Ultrasound-guided fine needle aspiration of thyroid nodules: factors affecting diagnostic outcomes and confounding variables

Abstract

Background

The incidence of thyroid cancer is increasing in men and women. Fine needle aspiration (FNA) is an accepted technique to assess thyroid nodules but is associated with a high rate of non-diagnostic sampling.

Purpose

To assess the diagnostic performance of ultrasound-guided FNA of thyroid nodules and identify factors associated with non-diagnostic sampling.

Material and Methods

A retrospective review of thyroid FNAs was performed between 2006 and 2013. Patient demographics, nodule characteristics, procedural technique, cytology, and complications were recorded. Cytology was categorized THY1-5 based on the British Thyroid Association guidelines. Descriptive and multivariable analysis were conducted to identify factors associated with non-diagnostic sampling.

Results

A total of 724 procedures were identified with 597 (82.5%) in women, and an overall mean age of 40 years (age range, 17–87 years). Factors associated with a non-diagnostic outcome in the multivariable regression analysis included increasing lesion depth (OR, 1.05 per mm; 95% confidence interval [CI], 1.007–1.10), age (OR, 1.012 per year; 95% CI, 1.0–1.025) and number of FNA passes (1 vs. 4+; OR, 6.07; 95% CI, 2.27–16.21). The complication rate was 1.1% related to perilesional hematomas and vaso-vagal episodes.

Conclusion

Thyroid FNA is a safe and reliable procedure for cytological assessment of thyroid nodules. Deeper nodules and older patients are more likely to have non-diagnostic samples.

Keywords

Thyroid nodules fine needle aspiration cytology non-diagnostic

Introduction

Thyroid carcinoma is a relatively uncommon cancer with a low mortality; however, the incidence is increasing in both men and women (1,2). Benign thyroid nodules are extremely common and can be detected by ultrasound in up to 68% of the general population (3 –6). Differentiation of benign and malignant thyroid nodules is not possible clinically and most professional society guidelines recommend ultrasound to help characterize thyroid nodules (7). The sensitivity and specificity of ultrasound for predicting a malignant nodule is highly variable and cytological sampling of solid nodules is required (7). Fine needle aspiration (FNA) has been shown to be a safe, accurate, and cost-effective method for the evaluation of thyroid nodules. A major limitation of the procedure is the high rate of inadequate or “non-diagnostic” specimens, in the range of 5–53.8% in the literature (8 –14). Non-diagnostic FNA samples result in an increase in overall FNA cost and can increase patient anxiety (15). Ultrasound-guided FNA improves the diagnostic accuracy; however, a significant non-diagnostic rate remains (16).

Our institution performs ultrasound-guided FNA of thyroid nodules based on the Society of Radiologists in Ultrasound guidelines which recommends FNA for nodules >10 mm unless cystic in nature (7). Nodules <10 mm undergo FNA if there are suspicious imaging features such as an irregular border of the nodule or internal micro calcifications (7). The aim of this study was to assess the diagnostic performance of ultrasound-guided thyroid FNA and to identify factors associated with non-diagnostic sampling.

Material and Methods

Patients

Institutional ethics board approval was granted for this retrospective study and consent was not required. A retrospective review of the hospital Electronic Patient Record (EPR) was performed from July 2006 to March 2013 to identify all patients who had undergone a thyroid FNA. Patients were included if they had a thyroid FNA performed in the study period and the ultrasound imaging and cytology result were available for review.

Nodule imaging

Patient imaging was performed with a high frequency linear array 18-5 MHz transducer on either a Philips iU22 (Philips Healthcare, Best, The Netherlands) or Hitachi Hi Vision Preirus (Hitachi Medical Imaging, Yokohama, Japan) machine by a trained sonographer, trainee radiologist, or staff radiologist. The imaging was stored on the hospital picture archiving and communication system (PACS). All imaging was retrospectively reviewed by JK (with 4 years of experience in thyroid ultrasound imaging) to assess the nodules for the following imaging characteristics; size, echogenicity, internal calcifications, nodule vascularity, internal architecture (cystic or solid), percentage cystic composition of the nodule, distance from skin surface, and depth in thyroid.

