Comparison of Fine Needle Aspiration Cytology in Thyroid Lesions using The Bethesda System for Reporting Thyroid Cytopathology with Ultrasonography using Thyroid Imaging Reporting and Data System

Corresponding Author: Sudeep Regmi e-mail: sudeepregmi10@gmail.com ORCID: http://orcid.org/0000-0002-3207-515X _______________________________________________________ —–————————————————————————————————————————————— ABSTRACT: Introduction: Thyroid nodule is a common presentation. The estimated prevalence of thyroid nodules is 4-7% by clinical examination and 50-60% on ultrasonographic (USG) evaluation. Most are benign without any symptoms or cosmetic concerns. Only around 5% are found to be malignant. Methods: This prospective study evaluated a total of 54 patients with thyroid lesions presenting to Otorhinolaryngology, Surgery and Internal Medicine out-patient departments of a tertiary hospital for a period of nine months. The thyroid lesions were categorized into different categories using Thyroid Imaging Reporting and Data System (TIRADS) by USG and The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) by fine needle aspiration (FNA). The agreement between TIRADS and TBSRTC was evaluated using Cohen's kappa statistics. Results: By FNA, 68.5% were benign lesions whereas 7.4% were malignant. Follicular Neoplasm (FN) or Suspicious for FN and Suspicious for Malignancy category comprised 5.6% each. 1.9% of the lesions showed Atypia of Unknown Significance (AUS). 11.1% of the lesions were nondiagnostic or unsatisfactory for evaluation. Overall agreement between the cases by USG and FNA using the TIRADS and TBSRTC respectively was 77.77%. There was a substantial agreement between the diagnosis made by these systems, kappa (κ)= .633 (95% CI, 0.41 to 0.85, p<0.05). Conclusion: This study observed a substantial agreement between the diagnosis made by TIRADS on USG and TBSRTC on FNA. Our study advocates the stratification of thyroid lesions according to TIRADS so that only suspicious lesions undergo FNA.


Original Research Article
Thyroid nodule is a common presentation.The prevalence of thyroid nodules is estimated to be 4-7% by clinical examination and 50-60% on ultrasonographic (USG) evaluation.[1,2] Most of these nodules are benign without any symptoms or cosmetic concerns.[1] Out of these thyroid nodules, only around 5% are found to be malignant.[3] Therefore, the major concern about thyroid nodule is to rule out or confirm malignancy by USG and Fine Needle Aspiration (FNA).Surgical removal of all thyroid nodules is unnecessary and should be stratified by the use of USG and FNA.[4] The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC) is a standardized, category based evaluation of thyroid FNAs to provide a uniform diagnostic terminology for pathologists.[3,5,6] TBSRTC also provides an implied risk of malignancy and recommends clinical management of the thyroid lesions according to the category.[3,4,5] The six diagnostic categories suggested by TBSRTC are: I) Non-diagnostic or unsatisfactory; II) Benign; III) Atypia of Undetermined Significance (AUS) or Follicular Lesion of Undetermined Significance (FLUS); IV) Follicular Neoplasm (FN) or Suspicious for Follicular Neoplasm; V) Suspicious for Malignancy; VI) Malignant.[5] The Thyroid Imaging Reporting and Data System (TIRADS), first proposed by Horvath et al. [7] was established to standardize the scoring system of thyroid USG and provide recommendations for using FNA and to improve appropriate patient management.[4,8,9] The TIRADS categories as suggested by American College of Radiology (ACR) and known as ACR TI-RADS are: I) Benign; II) Not Suspicious; III) Mildly Suspicious; IV) Moderately Suspicious; V) Highly Suspicious.[4] This study attempts to observe the concordance between the TIRADS and TBSRTC on thyroid lesions of patients presenting to our hospital.

METHODS:
This was an observational cross-sectional study conducted in the Department of Pathology, Lumbini Medical College and Teaching Hospital, Palpa, Nepal.We evaluated a total of 54 patients with thyroid lesions presenting to the Otorhinolaryngology, Surgery and Internal Medicine 0ut-patient Departments from March, 2018 to November, 2018 for a period of nine months.Seven patients who were referred to the Pathology Department for FNA had not undergone USG evaluation, hence were not included in the study.
Ethical approval was obtained from the Institutional Review Committee (IRC-LMC) prior to commencement of the study.
The USG evaluation was done by a single radiologist, and the lesion was categorized as per TIRADS.If the lesion was primarily evaluated by another radiologist, the radiologist (co-investigator) categorized the lesion into TIRADS.
All the FNAs were performed by either of the two pathologists (principal Investigator and co-investigator), and the lesions were categorized according the TBSRTC.All the cases were reviewed by both pathologists.
Statistical analysis was done using kappa statistics with a 95% confidence, and concordance between TIRADS and TBSRTC was evaluated.
For the purpose of statistical analysis, category I was omitted from both TIRADS and TBSRTC, and cases of TBSRTC VI category were retained in the TBSRTC V as there is no corresponding category i.e.TIRADS VI.The categories for analysis were as follows: • TBSRTC II and TIRADS II: "BENIGN" • TBSRTC III and TIRADS III: "PROBABLY BENIGN" • TBSRTC IV and TIRADS IV: "SUSPICIOUS FOR MALIGNANCY" • TBSRTC V and TIRADS V: "PROBABLY MALIGNANT" All the data management and statistical analysis were done using Statistical Package for Social Sciences (SPSS TM ) software, version 16.

