Fine-needle aspiration (FNA) of thyroid bed (TB) lesions is a common diagnostic modality in monitoring patients for recurrent cancer after a thyroidectomy. To elucidate the value of TB FNA, we reviewed our experience at The Johns Hopkins Hospital, Baltimore, MD. We identified 57 TB FNA specimens from 50 patients. Of the patients, 36 were being followed up for papillary carcinoma, 7 for medullary carcinoma, 4 for follicular carcinoma (1 also had papillary carcinoma), and 1 for poorly differentiated neuroendocrine carcinoma; 3 had previous benign diagnoses. TB FNA yielded diagnostic material in 49 of 57 cases. Of 37 malignant or atypical FNA samples, 32 had surgical follow-up; 30 of 32 were confirmed malignant. The FNA result was benign in 12 of 57, including 6 cases of benign thyroid and 1 case of parathyroid tissue. Immunohistochemical staining was contributory in 5 of 57 cases. TB FNA is a highly reliable tool for diagnosing recurrent thyroid carcinoma. Residual benign thyroid and parathyroid tissue are potential pitfalls; awareness of these and judicious use of immunohistochemical staining can prevent misdiagnoses.
Total or near-total thyroidectomy followed by surveillance of the neck and thyroid bed (TB) for recurrent disease is the mainstay of treatment for primary thyroid cancers, including papillary carcinoma, follicular carcinoma, and medullary carcinoma.1,2 Any new lesion detected in the TB radiographically or by physical examination requires further investigation. The distinction between recurrent tumors and benign causes is not always possible with imaging techniques.3 Tissue confirmation of recurrent tumor is necessary because recurrent tumor warrants additional treatment, and a benign condition does not need additional therapy. Fine-needle aspiration (FNA) is often the first tissue study performed in this setting.2
FNA of primary thyroid lesions is an excellent diagnostic modality for triaging patients to definitive surgical therapy for potential cancers or to conservative follow-up for benign lesions.4 Diagnostic criteria for the various primary thyroid neoplasms are well established for cytologic preparations.5 Although ultrasound-guided FNA is known to be a sensitive and specific modality for examining a new lesion in the TB after thyroidectomy for a primary thyroid cancer, to our knowledge, only 1 prior report has detailed the cytologic findings in this setting.6 We present herein our experience at The Johns Hopkins Hospital, Baltimore, MD, to better define the usefulness of FNA of TB lesions in patients who have undergone thyroidectomy and to address possible diagnostic dilemmas.
Materials and Methods
The cytopathology archives at The Johns Hopkins Hospital were searched, and all cases of FNA of TB lesions were identified for a 17-year period (1993–2009). Search parameters were restricted to lesions in the TB; cases in the nearby neck and regional lymph nodes were excluded. For every case, the cytology reports, available cytology slides, precedent surgical pathology slides, and reports and any follow-up surgical pathology slides and reports were reviewed. Clinical information, including patient demographics, TB nodule characteristics (size, laterality, and number of lesions), interval from thyroidectomy to TB FNA, and interval treatment, was tabulated.
FNA was performed under ultrasound guidance with 25- or 23-gauge needles. A cytopathologist or cytotechnologist was on site for assessment of specimen adequacy in all cases. Direct smears were air dried for rapid Romanowsky staining and fixed in 95% ethanol for Papanicolaou staining. The rinses from the syringe in each case were rinsed into Hank balanced salt solution to make cytocentrifuged preparations or a cell block. Cell-block material was stained with H&E, and, in some cases, immunohistochemical staining was performed.
The results of the TB FNA studies are summarized in Table 1. The study included 57 specimens from 50 patients that were identified and included in the analysis. Of the 50 patients, 36 were followed up for diagnoses of papillary carcinoma, 7 had medullary carcinoma, and 4 had follicular carcinoma (including 1 Hürthle cell carcinoma). One of those patients had a history of papillary and follicular carcinomas. In addition, 1 patient had a history of poorly differentiated neuroendocrine carcinoma, 1 had Hürthle cell adenoma, 1 had Hashimoto thyroiditis, and 1 had nodular hyperplasia on the respective precedent thyroidectomy specimens. Mean age at time of TB aspiration was 52.2 years (range, 15–86 years), and the male/female ratio was 1:2.2. The average interval from thyroidectomy to TB FNA was 95 months (range, 3–410 months).
