cancer Groups

	CLN present	CLN absent
Benign cytopathology	20	193
	8 lymphocytic thyroiditis	158 benign
	5 follicular lesion	25 lymphocytic thyroiditis
	3 lymphocytic thyroiditis	10 follicular lesion
	2 benign	All benign on surgery
	1 PTC
	Benign on surgery
	6 benign cytopathology
Malignancy on surgery report	43	9
	6 palpated positive	3 No comment on CLN
	1 lymphoma	3 FVPTC
	3 invasive PTC IVc	3 PTC
	2 stage III PTC
	37 benign but >1 cm in size
	1 follicular thyroid cancer
	2 FVPTC
	34 PTC
	8 stage II PTC or higher

Data presented as number of patients.

CLN, cervical lymph node; PTC, papillary thyroid cancer; FVPTC, follicular variant of papillary thyroid cancer.

In 9 of the 52 patients, no enlarged benign-appearing CLN were identified (false-negatives) (17%). In six of these nine patients, three had a follicular variant of PTC and three had stage I PTC. In the remaining three patients, there was no comment on lymph node status in their reports, but the data are included here for completeness sake.

Nonpalpable, enlarged benign-appearing lymph nodes (observed by US) were found in the remaining 37 patients with malignancies on surgical excision. All of the lymph nodes had a hilus, polar vascularity and S:L <0.5. None of the CLN had cystic degeneration, heterogeneity, peripheral vascularity, or microcalcifications.

When a simple logistic regression was used, the presence of enlarged benign-appearing lymph nodes (detected by US before aspiration biopsy) proved to be a significant predictor of malignancy (CLN coefficient = 3.80; standard error 0.43; X ² [df = 1] of 107.58; p < 0.001; odds ratio 45; 95% confidence interval = 19–105). After adjusting for all variables measured, the multiple logistic regression model maintained enlarged benign-appearing lymph nodes as a statistically significant predictor of malignancy (coefficient = 3.98; standard error 0.49; X ² [df = 1] of 8.09; p < 0.001; odds ratio 53.82; 95% confidence interval = 20.49–141.33). Age (53 [51–55] in the benign group vs. 45 [41–49] in the cancer group, p < 0.001) and TSH values (1.76 [1.49–2.03] in the benign group vs. 3.66 [1.26–6.05] in the cancer group, p = 0.003) were also significant predictors of malignancy. All of the logistic regression assumptions were met, and no colinearity or correlation was found between the values. To validate the statistical model used, the Jackknife procedure (19) was conducted; after 260 replications, only age and enlarged benign-appearing lymph nodes remained statistically significant. On the basis of the cytology and surgical excision reports, enlarged benign-appearing CLN as a diagnostic tool had an 82% sensitivity, a 90% specificity, a 68% positive predictive value, and an 80% negative predictive value.

Discussion

During evaluation of thyroid nodules by guided FNA biopsy, the presence of enlarged benign-appearing lymph nodes increased the sensitivity for discovering thyroid malignancy significantly better than thyroid nodule characteristics alone. This evaluation takes little time, helps determine which nodules to target for biopsy when multiple nodules exist, and provides information for the surgeon on surgical approach.

Of the 265 consecutive patients referred for thyroid nodule biopsy, there were 52 confirmed malignancies (51 PTCs and 1 lymphoma), and 43 of the 52 patients had evidence of lymph node size >1 cm in any dimension (81%). Six of these 43 patients had palpable CLN that had malignant features on US. In these six patients, the diagnosis was confirmed by FNA and Tg wash of the CNL. These patients underwent appropriate neck dissection at the time of initial surgery with five displaying advanced PTC and one displaying lymphoma (Table 2).

