Brought to you by
ECR 2019 / C-0774
Certificate of Merit
Diagnostic accuracy of TIRADS in evaluation of thyroid nodules
Congress: ECR 2019
Poster No.: C-0774
Type: Scientific Exhibit
Keywords: Head and neck, Thyroid / Parathyroids, Ultrasound, Diagnostic procedure, Calcifications / Calculi, Cysts, Neoplasia
Authors: M. Mohandas, S. Nadarajan, N. Hubert, L. Jose; Thiruvananthapuram/IN
DOI:10.26044/ecr2019/C-0774

Results

Out of the total 104 patients studied, 93.3% were benign and 6.7% were malignant.

 

Fig. 15: Percentage distribution of sample according to cyto/ histopathology (benign/ malignant)
References: Dr. Minu Mohandas

Benign cases were colloid nodules (55 cases), lymphocytic thyroiditis (8 cases), benign follicular adenoma (33 cases) and Hurthle cell adenoma (1 case). The most common benign nodule was colloid nodule. Malignant cases were Papillary carcinoma (2 cases), Follicular carcinoma (3 cases), Medullary and Hurthle cell carcinoma (1 case each). The most common malignancy in our study was Follicular carcinoma, followed by Papillary carcinoma. 

 

TIRADS 2 Features

In our study, 19 nodules had typical benign TIRADS 2 features which included simple cysts (3.8%) and spongiform nodules (14.4%). All 19 nodules with typical TIRADS 2 features turned out to be benign.

 

Fig. 8: Percentage distribution of sample according to TIRADS 2 features
References: Dr. Minu Mohandas

 

Fig. 16: Transverse sonogram of thyroid gland showing spongiform nodule with echogenic foci showing comet tail artefacts categorised as TIRADS 2, which turned out to be colloid nodule on cytopathological examination
References: Dr. Somervell Memorial CSI Medical College - Thiruvananthapuram/IN

 

Rest of the 85 nodules without TIRADS 2 features were assigned appropriate TIRADS category (3, 4A, 4B and 5) based on the echogenicity, shape and margin of nodules.

 

Echogenicity of nodules

Incidence of malignancy increased from 0% in hyperechoic to 50% in markedly hypoechoic nodules. Most of the hyperechoic and isoechoic nodules were benign except one which turned out to be Hurthle cell carcinoma. Out of the 7 malignant nodules, 4 were markedly hypoechoic. Statistically significant association was noted between echogenicity of nodules and risk of malignancy with p value <0.01.

 

Fig. 9: Percentage distribution of sample according to echogenicity in TIRADS 3, 4 & 5 nodules.
References: Dr. Minu Mohandas

 

Table 1: Distribution of sample according to echogenicity and percentage of malignancy in TIRADS 3, 4 and 5 nodules.
References: Dr. Minu Mohandas

 

Fig. 17: Transverse thyroid sonogram showing a predominantly solid wider than tall isoechoic nodule with well circumscribed margin in left lobe, which was categorised as TIRADS 3 and turned out to be benign follicular adenoma on histopathological examination
References: Dr. Somervell Memorial CSI Medical College - Thiruvananthapuram/IN
Fig. 18: Photomicrograph demonstrating histologic features of follicular adenoma with closely packed follicles and no capsular invasion
References: Department of Pathology, Dr. Somervell Memorial CSI Medical College, Thiruvananthapuram/IN
Fig. 19: Transverse thyroid sonogram showing a wider than tall isoechoic nodule with well circumscribed margins which was assigned TIRADS 3 score, and on histopathology, turned out to be Hurthle cell carcinoma. This was the only case which was assigned benign TIRADS and turned out to be malignant.
References: Dr. Somervell Memorial CSI Medical College - Thiruvananthapuram/IN
Fig. 20: Photomicrograph demonstrating histologic features of Hurthle cell carcinoma with large oncocytes (Hurthle cells).
References: Department of Pathology, Dr. Somervell Memorial CSI Medical College, Thiruvananthapuram/IN

 

Shape of nodules

Most of the thyroid nodules (97.6%) had wider than tall shape. Only 2 cases had taller than wide shape, of which 1 case was malignant which turned out to be follicular carcinoma. The other case with taller than wide shape turned out to be benign nodular colloid goitre with thyroiditis. There was statistically significant association between shape of nodule and risk of malignancy with p value < 0.05.

 

Fig. 10: Percentage distribution of sample according to shape in TIRADS 3, 4 & 5 nodules
References: Dr. Minu Mohandas

 

Table 2: Distribution of sample according to shape and percentage of malignancy in TIRADS 3, 4 and 5 nodules.
References: Dr. Minu Mohandas

 

Fig. 21: Transverse sonogram of thyroid showing a taller than wide markedly hypoechoic nodule with irregular margins and microcalcification assigned as TIRADS 5 nodule which turned out to be Follicular carcinoma
References: Dr. Somervell Memorial CSI Medical College - Thiruvananthapuram/IN
Fig. 22: Photomicrograph demonstrating histologic features of follicular carcinoma with capsular invasion
References: Department of Pathology, Dr. Somervell Memorial CSI Medical College, Thiruvananthapuram/IN

 

Margin of nodules

Only 6 nodules had irregular margins, of which 4 turned out to be malignant. Statistically significant association was noted between margin of the lesion and risk of malignancy with p value <0.05.

