Histopathological Analysis of Thyroid Nodules with Taller-Than-Wide Shape in Adults

Introduction

The presence of a taller-than-wide (TTW) shape on ultrasound examination (USE) is closely associated with malignancy, exhibiting a higher diagnostic odds ratio than other malignant features such as irregular margin, ill-defined, markedly hypoechoic, and microcalcification.^1,2 This feature was first recognized in 2002 because of the observed growth of benign nodules in normal tissue planes, while malignant nodules expand across tissue planes.³ Previous studies showed that the diagnostic specificity of TTW in the diagnosis of thyroid carcinoma was as high as 85.2%–96.85%, but the sensitivity varied from 0% to 81.4%,^4,5 more investigation is needed to deeper understanding this special USE feature.

USE is the preferred imaging modality for screening thyroid nodules, while pathological testing is the gold standard for definitive diagnosis. Both of them analyze structures of thyroid from different angles with different principles. Most popular guidelines list microcalcification, irregular margin, ill-defined, markedly hypoechoic and TTW shape as malignancy USE features.^6–11 Investigations have verified the relations between some of these features and pathological findings. For instance, ill-defined boundaries on USE exhibit infiltration into the surrounding tissue or irregular margins by histology.¹² Microcalcifications in thyroid nodules often reflect the presence of psammoma bodies, a specific histopathological structure diagnosing papillary thyroid carcinoma (PTC).^13,14 However, well-defined boundary was also found in PTC, and microcalcifications of USE might often detect in benign nodules, for some of them might be dystrophic calcifications or inspissated colloid.^12,13 The relationship between USE features and histopathological structures is still ambiguous.^15,16

PTC is the most common malignancy of the thyroid and the most studied of USE. The diagnosis of PTC is mainly based on the characteristic changes of the cell nucleus such as “ground-glass” nuclei, nuclear overlap, nuclear sulcus and nuclear inclusions.¹⁷ In addition, neoplastic papillae and psammoma bodies are also specific diagnostic indicators of PTC.¹⁷ Recently, pathological studies showed that dense fibrosis in stromal may be another vital indicator of PTC.^18,19 A few studies focused on the relationship between fibrosis content and USE features found that the amount of fiber in a nodule correlated well with the degree of echogenicity.^14,20 The more fiber in a nodule, the lower was the echogenicity, and the likelihood of malignancy improved.^21,22 As for the TTW shape, most studies mainly focused on analyzing its correlation with malignancy, few on its histopathological structures.^15,16 We hypothesized that there would be a significant histopathological difference between benign and malignant thyroid nodules with a TTW shape. Therefore, we retrospectively analyzed the histopathologic structures of 85 nodules with TTW shapes to further understand this specific US feature.

Methods

Study Population

The study complied with the Declaration of Helsinki and was approved by the ethic committee of Guigang People’s Hospital (GYYXLL-20211229-41). A retrospective analysis was conducted on 1680 individuals who underwent partial or complete thyroidectomies at our facility between September 2019 and July 2023. These were the conditions for inclusion: (1) patients had thyroid ultrasound scans three months prior to surgery in our hospital, and the archived USE images (in both transverse and longitudinal plains) were clearly reviewed by sonologists (including the location, shape, size, boundary, echotexture, echogenicity of the nodule and the anterior neck muscles were clearly identified); (2) patients had thyroidectomies and received a definitive histopathological diagnosis; (3) nodules with TTW shape, which was defined as a greater anteroposterior diameter than the transverse diameter or longitudinal diameter on US image;⁵ and (4) the number and location of nodules found on ultrasound were consistent with the postoperative section information. All data were collected from the hospital electronic archive system and de-identified in the study. The institutional review board of Guigang People’s Hospital approved this retrospective study and waived the requirement for informed consent from the patients (NO. GYYXLL-20210601-15) for no intervention was done. The study was done in accordance with the principles of the Declaration of Helsinki. A flowchart illustrating the patient selection process is shown in Figure 1.

Figure 1 Flowchart of criteria for patient selection.

Ultrasound Examination and Ultrasound Image Analysis

Ultrasound scans of the thyroid and neck were performed using a 5–12-MHz linear array transducer (LOGIQ S8 or LOGIQ E9, GE Healthcare Wauwatosa, WI, USA). At least two sonograms of each nodule were preserved in both transverse and longitudinal plains and stored numerically in the archived USE. An experienced sonologist with 15 years of experience in thyroid ultrasound retrospectively viewed the sonograms. The reviewer, who had no previous knowledge of the pathologic diagnosis, determined the presence of TTW shape in the nodules and recorded the ultrasound features of the nodules according to the China Thyroid Imaging Reporting and Data System (C-TIRADS).¹¹ The recording ultrasound features included location (upper, middle, and lower), size (maximum longitudinal, anteroposterior, and transverse diameters), boundary (defined or ill-defined), margin (regular or irregular), composition (solid, predominantly solid, predominantly cystic, cystic, or spongiform), calcification (microcalcification or non-microcalcification), and echogenicity (hyperechoic, isoechoic, hypoechoic, or anechoic). The degree of hypoechogenicity was further categorized into mild (hypoechoic relative to the thyroid parenchyma but hyperechoic relative to the anterior neck muscles), moderate (similar echogenicity to the anterior neck muscles), and markedly hypoechogenicity (hypoechoic relative to the anterior neck muscles), with the reference standard of adjacent thyroid tissue and anterior neck muscles (Figure 2A–D).²³ When the echotexture of a nodule was inhomogeneous, main echogenicity was considered.

