A Case-Based Comparison of the American Thyroid Association and Thyroid Imaging Reporting & Data System Guidelines

Abstract

Thyroid nodules are a common clinical finding. Approximately 4–7% of the population have a palpable nodule on physical exam,¹ while up to 70% of the population have a nodule detected incidentally on ultrasound.² The vast majority of nodules are benign, however, approximately 5–13% of thyroid nodules detected on imaging are at risk of malignancy.³ Some malignant nodules, especially those smaller than 1 cm, can exhibit indolent behavior and do not require aggressive treatment.⁴ Therefore, thyroid nodules need to be accurately assessed to avoid overdiagnosis and overtreatment of nodules which would not otherwise affect patient morbidity. The American Thyroid Association (ATA) addressed this challenge by developing a set of ultrasound pattern-based guidelines for thyroid nodule management in 2009, which were updated in 2015.⁵ Other societies have since published similar guidelines, such as the Thyroid Imaging Reporting & Data System (TI-RADS) by the American College of Radiology in 2017. TI-RADS was similarly intended to risk-stratify nodules based on ultrasound appearance, but uses a points-based approach. The purpose of this review is to provide an overview of thyroid nodule evaluation and management through a case-based comparison using the ATA and TI-RADS guidelines.

Overview of Thyroid Nodule Workup

Thyroid nodule assessment begins with a thorough patient history and physical exam. When taking the patient history, the physician should aim to identify risk factors for thyroid malignancy, such as age less than 20, age greater than 70, male sex, history of childhood head and neck radiation, family history of thyroid cancer, or family history of hereditary syndromes involving the thyroid (MEN2, Cowden syndrome, FAP, etc.).⁵ On physical exam, findings such as a firm and immobile nodule are suggestive of malignancy, while findings of hoarseness or dysphagia suggest mass effect or invasive disease.¹ A thyroid ultrasound should be performed on all patients with a suspected thyroid nodule or goiter. Nodule size, composition and other ultrasound findings then determine management as outlined by the ATA or TI-RADS guidelines.⁴ Management can range from clinical observation to ultrasound surveillance or fine needle aspiration (FNA).

ATA Guidelines

The ATA guidelines are all-encompassing, from the initial evaluation of thyroid nodules to the treatment of thyroid cancers.⁵ This review will focus on the ATA’s recommendations for initial ultrasound evaluation and decision to perform fine needle aspiration. The ATA guidelines places nodules into risk “patterns” based on ultrasound features and recommends biopsy based on pattern and size. Nodules are assessed for their composition, echogenicity, margins, and the presence or absence of additional features such as microcalcifications. The combination of these ultrasound findings then determines placement into one of five patterns regarding risk of malignancy: High suspicion, intermediate suspicion, low suspicion, very low suspicion, or benign.⁵ Features such as microcalcifications, irregular margins, and a shape that is taller-than-wide in the transverse plane are more associated with thyroid cancer, therefore, a nodule which exhibit these features will place it in a pattern which is at higher suspicion for malignancy.⁵ Conversely, a nodule which is hyperechoic, partially cystic, or spongiform, is often placed in a pattern which is less suspicions for malignancy.⁵ FNA is recommended if any dimension of the nodule exceeds the size cutoff per pattern. A summary of the sonographic patterns and their ultrasound features are listed in Table 1.

Table 1.

Summary of the Sonographic Patterns and Their Ultrasound Features

Adapted from the 2015 ATA guidelines.

Sonographic Pattern	Ultrasound Features	FNA Threshold
High Suspicion	Solid hypoechoic nodule or solid hypoechoic component of a part cystic nodule with any of the following: - Irregular margins (infiltrative, microlobulated) - Microcalcifications - Taller-than-wide shape - Rim calcification - Small extrusive soft tissue component - Extra-thyroidal extension (ETE)	≥ 1 cm
Intermediate Suspicion	Hypoechoic solid nodule with smooth margins without: - Microcalcifications - Taller-than-wide shape - Extra-thyroidal extension (ETE)	≥ 1 cm
Low Suspicion	Isoechoic or hyperechoic solid nodule, or part cystic nodule with eccentric solid component without: - Microcalcifications - Irregular margins - Taller-than-wide shape - Extra-thyroidal extension (ETE)	≥ 1.5 cm
Very Low Suspicion	Spongiform or part cystic nodules without any sonographic features described in the low, intermediate, or high suspicion patterns.	Observation - vs - FNA ≥ 2 cm
Benign	Purely cystic nodules (no solid component).	No Biopsy

Thyroid Imaging Reporting & Data System (TI-RADS)

TI-RADS was developed by the American College of Radiology with the intent to “identify most clinically significant malignancies while reducing the number of biopsies performed on benign nodules.”⁴ The system also includes a lexicon of standardized terminology for describing thyroid nodules and provides a structured method for nodule evaluation to improve reporting in clinical practice.

