記住我
ORIGINAL ARTICLE
Year : 2022 | Volume
: 21
| Issue : 4 | Page : 377-382
American college of radiology thyroid imaging reporting and data system score has high diagnostic value in the diagnosis of malignant thyroid nodules: A prospective single-center cross-sectional study
Souvik Biswas1, Binit Sureka1, Darwin Kaushal2, Poonam Elhence3, Amit Goyal2, Taruna Yadav1, Akhil Goel4, Pushpinder Singh Khera1
1 Department of Diagnostic and Interventional Radiology, All India Institute of Medical Sciences, Basni, Jodhpur, Rajasthan, India
2 Department of E.N.T. Otorhinolaryngology, All India Institute of Medical Sciences, Basni, Jodhpur, Rajasthan, India
3 Department of Pathology and Lab Medicine, All India Institute of Medical Sciences, Basni, Jodhpur, Rajasthan, India
4 Department of Community Medicine and Family Medicine, All India Institute of Medical Sciences, Basni, Jodhpur, Rajasthan, India
Correspondence Address:
Binit Sureka
Department of Diagnostic and Interventional Radiology, All India Institute of Medical Sciences, Jodhpur - 342 005, Rajasthan
India
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/aam.aam_123_21
Abstract
Introduction: A palpable thyroid swelling is a very common finding and is seen in almost 12% of Asian Indian population. Thyroid imaging reporting and data system (TI-RADS) can be used as a risk stratification system to determine malignant or benign thyroid nodules and necessity of further intervention. Objective: The objective of this study was to determine the positive predictive value (PPV) of TI-RADS category/ultrasound using TI-RADS categories in the diagnosis of malignancy in clinically suspected thyroid nodule and necessity for further intervention in the case of malignant thyroid nodules. Materials and Methods: We conducted a prospective study evaluating 110 patients (140 thyroid nodules) from March 2018 to April 2020 including patients with thyroid swelling. Ultrasound was performed by a radiologist on the patients, and targeted fine-needle aspiration cytology from thyroid nodules was interpreted by an experienced pathologist. Ultrasound features and TI-RADS category were compared with cytology and surgical histopathology. Sensitivity, specificity, PPV, and positive likelihood ratio in different categories of TI-RADS nodules were calculated. Results: A total of 113 thyroid nodules were assessed. Out of 113 nodules (right lobe – 64, isthmus – 6, and left lobe – 43), 84 nodules were benign and 29 nodules were malignant. Eleven (10%) patients were found to have metastatic cervical lymphadenopathy. There was no significant difference in the occurrence of malignant nodules according to gender, location of the nodule, or size of the nodule. The mean anteroposterior and transverse diameter of benign nodule was 14.1 ± 6.9 mm and 20.9 ± 9.9 mm, respectively, whereas in the case of malignant nodules, it was 15.6 ± 7.1 mm and 19.5 ± 9.0 mm, respectively. A TI-RADS score of ≥4 had 84% PPV for malignancy. The PPV for malignancy was 32.2%, 49.1%, and 100% for TI-RADS 2, 3, and 5 categories. Conclusion: TI-RADS is a simple, practical, and cost-effective tool for assessing the malignancy rates of thyroid nodules. TI-RADS categories 4 and 5 have high PPV for malignancy in thyroid nodules.
Abstract in French Résumé
Introduction: Un gonflement palpable de la thyroïde est une constatation très courante et est observé chez près de 12 % de la population indienne d'Asie. Le système de rapport et de données d'imagerie thyroïdienne (TI-RADS) peut être utilisé comme système de stratification des risques pour déterminer les nodules thyroïdiens malins ou bénins et la nécessité d'une intervention supplémentaire. Objectif: L'objectif de cette étude était de déterminer la valeur prédictive positive (VPP) de la catégorie/échographie TI-RADS à l'aide des catégories TI-RADS dans le diagnostic de malignité dans un nodule thyroïdien cliniquement suspecté et la nécessité d'une intervention supplémentaire dans le cas d'une tumeur maligne de la thyroïde. nodules. Matériels et méthodes: Nous avons mené une étude prospective évaluant 110 patients (140 nodules thyroïdiens) de mars 2018 à avril 2020 incluant des patients présentant un gonflement de la thyroïde. L'échographie a été réalisée par un radiologue sur les patients, et la cytologie par aspiration à l'aiguille fine ciblée des nodules thyroïdiens a été interprétée par un pathologiste expérimenté. Les caractéristiques échographiques et la catégorie TI-RADS ont été comparées à la cytologie et à l'histopathologie chirurgicale. La sensibilité, la spécificité, la VPP et le rapport de vraisemblance positif dans différentes catégories de nodules TI-RADS ont été calculés. Résultats: Au total, 113 nodules thyroïdiens ont été évalués. Sur 113 nodules (lobe droit - 64, isthme - 6 et lobe gauche - 43), 84 nodules étaient bénins et 29 nodules malins. Onze (10%) patients présentaient une lymphadénopathie cervicale métastatique. Il n'y avait pas de différence significative dans la survenue de nodules malins selon le sexe, la localisation du nodule, ou la taille du nodule. Le diamètre moyen antéropostérieur et transversal du nodule bénin était de 14,1 ± 6,9 mm et 20,9 ± 9,9 mm, respectivement, alors que dans le cas des nodules malins, il était de 15,6 ± 7,1 mm et 19,5 ± 9,0 mm, respectivement. Un score TI-RADS ≥ 4 avait une VPP de 84 % pour la malignité. La VPP pour la malignité était de 32,2 %, 49,1 % et 100 % pour les catégories TI-RADS 2, 3 et 5. Conclusion: TI-RADS est un outil simple, pratique et rentable pour évaluer les taux de malignité des nodules thyroïdiens. Les catégories TI-RADS 4 et 5 ont une VPP élevée pour la malignité des nodules thyroïdiens.