Figure 2 Nodules with different echogenicity and histopathological specimens (hematoxylin-eosin stain, ×20) illustrations. The nodule echogenicity was compared to that of anterior strap muscle (white arrows). (A) An isoechoic nodule (indicated by arrowhead and identified between the calipers “+”) with the dominant echoic closed to the anterior strap muscle, histopathological specimen shows a benign nodular goiter (B). (C) A mild hypoechoic nodule (indicated by arrowhead and identified between the calipers “+”) and histopathological slice also shows a benign nodular goiter (D). (E) A moderate hypoechoic nodule (indicated by arrowhead and identified between the calipers “+”) with histopathological slice showing a papillary carcinoma (F). (G) A markedly hypoechoic nodule (indicated by arrowhead and identified between the calipers “+”) with histopathological slice also showing a papillary carcinoma (H).

Nodule Histology Analysis

The pathological structural analysis of the selected nodules was conducted by an experienced pathologist via light microscopy. The pathologist was unaware of the pathologic diagnosis and the USE features of the nodules. The interpreter evaluated the histological features of the nodule in five categories: arrangement of follicular cells (normal follicle or microfollicle), presence of papillary structures or psammoma bodies, degree of fibrosis in tumor stroma (focal or widespread), and the amount of lymphoid infiltration in tumor stroma. Widespread fibrosis and lymphoid infiltration were defined as proportions of fibrosis or lymphoid infiltration >30%.¹⁴

Data Analysis and Statistics

Statistical analyses were performed using SPSS version 27.0. Data normality was assessed via Kolmogorov–Smirnov (K-S) test. Normally distributed measurement data were presented as , non-normally distributed data were presented as median (interquartile range), and categorical data were expressed as counts (percentages). The comparison of normally distributed data between the two groups was performed using an independent sample t-test, and the Mann–Whitney U-test was used to compare two non-normally distributed data. The composition ratios for the counted data were compared using the chi-square test or Fisher’s exact test. Multivariable ordinal logistic regression analysis was performed with nodule echogenicity as dependent variable and histological features as independent variables. The results were presented as odds ratio (OR) with 95% confidence interval (CI). Statistical significance was set at P < 0.05.

Results

85 TTW nodules from 80 patients (25 men and 55 women) were included in this study. The average age was 47.45 ± 13.18y(range 14 ~ 76y). The mean maximum diameter was 6.9[5.1, 10.3]mm, and the TTW ratio was 1.20[1.10, 1.33]. Among the 85 nodules, 72 (84.71%) were malignant, and 13 (15.29%) were benign. Histopathological diagnoses of the nodules are presented in Table 1. No statistical differences in sex, age, tumor size, or TTW ratio were observed between the benign and malignant groups (Table 2).

Table 1 Histopathologic Findings of the Nodules

Table 2 Distribution of Nodules Between Malignant and Benign, Regarding Age, Sex, Maximum Diameter and TTW Ratio

All nodules were classified as 4B or higher according to the C-TIRADS classification and all were solid components. Except for echogenicity, neither US characteristics nor C-TIRADS classification differed significantly between the benign and malignant nodules. Most nodules were mild or moderately hypoechogenic (83.6%), and only a small portion were iso/hyperechogenic or markedly hypoechogenic (16.4%). A comparison of the results is presented in Table 3.

Table 3 Ultrasound Characteristics of the Nodules

Nodular goiters accounted for a large part of benign lesions in this cohort. The typical histopathologic features of nodular goiter were thyroid tissue segmented by proliferating fibrous tissue into lots of follicles of various sizes. The follicles were covered by flattened epithelium and filled in with goiter (Figure 2E and F). Otherwise, PTC accounted for the vast majority of malignancies, which characterized by the presence of micro-papillae, composed of a central axis of fibro-vascular, branched, and coated by axes rich in collagen fibers. Psammoma bodies were also common in PTC (Figure 2G and H). Apart from micro-papillae, axis of fibro-vascular and psammoma bodies, the rate of fibrosis >30% was also higher in malignant nodules than in benign nodules. There was no difference in lymphoid infiltration between the two groups. Histological feature comparisons are presented in Table 4.

Table 4 Histological Features of the Nodules

Further logistic regression analysis showed a trend of decreasing nodule echogenicity with increasing fibrosis frequency (>30%) (P = 0.018, OR = 4.500, 95% CI (1.299–15.580). There was no trend for other histological features with changes in nodule echogenicity (all P > 0.05) (Table 5).