TI-RADS is a point-based system organized into five categories which are to be evaluated by ultrasound: Composition, echogenicity, shape, margin, and echogenic foci. A single best descriptor is chosen for the first four categories, and all applicable descriptors are chosen for the final category (echogenic foci). Each descriptor is associated with a specific number of points based on its risk of malignancy.

Points are then summed to classify nodules into one of five TI-RADS “levels.” Zero points is TI-RADS level 1 or TR1 (benign), 2 points is TR2 (not suspicious), 3 points is TR3 (mildly suspicious), 4–6 points is TR4 (moderately suspicious), and ≥7 points is TR5 (highly suspicious).⁴

A summary of the point system is provided in Table 2. Suggestions for management, which include no further workup, serial ultrasound surveillance or FNA, are provided according to the TI-RADS level and nodule size.⁴ A summary of these recommendations are provided in Table 3.

Table 2.

A Summary of the American College of Radiology TI-RADS Point System.

Adapted from the ACR TI-RADS guidelines.

Composition (Choose 1)		Echogenicity (Choose 1)		Shape (Choose 1)		Margin (Choose 1)		Echogenic Foci (Choose all that apply)
	Points		Points		Points		Points		Points
Cystic	0	Anechoic	0	Wider-than-tall	0	Smooth	0	None	0
Spongiform	0	Hyperechoic	1	Taller-than-wide	3	Ill-defined	0	Large Comet-tail Artifacts	0
Mixed Cystic & Solid	1	Isoechoic	1			Lobulated	2	Macrocalcifications	1
Solid	2	Hypoechoic	2			Extra-thyroidal extension (ETE)	3	Peripheral or Rim Calcifications	2
		Very Hypoechoic	3					Punctate Echogenic Foci	3

Table 3.

American College of Radiology TI-RADS Management Recommendations from Point System.

Adapted from the ACR TI-RADS guidelines.

TI-RADS Level	Points	Suggested Management
TR1	0	No FNA
TR2	2	No FNA
TR3	3	FNA if ≥ 2.5 cm	Follow if ≥ 1.5 cm (US at 1, 3, 5 yrs.)
TR4	4–6	FNA if ≥ 1.5 cm	Follow if ≥ 1 cm (US at 1, 2, 3, 5 yrs.)
TR5	≥ 7	FNA if ≥ 1 cm	Follow if ≥ 0.5 cm (yearly US for 5 yrs.)

Applying ATA and TI-RADS Guidelines to Thyroid Nodule Management

The following cases will illustrate application of the ATA and TI-RADS guidelines to nodules with various levels of suspicion for malignancy.

Case 1: Nodule with high suspicion of malignancy on ultrasound. Cytology consistent with undifferentiated carcinoma

ATA Approach

This nodule is solid in composition, hypoechoic in echogenicity, taller-than-wide in shape, has lobulated margins and contains microcalcifications (white arrow). This combination of ultrasound features places the nodule within the “high suspicion” pattern. ATA recommends FNA for nodules in the high suspicion pattern beginning at 1.0 cm. This nodule measured 1.7 cm, so it qualifies for FNA.

TI-RADS Approach

This nodule is solid in composition (+2 points), very hypoechoic in echogenicity as it is darker than the adjacent strap muscles (+3 points), is taller-than-wide in shape (+3 points), has lobulated margins (+2 points) and contains punctate echogenic foci (+3 points). This nodule receives a total of 13 points, placing it in TI-RADS level 5 (7+ points, highly suspicious). TI-RADS recommends FNA for TR5 nodules beginning at 1.0 cm. This nodule measured 1.7 cm, so it qualifies for FNA.

Comparison

Both ATA and TI-RADS recommend biopsy beginning at 1.0 cm for nodules at high suspicion of malignancy.

Case 2: Nodule with low suspicion of malignancy on ultrasound. Cytology consistent with follicular adenoma

ATA Approach

This nodule is solid in composition, isoechoic in echogenicity, and has smooth margins. It does not have additional suspicious sonographic features such as taller-than-wide shape or microcalcifications. This combination of ultrasound features places the nodule within the “low suspicion” pattern, for which FNA is recommended beginning at 1.5 cm. This nodule measured 2.0 cm, so it qualifies for FNA.

TI-RADS Approach

This nodule is solid in composition (+2 points), isoechoic in echogenicity (+1 point), has a wider-than-tall shape (+0 points), and has smooth margins (+0 points). There are no echogenic foci (+0 points). This nodule receives a total of 3 points, placing it in TI-RADS level 3 (3 points, mildly suspicious). TI-RADS recommends FNA for TR3 nodules beginning at 2.5 cm and ultrasound surveillance for nodules measuring 1.5 – 2.4 cm. This nodule measured 2.0 cm, so follow-up ultrasound would be performed at 1, 3 and 5 years.

