"Onion Skin-liked Sign" in Thyroid Ultrasonography: A Characteristic Feature of Benign Thyroid Nodules.

BACKGROUND: Some ultrasonographic (US) signs overlap between benign and malignant nodules. The purpose of this study was to raise a special US sign of benign thyroid nodules, termed the "onion skin-liked sign."
METHODS: Twenty-seven patients with 27 nodules who shrank naturally and the "onion skin-liked sign" appeared on the final US images were enrolled in the study. The ultrasound characters and risk stratifications at the start and end of observation were compared. Then, thirty goiters with fibrosis and thirty papillary thyroid carcinomas (PTC) were randomly selected from the database of our hospital, matched the sizes of 27 nodules at the end point of observation. The differences of "onion skin-liked sign" between the two groups were analyzed.
RESULTS: The average duration of follow-up of 27 nodules was 24.0 ± 12.2 months (range, 12-65 months). At the end of the follow-up, the size of the nodules decreased on average by 1.26 ± 0.82 cm (range, 0.3-3.4 cm) and calcification was found in 21 nodules, compared with only 2 nodules with calcification at the start of the follow-up. In addition, only negligible or no blood flow signal could be detected at the periphery of all the nodules and 100% (27/27) were high suspicion at the end of observation. In matched groups, all PTC showed high suspicion of malignancy, 18/30 (60%) goiters with fibrosis were high suspicion and 11/30 (37%) were intermediate suspicion. Twenty-two patients in the group of nodular fibrosis presented "onion skin-liked sign," which was not shown in any patient of PTC group. The sensitivity, specificity, positive predictive value, and negative predictive value of "onion skin-liked sign" in predicting nodular goiter with fibrosis were 73.3%, 100%, 100%, and 78.9%, respectively.
CONCLUSIONS: The "onion skin-liked sign" was a characteristic US feature of benign thyroid nodules detected in the follow-up of thyroid nodules. It is useful to differentiate PTCs and nodular goiters with fibrosis.

INTRODUCTION

Thyroid nodules are quite common. Approximately 4–8% of adults are diagnosed with thyroid nodules by palpation while 40–50% are diagnosed by ultrasound. However, approximately 5.0–6.5% of these nodules are malignant. In recent years, the incidence of thyroid carcinoma is increasing fast.[123] Therefore, the key for the diagnosis and treatment of thyroid nodules is to identify the very few malignant nodules from the tremendous amount of thyroid nodules.[4]

As is known to us, microcalcifications, hypoechogenicity, increased nodular vascularity, infiltrative margins, and taller than wide dimensions on the transverse view are suspicious ultrasonographic (US) features of malignant nodules.[5] However, some benign nodules can also display one or more of the above features.[6] In other words, some US signs overlap between benign and malignant nodules. The purpose of the study was to raise a special US sign of benign thyroid nodules, termed “onion skin-liked sign,” which can help us recognize some benign nodules with suspicious malignant US features. The “onion skin-liked sign” was defined as a typical layered structure with hypoechogenicity and hyperechogenicity alternatively distributing from the edge to the center of the nodule.

METHODS

This study was approved by the Ethics Committee of Peking Union Medical Hospital, and written informed consents were obtained from all patients to use their data in this study. The study included two parts. In the first part, we reviewed the outpatient database including patients who were diagnosed as benign thyroid nodules from June 2009 to January 2015 and 27 nodules in 27 patients were enrolled in the study. The inclusion criteria were that nodules shrank naturally and displayed the “onion skin-liked sign” on the final US image during US follow-up. Eighteen cases demonstrated normal follicles and fiber connective tissue pathologically by fine needle aspiration (FNA) or were diagnosed as a nodular goiter with fibrosis by surgical pathology; others were diagnosed benign nodules by clinical manifestations. In the second part, we randomly selected 30 goiters with fibrosis and thirty papillary thyroid carcinomas (PTC) from the inpatient database at the same period, taking the size of 27 nodules at the end point of observation as the matched condition. The differences of “onion skin-liked sign” and risk stratification between the two groups were analyzed.

