Ethem Turgay Cerit M.D., Gazi University Faculty of Medicine, Endocrinology Department, Beşevler/Ankara/Turkey 06590, phone: 505 540 6182, fax: 312 215 42 04, email address: ceritturgay1@yahoo.com

Introduction

Thyroid nodules are among the most common endocrine disorders and are present in 10-70% of adults [1]. Fine-needle aspiration biopsy (FNAB), a minimally invasive and safe procedure usually performed on an outpatient basis, is considered the first-line test in the diagnosis of thyroid nodules as well as in distinguishing benign from malignant thyroid nodules [2, 3]. Ultrasonography (US)-guided FNAB was recently shown to have a sensitivity of 76-98% and a specificity of 71-100% in detecting thyroid cancers [4]. The specimens obtained were cytopathologically classified as benign, lesion (atypia) of undetermined significance, follicular neoplasm, suspicious, malignant, or nondiagnostic [5]. The major limitation of US-guided thyroid FNAB is non-diagnostic aspiration due to specimen inadequacy, which has been reported to occur at rates ranging from 1% to 25% [6-8]. Thyroid nodules can be biopsied by fine-needle aspiration technique (AT) or fine-needle capillary technique (CT). However, there is controversy about whether either technique is superior to the other [9-13]. The size of the needle used for biopsy also affects the accuracy of diagnosis. Although 20-27 gauge (G) needles are usually used for biopsy, few studies to date have compared the adequacy of samples obtained with different sized needles [10, 14-17]. Moreover, onsite evaluation of biopsy samples by a cytopathologist has been reported to increase the success rate of biopsy procedures, reducing the need to repeat the biopsy and decreasing the rate of false negative results [8, 18]. By reducing the number of interventions, on-site cytopathological evaluation may also reduce the risk of patient complications [8,19].

To the best of our knowledge, no previous study has simultaneously evaluated the effects of different needle sizes, biopsy techniques and the presence of onsite cytopathological analysis (OCA) on biopsy adequacy. This study therefore compared the adequacy of biopsy samples obtained by aspiration or capillary biopsy techniques, using 22G or 27G needles, and with/without OCA.

Material and methods

This study enrolled 400 consecutive patients, each with one thyroid nodule, who underwent US-guided fine-needle biopsy. Patients were divided into 8 groups of 50 patients each according to needle size (22 G vs. 27 G), sampling technique (aspiration vs. capillary), and OCA (performed vs. not performed) (Table 1). All biopsy procedures were performed by the same researcher, using a Hitachi EUB 8500 US system and a 10 MHz linear probe. All patients provided written informed consent, and the study protocol was approved by the Local Ethics Board.

The patient was placed in the supine position with the neck slightly extended. The dimensions, location, depth, and vascularity of each nodule were evaluated, as was its relationships with neighbouring structures such as the carotid artery. Nodule vascularity was evaluated by colour Doppler sonography and classified as type 1 (no vascularity), type 2 (peripheral vascularity alone), type 3 (peripheral and minimal central vascularity), or type 4 (prominent central vascularity) [20]. The neck was cleaned with povidoneiodine solution. Coagulation screening was routinely performed. The patients were instructed not to swallow during the needle insertion and aspiration. The biopsy needles and techniques for each patient were selected sequentially, independent of the sonographic characteristics of the nodules.

After localisation of the nodule, a biopsy specimen was acquired by the aspiration or capillary method, independent of the sonographic characteristics of the nodules. The aspiration method utilised a freehand biopsy technique. The needle was attached to a 10-mL syringe and inserted obliquely into the nodule parallel to the scanning plane. Once the needle tip was visualised inside the nodule, material was aspirated with a to-and-fro motion and smeared onto glass slides. In the capillary method, once the needle tip was visualised inside the nodule the syringe was detached and the needle advanced using a to-and-fro motion within the nodule while being rotated on its axis until blood was seen in the hub of the needle. Biopsy materials of each group of patients were air-dried. Slides of patients in Group I through IV (OCA negative groups) were sent to the pathology laboratory for cytopathological evaluation. In those groups, at least two needle passes were planned, although the number of needle passes was affected by the amount of material obtained, patient cooperation, and patient comfort. Half of the slides obtained from patients in Group V-VIII (OCA positive groups) were sent to the pathology lab, while the other half was stained with May-Grunwald Giernsa quick stain and evaluated onsite for material adequacy by an experienced cytopathologist. Adequacy was defined as six or eight groups of well-preserved follicular cells with 10 or more cells per group. Needle passes were repeated until material adequate for two glass slides was obtained, with a minimum of two needle passes. The cytopathologist evaluating the aspirates informed the researcher directly about ‘acceptance’ of any given slide as a diagnostic one by on-site assessment. This enabled evaluation of whether on-cite cytological assessment reduced the number of interventions. For partial cystic nodules, biopsies were taken from the solid parts. Pure cysts were aspirated and emptied; their contents were centrifuged and the pellet was examined for the presence of thyroid cells.