Mots-clés : Nodule, système de rapport et de données d'imagerie thyroïdienne, thyroïde, échographie
Keywords: Nodule, thyroid imaging reporting and data system, thyroid, ultrasound
Introduction 
Thyroid swelling is a common finding, with an incidence of 12% in Indian population.[1] The ultrasound prevalence of nodules can be as high as 80% in iodine-deficient regions of India.[2] Ultrasonography is the most accurate imaging modality for evaluation of thyroid nodules, their characteristics, shape, dimension, and vascularity. Ultrasonography accurately evaluates the internal structure, morphology, and vascularity of the gland. High-resolution ultrasonography of the neck and thyroid accurately detects lesions as small as 1–2 mm.[3] High-resolution ultrasound helps to differentiate between benign and malignant thyroid nodules, detects occult carcinoma of thyroid, invasion of surrounding anatomical structures and guide in fine-needle aspiration.[4],[5],[6],[7],[8],[9],[10],[11] On ultrasonography of the thyroid, about 30%–50% of the population has thyroid nodules, but only 5%–6.5% of these are known to be malignant.[9],[10]
Thyroid imaging reporting and data system (TI-RADS) was first developed in 2009 by Horvath et al.[11] Horvarth classified thyroid nodules into TI-RADS 1 (normal thyroid gland), TI-RADS 2 (benign conditions), TI-RADS 3 (probably benign, <5% chance of malignancy), TI-RADS 4 (suspicious nodules, 5%–80% chance of malignancy – subdivision into 4a and 4b was optional), TI-RADS 5 (probably malignant with malignancy rate more than 80%), and TI-RADS 6 (biopsy-proven malignant nodules). Kwak et al.[12] in 2011 modified TI-RADS classification system and created TI-RADS categories 3, 4a, 4b, 4c, and 5 according to the risk of malignancy.
In 2017, the TI-RADS Committee of the American College of Radiology (ACR) published a white paper in order to classify thyroid nodules on the basis of their appearance at ultrasonography.[13] In the ACR TI-RADS, thyroid nodules are evaluated on the basis of composition, echogenicity, shape, and margin, and each feature is assigned 0–3 points. Thyroid nodules are categorized from TI-RADS 1 to TI-RADS 5 on the basis of total number of points scored for the purpose of risk assessment for malignancy.
The aim of this study was to determine the positive predictive value (PPV) of the TI-RADS score in thyroid nodules with a score of 3 and above.
Materials and Methods Study design, setting, and location
It was a prospective single-center cross-sectional study in a tertiary care referral hospital from western Rajasthan from March 2018 to April 2020.
Study population
All patients aged more than 18 years with thyroid nodules who presented in the Department of Otorhinolaryngology were included in the study. Nonconsenting patients and thyroid nodule with known histopathological diagnosis were excluded from the study. The sample size for the study was 119 (considering the sensitivity of TI-RADS as 94%, the estimated 95% confidence interval, 5% relative precision, and 20% contingency). A total of 110 patients with thyroid nodule were enrolled into the study cohort. All patients of neck swelling, clinically diagnosed as thyroid nodule referred from Otorhinolaryngology OPD, were included in the study. Histopathologically proven thyroid nodules, nonconsenting patients, and patients with contraindication to fine-needle aspiration cytology (FNAC) were excluded from the study.
Data collection
Any patient (male or female) with neck swelling diagnosed as thyroid nodule was enrolled for the study with their written consent. Data collection included patient's particulars such as name, age, gender, and clinical findings. Patients with thyroid nodules were selected and reported according to ACR-TI-RADS 2017. Correlation was done with the assigned TI-RADS category of the nodule, and cytopathological/histopathological findings were documented. Data were collected according to the location, size, consistency, echogenicity, orientation, margins of the nodule, presence or absence of echogenic foci within the nodule, and its TI-RADS grade.