Conventional US imaging scans were obtained using a Philips iU22 scanner (Philips Medical Systems, Bothell, WA, USA) with a 5–12 MHZ linear transducer. The images were obtained under optimum instrument conditions. Standard machine settings were used with a mechanical index of 0.05–0.08, compression of 33–35, and dynamic spatial reconstructor middle. Once set, the US parameters remained unchanged in each patient. Doppler US was performed to evaluate the vascularity of the nodule. Observation began at the first US examination of the thyroid nodule and ended at the appearance of “onion skin-liked sign.” The nodule's position, size, shape, margin, short-axis diameter/long-axis diameter ratio, echogenicity, calcification, and blood flow were recorded with details. The blood flow distribution of the lesion was categorized into four levels as follows: Level 0 - no blood flow signal was detected in or around the nodule; Level 1 - most of the blood flow was detected around the nodule with few blood flow signals or no blood flow signal in the nodule; Level 2 - blood flow was detected around and in <1/2 of the nodule; and Level 3 - colored blood flow was detected in more than 1/2 of the nodule. By comparing the maximum diameter change in the follow-up visits of one nodule, we assumed that if two continuous examinations of one nodule were conducted within 3 months and the diameter change was 1–2 mm, the result was calculated as measuring error and was not recorded. The nodules were also classified as solid or predominantly solid, with either peripheral blood flow or no blood flow. All nodules were defined as one of the four risk stratifications based on the “2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer.”[7]

Statistical analysis was performed using the SPSS statistical package (version 19.0, SPSS Chicago, IL, USA). Quantitative data are expressed as the mean ± standard deviation (SD). Groups were compared using the Mann–Whitney test. The Chi-square test was used to compare categorical data. A value of P < 0.05 was considered statistically significant.

RESULTS

Of all the 27 patients (overall mean age, 50.5 years; range from 32 to 67 years), there were 22 women and 5 men. Among 27 thyroid nodules, 16 were located in the left lobe (59%) and 11 were located in the right lobe (41%). The mean number of US examinations performed for one nodule was 3.7 ± 1.4 (range, 2–8) within 12–65 months of follow-up (average, 24.0 ± 12.2 months). At the end of the follow-up, the size of the nodules decreased on average by 1.26 ± 0.82 cm (range, 0.3–3.4 cm). The US features of the nodules are shown in Table 1 and the “onion skin-liked sign” is shown in Figures 1–3.

Table 1

US features of the 27 thyroid nodules and risk stratification at the start and the end of observation

Ultrasonic features	Start of observation (n = 27)	End of observation (n = 27)
Maximal diameter in cm*, mean ± SD	2.10 ± 0.86	1.05 ± 0.49
Shape, n (%)
Regular	27 (100)	20 (74)
Irregular	0	7 (26)
Margin*, n (%)
Defined	27 (100)	18 (67)
Ill-defined	0	9 (33)
Internal components*, n (%)
Cystic or predominantly cystic	18 (67)	0
Solid or predominantly solid	9 (33)	27 (100)
Echogenicity*, n (%)
Isoechoic	10 (37)	2 (7)
Hypoechoic	17 (63)	25 (93)
S/L ratio†, n (%)
≥1	0	2 (7)
<1	27 (100)	25 (93)
Calcification*
No calcification	25 (93)	6 (22)
Calcification	2 (7)	21 (78)
Color doppler*, n (%)
Level 3	4 (15)	0
Level 2	6 (22)	0
Levels 1 and 0	17 (63)	27 (100)
Risk stratification*, n (%)
High suspicion	0	19 (70)
Intermediate suspicion	3 (11)	8 (30)
Low suspicion	12 (44)	0
Extremely low suspicion	12 (44)	0

*P<0.05; †S/L ratio: Short-axis diameter/long-axis diameter at a cross-sectional ultrasound image. US: Ultrasonic.