All slides were stained with haematoxylin and eosin for final cytopathological evaluation. All cytology specimens were finally interpreted by the same cytopathologist and were classified as benign, lesion (atypia) of undetermined significance, follicular neoplasm, suspicious, malignant, or nondiagnostic.

Statistical analysis

Categorical variables (sample adequacy, haematoma existence) were compared in sampling methodology sub-groups by using chi-square or Fisher’s exact tests whereas continuous variable (needle pass number) was compared in needle size sub-groups with Mann-WhitneyU test. Effects of sampling methods to estimate having adequate sample were assessed by univariate evaluation and mutually adjusted logistic regression models. Type 1 error was set to 0.05.

Results

The study population consisted of 400 patients (85.8% female, mean age: 48.1 years; min-max: 17-80 years) with thyroid nodule and who underwent US-guided fine-needle biopsy.

Nodule and biopsy characteristics

Of the 400 nodules, 205 (51.2%) were solid, 185 (46.3%) were partially cystic, and 10 (2.5%) were predominantly or purely cystic.

Nodule diameters ranged from 5 to 46 mm (mean: 16.1 mm) while 23% were larger than 20 mm, 58.5% between 10 and 20 mm, and 18.5% smaller than 10 mm. Study groups (definitions given in Table I) were similar in terms of nodule size (p = 0.227).

OCA sub-groups (used vs. not used) and needle size subgroups (22G vs. 27G) were similar in terms of nodule sonographic characteristics (p values = 0.679 and 0.438, respectively). Flowever dominantly or pure cystic nodule percentage was higher in the capillary technique subgroup than in the aspiration one (56.0% vs. 41.5%, p = 0.008).

None of tire patients had a history of thyroid surgery or radioactive iodine therapy. Chronic thyroiditis was reported in 7.3% of the patients, and all patients were euthyroid at the time of thyroid biopsy.

Using OCA significantly reduced the number of needle passes compared to not using it (median pass number 2 vs. 3 respectively, p < 0.001).

In terms of complications, 2.5% of patients in whom a 22G needle had been used and 0.0% of patients in whom a 27G needle had been used developed haematoma, respectively. Needle size sub-groups were statistically similar in terms of haematoma incidence (p = 0.061).

Sample adequacy

Adequacy rates of the study groups varied from 88.0% to 100.0%, and no statistically significant difference between groups was found (P = 0.054) (Table I).

Nodule size sub groups (< 10 mm, 10-20 mm, and > 20 mm) were similar in terms of sample adequacy rate (90.5%, 95.7%, and 93.5%, respectively; p = 0.232). The nodular vascularity pattern degree did not affect the sample adequacy rate in either capillary or aspiration technique subgroups (p value = 0.714 and 0.238, respectively).

Univariate analyses revealed that the capillary technique provided a better adequacy rate than aspiration technique (97.0% vs. 91.5%, p = 0.032) whereas using OCA was better than not using it (97.0% vs. 91.5%, p = 0.032). On the other hand, needle size did not have a significant effect on sample adequacy (94.5% vs. 94.0%, p = 0.830) (Table II).

Since nodule sonographic characteristics varied significantly in biopsy technique sub-groups, univariate analysis was repeated to include solid nodules only (n = 205). Similar adequacy rates were obtained for the analysis for solid nodules, per se. The capillary technique provided better adequacy rates than aspiration technique (98.9% vs. 89.7%, p = 0.008) whereas using OCA was better than not using it (97.9% vs. 89.6%, p = 0.014). On the other hand, needle size had no significant effect on sample adequacy (93.1% vs. 94.2%, p = 0.733) (Table II).