The average size of the TB nodule was 2.2 cm (range, 0.4–5 cm). Of the 45 cases in which the side of the original lesion was known, 19 had TB nodules on the ipsilateral side, 15 had TB nodules lateralizing to one side following excision of a midline primary or multiple bilateral primary disease, 2 cases had bilateral TB nodules following a unilateral primary tumor, 1 case had bilateral TB nodules following an original diagnosis of Hashimoto thyroiditis, 7 cases had TB nodules on the contralateral side from the primary lesion, and 1 case had a midline TB nodule following a unilateral primary tumor.
Of the 57 total cases, recurrent papillary thyroid carcinoma was diagnosed in 26 (46%), medullary carcinoma in 7 (12%), and poorly differentiated neuroendocrine carcinoma in 1 (2%). Of the 57 cases, 3 (5%) were regarded as atypical but not diagnostic for papillary carcinoma. The remaining 20 cases had benign findings (n = 12) or specimens were unsatisfactory (n = 8). The 12 benign cases included 6 cases of benign thyroid tissue, 5 cases of benign inflammatory tissue, and 1 case of parathyroid tissue. Of the 7 cases of TB FNA contralateral to the original thyroid primary tumor, 2 were papillary thyroid carcinoma, 2 cases were atypical but not diagnostic for papillary carcinoma, and the other 3 cases were benign inflammatory cells. Of the 34 total cases in which recurrent carcinoma was definitively diagnosed on TB FNA, 30 cases had surgical pathology follow-up. In 29 of 30, recurrent tumor was confirmed by subsequent excision. There was 1 case of recurrent papillary carcinoma diagnosed on TB FNA in which the surgical pathology material found only skeletal muscle and scarring.
Of the 3 atypical cases, 2 had surgical pathology follow-up; 1 of 2 had confirmed recurrent papillary carcinoma. The other case consisted of hyperplastic thyroid tissue on subsequent resection. None of the inadequate specimens or the cases of benign findings had surgical pathology follow-up. Clinical follow-up for the cases with benign TB FNA diagnoses was also benign (ie, no changes in the TB nodule on clinical or radiographic examinations) in all but 1 case in which recurrent medullary thyroid carcinoma had been diagnosed 11 months previously and was found again on TB FNA 9 months later.
The TB FNA samples showed cytomorphologic features characteristic of their respective primary diagnoses Image 1 and Image 2. Immunohistochemical studies were performed on 6 cases: 4 medullary carcinomas (calcitonin in all 4; thyroid transcription factor-1, thyroglobulin, and CD56 in 1), the poorly differentiated thyroid carcinoma with neuroendocrine features (synaptophysin, chromogranin, and CD56), and the 1 case with parathyroid tissue (synaptophysin). These studies were confirmatory in 5 of 6 cases.
Our results indicate that FNA of TB lesions in patients who have undergone thyroidectomy is an excellent diagnostic modality for confirming recurrent tumor and also for establishing benign diagnoses. Reoperative thyroid surgery is technically difficult because of scar tissue formation and distortion of normal anatomy and carries a greater risk of iatrogenic injury to recurrent laryngeal nerves and parathyroid glands.7 Because of the increased risk of surgery, obtaining a definitive diagnosis via less invasive FNA in TB nodules is desirable so that benign lesions can be observed without subjecting a patient to an unnecessary surgery that has significant risk. The issue of increased scarring in the area targeted for FNA also raises a concern in regard to anticipated specimen adequacy. In other sites, tumors associated with intense desmoplastic response often yield hypocellular aspirates resulting in inadequate, equivocal, or even falsely negative cytology results.2,7
The rate of obtaining an adequate specimen in our study was high with 49 (86%) of 57 TB FNA samples yielding adequate cellularity for definitive interpretation. In 29 of 30 cases with positive cytology results, follow-up surgical pathology results were concordant, for a positive predictive value of 97%. In the only discrepant case, a true surgical resection was not attempted (only biopsies were done after the surgeon decided there was no resectable mass), so in our series, there were no false-positive results related to interpretive error of the cytopathologist. In the 1 discrepant “atypical” case (with benign hyperplastic tissue found on resection), reactive atypia was actually favored cytologically.