FNA technique

A consultant radiologist or a trainee radiologist under the supervision of a consultant performed the FNAs. All trainee radiologists had at least 3 years of experience in thyroid FNA (16 trainees). Consultants (5 radiologists) had a minimum of 8 years of experience in thyroid imaging and FNA. The FNA technique employed has been previously published and performed after informed patient consent (17). Routine anticoagulation with aspirin and clopidogrel was not discontinued prior to FNA; however, Warfarin was adjusted to an INR <2 as per departmental protocol (18). Initially the nodule was localized with ultrasound and the skin site overlying the nodule was marked. The skin and subcutaneous tissue were infiltrated with up to 5 mL of 1% Lignocaine. Fine gauge needles (23 G or 20 G) were attached to a 5 cc or 10 cc syringe and inserted into the nodules under real-time ultrasound guidance using a freehand technique. Capillary technique was used with a small amount of suction at the end of the FNA (prior to removal of the needle). The size of needle used and number of passes was operator-dependent. The needles were then rinsed in an alcohol-based preservative and the samples were transferred to the cytology laboratory for analysis. Complications were recorded from the radiology report. The following information was recorded for each procedure: operator, needle gauge, number of FNA samples obtained, and any reported complications.

Cytology result

The cytology result was obtained from the hospital EPR system. Cytological samples were categorized THY1-5 based on the British Thyroid Association/Royal College of Physicians guidelines (19). THY1 classification describes non-diagnostic nodules that have inadequate cellular sample or where technical artefact precludes interpretation. THY2 classification defines non-neoplastic nodules which have features consistent with thyroiditis or colloid nodules. Cysts containing benign epithelial cells are classified as THY2. Cysts may be classified as THY2 if benign epithelial cells are present. THY3 classification describes follicular lesion or suspected follicular neoplasm. While many THY3 nodules will be hyperplastic nodules, some will be tumors. The differentiation between the two is based on the nodule histology. THY3 classification can also describe those nodules which show features of malignancy, but cell numbers are too few to qualify the nodule as THY4. THY4 classification describes nodules suspicious, but not diagnostic, of papillary, medullary, or anaplastic carcinoma, or lymphoma. THY5 classification describes nodules that have unequivocal features of papillary, medullary, or anaplastic carcinoma, lymphoma, or metastatic tumor and is diagnostic of malignancy.

Statistical analysis

Descriptive analyses are presented as proportions, means, standard deviations (SD), median, ranges, or inter-quartile range (IQR). Univariate analysis was performed to compare characteristics between diagnostic and non-diagnostic groups. Chi-square tests were used for comparison of proportions and non-parametric Mann–Whitney U tests for comparisons of age, nodule size, depth to skin nodule, and depth to gland, due to being non-normal data. Multivariable logistic regression was used to predict non-diagnostic outcome from the factors described above, and considered in the univariate analyses. Odds ratios (OR) and 95% confidence intervals are presented, with P values, as an estimate of the effect of each factor on the outcome. Data were collated using Excel 2010 (Microsoft, Redmond WA, USA) and statistical analysis was performed using SPSS software (version 14.0, SPSS Inc., Chicago, IL, USA). Significance at P < 0.05 was assumed.

Results

Patients

A total number of 734 thyroid FNA procedures were identified from the EPR system. Ten procedures were excluded due to the lack of an available cytological report. This resulted in a study population of 724 participants (597 women; 82.5%) and a mean age of 40 years (age range, 17–87 years). Of these, 93 were repeat procedures (13%).

Nodule characteristics

The mean nodule size was 22.5 mm (range, 5–110 mm). A total of 48.3% (n = 349) of nodules were solid, 48.0% (n = 348) were mixed (solid and cystic), and 3.7% (n = 27) were purely cystic. The mean percentage cystic component for mixed nodules was 40.4%. Internal flow was noted in 14.0% (n = 101) of nodules. A total of 6.2% (n = 45) of nodules were hyperechoic to thyroid parenchyma, 40.3% (n = 292) were isoechoic, and 53.5% (n = 387) were hypoechoic to thyroid parenchyma. Macrocalcifications were present in 8.3% (n = 60) of nodules and microcalcifications were identified in 11.9% (n = 86) of nodules. The mean depth of the nodules from the skin was 13.5 mm (range, 3–34 mm) and the mean depth of the nodules from the anterior margin of the thyroid gland was 2.0 mm (range, 0–23 mm). The mean number of FNA passes was 2.8 (range, 1–5) with a mean gauge size of 22.9 G (range, 20–23).

Cytological correlation

In total, 156 (21.5%) samples were THY1 (non-diagnostic). The mean size for THY1 was 16 mm (range, 5–46 mm) with average skin to lesion depth of 13 mm (range, 8–18 mm) and mean cystic component in THY1 nodules of 30.0%. A total of 471 (65.0%) samples were THY2 nodules. The mean nodule size for THY2 nodules was 23 mm (range, 5–57 mm) and mean depth was 8 mm (range, 4–23 mm), the mean nodule cystic component was 21.9%. Ninety-seven (13.5%) were THY3-5 nodules (47 THY3/a/f; 17 THY4; 33 THY5). The mean nodule size for THY3-5 was 21 mm (range, 5–70 mm). The average depth was 11 mm (range, 5–20 mm) and the mean cystic component was 14.%.