Data on univariate and multivariate logistic regression analysis of parameters for sample adequacy rates in solid nodules are presented in Figure 1. Mutually adjusted multivariate logistic regression model for solid nodules (n = 205) revealed that using onsite cytological analysis and using capillary biopsy technique increases the likelihood of having an adequate sample (by 5.81 and 10.28 times, respectively). However, needle size did not significantly affect the likelihood of having an adequate sample (Fig. 1).

Figure 1. Odds ratios of selected parameters for sample adequacy rates in solid nodides (residts of univariate and midtivariate logistic regression) (n = 205)
Rycina 1. Ilorazy szans wybranych parametrów dla wskaźnika wydajności próbki w guzkach litych (wyniki jednoczynnikowej i wieloczynnikowej regresji logistycznej) (n = 205)

Discussion

Thyroid nodules can be biopsied by fine needle aspiration technique or capillary technique, but there is some controversy about whether either technique is superior to the other [9-13]. Aspiration during biopsy may lead to a number of reactive changes, such as papillary endothelial hyperplasia, haemorrhage, vascular thrombosis and proliferation, fibrosis, cystic transformation, and infarction [18]. Trauma-related consequences may be minimised by using a fine-needle biopsy procedure without aspiration (fine-needle capillary technique). In this method, cells are detached by the cutting edge of the needle and are conducted into the lumen by capillary forces [18]. The absence of the vacuum effect of aspiration results in a significant reduction in trauma to the nodule and surrounding thyroid tissues and thus reduces blood contamination and yields higher quality material [18, 21]. The capillary technique is therefore recommended, especially for nodules found to be highly vascular on colour Doppler imaging [4, 10, 11]. However, no previous study has investigated the effects of nodular vascularity patterns on the adequacy rates of the capillary technique, and few studies have evaluated the effects of capillary technique on diagnostic adequacy. We found that inadequacy rates were significantly lower with the capillary than with the aspiration sampling technique. Moreover, adequacy rates with the capillary technique were independent of nodular vascularity. These findings suggest that the capillary technique is an easily applied method that causes less blood contamination and provides higher amounts of cytological material than the aspiration technique.

On-site evaluation of samples by a cytopathologist has been reported to increase biopsy success rates and reduce the rates of biopsy repetition and false-negative results [8, 18]. Data from a retrospective study lasting 10 years revealed that on-site cytopathological evaluation reduced inadequacy rates from 14% to 4% [22]. Another study, however, showed that adequacy rates of US-guided thyroid nodule FNAB were similar with and without immediate cytological assessment; moreover, on-site cytopathological evaluation prolonged the duration of biopsy [19]. On-site cytopathological evaluation can potentially reduce the number of needle passes, thus reducing patient discomfort and risk of complications [8, 19]. We found that inadequacy rates and the median number of needle passes were significantly lower with than without OCA. Thus, our findings seem to indicate that on-site cytopathological evaluation increases adequacy rates and patient comfort, while avoiding repeat biopsies and unnecessary interventions.

It is unclear whether the cellularity of a biopsy specimen correlates with the size of the needle [10, 14-17], In most studies 20-27G needles are used for thyroid fine needle biopsies [10, 14, 15, 19]. If the lumen of the needle is thinner, sufficient material for diagnosis may not be obtained; whereas, if the needle is thicker, the material obtained may be contaminated with blood, making microscopic evaluation more difficult [4, 10, 14, 17]. Blood contamination is especially prevalent in hypervascular nodules [4]. We found no significant difference in adequacy rates between specimens obtained with 22G and 27G needles, and the needle size did not affect the complication rate.

Inclusion of a large cohort with 400 cases and combined evaluation of biopsy technique, needle size, and on-site cytological analysis along with nodular vascularity pattern in terms of their effects on sample adequacy rate are the major strengths of the present study. The major limitation of the present study is the lack of homogeneity between groups in terms of nodular characteristics. Nevertheless, the large sample size enables us to repeat analysis to include solid nodules, per se (n = 205), after exclusion of dominantly cystic and partially cystic nodules, which also revealed similar findings related to the effects of biopsy technique, needle size, and on-site cytological analysis on adequacy.

Conclusions

In conclusion, this study compared the adequacy of thyroid biopsy samples obtained with different sized needles (22G vs. 27G), different sampling techniques (aspiration vs. capillary), and with/without onsite cytopathological evaluation. Optimal results were obtained using the capillary sampling technique and onsite cytological evaluation. Adequacy rates and patient comfort may both be optimised by using a capillary sampling technique and onsite cytopathological evaluation.