It is also important to note that in 11 of 12 cases with adequate cellularity and benign findings, clinical and radiographic follow-up supported benignancy. In the remaining case, findings consistent with a reactive lymph node and a foreign body giant cell reaction were seen; however, this patient was being followed up closely for previously recurrent medullary thyroid carcinoma, which was found again in the TB months later.
In our series, patients being followed up for papillary and medullary thyroid carcinoma had TB FNA samples positive for malignancy far more often (65% and 70%, respectively) than patients followed up for follicular carcinoma or Hürthle cell carcinoma (0/4). Although our sample was admittedly small, this may be reflective of the fact that papillary and medullary thyroid carcinomas are more prone to recur locally than is follicular carcinoma.8
Analysis of the clinical features of the patients in our study showed that most of the time, nodules that proved to be recurrent tumor localized to one side, usually the side ipsilateral to the primary thyroid lesion (although this was not uniformly true). This finding is not surprising considering the lymphatic drainage of the anterior part of the neck. There was quite a range in the duration between thyroidectomy and FNA of TB in our study. Notably, 1 patient had a malignant FNA 341 months after thyroidectomy, so an extremely long interval should not dissuade one from diagnosing recurrent disease in this setting.
There are several benign causes of TB nodules in patients who have undergone thyroidectomy. In our series 6 (50%) of 12 cases with benign findings had benign thyroid tissue. In each case, unremarkable follicular epithelial cells and colloid were identified. Although it is counterintuitive to encounter benign tissue from an organ that has been resected, it has been shown that most patients actually have thyroid remnants after “total” thyroidectomy for thyroid cancers, with just over half of the remnants located in the TB.9 Therefore, the mere presence of thyroid epithelial cells in the setting of TB FNA is not by itself diagnostic of cancer. The cells must have diagnostic features that would be accepted in the setting of the FNA of a primary thyroid nodule. This is somewhat different from the case when aspirating lateral cervical lymph nodes in the setting of thyroid cancer, in which any epithelial group identified is “suspicious,” regardless of the cytologic features.
As 1 case in this series highlighted, another potential diagnostic pitfall lies in aspirating benign parathyroid tissue when targeting a TB lesion. An enlarged parathyroid gland in the TB can imitate local tumor recurrence or be inadvertently aspirated when targeting an adjacent TB lesion. In either case, the aspirates are often cellular with features that can be confused with tumor recurrence. Parathyroid tissue and lesions are notorious for mimicking thyroid follicular neoplasms because of high cellularity and microfollicular arrangements.10,11 Immunohistochemical staining for neuroendocrine markers (as performed in this study) or parathyroid hormone can distinguish parathyroid tissue from thyroid follicular tissue. Chemical analysis of the needle rinse solution for parathyroid hormone can also be helpful for this differential.11
FNA of TB lesions remains an excellent, minimally invasive diagnostic modality with high rates of achieving a satisfactory specimen and high positive and negative predictive values. Definitive therapy can follow a positive FNA result without need for confirmation by tissue biopsy. Similarly, an adequate and negative FNA case can be triaged to conservative follow-up. A small percentage of cases will be inadequate or equivocal. There are not enough data to predict behavior in this setting, and, as such, further workup is warranted in this small subset of cases. Finally, it is important to remember the benign causes of TB nodules that lead to cellular specimens, specifically residual thyroid tissue and parathyroid tissue, to avoid a false-positive diagnosis of recurrent tumor. Awareness of these potential pitfalls and adherence to the usual criteria for making cytologic thyroid diagnoses can prevent misdiagnoses. In selected cases, immunohistochemical staining can be helpful.
- fine needle biopsy
- follicular thyroid carcinoma
- medullary carcinoma
- papillary carcinoma
- surgical procedures, operative
- poorly differentiated neuroendocrine carcinoma
- tertiary care hospital
- thyroid cancer, recurrent
- patient monitoring