The more interesting cases are the 37 patients with confirmed malignancy who had enlarged benign-appearing CLN (observed by US). In 5 of these 37 patients, a malignancy was discovered in the nondominant thyroid nodule (ranging from 0.82 to 1.14 cm). These nodules were on the contralateral side from the dominant thyroid nodule requested for biopsy, but were on the ipsilateral side of the enlarged lymph node. In all five cases, the requested dominant nodule was benign. In three other cases, a microcancer (0.2, 0.4, and 0.7 cm) was discovered incidentally while the patients had thyroidectomy for other reasons (i.e., compressive symptoms or large benign thyroid nodule). All three of these patients were observed to have ipsilateral enlarged but benign-appearing lymph nodes by US on the same side as the microcancer with a report stating “significance unknown.” Therefore, in 8 cases from the 37 patients (21.6%), the primary dominant thyroid nodule was negative on cytologic evaluation, but malignancies were confirmed in other nodules on the same side as the enlarged benign-appearing lymph nodes.

In these eight patients, the smaller malignancies would have been missed if only the largest nodule had been biopsied. Kim and his colleagues have found that diagnosing small thyroid cancers is important because thyroid carcinomas with a maximum diameter of <1 cm may manifest with early lymph node metastasis or extra nodal invasion (20). It is, however, still controversial if mortality benefit is improved in finding subcentimeric tumors, since clinically significant disease may never have developed. Even so, in patients who have the subcentimeric microcancers surgically removed, they may (i) reduce the use and need for I 131, further reducing morbidity, (ii) increase survival in an already indolent cancer, and (iii) reduce the risk of metastatic disease.

In this review, there were no consistent US criteria in the thyroid nodules that helped predict which nodule would be malignant. There was no increased vascularity, irregular borders, presence of microcalcifications, or size differences, but, in each case, they were found solely on the observation of enlarged benign-appearing lymph nodes on the same side below level II. This feature alone modified the prevalence of thyroid cancer from 16.04% (44/265) to 19.03% (52/265).

Many authors propose that guided FNA biopsy is best limited to nodules with high-risk US characteristics (2,3). Unfortunately, as seen above, some thyroid cancers do not present with typical US characteristics (20,21) and it is not known if a delayed diagnosis without aspiration biopsy increases the risk for cancer. In addition, the thyroid nodule to target remains controversial (22,23).

Using tools that increase the ability to predict malignancy more accurately, such as adding lymph node assessment to all thyroid nodule assessments, may allow for better diagnosis of thyroid malignancies. Since absence of enlarged benign-appearing lymph nodes had a predictive value, avoidance of unnecessary aspiration and surgery may ensue and thyroid nodules can be followed without immediately initiating diagnostic investigation. This may allow a more conservative approach, since absence of enlarged lymph nodes co-notates a low risk of malignancy. Since the majority of recurrences involve the CLN, US has proven to be the most sensitive imaging technique for detecting recurrence. Therefore, becoming competent in detection of lymph nodes by US is a useful tool, both before guided FNA biopsy of nodules and in postoperative neck surveillance for recurrence. In addition, these data provide stronger support for the American Thyroid Association guidelines (3) recommending preoperative cervical US in all patients undergoing thyroidectomy. This may potentially alter the surgical approach in many patients where routine central neck dissection is not performed.

One limitation of the present study is its retrospective design that does not allow affirmation of causation. It cannot be concluded that a better diagnosis of thyroid cancer would be achieved based solely on enlarged lymph node identification without malignant features, although a study using it alone while disregarding other items, like thyroid nodule features and demographic data, would not be ethically feasible. Since March 2008, a prospective study has examined the enlarged lymph node/thyroid cancer relationship and to date continues to show a positive correlation with >70% of thyroid malignancies containing ECLN.

A second limitation of the study is that enlarged lymph nodes cannot be considered pathognomonic for thyroid cancer; moreover, it is a common and transient finding in other situations like infections or aging. Since the enlarged benign-appearing lymph nodes were not biopsied, it is not known if they were infiltrated by cancer or merely reactive.