Factors associated with a non-diagnostic sampling

In the univariate analysis comparing non-diagnostic and diagnostic groups, there was a significant association between non diagnostic sampling and the patient’s age (P = 0.02), depth of the lesion within the gland (P < 0.01), and the number of FNA samples performed (P < 0.01) (Table 1). In addition, the experience of the operator (trainee radiologist versus consultant radiologist) was also associated with greater likelihood of a non-diagnostic sample (P = 0.047), with a trainee radiologist more likely to have a non-diagnostic sample. Nodule appearance, size, and echogenicity were not associated with the diagnostic outcome (Table 1). Results from multivariable logistic regression analysis are presented in Table 2. After adjusting for technical and nodule characteristics, the depth to the target lesion remained significantly associated a with non-diagnostic outcome (OR, 1.05; 95% CI, 1.007–1.10). Other factors remaining significant after multivariate analysis were the age of the patient and the number of FNA passes performed (Table 2).

Table 1.

Univariate associations between patient characteristics and non-diagnostic/diagnostic outcome.

	Non-diagnostic (n = 158)	Diagnostic (n = 566)	P value
Repeat procedure (% Yes)	21 (13.3%)	72 (12.7%)	0.85
Gender (% Male)	31 (19.6%)	95 (16.8%)	0.41
Appearance (%)			0.14
Cystic	10 (6.3%)	17 (3%)
Solid	76 (48.1%)	274 (48.4%)
Mixed	72 (45.6%)	275 (48.6%)
Echogenecity (%)			0.17
Hypoechoic	93 (59.2%)	294 (51.9%)
Isoechoic	53 (33.8%)	238 (42.0%)
Hyperechoic	11 (7%)	34 (6.%)
Operator (% Consultant)	36 (22.8%)	175 (30.9%)	0.047
Size of needle (% 23 G)	140 (88.6%)	528 (93.3%)	0.052
FNA passes (n)			<0.0001
1	13 (8.3%)	11 (1.9%)
2	38 (24.2%)	115 (20.3%)
3	74 (47.1%)	257 (45.4%)
4+	32 (20.4%)	183 (32.3%)
Age (years) (median, IQR)	56.5 (44–68)	53 (41–64)	0.021
Nodule size (mm)(median, IQR)	21 (15–30)	20 (14–29)	0.39
Depth to skin nodule (mm) (median, IQR)	15 (11 –17)	13 (10 –16)	0.003
Depth in gland (median, IQR)	0 (0–3)	0 (0–3)	0.93

IQR, interquartile range.

Table 2.

Multivariable logistic regression analysis predicting non-diagnostic outcome (OR and 95% CI).

	OR	95% CI	P value
Appearance			0.93
Cystic	1.11	0.42–2.95
Solid (reference)	1
Mixed	1.08	0.72–1.60
Echogenecity			0.20
Hypoechoic	0.85	0.40–1.81
Isoechoic	0.61	0.28–1.32
Hyperechoic (reference)	1
Operator (Registrar vs. Consultant)	1.27	0.82–1.96	0.29
FNA passes (n)			0.002
1	6.07	2.27–16.21
2	1.90	1.10–3.30
3	1.64	1.02–2.64
4+ (reference)	1
Age (years)	1.012	1.0–1.025	0.049
Nodule size (mm)	1.012	0.996–1.03	0.15
Depth to skin nodule (mm)	1.05	1.007–1.10	0.022
Size of needle (23 G vs. 20 G)	0.64	0.35–1.19	0.16

CI, confidence interval; OR, odds ratio.

Complications

The complication rate was 1.1% consisting of six post-procedural hematomas (five occurred in THY1 nodules) and two vasovagal episodes. No significant morbidity or mortality was recorded. The hematomas were self-limiting and no admission was required. The incidence of complications was too low to assess for any significant association with technical factors.

Discussion

This study demonstrates that FNA of deeper thyroid nodules in older patients is more likely to be associated with non-diagnostic sampling. In addition, the greater the number of FNA passes, the more likely an FNA is to be diagnostic. Thyroid nodules are a frequent clinical problem with nodule detection rates in the range of 19–67% of randomly selected individuals; however, the incidence of malignancy is relatively low (20). A number of ultrasound imaging features have been described as being associated with malignancy such as internal vascular flow and microcalcifications; however, the diagnostic accuracy is low (21). The majority of non-cystic nodules will require FNA for definitive assessment (7,19).