Outcome measures
The primary outcome measures of this study are to determine whether TI-RADS score of 3 and above is significantly predictive for malignancy.
Equipment
All ultrasound scans and power Doppler US examinations were performed in GE ultrasound machine, LOGIQ S8, Boston, USA, with a linear-array transducer (5–12 MHz) using an optimized gain.
Patient positioning and preparation
All patients were examined in supine position with mild hyperextension of neck. Scan was done with a high-frequency linear transducer (5–12 MHz). Images were acquired on grayscale as well as color/power Doppler. Scanning of the neck and thyroid was done in both longitudinal and transverse planes. Both lobes of the thyroid and isthmus were evaluated for any lesion, symmetry, echogenicity, and size. Sizes of both lobes were measured in three planes. Based on these ultrasound features, nodules were categorized into TI-RADS 1 to TI-RADS 5 using ACR TI-RADS 2017 criteria. In our study, one radiologist (S.B. with 3 years of experience) performed real-time USG of thyroid nodules and later it was reinterpreted by experienced second radiologist (B.S. with 11 years of experience) from stored pictures/data. Overall agreement between observers was good.
Thyroid fine-needle aspiration cytology
Consent was obtained from the patients after thorough explanation of the procedure. The anterior aspect of the neck was thoroughly cleaned with 5% w/v povidone-iodine solution and sterile draping was done. Under ultrasound guidance, FNAC was done from the targeted nodule with 23G needle. No local anesthesia was used. Materials obtained from aspiration were sent to the Department of Pathology for cytopathological evaluation using Bethesda system. Histopathology reports of the large biopsies were obtained if the patient underwent surgery. TI-RADS 1 and 2 lesions would be subjected to aspiration and FNAC on request by the clinician, especially when the nodules were painful. Written consent was obtained in all cases, and the procedure was thoroughly explained to the patient.
Statistical analysis
Data were analyzed using the Statistical Package for the Social Sciences (SPSS) v. 23.0 (Armonk, NY, USA: IBM Corp.) All nominal data like distribution of malignant and benign lesions were described using frequencies and percentages and analyzed using Chi-square test or Fisher's exact test as applicable. The diagnostic performance of TI-RADS for detecting malignancy was described using sensitivity, specificity, PPV, negative predictive value, and positive likelihood ratio. P < 0.05 was taken as statistically significant.
Ethical statement
This study was approved by the Institutional Ethics committee, All India Institute of Medical Sciences, Jodhpur (approval number: AIIMS/IEC/2018/1450) on September 28, 2018, and conducted in the Department of Radiology. Informed consent was taken from all the patients before enrolling in the study. The procedures followed were in accordance with the ethical standards of the institutional committee on human experimentation and with the Helsinki Declaration of 1964, as revised in 2013.
Results In our study, we found that 29 nodules (25.7%) were malignant and 84 nodules (74.3%) were benign out of a total of 113 nodules.
Our study consisted of 140 nodules in 110 patients. Cases of multinodular goiter with similar nodules were considered one nodule. Twenty-seven nodules were excluded from study as they were nondiagnostic and final benign or malignant etiology was not proved (16 – unsatisfactory for evaluation on FNAC, 3 – atypia of undetermined significance, 5 – follicular lesion of undetermined significance, and 3 – follicular). The flowchart of the study of the thyroid nodules is shown in [Figure 1]. Out of 113 nodules (right lobe – 64, isthmus – 6, and left lobe – 43), 84 nodules were benign and 29 (25.7%) nodules were malignant, details of which are described in [Table 1]. Twenty-seven nodules were excluded from the study since there was no definitive categorization of the nodules. There was no significant difference in the occurrence of malignant nodules between males and females (P = 0.396). For malignant nodules, the mean age for males was 44 ± 18.4 years, and for females, it was 35.2 ± 12.3 years. The mean age was 40.3 ± 15 years in males and 38 ± 12.7 years in females, and no significant difference in mean age was found in the case of benign nodules (P = 0.579). In the case of benign nodule, the mean age was 38.3 ± 13 years, whereas in the case of malignant nodule, it was 36.9 ± 13.9 years. The mean age of our study cohort was 37.7 years. Total 110 patients participated in our study. Among which, 89 were female and 21 were male. Eleven (10%) patients were found to have metastatic cervical lymphadenopathy confirmed by histopathology or FNAC. One patient with benign nodule was found to have enlarged supraclavicular lymph nodes. The mean anteroposterior and transverse diameter of benign nodule was 14.1 ± 6.9 mm and 20.9 ± 9.9 mm, respectively, whereas for malignant nodules, it was 15.6 ± 7.1 mm and 19.5 ± 9.0 mm, respectively.