 

Fig. 11: Percentage distribution of sample according to margin in TIRADS 3, 4 & 5 nodules
References: Dr. Minu Mohandas
Table 3: Distribution of sample according to margin and percentage of malignancy in TIRADS 3, 4 and 5 nodules
References: Dr. Minu Mohandas
Fig. 23: Transverse sonogram of thyroid showing a markedly hypoechoic nodule with irregular margins and microcalcification which was categorised as TIRADS 5 nodule and turned out to be Papillary carcinoma on histopathological examination
References: Dr. Somervell Memorial CSI Medical College - Thiruvananthapuram/IN
Fig. 24: Photomicrograph demonstrating histologic features of papillary carcinoma with papillary pattern and nuclear clearing
References: Department of Pathology, Dr. Somervell Memorial CSI Medical College, Thiruvananthapuram/IN

 

Microcalcification

Microcalcification was present only in 6 nodules, of which 4 turned out to be malignant. Of the 4 malignant nodules with microcalcification, 2 were papillary carcinoma and 2 were follicular carcinoma. 2 nodules with microcalcification  turned out to be benign colloid nodules, with associated thyroiditis present in one case. Statistically significant association was noted between presence of microcalcification and risk of malignancy with p value < 0.01.

 

Fig. 12: Percentage distribution of sample according to microcalcification
References: Dr. Minu Mohandas
Table 4: Distribution of sample according to microcalcification and percentage of malignancy in TIRADS 3, 4 and 5 nodules
References: Dr. Minu Mohandas
Fig. 23: Transverse sonogram of thyroid showing a markedly hypoechoic nodule with irregular margins and microcalcification which was categorised as TIRADS 5 nodule and turned out to be Papillary carcinoma on histopathological examination
References: Dr. Somervell Memorial CSI Medical College - Thiruvananthapuram/IN
Fig. 28: Transverse sonogram of thyroid gland showing a taller than wide markedly hypoechoic nodule with irregular margin and punctate echogenic foci interpreted as microcalcification. This nodule was assigned TIRADS 5 score and the pathological diagnosis came as nodular colloid goitre with thyroiditis. This was the only TIRADS 5 nodule that turned out to be benign.
References: Dr. Somervell Memorial CSI Medical College - Thiruvananthapuram/IN

 

Lymph node without fatty hilum

Lymph node without fatty hilum was noted only in 1 case which turned out to be malignant- Papillary carcinoma. Statistically significant association was noted between lymph node without fatty hilum and risk of malignancy with p value <0.01.

 

Fig. 13: Percentage distribution of sample according to lymph node without fatty hilum
References: Dr. Minu Mohandas
Table 5: Distribution of sample according to presence of lymph nodes without fatty hilum and percentage of malignancy in TIRADS 3, 4 and 5 nodules.
References: Dr. Minu Mohandas
Fig. 27: Transverse neck sonogram of patient shows a cervical lymph node which had round shape with absence of fatty hilum, confirmed to be metastatic lymph node on histopathology
References: Dr. Somervell Memorial CSI Medical College - Thiruvananthapuram/IN

 

Risk of malignancy

The risk of malignancy was 0%, 2.1%, 8.3%, 20% and 66.7% for TIRADS 2, 3, 4A, 4B and 5 categories respectively. 

 

Table 6: Association of TIRADS with cyto/histopathological result
References: Dr. Minu Mohandas

 

Fig. 14: Bar diagram showing ssociation of TIRADS with cyto/histopathological result
References: Dr. Minu Mohandas

 

Diagnostic test properties of TIRADS

The diagnostic performance of TIRADS considering categories 4A, 4B and 5 as malignant and categories 2 and 3 as benign is as follows: Sensitivity = 85.7%, Specificity = 68%, positive predictive value (PPV) = 16.2%, negative predictive value (NPV) = 98.5%, and accuracy = 69.2%. 

 

Table 7: Statistical indices showing comparison of TIRADS with histopathology.
References: Dr. Minu Mohandas

 

High specificity, negative predictive value and accuracy for detecting malignancy were noted with the following parameters: Marked hypoechogenicity, taller than wide shape, irregular margin and presence of microcalcification. Highest specificity (98.7%) was noted with taller than wide shape. 

 

Table 8: Diagnostic test properties of sonological findings of malignant nodules
References: Dr. Minu Mohandas

 

Retrospective comparison of pathological diagnosis with assigned TIRADS categories

 

On retrospective comparison of final pathological diagnosis with the assigned TIRADS score, it was found that majority of benign nodules were assigned benign TIRADS (2 & 3) and all malignancies except 1 were assigned malignant TIRADS score (4A, 4B & 5).

 

Table 9: Retrospective comparison of pathological diagnosis with assigned TIRADS categories
References: Dr. Minu Mohandas

 

 

POSTER ACTIONS Add bookmark Contact presenter Send to a friend Download pdf
SHARE THIS POSTER
2 clicks for more privacy: On the first click the button will be activated and you can then share the poster with a second click.

This website uses cookies. Learn more