The number of ultrasound-guided FNAs have risen dramatically over the past two decades with increased acceptance of the additional benefit of ultrasound guidance when performing the procedure (10). The non-diagnostic rate remains highly variable with reports in the range of 5–58.3% (8 –14). The non-diagnostic rate in our study was 21.5%, which is 6.5% above the range recommended rate in the American Association of Endocrinologists/ Associazione Medici Endocrinologi/European Thyroid Association (AACE/AME/ETA) guidelines, however, compares favorably to recently published large series (22,23). The higher non-diagnostic rate in our cohort may reflect the lack on onsite cytological assessment at the time of the FNA, which has been shown to improve diagnostic performance (24). In addition, some of our referring surgeons perform clinical FNAs on larger palpable nodules and only refer the more technically challenging deeper and smaller nodules for ultrasound-guided FNA.

The study set out to establish technical factors and nodule characteristics associated with non-diagnostic sampling. Nodules that were located further from the skin surface were more likely to be non-diagnostic. Deeper nodules may be more difficult to FNA as the needle may be more difficult to manipulate at greater depths and image resolution deteriorates at increasing depth (23). In addition, deeper nodules are more difficult to achieve complete anesthesia which may increase patient discomfort and increase nodule movement. The number of FNA passes had a significant effect on diagnostic outcome. The operator technique was heterogeneous with variation in the number of FNA passes and needle size. The needle gauge did not alter the outcome, however, reducing the number of passes from four to one increased the likelihood of a non-diagnostic result by a factor of six.

The incidence of malignant thyroid cytology diagnosed on FNA decreases with age with incidences of 2.3% in men and 1.9% in women aged over 70 years reported in a recent series and the highest incidence in men aged less than 45 years (25). The incidence of malignant nodules (THY4 or 5) in our cohort was 6.9%, which may reflect our institution being a national referral center for head and neck cancer. The incidence of thyroid nodules increases with age and our study supports the association of increasing age with non-diagnostic cytology (26,27). The reason for this is uncertain but may be due to associated co-morbidities in older patients resulting in difficulty with stable neck positioning.

The overall complication rate was low at 1.1%, with no patients requiring hospital admission and no serious morbidity or mortality. Six patients had a post-procedure hematoma, which all settled with conservative management. We do not routinely discontinue anti-coagulation prior to FNA and this practice is supported by the very low rate of significant hematoma reported and none resulting in airway compromise. Unfortunately, data were not available on whether patients who developed a hematoma were taking anticoagulants. Interestingly, five of the six patients with a hematoma had a THY1 sample which may reflect a more difficult procedure resulting in a hematoma or an immediate hematoma resulting in a more difficult FNA. Two patients developed a vasovagal episode, which required no treatment.

This study has some limitations. The factors associated with a non-diagnostic FNA were limited to technical factors and nodule characteristics at the time of FNA. The influence of cytological analysis was not assessed on the outcome of the FNA. There was also variation in technique with certain operators performing more FNA passes and some operators having significantly more experience. These factors are undergoing further analysis. The data on complications relied on the radiology reports and the electronic patient record which may not include all minor complications reported by a patient.

In conclusion, ultrasound-guided FNA of thyroid nodules is a very safe procedure; however, FNA of deeper thyroid nodules in older patients are more likely to be non-diagnostic. It may be more prudent to complete imaging follow-up of a deep nodule in an elderly patient due to the increased likelihood of a non-diagnostic result.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

Assessment UENC for E. SEER cancer statistics review, 1975–2005. 2009. Available at: http://hero.epa.gov/index.cfm/reference/details/reference_id/730406 (accessed 17 January 2015).

Chen

Jemal

Ward

. Increasing incidence of differentiated thyroid cancer in the United States, 1988–2005. Cancer 2009; 115: 3801–3807.

Desforges

Mazzaferri

. Management of a solitary thyroid nodule. N Engl J Med 1993; 328: 553–559.

Gharib

Papini

Paschke

, et al. Medical guidelines for clinical practice for the diagnosis and management of thyroid nodules. Endocr Pr 2006; 12: 63–93.

Cooper

Doherty

Haugen

, et al. Revised American Thyroid Association management guidelines for patients with thyroid nodules and differentiated thyroid cancer. Thyroid 2009; 19: 1167–1214.

Guth

Theune

Aberle

, et al. Very high prevalence of thyroid nodules detected by high frequency (13 MHz) ultrasound examination. Eur J Clin Invest 2009; 39: 699–706.