Table 5 Multivariable Logistic Regression Analysis of Nodule Echogenicity and Histological Features

Discussion

The mechanism by which malignant thyroid nodules grow across normal tissue plains has not been verified. One of the convincing mechanisms was that malignant nodules were often stiff and infiltrated into the surrounding tissues, thus not easily compressed by the probe during US examination.²⁴ Contrarily, benign nodules were often much softer and more easily compressed than malignant ones, thus TTW shape were not common in them. Otherwise, dense or diffuse fibrosis was often observed in thyroid carcinomas, which would also make malignant nodules less compressible and result in a TTW shape.²⁵ In the present study, the rate of extensive fibrosis in malignant nodules was most higher than benign ones (88.89% vs 38.46%), consisted with results of Wang et al.²⁶ Infiltration growth and diffuse fibrosis accounted for most of the TTW shape.

Several studies have shown that the TTW shape is a strong predictive sign of malignancy, either as a sole criterion or in combination with other malignant features.^27,28 The rate of benign TTW shape in the present study was 15.29%, which indicated a false-positive rate for this US character. In the present study, the difference in the US characteristics between benign and malignant nodules was not significant, except for the degree of echogenicity. The echogenic degree of malignant nodules was lower than that of the benign nodules. In the benign group, most nodules (69.2%) were mildly hypoechogenic, whereas in the malignant group, most nodules (50%) were moderately hypoechogenic and markedly hypoechogenic (13.9%), as previously reported by Lee et al and Delfim et al.^22,23

Previous studies showed that the malignancy risk of hypoechoic nodules was higher than that of isoechoic or hyperechoic ones.^2,6,29 However, this USE feature is less specific for malignancy. Investigators proposed that subdivision of hypoechoic be great help in differentia malignant from benign.^22,23 In their study, hypoechoic was divided into three degrees: mild, moderate, and markedly, and results showed that no significant difference was seen between the malignancy risk of mildly hypoechoic and iso-hyperechoic nodules, between moderate and markedly hypoechoic nodules. However, malignancy risk of moderate and markedly hypoechoic nodules was significantly higher than the mild and iso-hyperechoic ones. The result of this present study was consisted with their results. C-TIRADS uses only markedly hypoechoic as a malignant feature for higher specificity, which might result in lower sensitivity.¹¹ We proposed that both moderate and markedly hypoechogenicity be grouped as malignant risk factors for better diagnostic accuracy.

In the present study, there was a trend of decreasing nodule echogenicity with increasing frequency of fibrosis, which is consistent with the studies of Chen et al and Kim et al^14,20 Chen et al first evaluated the relationship between sonographic textural features and histopathologic components.²⁰ They confirmed the echogenicity of fibrosis, papillary cancer cells, follicular cells, and follicles was from low to high. These components are often mixed in different proportions, leading to varied echo patterns in thyroid nodules. Kim et al also found that nodule echogenicity decreased when the percentage of fibrosis increased.¹⁴ They advocated that FNA or CNB should be performed at different sites in nodules with heterogeneous echogenicity, as different echogenicity might represent different histologic features.

Although our results showed that mild hypoechoic or isoechoic thyroid nodules with a TTW shape may often be found in benign nodules, special attention should be paid to follicular neoplasms. Two follicular neoplasms were observed in this study, both of which were isoechoic (Figure 3). This is consistent with the results of previous studies,^30,31 which showed that follicular neoplasms are often isoechoic or mildly hypoechoic. Although follicular nodules were rich with micro-papillae or solid components which are also seen in PTC, extensive fibrosis was rarely in follicular nodules. This could explain why the echogenicity of PTC lower than the follicular nodules. The same goes for the follicular subtype thyroid carcinoma and classic ones. We also proposed that when there is an isoechoic or mild hypoechoic nodule with a TTW shape, follicular neoplasm should not be neglected, and the biopsy specimen should contain the nodule boundary to determine the existence of a capsule.¹⁴

Figure 3 Two follicular nodules with isoechoic and histopathological specimens (hematoxylin-eosin stain, ×20) illustrations. (A) An isoechoic nodule (indicated by arrowhead and identified between the calipers “+”) with histopathological slice showing an eosinophilic follicular adenoma (B). (C) An isoechoic nodule (indicated by arrowhead and identified between the calipers “+”) with histopathological slice showing a follicular carcinoma (D).

Limitations of the Study

Our study has several limitations. First, selection bias existed only for patients who underwent surgery, and the pathological results were confirmed. Second, this retrospective study had some limitations in interpreting the TTW shape. Third, only a few cases of follicular and inflammatory nodules were included in this study.

Conclusion

In conclusion, extensive fibrosis was the most common histopathological feature of thyroid cancers, which appeared in the form of hypoechogenicity on US. TTW-shaped thyroid nodules are highly suggestive of malignancy, especially those with moderate or marked hypoechogenicity. However, isoechoic or mild hypoechoic nodules with TTW shapes often indicate benign or follicular neoplasms.