Cytomorphological patterns of thyroid lesions among 1646 Sudanese patients: what we can learn from fine needle aspiration cytology retrospective analysis? [version 1; peer review: 2 approved with reservations]

Introduction: Fine-needle aspiration (FNA) cytology biopsy of the thyroid gland is an accurate and useful diagnostic tool in the initial evaluation of nodular thyroid lesions. We aimed in this study to determine the cytomorphological patterns of thyroid lesions diagnosed by FNA among Sudanese patients. Methods: A descriptive retrospective, clinic-based study was performed. Cytopathological records of patients that attended the Total Lab Care Clinic in Khartoum-Sudan between January 2016 and December 2017 were reviewed. Results: A total of 1646 patients records were reviewed; 1385 (84.1%) were females and 261 (15.9%) males. A total of 1563 (94.9%) were negative for malignancy, 39 (2.4%) were positive for malignancy, 42 (2.6%) were indeterminate for malignancy and 2 (0.1%) were non-diagnostic. Colloid goiter was seen in 1147 patients (73.4%), benign hemorrhagic cysts were seen in 257, Hashimoto thyroiditis was seen in 77, benign thyroid nodules were seen in 76, keratocysts were seen in 2, thyroglossal duct cysts were seen in 2, thyroid follicular adenoma was seen in 1 and myxedema was seen in 1. For malignant patients; 11 had anaplastic thyroid cancer, 8 had papillary thyroid cancer, 7 had follicular thyroid cancer, 5 had metastatic thyroid cancer, 4 had medullary thyroid cancer, 3 had non-Hodgkin lymphoma and 1 had thyroid follicular adenoma. For those indeterminate for malignancy, 24 had follicular neoplasm and 18 had Hurthle cell neoplasm. Conclusions: Fine needle aspiration cytology for thyroid nodules provides a rapid and non-invasive technique for the evaluation and differentiation between benign and malignant lesions. This study also addresses the increased predominance of benign thyroid lesions among young patients and thyroid malignancy among the 4th decade of life. This study illustrated the FNA works in Thyroid Nodule Center at the Total Lab Care Clinic in Khartoum state, Sudan, which might represent prevalence of thyroid diseases in Sudanese patients. The information in the research was useful and actual. I approved the article to be indexed with reservations. The article describes the results of large number of thyroid FNA in Sudan where similar studies are scarce. Therefore, this study is useful to give us insight about that area. However, it's not clear to me how the FNA categories were made. The authors talk in the beginning about the use of Bethesda system; which is a 6-tier system (unsatisfactory, benign, atypia of uncertain significance, follicular/Hurthle cell lesions, suspicious and malignant). Then, authors divide their results into 4-tiers only excluding both atypical categories (AUS and susp.). Does that mean they always have a definite diagnosis in cytology? Or these are final results correlating with other parameters in these cases? It is also a well known face that cytology can not discriminate in follicular or Hurthle between this clear that distinguishing


Introduction
Fine-needle aspiration cytology (FNA) biopsy of the thyroid gland is an accurate and useful diagnostic tool in the initial evaluation of the nodular thyroid lesions [1][2][3][4] . Since ≤5% of thyroid nodules are malignant, most thyroid swellings are non-neoplastic and may not require surgical intercession 5-7 . Although surgical management of thyroid nodules is a feasible option, it is associated with many adverse effects, including development of hypoparathyroidism and an enduring need for thyroid hormone replacement therapy 8 , and also risk during surgical intercession such as injury of the recurrent nerves, bacterial infections of the surgery wound, and other related factors 9 . Therefore, to avoid unnecessary thyroid surgery, the standardized Bethesda scoring system and the ACR Thyroid Imaging Reporting and Data System (ACR TI-RADS) have been developed to empower clinicians to perform suitable therapeutic intercessions 10,11 . When an accurate preoperative diagnosis can be made, and no subsequent risk of malignant changes was suspected, surgery can be avoided 12,13 . FNA is one of the initial preoperative screening procedures for the diagnosis of nodular thyroid disease and considered as the most accurate diagnostic modality 4,13-16 , with a sensitivity ranging between 43 to 95%, and specificity of 47 to 100% for thyroid lesions 16,17 . This wide range of sensitivity and specificity is attributed to the testing of suspicious and rare cases and the inclusion of occult papillary carcinoma in the category of false negative diagnosis 18,19 . Also, FNA is a safe and useful procedure for the differential diagnosis of thyroid malignancies, such as medullary thyroid carcinoma 20 .
In Sudan, ultrasound-guided FNA is considered to be an expensive technique in order to be either provided by the hospital or requested by the physician managing the patient. Lacking the epidemiological data addressing the prevalence of thyroid malignancy in Sudan, besides, there is no recent descriptive study updating the status of thyroid malignancy through the past six years. Therefore, this study aimed at determining the cytomorphological patterns of thyroid lesions diagnosed by FNA among Sudanese patients attending a cytology clinic in Khartoum state, Sudan from January 2016 to December 2017.