Table 1: Frequency of ultrasound findings in thyroid nodules according to thyroid imaging reporting and data system descriptorsIn our study, we found that out of 84 (74.3%) benign thyroid nodules, the most common pathological category was colloid nodule (45.1%), thyroiditis (9.7%), and adenoma (8.8%). Papillary thyroid carcinoma comprised 15% of all thyroid nodules and is the most common pathological entity among malignant nodules. The correlation of TI-RADS score and histopathology is depicted in [Table 2]. The PPV, positive likelihood ratio, and specificity for malignancy were high for TI-RADS category 4 and 5 nodules [Table 3]. [Figure 2] shows an example of TI-RADS 4 malignant nodule.
Figure 2: A 30-year-woman presented with swelling for 7 months. (a) Transverse and (b) longitudinal axis sonogram of the neck and thyroid – A well-defined (0), wider than taller (0), hyper to isoechoic (1), completely solid (2) nodule in the right lobe of the thyroid (solid yellow arrows) with the presence of punctate echogenic foci (3) (red dotted arrows) suggestive of TI-RADS 4 lesion (total point 6). (c) Histopathology image shows follicular carcinoma thyroid with capsular invasion, ×4, H and E stain
Table 2: Correlation between histopathology and thyroid imaging reporting and data system of 29 nodules underwent surgery
Table 3: Sensitivity, specificity, positive predictive value, and positive likelihood ratio in different categories of nodules Discussion Although several classification systems were developed to stratify the risk of malignancy based on multiple ultrasound features, these are not convenient to apply in a day-to-day basis due to complexity. Though TI-RADS classification system was initially developed by Hovarth et al[11] and later modified by Kwak et al[12], ACR TI-RADS is simple and familiar to the radiologists. It is also useful to apply in a day-to-day basis to stratify the risk of malignancy in a thyroid nodule. [Table 4] shows the comparison between different studies for malignant thyroid nodules. The PPV for malignancy was 36%, 64%, 59%, and 91% for TI-RADS 4a, 4b, 4c, and 5 categories, respectively. Kwak et al. described 0%, 1.7%, 3.3%, 9.2%, 44.4%–72.4%, and 87.5% risk of malignancy for TI-RADS 2, 3, 4a, 4b, 4c, and 5 categories, respectively. There was improvement in the PPV (from 60% to 92%) when TI-RADS category 3 nodules were reassigned as not to be benign as compared from previous study by Horvarth et al. The PPV for diagnosing malignant nodule and subsequently risk of malignancy is also improved in the current (our) study. In our study PPV for nodules TIRADS 3 and above was 49.15%, for nodules 4 and above was 84.38% and for nodules TIRADS 5, it was 100%. Recent studies have shown that ACR TI-RADS score is helpful in the management of cytologically indeterminate thyroid nodules that present a challenge in clinical decision-making.[19],[20],[21] In practice, it is recommended to follow up thyroid nodules that are cytologically indeterminate on FNAC.
Table 4: Comparison of diagnostic performance in terms of positive predictive percentage value of the various ultrasound classification systems available to assess thyroid nodulesOur study had few limitations. Firstly, 27 thyroid nodules in our study were inconclusive by FNAC which includes nondiagnostic on cytopathology, AUS, FLUS, and follicular neoplasms. In these cases, proper follow-up/repeat FNAC was not done. Secondly, FNAC is less reliable than histopathology to characterize a thyroid nodule. Hence, it is possible that a benign nodule may harbor focus of malignancy which can be missed in FNAC. These patients were not operated for histopathology. However, as, in FNAC, they were proven to be benign, further histopathological evaluation was warranted for confirmed diagnosis which could not be done due to lost to follow-up. Thirdly, a number of malignant nodules were less in our study. A number of situations are not addressed in TI-RADS. Currently, more focus is given on the ultrasound appearance of a lesion. However, other clinical and radiological parameters (vascularity pattern and elastography) can be included to increase the diagnostic accuracy and more specific guidelines for management.
Conclusion Ultrasonography of the neck and thyroid is a reliable and safe imaging modality to predict malignancy in thyroid nodules. TI-RADS is a convenient method that can be applied in a day-to-day practice to categorize thyroid nodules with good interobserver agreement. PPV for malignancy is very high for TI-RADS 4 and TI-RADS 5 category lesions. High PPV in higher TI-RADS nodules (TI-RADS 4 and 5) indicates higher probability for a nodule to be malignant and thus warrants further intervention.
Financial support and sponsorship
Nil.
Conflicts of interest
There are no conflicts of interest.
References 
留言 (0)