At the time of surgery, in a center where central neck dissections (level VI) are not routinely performed, the surgeon was informed of the lymph node status and central neck dissection was performed in 12 of the 37 patients with enlarged benign-appearing lymph nodes due to suspicious central compartment disease. No lateral neck dissections were performed in any of the 37 patients with ECLN. In 4 cases of these 12 patients with enlarged benign-appearing nodes, there was metastatic disease in the central compartment and those patients received I 131 ablation. All of these patients had stage I disease. Those patients without metastatic disease or without dissection did not receive I 131 ablation, initially, since only high-dose I 131 is used for remnant ablation in this institution. Of the 25 patients who did not undergo central neck dissection, all but 5 have had regression of the enlarged lymph nodes on 6 month and 1 year follow-up neck surveillance US and maintain an undetectable Tg on suppression. Since they did not receive I 131 and did not have any nodes removed at the time of surgery, it could be assumed that the enlarged lymph nodes had undergone a localized reactive process. One year postoperatively, five patients had persistent Tg positivity. US reassessment of CLN showed persistence of enlarged but non-malignant-appearing LN. All underwent FNA of the LN with Tg wash and were all positive for disease. Four have undergone a second surgery with conversion to undetectable Tg with suppression; the fifth has opted for close observation and no surgery or radioactive iodine treatment.

At present, ECLN (below level II) are certainly as useful as other factors (microcalcifications, irregular shape, hypervascularity, and hypoechoic echotexture) in identifying thyroid nodules for biopsy. They are, however, not diagnostic of thyroid cancer until future studies present more evidence to clarify their clinical significance regarding thyroid cancer risk.

Conclusion

In this study of 265 patients evaluated for thyroid nodules, those patients with ECLN provided an 82% sensitivity and 90% specificity for malignancy, significantly higher than the thyroid nodule characteristics that are currently used. Routine surveillance with US of bilateral jugular chain lymph nodes takes an additional 1–2 minutes and provides valuable information to the physician concerning the presence of malignancy. In cases of multiple thyroid nodules, the nearby-enlarged lymph nodes can also help direct aspiration biopsy to the nodule most likely to carry malignancy as opposed to the largest nodule, especially in the absence of other suspicious nodule characteristics. This study clearly demonstrates that observing seemingly benign ECLN during a thyroid US yields important information about the risk of malignancy. As part of the routine evaluation, practitioners specializing in neck US should evaluate, recognize, and comment on all CLN presence as a criteria in evaluating thyroid nodules. Presence of nodes without malignant features may point to an underlying thyroid malignancy. All lymph nodes with suspicious features should undergo FNA cytology with Tg wash before surgery. If a thyroid nodule is suspicious on FNA, biopsy and Tg wash of an adjacent enlarged non-malignant-appearing LN should strongly be considered.

Footnotes

Acknowledgments

The authors will like to acknowledge Dr. Thomas H. Champney, Dr. Jerry Stock, Dr. Joaquin Joya-Galeana, and Dr. Alejandro Sosa-Caballero for reviewing and commenting on the article.

Disclosure Statement

The authors declare that no competing financial interests exist.

References

Frates

, Benson

, Charboneau

, Cibas

, Clark

, Coleman

, Cronan

, Doubilet

, Evans

, Goellner

, Hay

, Hertzberg

, Intenzo

, Jeffrey

, Langer

, Larsen

, Mandel

, Middleton

, Reading

, Sherman

, Tessler

. 2005. Management of thyroid nodules detected at US: Society of Radiologists in Ultrasound Consensus Conference Statement. Radiology, 237:794–800.

Gharib

, Papini

, Valcavi

, Baskin

, Crescenzi

, Dottorini

, Duick

, Guglielmi

, Hamilton

Jr , Zeiger

, Zini

. 2006. AACE/AME task force on thyroid nodules American Association of Clinical Endocrinologists and Associazione Medici Endocrinologi medical guidelines for clinical practice for the diagnosis and management of thyroid nodules. Endocr Pract, 12:63–102.

Cooper

, Doherty

, Haugen

, Kloos

, Lee

, Mandel

, Mazzaferri

, McIver

, Pacini

, Schlumberger

, Sherman

, Steward

, Tuttle

. 2009. Revised American Thyroid Association guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid, 19:1167–1214.

Rosen

, Azadian

, Walfish

, Salem

, Lansdown

, Bedard

. 1993. Ultrasound-guided fine-needle aspiration biopsy in the management of thyroid disease. Am J Surg, 166:346–349.

Filie

, Asa

, Geisinger

, Logani

, Merino

, Nikiforov

, Clark

. 2008. NCI Special Conference Utilization of ancillary studies in thyroid fine needle aspirates: a synopsis of the National Cancer Institute Thyroid Fine Needle Aspiration State of the Science Conference. Diagn Cytopathol, 36:438–441.