Patients and setting
A descriptive retrospective, clinic-based study in which we reviewed records of all patients referred to the Thyroid Nodule Center at the Total Lab Care Clinic in Khartoum state, Sudan for the evaluation of nodular thyroid disease from January 2016 to December 2017. Other than the presence of thyroid disease, no other eligibility criteria were used. The retrospective data, including type of disease, macroscopic appearance and malignancy status, were collected from the cytopathological records and sorted synonymously for analysis.
FNA and diagnosis FNA was performed previously by two expert physicians using the palpation technique 21 . In cases of multinodular lesions and complex nodules consisting of solid and fluid matters, FNA was done multiple times to avoid sampling from single nodule or aspirating cystic fluid only. Uninodular lesions with solid composition were aspirated four times to avoid bias due to sampling from single area and to insure coverage of the whole nodule. Meanwhile, in case of small nodules which were difficult to palpate (<1 cm), cystic nodules and/or nodules located posteriorly in the thyroid gland, ultrasound guided FNA were done by experienced radiologist using a LOGIQ P5 Trackball device (Guangzhou Rongtao Medical Technology Co., Ltd, China). Four smears for each patient were taken from different regions of the thyroid nodule in order to prevent misdiagnosis. The four FNA smears of each patient were fixed in 95% ethanol according to procedures outlines by Safneck et al. 22 . Then, two smears were stained using Papanicolaou (Pap) stain and the other two were stained using May-Grünwald Giemsa (MGG) stain, as described previously 23,24 . Cytopathologic diagnosis was performed by two experienced cytopathologists using microscopy with a high-power magnification lenses (X40). Both cytopathologists' results were compared with each other to confirm the diagnosis. Results of diagnosis of smears were categorized as non-diagnostic if insufficient cellular material was present and no evidence of cellular atypia was found (<6 groups of cells containing ≥15 cells each), and into negative for malignancy, positive for malignancy and indeterminate for malignancy according to the Bethesda System for Reporting Thyroid Cytopathology 10 .
A total of 1646 patients were identified and their smears were reviewed. The main cytomorphological features used for classification of thyroid nodules used included; cellularity of smears, cellular architectures (microfollicles and macrofollicles), cohesiveness, nuclear to cytoplasmic ratio, intactness of cell membrane, size of nuclei, pleomorphism, shape of nuclei, condensation of chromatin, alteration of nuclear contours, prominence of conspicuous nucleoli, and presence of pseudoinclusions, nuclear grooves, and nuclear enlargement. Additionally, the presence or absence of Hurthle cells, amyloid, colloid, psammoma bodies, inflammatory cells, necrotic material, and tumor diathesis. The criteria of the Bethesda System for Reporting Thyroid Cytopathology were followed 25 .

Statistical analysis
Descriptive data were analyzed using the Statistical Package for the Social Sciences (SPSSv16). Pearson Chi Square Test was used to test the relationship between the different categorical variables. P value < 0.05 was considered a statistically significant.

Ethical approval
Ethical approval was obtained from the ethics committee of the Faculty of Medical Laboratory Science, University of Khartoum-Sudan (UofK, FMLS 01/2017). Written informed consent for the procedure and for use of patient data for publication were obtained previously by the Thyroid Nodule Center at the Total Lab Care Clinic.

Results
The study population consisted of 1385 (84.1%) females with mean age of 42.46±14.19, and 261 (15.9%) males with mean age of 48.29±16.84. The male to female ratio was 1:5.3. Age of patients was grouped based on interval of 10 years, age group with peak thyroid lesions was 31 to 40 years followed by 41 to 50 years.
Based on the Bethesda system, results of both cytopathologists were compared with each other to obtain the final diagnosis. Each smear was diagnosed by the first cytopathologist was matching to the diagnosis of the second cytopathologist. The retrospective results of the cytomorphological features of the FNA and data recorded about the size of swellings, site of swelling distributed based on age group were shown in Table 1. Size of swilling was significantly associated with malignancy (P=0.006), whereas site of swelling was not. Individual results for each patient are available as Underlying data 26 .