Comparison

ATA and TI-RADS set different thresholds for biopsy of nodules at low and intermediate risk of malignancy (1.5 cm versus 2.5 cm and 1.0 cm versus 1.5 cm, respectively). This 2.0 cm nodule at low suspicion for malignancy would be biopsied under the ATA guidelines but would receive ultrasound surveillance under TI-RADS. Implications of this difference are discussed later.

Case 3: Nodule with very low suspicion of malignancy on ultrasound. FNA has not been performed

ATA Approach

This nodule has a classic spongiform appearance without any sonographic features seen in high, intermediate, or low suspicion patterns. This places the nodule in the “very low suspicion” pattern. ATA guidelines give the clinician the option of observation without FNA versus FNA beginning at 2.0 cm.

TI-RADS Approach

This nodule is spongiform in composition (+0 points) since at least 50% of the nodule volume is composed of small cystic spaces.⁴ Spongiform nodules are presumed benign in TI-RADS and are not subject to further feature assignment. Small echogenic foci which represent the posterior wall of tiny cysts (white arrow) should not be misinterpreted as punctate echogenic foci, which would otherwise add 3 points. Nodules are not allowed be classified as spongiform if other features such as macrocalcification or rim calcification are present.⁴ This nodule receives a total of 0 points, placing it in TI-RADS level 1 (benign). No FNA or ultrasound follow-up is recommended for this nodule.

Comparison

ATA classifies spongiform nodules as “very low suspicion” for malignancy while TI-RADS classifies them as “benign.” Management under both systems is expectant, with ATA giving the clinician the option of biopsy beginning at 2.0 cm.

Results of Fine Needle Aspiration

FNA results are reported through the Bethesda system, a cytology classification system with six diagnostic categories: Nondiagnostic (Bethesda I), benign (Bethesda II), indeterminate (Bethesda III and IV), suspicious for malignancy (Bethesda V), and malignant (Bethesda VI).⁶ This standardized system also provides recommendations for nodule management which include ultrasound surveillance, repeat aspiration or surgery.

Nodules with indeterminant cytology (Bethesda III or IV) usually prompt additional work-up such as a repeat FNA, repeat ultrasound risk stratification, or diagnostic lobectomy.⁷ Molecular testing is used as an adjunct to reduce the number of unnecessary lobectomies as analysis of abnormal gene expression on FNA samples often reclassifies nodules as malignant or benign.⁷

Studies have shown that molecular testing was highly sensitive (94%) in differentiating benign from malignant nodules in samples originally classified as Bethesda III or IV.⁷

Conclusion

Evaluation of thyroid nodules requires a careful assessment of patient risk factors, physical exam findings, and visual inspection by ultrasound. Both ATA and TI-RADS guidelines drive nodule management though ultrasound findings, but do so with slight differences as highlighted in the cases above.

Studies comparing the diagnostic performance of these guidelines have found that both systems have a high sensitivity for detection of thyroid cancer and a negative predictive value ≥90%, a finding which reflects the overall parity of both systems and consensus in management of nodules at very low and high risk of malignancy.^{8, 11}

Recommendations diverge, however, with nodules at low and intermediate risk of malignancy as reflected by the thresholds set for performing FNA. The lower size cutoff set by the ATA gives its system a higher sensitivity and negative predictive value^, while the higher size cutoff set by TI-RADS gives its system a higher specificity and positive predictive value.^{8, 9} This trade-off has important clinical implications regarding detection of disease and overdiagnosis.

Studies have shown that a higher percentage of malignancies do not receive recommendation for FNA when TI-RADS is used compared to the ATA guidelines.^{9, 12, 13} A recent paper has shown that most of these nodules: 1) undergo a five-year ultrasound surveillance period during which it is presumed that clinically-significant cancer will be detected; 2) are smaller than 1 cm and would not be routinely-biopsied under either guideline; or 3) are of low malignant potential and theoretical risk to the patient.¹²

Authors of this paper suggest the low rate of malignancy missed by TI-RADS to be acceptable and justifiable by the reduction in unnecessary biopsies and need to prevent overdiagnosis of thyroid cancer.¹² Conversely, ATA guidelines recommend FNA for a higher percentage of malignancies, but the simultaneous increased rate of biopsy on benign nodules sacrifices overall accuracy and cost effectiveness.^{13, 14} Regardless of which set of guidelines are chosen, thyroid nodule management should ultimately be focused on efficient and effective detection of thyroid cancer while preventing overdiagnosis and incorporating goals of patient care.