Frates

Benson

Charboneau

, et al. Management of thyroid nodules detected at US: Society of Radiologists in Ultrasound consensus conference statement. Ultrasound Q 2006; 22: 231–238. discussion 239–240.

H-M

Jhu

I-K

Yang

S-H

, et al. Clinicopathologic analysis of fine needle aspiration cytology of the thyroid. Acta Cytol 2003; 47: 727–732.

Carmeci

Jeffrey

McDougall

, et al. Ultrasound-guided fine-needle aspiration biopsy of thyroid masses. Thyroid 1998; 8: 283–289.

10.

Danese

Sciacchitano

Farsetti

, et al. Diagnostic accuracy of conventional versus sonography-guided fine-needle aspiration biopsy of thyroid nodules. Thyroid 1998; 8(1): 15–21.

11.

Hatada

Okada

Ishii

, et al. Evaluation of ultrasound-guided fine-needle aspiration biopsy for thyroid nodules. Am J Surg 1998 Feb 1; 175(2): 133–6.

12.

Tabaqchali

Hanson

Johnson

, et al. Thyroid aspiration cytology in Newcastle: a six year cytology/histology correlation study. Ann R Coll Surg Engl 2000; 82: 149–155.

13.

Morgan

Serpell

Cheng

MSP

. Fine-needle aspiration cytology of thyroid nodules: how useful is it? ANZ J Surg 2003; 73: 480–483.

14.

De Meer

SGA

Schreinemakers

JMJ

Zelissen

PMJ

, et al. Fine-needle aspiration of thyroid tumors: Identifying factors associated with adequacy rate in a large academic center in the Netherlands. Diagn Cytopathol 2012; 40: E21–26.

15.

Borget

Vielh

Leboulleux

, et al. Assessment of the cost of fine-needle aspiration cytology as a diagnostic tool in patients with thyroid nodules. Am J Clin Pathol 2008; 129: 763–771.

16.

Alexander

Heering

Benson

, et al. Assessment of nondiagnostic ultrasound-guided fine needle aspirations of thyroid nodules. J Clin Endocrinol Metab 2002; 87: 4924–4927.

17.

Kim

E-K

Park

, et al. US-guided fine-needle aspiration of thyroid nodules: indications, techniques, results. Radiographics 2008; 28: 1869–1886.

18.

Patel

Davidson

Nikolic

, et al. Addendum of newer anticoagulants to the SIR consensus guideline. J Vasc Interv Radiol 2013; 24: 641–645.

19.

Perros

Boelaert

Colley

, et al. Guidelines for the management of thyroid cancer. Clin Endocrinol (Oxf) 2014; 81: 1–122.

20.

Tan

Gharib

. Thyroid incidentalomas: management approaches to nonpalpable nodules discovered incidentally on thyroid imaging. Ann Intern Med 1997; 126: 226–231.

21.

Frates MC, Benson CB, Doubilet PM, et al. Likelihood of thyroid cancer based on sonographic assessment of nodule size and composition. In: Radiological Society of North America Scientific Assembly and Annual Meeting Program. Chicago, IL, 28 November–3 December 2004. Radiological Society of North America. Available at: http://archive.rsna.org/2004/4409626.html (accessed 2 June 2016).

22.

Isaac

Jeffery

Seikaly

, et al. Predictors of non-diagnostic cytology in surgeon-performed ultrasound guided fine needle aspiration of thyroid nodules. J Otolaryngol Head Neck Surg 2014; 43: 48–48.

23.

Grani

Calvanese

Carbotta

, et al. Intrinsic factors affecting adequacy of thyroid nodule fine-needle aspiration cytology. Clin Endocrinol (Oxf) 2013; 78: 141–144.

24.

Baloch

Tam

Langer

, et al. Ultrasound-guided fine-needle aspiration biopsy of the thyroid: Role of on-site assessment and multiple cytologic preparations. Diagn Cytopathol 2000; 23: 425–429.

25.

Bessey

Lai

NBK

Coorough

, et al. The incidence of thyroid cancer by fine needle aspiration varies by age and gender. J Surg Res 2013; 184: 761–765.

26.

Baier

Hahn

Gervais

, et al. Fine-needle aspiration biopsy of thyroid nodules: experience in a cohort of 944 patients. Am J Roentgenol 2009; 193: 1175–1179.

27.

Hambly

Gonen

Gerst

, et al. Implementation of evidence-based guidelines for thyroid nodule biopsy: a model for establishment of practice standards. Am J Roentgenol 2011; 196: 655–660.