RESULTS:
We studied a total of 54 patients with thyroid lesions presenting to the Department of Pathology who underwent USG evaluation.Age of the patients ranged from 17 to 88 years with a mean of 50.74±17.8years.92.6% were females and 7.4% were males.Most of the patients presented with Right sided thyroid lesion (40.7%) followed by Bilateral (31.5%),Left (20.4%), and Midline (4.4%).A majority of the lesions were nodular (85.2%), and the rest 14.8% were diffuse.The maximum diameter of nodular thyroid lesions ranged from 10 to 86 mm with a mean of 30.15±15.63 mm by USG.
By FNA, 68.5% were benign lesions whereas 7.4% were malignant.Follicular Neoplasm (FN) or Suspicious for FN and Suspicious for Malignancy category comprised 5.6% each.1.9% of the lesions showed AUS.11.1% of the lesions were nondiagnostic or unsatisfactory for evaluation.The distribution of TIRADS and TBSRTC categories in all thyroid lesions is shown in Figure 1.
The malignant lesions (TBSRTC-VI) comprised 7.4% cases.These cases were revised as TBSRTC-V for statistical purpose, as TIRADS has no corresponding category and TIRADS-V shows the highest risk of malignancy by USG.The non-diagnostic or unsatisfactory cases comprising 11.1% of the cases were omitted from the statistical calculations.Overall agreement between the cases by USG and FNA using the TIRADS and TBSRTC respectively was 77.77%.Cohen's kappa was run to determine the agreement between the TIRADS and TBSRTC categories.There was a substantial agreement between the diagnosis made by these systems, kappa (κ)= .633(95% CI, 0.41 to 0.85, p<0.05).

DISCUSSION:
This study intended to analyze the agreement between USG evaluation of thyroid lesions using TIRADS and cytological evaluation using TBSRTC.Management of thyroid lesions depends on clinical history, thyroid function test (TFT), USG and FNA.[1] FNA is a tool to safely, accurately and costeffectively evaluate the thyroid lesions and to decide a management protocol.[10,11] However, not all thyroid lesions need surgical interventions.Ideally, after a clinical evaluation and TFT, USG should be performed to stratify the thyroid lesions and FNA should be performed only on suspicious lesions.[12,13,14] Majority of the patients in our study were females.Other similar studies also reported predominance of female population presenting with thyroid lesions.[8,9,12] Increased prevalence of thyroid lesions in females, according to studies, might be related to hormonal influences of estrogen and progesterone.Rate of thyroid nodules increases proportionately with ionizing radiation or occupational exposures.[13] Our study observed 11.1% non-diagnostic (ND) or unsatisfactory (USF) i.e.TBSRTC-I cases.ND/USF cases range from 1.8% to 23.6% of all thyroid FNAs.[3] Thyroid FNAs are ND/USF mostly due to cystic nature, sampling error or poor preparation techniques.Around 7% of thyroid lesions will be ND/USF even on repeat FNAs.[12,13,14] To reduce the chances of false-negative rate, TBSRTC advise  that FNA smears should have at least six group of well-preserved, well-stained, and well-visualized thyroid follicular epithelial cells and each group should be composed of ten cells, preferably on a single slide.[5,15] Periakaruppan et al. [12] observed 8% unsatisfactory cases on FNA in their study.The implied risk of malignancy on ND/USF is 5-10%, as laid down by TBSRTC.[5] However, a systematic review shows malignancy rates of 0-26% on followup FNAs.[16] American Thyroid Association (ATA) recommends using a USG guidance for repeat FNA and correlate with suspicious features on USG.ATA suggests that partially cystic nodules yielding ND aspirate need close observation or surgical excision, whereas surgery should be the considered if lesions are solid.[3,14] In our study, 7.4% of the cases were malignant on FNA.If we include the cases in TIRADS-V and TBSRTC-V, the prevalence rate of malignant lesions on thyroid is 9.25%, including the highly suspicious lesions for malignancy.However, no histopathological correlation was made for the suspicious lesions on FNA.According to Kilfoy BA et al. [17], thyroid malignancies constitute 1-5% of thyroid lesions among women and around 2% in men.There is an increasing trend of malignancy of thyroid lesions all over the world.This can be attributed to the availability of diagnostic facilities.
[18] A capital based study from Nepal by Bista M et al.
Overall, our study observed 77.77% agreement in diagnosis between TIRADS and TBSRTC.Study of Singaporewalla RM et al. [20] suggested an agreement of 83% between these two systems of thyroid lesion evaluation.In our study, there was a substantial agreement between TIRADS and TBSRTC as suggested by kappa statistics.Similar to our study, Vargas-Uricoechea H et al. showed a good concordance between TIRADS and TBSRTC.[8] However, these findings of our study should be validated by a larger scale study.
To establish an uniformity in reporting purpose and for a clear communication between the pathologists and clinicians, TBSRTC has proposed different diagnostic categories for thyroid lesions on FNA.For each diagnostic category, TBSRTC provides implied risk of malignancy and recommended clinical management (Table 1).[5] TIRADS endorsed by the ACR recommends FNA to be performed in lesions which are suspicious for malignancy.They have suggested FNA as per the TIRADS criteria as follows [4]

CONCLUSION:
Our study observed a substantial agreement between the diagnosis made by TIRADS on USG and TBSRTC on FNA.Thyroid lesions should be stratified according to TIRADS and the suspicious lesions should only undergo FNA.This protocol will be beneficial to avoid unnecessary FNAs on thyroid lesions.
https://doi.org/10.22502/jlmc.v6i2.274Sudeep Regmi a,d Archana Tiwari b,d Rupesh Sharma c,d Comparison of Fine Needle Aspiration Cytology in Thyroid Lesions using The Bethesda System for Reporting Thyroid Cytopathology with Ultrasonography using Thyroid Imaging Reporting and Data System INTRODUCTION: