Thyroid Fine-Needle Aspiration under Ultrasound Guidance : Experience from an Academic Tertiary Care Center in Lebanon

Thyroid nodules are common. Malignancy was reported in only 5-15% of cases. Fine needle aspiration under US guidance (USG-FNA) proved to be accurate for the detection of thyroid cancer. This is a retrospective review of 400 radiology and cytology USG-FNA reports, randomly selected among those done at the radiology department at a tertiary care center in Beirut during the last five years. The specimen was inadequate in 60 (12%) of cases but FNA was repeated in 10 cases only. The final diagnosis was benign in 76.7% of cases, mostly in women. 35% of the malignant and 19% of the benign nodules were hypoechoic, p=0.03. No significant correlation was observed between malignancy and other ultrasonic characteristics. Hypoechogenecity was also more common in nodules with inadequate specimen (40% versus 21.4%, p=0.01). Age, gender, location and size of the nodule did not differ between groups of adequate and inadequate specimen. In conclusion, 3 out of 4 thyroid nodules referred for USG-FNA are benign, mostly in women. Inadequate specimen was observed in 12% of cases. Hypoechogenecity but not other ultrasonic characteristics was associated with malignancy and with test failure. Thyroid Fine-Needle Aspiration under Ultrasound Guidance: Experience from an Academic Tertiary Care Center in Lebanon ORIGINAL Rita Hajj Boutros1, Maurice Haddad2, Fouad Boulos3, Asma Arabi1

Thyroid Fine-Needle Aspiration under Ultrasound Guidance: Experience from an Academic Tertiary Care Center in Lebanon Introduction Thyroid nodules are very common: found in 4%-8% of adults by means of palpation [1], in 10%-41% by means of US [1], and in 50% by means of pathologic examination at autopsy [1].The detection of non palpable thyroid nodules is increasing due to improved ultrasound imaging techniques [2].
The prevalence of thyroid nodules increases with age.They are usually benign; however 5% to 15% prove to be malignant [3].Several characteristics have been identified as risk factors for malignancy, these include age younger than 20 or older than 60 years, firmness of the nodule, rapid growth, fixation to adjacent structures, vocal cord paralysis, enlarged regional lymph nodes, a history of neck irradiation or a family history of thyroid cancer [1].
Compared with the very high prevalence of nodular thyroid disease, thyroid cancer is not very common.Papillary thyroid cancer is the most common type of thyroid cancer.In the US, 75%-80% of new cases of thyroid cancer in 2005 were papillary [1] with a 30 year survival rate of 95% [1].
Few ultrasonic features have been associated with an increased risk of thyroid cancer, including presence of microcalcifications, hypoechogenecity, irregular margins, absence of a halo, predominantly solid composition, and intranodular vascularity [1].In addition, risk of malignancy is higher in solitary as opposed to multiple nodules.However the sensitivity, specificity, negative and positive predictive values for these criteria are extremely variable from study to study, and no single feature has both a high sensitivity and a high positive predictive value for thyroid cancer [1].
Fine needle aspirate (FNA) with cytologic evaluation has become the accepted method for screening a thyroid nodule for cancer, and, in the hands of an experienced cytologist, FNA has a high accuracy rate.It is recommended to do FNA for nodules of ≥ 10 mm as well as those with suspicious malignant US features, even if the nodule size is less then 10mm.As a triage tool, FNA biopsy can be used to distinguish thyroid nodules that might have a higher risk of malignancy (neoplasm), and would thus require surgical excision, from goitrous nodules or thyroiditis, which can be managed medically.The use of US guidance for FNA (USG-FNA) ensures that the sample is obtained from the nodule in question and permits direction of the needle into the solid portions of partially cystic nodules, which will improve the diagnostic yield.It is minimally invasive, provides fast and accurate results and has a diagnostic accuracy of 85-94% [2].However, nondiagnostic cytology results (10-21% of cases) are a major limitation of USG-FNA requiring a repeat FNA because of the high malignancy rate, which is variable depending on different studies (reaching 14% in one study) [4].Therefore, determining the clinical and ultrasonographic characteristics that predict or influence non-diagnostic results would be helpful to improve the yield of the test and avoid repetitive USG-FNA or unnecessary surgical procedures.
To our knowledge, the prevalence of thyroid cancer among Lebanese patients with thyroid nodules referred for FNA has not been previously reported.Also no study evaluated the satisfactory rate of USG-FNA in Lebanese centers.In this study we aimed to determine the prevalence of thyroid cancer, the frequency of different types and the prevalence of atypia of undetermined significance in patients referred for FNA of thyroid nodules under US guidance at a tertiary care center in Lebanon.We also aimed to determine the rate of unsatisfactory tests and the predictors of test failure.

Study design
This was a retrospective observational study.

Data collection
We conducted review of 400 reports that were randomly selected among cases referred to the diagnostic radiology department at the American University of Beirut for USG-FNA of thyroid nodule between 2009 and 2014.Radiology and cytology reports were accessed electronically using the health information system of the medical center.The following information were retrieved: Age and gender, the radiologic characteristics of the thyroid

Statistical analyses
Continuous variables [age, size of the nodules] were not normally distributed, therefore the values were presented as median [min-max] and non parametric test (Mann-Whitney) was performed to check for statistical significance in the median of continuous variables (i.e age, size of the nodules) between groups.Categorical variables were reported as number and percentages.Chi-square or Fisher's exact-test were used to compare percentages of categorical variables between groups.Data was analyses using SPSS 20.P-value < 0.05 was considered statistically significant.

Ethics
The study was approved by the Institutional Review Board of our institution.Data was reviewed and entered for analyses anonymously.

General characteristics
The age and general findings of the ultrasound reports are shown in Table 1.The median age of the study population was 58 [range 20-92].Men were older than women [63 versus 57 years, p=0.005].Solitary nodule was reported in 22.2% of cases, without difference between genders.The median size of the solitary nodule however was larger in men than in women [27.5 mm versus 16 mm].FNA was performed in one nodule in the majority of cases (75.5%), and in both genders.Cervical lymphadenopathy and hypervascularity were reported in 5% of the cases, without difference between genders.

Radiological characteristics
FNA was performed on one nodule in 302 subjects and on two nodules in 98 subjects, thus 498 nodules were FNA.The decision to do FNA was based on ultrasonic suspicious features and/or dominant nodule in case of multinodular goiter.
The median size of the nodules on which FNA was performed was 17 [range .This median was similar in the first and the second nodules and across genders.Most of the nodules were located in the right or the left lobe of the thyroid bed.Only a minor percentage (5%) was located in the isthmus.70% of the nodules were solid and almost 50% were isoechoic in the overall group and across genders.Irregular borders were reported in less than 10% of the nodules, without difference between genders.The presence of calcifications or of a halo was not reported in the majority of nodules, without difference between genders

Cytology findings
Table 3 shows the cytology findings and the final diagnosis of the nodules in which FNA was performed.75% the nodules in which FNA was performed (377 out of 498 nodules) were benign.The specimen was not adequate for cytological diagnosis in 60 (12%) of cases but FNA was repeated in 10 cases only and the repeat specimen was adequate for diagnosis in 9 out of 10 cases.Of these 9 cases, 5 were benign, two were papillary cancer, and one was suspicious for follicular neoplasm, one showed atypia for undetermined significance.
The final diagnosis after FNA repeat was as follows: benign in the majority of cases (76.7%).8% were suspicious for follicular cancer, 3.8% showed changes of papillary cancer, less than 1% were consistent with medullary cancer, 1% showed atypia of undetermined significance.The diagnosis remained unknown in the 50 nodules in which the first FNA was inadequate for diagnosis and was not repeated.

Relationship between adequacy of the FNA specimen and potential predictors
There was no relationship between the adequacy of the test and age of the patient, size of the nodule or gender.Hypoehogenicity was the only ultrasonic characteristic that showed association with test inadequacy (observed in 40% of nodules with inadequate test versus 21.4% of nodules with adequate test, p=0.01).(Table 4

Relationship between the final diagnosis by cytology, and potential predictors
66% of nodules in men versus 79% of nodules in women were benign (Table 3) with a significant difference in the distribution of malignant and benign nodules between genders (p=0.05,Table 5).A significant difference was also observed in the echogenecity between malignant and benign nodules (35% of the malignant nodules versus 19% of the benign nodules were hypoechoic, p=0.03).No significant correlation was observed between other reported radiological characteristics such as micro or macrocalcifications, nodule component, margins, presence of halo or hypervascularity

Discussion
In this study where 400 cases of FNA of thyroid nodules performed under ultrasound guidance at a tertiary care center in Lebanon, 75% of the nodules were benign, only one percent showed atypia of undetermined significance and the test failure rate was 12%.Although nodular thyroid disease is very common and its prevalence increases with age [3], thyroid cancer is not common and most of the thyroid nodules that undergo FNA are benign and only 5-15% have been shown to be malignant [3], Papillary thyroid cancer being the most common type of thyroid cancer.The findings of the current study are in line with what has been reported in the literature.
Ultrasound-guided fine needle aspiration (USG-FNA) is considered to be the most effective and reliable procedure for the evaluation of thyroid lesions.It is minimally invasive, provides fast and accurate results and has a diagnostic accuracy of 85-94% [2].However, non-diagnostic cytology results are a major limitation of USG-FNA requiring a repeat FNA because of the high malignancy rate, which is variable depending on different studies, reaching 14% in one study [4].The proportion of inadequate/nondiagnostic cytology results observed in our study was 12%, and this is a very acceptable rate since it falls within the range that was reported by others (10-21%).
Several studies have been conducted to identify predictors of failure in obtaining adequate specimen of USG-FNA.These included size less then 6mm [5,6], position of the nodules [7,8], cystic component [9][10][11] and the experience of the radiologist and pathologist [12].In our series, only hypoechogenecity was found to be associated with unsatisfactory results, whereas the age of the patient, the size of the nodule in which FNA was performed, the gender, the location of the nodule and other ultrasonic characteristics were not related.It has been suggested that hypoechoic nodules are likely to contain fibrotic changes, thus aspiration becomes difficult and the likelihood of specimen inadequacy increases.
A non diagnostic thyroid FNA result presents a clinical challenge.Because of the considerably high rate of malignancy in nodules with nondiagnostic cytology, which was reported in up to 11%, the American Thyroid Association (ATA), the American Association of Clinical Endocrinologists (AACE) and the European Thyroid Association (ETA) recommend a second FNA on the nodule that had non diagnostic results for the first FNA and further suggest that such a nodule should be considered for surgical resection in the case of repeated nondiagnostic US-FNAs [12].In our series, FNA was repeated in only 10 cases out of the 60 cases and the repeat specimen was adequate in 9 out of 10 cases, most of them being benign which might propose that these nodules with initial nondiagnostic results should be followed with close clinical observation and repeat imaging rather then repeating the FNA, an invasive procedure.However, this recommendation cannot be adopted because the test was repeated in 10 out of 60 test failure cases.Thus the diagnosis remained unknown in the majority of cases with test failure.
Many studies have been published in which the ability to predict whether a thyroid nodule is benign or malignant on the basis of US findings was assessed.Several US features have been found to be associated with an increased risk of thyroid cancer, including the presence of microcalcifications, hypoechogenecity, irregular margins, absence of a halo, predominantly solid composition, and intranodular vascularity [1].In addition, risk of malignancy is higher in solitary as opposed to multiple nodules.These results have not been shown in our series where only hypoechogenecity and male gender were found to be predictors of malignancy.
This study has some limitations.The experience of the radiologist performing the test was not adjusted for.USG-FNA is highly dependent on the skill of the performer.Most of the non-diagnostic cytology tests were not repeated, and the pathological reports were not checked, which could have affected the results.Nevertheless, this is to our knowledge, the first report studying the prevalence of thyroid cancer among patients referred for FNA under US guidance in Lebanon.The prevalence and the rate of non-diagnostic test were similar to those reported in the literature.

Conclusion
3 out of 4 thyroid nodules referred for USG-FNA are benign, mostly in women.Inadequate specimen was observed in 12% of cases.Hypoechogenecity but not other ultrasonic characteristics was associated with malignancy and with test failure.
are N (%) **: Final diagnosis based on the total number of nodules in which FNA aspiration was performed.

Table 1 .
General ultrasound findings in the whole study population, and by gender.
*p-value for difference between genders by Mann-Whitney test for continuous variables (age and size), and by chi-square or Fisher's exact test for proportions (number of nodules, presence of lymphadenopathy and hypervascularity).** FNA=fine needle aspiration.

Table 3 .
Cytologic diagnosis in the overall population*

Table 2 .
Ultrasound characteristics of the nodules that were selected for fine needle aspiration.

Table 4 .
) Relationship between adequacy of the FNA specimen and potential predictors a,b .
*: Values are N (%), **: p-value for difference between groups by Mann-Whitney for age, and size, and by chi-square or Fisher's exacttest for other variables, as appropriate.a : Based on 498 nodules selected for FNA.b : FNA=fine needle aspiration

Table 5 .
Relationship between the final diagnosis by cytology, and potential predictors.
*: p-value for difference between groups by Mann-Whitney for age, and size, and by chi-square or Fisher's exact-test for other variables, as appropriate.Values for categorical variables are N (%) *