Patient history of thyroidectomy and family history of thyroid disease
In total, 13/1646 patients only reported past medical history of thyroidectomy of one thyroid lobe, while the remaining 1633/1646 were not having any previous surgical removal of any of the thyroid lobes.
Family history of thyroid diseases were only reported among the negative and positive for malignancy, 77/1646 and 39/1646 respectively. Having family history of thyroid diseases was statistically significant (p = 0.0001).
Family history with thyroid diseases among patients diagnosed with malignancies was mostly noted among anaplastic carcinoma 11/39, followed by papillary thyroid cancer and follicular thyroid cancer (8/39 and 7/39, respectively). However, the remaining malignancies were also reported: 5/39 patients with metastatic cancer, 4/39 with medullary thyroid cancer and 3/39 patients with non-Hodgkin lymphoma.
In patients indeterminate for malignancy, the most frequent ages were 21 to 50 years, the peak age group was between the 20-39 years. The FNA diagnosis of the indeterminate for thyroid malignancy were grouped into two diagnostic verdicts, 24 (57.1%) as follicular neoplasm and 18 (43.9%) as Hurthle cell neoplasm. Figure 1 shows the different cytomorphological patterns of the thyroid lesions diagnosed in this study.

Discussion
Thyroid nodules are a common clinical problem 27 . FNA of the thyroid is practiced worldwide and is reliable and cost-effective diagnostic procedure to diagnose thyroid lesions that may need surgical excision or conservative management 20,28 . The key factors to ensure informative thyroid FNA are having an adequate or representative cell sample and expertise of the healthcare professional performing the thyroid cytology 4,29,30 .
Thyroid cancer is the most common endocrine cancer (approximately 1.0%-1.5% of all new cancers diagnosed each year) 31 and its incidence has continuously increased in the last three decades all over the world, including Sudan 32-34 . In this study, thyroid cancers were seen predominantly among those aged 30-59 years with the peak age in those aged 50-59 years. These results are in agreement with previous reports 35, 36 . In this study, most of the diagnosed thyroid lesions were found to be benign in nature. However, in another study conducted by Caruso and his colleagues, they reported a lower percentage (74%) of benign thyroid lesions in more than 9000 patients 37 . Gharib et al. 38  Regarding thyroiditis, Hashimoto thyroiditis had the highest prevalence. Hashimoto thyroiditis is a common autoimmune disease characterized by marked lymphoid infiltrate destroying the thyroid follicles; it has a peak incidence between 40 and 60 years of age and a female predominance 10,48 . In the present study, Hashimoto thyroiditis were common among patients falling in the age between 21 and 50 years. But, Bhatia and his colleagues observed the commonest age group is those 20-39 years of age 49 . Also, female predominance has been observed in this study which agrees with previous published studies 48,50-52 .
With respect to malignant thyroid lesions, nearly 3% of the study cases were diagnosed as malignant and the majority of these were in females. Moreover, the malignant cases were reported to be in those between 30 and 59 years of age. Although anaplastic carcinoma occurs in approximately 2% of reported thyroid cancer cases 53 , this study depicted that anaplastic carcinoma accounted the highest proportion of thyroid malignancies followed by papillary thyroid carcinoma and follicular thyroid carcinoma in Sudan. This result is comparably different from other studies showed higher prevalence of thyroid malignancy 34,54 .
Our study has highlighted several potential benefits of FNA which is consistent with the previous studies addressing the FNA reimbursements 4,38,55,56 . The most important benefits are that FNA is a simple, safe and cost-effective first-line method to investigate thyroid lesions, particularly in low-resource settings such as Sudan, in which most of the patients suffering from thyroid lesions presented late to the clinics due to their low-income status. Furthermore, the diagnosis of benign lesions is 50-fold that of malignant ones; this can be interpreted as an increase in the community awareness about thyroid diseases as well as more clinicians recognizing the utility of FNA in evaluating thyroid nodule 57,58 . However, further studies considering this increase in benign lesions showed be well investigated in respect to all factors associated with developing benign thyroid lesions.

Limitations
Most of the patients underwent surgical removal in different clinics. Therefore, follow-up data of the negative for malignancy lesions that showed non-neoplastic features were not correlated with the clinical presentation and ultrasonographic findings to conclude the final diagnosis. Follow up data of the repeated palpation, FNA and ultrasound at 6-18 months intervals to detect the appearance of significant growth or suspicious sonographic changes were not feasible to obtain from all the study participants and were excluded to avoid bias.

Conclusions
This study showed the usefulness of FNA in the evaluation of thyroid lesions and a high diagnostic performance for detection and differentiation of benign lesions from malignancy, with low rates of technical failures and complications with respect to patients' economic status since ultrasound guided FNA were only used for unpalpable nodules. Also, this study addresses the increased predominance of benign thyroid lesions among young patients and thyroid malignancy among those aged 30-39. This project contains the following underlying data:

Data availability
• Thyroid data (underlying data, including demographic details and information about malignancy, in SAV format).
Data are available under the terms of the Creative Commons Zero "No rights reserved" data waiver (CC0 1.0 Public domain dedication).