Figure 1

US follow-up of a 55-year-old woman. Longitudinal US images in this woman showed a right thyroid nodule. Initial size at (a) the beginning of the observation was 3.7 cm × 2.1 cm with a cyst-solid component, while the size at (b) the end of the observation (after 38 months) was 1.5 cm × 1.2 cm, with the cyst component disappeared and the solid component evolved into a layer structure (arrow in d) distributed alternately with hypoechogenicity and hyperechogenicity. Figures c and d were sketches. US: Ultrasonographic.

Figure 3

US follow-up of a 28-year-old man. Transverse US images in this man show a right thyroid nodule. Initially, the nodule was cyst-solid (a), while after 6 months, the nodule (b) shrank obviously and became absolutely solid. A layer structure (arrow in d) distributed alternately with hypoechogenicity and hyperechogenicity could be observed. However, it could be misdiagnosed as a suspicious malignant nodule because of the ill-defined boundary and calcification (but was actually hyperechogenecity). US: Ultrasonographic.

US follow-up of a 62-year-old man. Longitudinal US images in this man showed a left thyroid nodule. Initially, the nodule was cyst-solid (a) and the solid component is hypoechoic, while after 13 months, the nodule (b) shrank obviously, the cyst component disappeared and the solid component displayed extremely hypoechoic with several fine hyperechogenicity distributed alternately. The “onion skin-liked sign” (arrow in d) appeared at the end of the observation. US: Ultrasonographic.

Twenty-seven patients were classified into three groups according to the duration of the follow-up visits. The first and second group received more than 1 year of follow-up. Within the year before the last US examination (the end of observation), the first group underwent US examination while the second group did not undergo US examination. The third group received <1 year of follow-up. The maximum diameter of the nodules in each group was calculated at the final US examination, within the year before the last US examination, and more than 1 year before the last US examination. The results are shown in Figure 4.

Figure 4

Change of the size of all 27 nodules during the follow up, endpoint of observation was defined as appearence of “onion-skin liked sign”, a special sign of US. Group 1: Nodules had been followed for more than 12 months and took one or more US examinations within one year before the endpoint. Group 2: Nodules had been followed for more than 12 months and didn’t take any US examination within one year before the endpoint. Group 3: Nodules had been followed for less than 12 months and took one or more US examination within one year before the endpoint. US: Ultrasonography.

In matched groups, there were 8 male and 22 female patients in goiters with fibrosis (50.3 ± 12.5 years) and PTC groups (45.5 ± 9.6 years), respectively. The size of benign nodules was 1.08 ± 0.30 cm, and that in PTC group was 1.06 ± 0.29 cm; no statistically significant difference was found between the two groups (P > 0.05). All PTC patients showed high suspicion of malignancy, 18/30 (60%) goiters with fibrosis were high suspicion, 11/30 (37%) were intermediate suspicion, and 1/33 (3%) were low suspicion [Table 2]. Twenty-two patients in the group of nodular fibrosis presented “onion skin-liked sign,” which was not shown in any patient of PTC group. Moreover, there was statistically significant difference between the two groups (P < 0.05). The sensitivity, specificity, positive predictive value, and negative predictive value of “onion skin-liked sign” in predicting nodular goiter with fibrosis were 73.3%, 100%, 100%, and 78.9% respectively.

Table 2

Ultrasound risk stratification of thyroid nodules in the two groups

Groups	High risk (%)	Intermediate risk (%)	Low-risk (%)
Goiter with fibrosis (n = 30)	18 (60)	11 (37)	1 (3)
PTC (n = 30)	30 (100)	0	0

PTC: Papillary thyroid carcinomas.

DISCUSSION

US is the preferred imaging method for thyroid nodular lesions and is highly sensitive and specific in identifying benign and malignant thyroid nodules.[8910] Some nodules show typical benign features, the most common of which include the “comet tail”[11] and “honeycomb sign/sponge sign.”[121314] However, some nodules do not display typical benign or malignant features, and a great number of goiters with fibrosis were evaluated as malignant nodules by mistake.[15] Hence, how to evaluate such nodules has become an issue of concern.

The natural history of benign nodular goiter is largely variable and unpredictable in a given patient because no specific predictive growth parameters exist.[16] The main pathological process of goiters is that proliferation and degeneration occur alternatively in thyroid follicles for a long period, thereby facilitating the proliferation of fibrous tissues in the thyroid.[17] In addition, lobules or follicles full of colloids are often surrounded by fibrous tissues and the fibro-capsule surrounding the nodule may reduce the blood supply of some follicles, resulting in necrosis, hemorrhage, cyst formation, scar formation, and calcification of follicles. This process can be clearly imaged with US. Hyperechogenicity corresponds to the interface between the fibrous tissues and lobules or follicles and the layer structure formed as a result of degeneration.

At follow-up observation, a decrease in nodule size is a feature of benign nodules.[17181920] In our study, the size of all nodules in the three groups decreased. At the start of the observation period, all of the nodules displayed a defined margin and a regular shape. However, at the end of the observation period, 9 nodules became ill-defined and 7 nodules became irregularly shaped. We speculated that these changes appeared in the US images because of scar formation after necrosis of the follicles or the stretch on normal tissues caused by the presence of fibrous tissue. However, thyroid carcinoma can also display an ill-defined margin and an irregular shape owing to tumor invasion to the surrounding thyroid gland. The pathology of these two diseases is different. Furthermore, the number of nodules with calcification ranged from 2 at the start of the observation to 21 at the end of the observation. Calcification of the walls of the nodule or fibrous band after focal bleeding, necrosis, hematoma absorption, and organization of the nodule may explain this process and may present as punctuate, strip-like or curved calcification in the US image. Colloids in follicles may become concentrated and form crystals, which display similarly in US images and need to be distinguished from microcalcification in thyroid carcinoma lesions. In addition, the blood flow feature of the nodules gradually evolved to Level 0 and Level 1, findings that are different from those of malignant nodules (abundant and disordered distribution of blood flow in the nodule). Indeed, the blood flow feature plays an important role in identifying benign and malignant nodules. We also found that the nodules in our study entered a steady state after the appearance of the “onion skin-liked sign,” a typical layered structure with hypoechogenicity and hyperechogenicity alternatively distributing from the edge to the center of the nodule, which may be indicator of the endpoint of natural development of nodular goiter.

According to the guideline of American Thyroid Association (2015 version), all thyroid nodules should be evaluated risk stratification and make a decision of FNA or not.[7] For the nodules with high and intermediate risk, suspicion should be treated positively, but for low-risk and extremely low-risk ones, following up may be advised. The study showed that within years of follow-up of benign nodules, the risk stratification by US might change, with some low-risk or extremely low-risk nodules turning into high-risk or intermediate-risk ones. During a certain period of follow-up, 70% (19/27) fibrosis goiters may present high-risk suspicion, 30% (8/27) intermediate risks, adding to the possibility of FNA. In surgical patients, 73% (22/30) patients with goiter fibrosis presented with “onion skin-liked sign” while this sign was not shown in PTC. Hence, the proper identification of these nodules can help avoid unnecessary intervention. When the nodule displays the “onion skin-liked sign” alone at the first examination, it is suggested that the US images from the previous 2 years be reviewed. The limitation of our study was that the goiters with fibrosis and PTC were not randomly selected. These nodules were size-matched. Further studies about the diagnostic value of “onion skin-liked sign” in all nodules may be necessary.

In conclusion, the “onion skin-liked sign” was a characteristic US sign of benign thyroid nodules detected in the follow-up of thyroid nodule imaging. It is useful to differentiate PTC and goiter with fibrosis.

Financial support and sponsorship

Ni. lThis study was supported by the grants from the National Natural Science Foundation of China (No. 81541131), the National International Science and Technology Cooperation Project (No. 2015DFA30440).

Conflicts of interest

There are no conflicts of interest.