Iannuccilli

, Cronan

, Monchik

. 2004. Risk for malignancy of thyroid nodules as assessed by sonographic criteria. J Ultrasound Med, 23:1455–1464.

Alexander

, Marqusee

, Orcutt

, Benson

, Frates

, Doubilet

, Cibas

, Atri

. 2004. Thyroid nodule shape and prediction of malignancy. Thyroid, 14:953–958.

Jun

, Chow

, Jeffrey

. 2005. The sonographic features of papillary thyroid carcinomas: pictorial essay. Ultrasound Q, 21:39–45.

Chow

, Law

, Chan

, Au

, Yau

, Lau

. 2003. Papillary microcarcinoma of the thyroid: prognostic significance of lymph node metastasis and multifocality. Cancer, 98:31–40.

10.

Pereira

, Jimeno

, Miquel

, Iglesias

, Munné

, Sancho

, Sitges-Serra

. 2005. Nodal yield, morbidity, and recurrence after central neck dissection for papillary thyroid carcinoma. Surgery, 138:1095–1100.

11.

Zaydfudim

, Feurer

, Griffin

, Phay

. 2008. The impact of lymph node involvement on survival in patients with papillary and follicular thyroid carcinoma. Surgery, 144:1070–1077.

12.

Bruneton

, Balu-Maestro

, Marry

, Melia

, Mourou

. 1994. Very high frequency (13 MHz) ultragraphic examination of the normal neck: detection of normal lymph nodes and thyroid nodules. J Ultrasound Med, 13:87–90.

13.

Ying

, Ahuja

, Yuen

. 2004. Grey-scale and power Doppler sonography of unusual cervical lymphadenopathy. Ultrasound Med Biol, 30:449–454.

14.

Ahuja

, Ying

, Yuen

, Metreweli

. 2001. Power Doppler sonography of metastatic nodes from papillary carcinoma of the thyroid. Clin Radiol, 56:284–288.

15.

Ying

, Ahuja

, Brook

, Metreweli

. 2000. Power Doppler sonography of normal cervical lymph nodes. J Ultrasound Med, 19:511–517.

16.

Ying

, Ahuja

, Metreweli

. 1998. Diagnostic accuracy of sonographic criteria for evaluation of cervical lymphadenopathy. J Ultrasound Med, 17:437–445.

17.

Ying

, Ahuja

. 2003. Sonography of neck lymph nodes. Part I: normal lymph nodes. Clin Radiol, 58:351–358.

18.

Boelaert

, Horacek

, Holder

, Watkinson

, Sheppard

, Franklyn

. 2006. Serum thyrotropin concentration as a novel predictor of malignancy in thyroid nodules investigated by fine-needle aspiration. J Clin Endocrinol Metab, 91:4295–4301.

19.

Concato

, Feinstein

. Holford TR 2006 Predicting values from one or more variables: reporting regression analyses. Lang

, Secic

. How to Report Statistics in Medicine: Annotated Guidelines for Authors, Editors, and Reviewers, second edition. American College of Physicians, AMA Press: Philadelphia, PA, 85–105.

20.

Kim

, Park

, Chung

, Oh

, Kim

, Lee

, Yoo

. 2002. New sonographic criteria for recommending fine-needle aspiration biopsy of nonpalpable solid nodules of the thyroid. Am J Roentgenol, 178:687–691.

21.

Papini

, Guglielmi

, Bianchini

, Crescenzi

, Taccogna

, Nardi

, Panunzi

, Rinaldi

, Toscano

, Pacella

. 2002. Risk of malignancy in nonpalpable thyroid nodules: predictive value of ultrasound and color-Doppler features. J Clin Endocrinol Metab, 87:1941–1946.

22.

Baskin

, Duick

. 2006. The endocrinologists' view of ultrasound guidelines for fine needle aspiration. Thyroid, 16:207–208.

23.

Mazzaferri

, Sipos

. 2008. Should all patients with subcentimeter thyroid nodules undergo fine-needle aspiration biopsy and preoperative neck ultrasonography to define the extent of tumor invasion? Thyroid, 18:597–602.