1 INTRODUCTION

Primary hyperparathyroidism (pHPT) is recognized as symptomatic or asymptomatic hypercalcemia and high to borderline high levels parathyroid hormone (PTH) in the blood. If secondary reasons for hyperparathyroidism have been ruled out (such as renal failure), the cause is one or several parathyroid neoplasms, such as an adenoma (single or multiple) or carcinoma (almost always single nodules).1, 2 Traditionally, patients were surgically treated using a bilateral neck exploration approach. However, since most cases are caused by single adenomas, less invasive approaches have replaced bilateral neck exploration, facilitated by preoperative localization of the abnormal gland using high resolution ultrasound and/or technetium sestamibi scintigraphy.3, 4 Even so, some of these lesions are still difficult to find by radiological methods. Sestamibi scans exhibit limitations when it comes to small tumors, as well as tumors with water clear cell appearance,5 and thyroid nodules may sometimes create false positive scan results. High-resolution ultrasound may be optimal for the identification of cysts and smaller lesions, but is generally sub-par in distinguishing parathyroid lesions from thyroid nodules and adjacent lymph nodes.6 When the patient exhibits concomitant thyroid disease, radiological modalities may not be entirely sufficient, and ultrasonography guided fine-needle aspiration cytology (FNAC) could in certain cases be used to pinpoint the location.3

FNACs of parathyroid tumors are not recommended in general, and are usually utilized as part of a work-up of a thyroid nodule. Indeed, it is estimated that 0.12% of all thyroid FNACs report intrathyroidal parathyroid glands.7 Normal parathyroid glands as well as adenomas usually contain conventional chief cells, with or without smaller populations of clear-type cell chief cells and oxyphilic cells. Thyroid lesions on the other hand contain follicular epithelial cells—but they are still traditionally difficult to distinguish from parathyroid cells in cytology preparations.2, 4, 7-13 The literature regarding cyto-morphological hallmarks for parathyroid cells is rather limited, and specifically meager in terms of studies that offer clues how to differ parathyroid cells from thyroid epithelium. If the cytopathologist, already while investigating a quick-stained smear, could raise the suspicion of a parathyroid lesion, part of the aspirate can be submitted for biochemical estimation of PTH to establish the diagnosis. Absher et al and Dimashkieh et al made the first attempts to find common hallmarks for parathyroid cells to differ them from thyroid cells using fine-needle aspiration biopsies from 12 to 20 parathyroid lesions, respectively.8, 10 Up until these studies were launched, contemporary scientific literature on the subject was restricted to occasional case reports, or were based on small studies performing aspirations or imprints from surgically resected specimens. Since then, several studies have been performed to define cyto-morphological criteria for parathyroid lesions,4, 7, 11-13 but no clear-cut definitions or recommendations in terms of morphological classification have been proposed. Generally, when listing subjective parameters based on pattern recognition, the interobserver variability is usually high. Moreover, different terminology and definitions for cytological patterns have been used, adding to the heterogeneous results obtained. In this study, we correlated cytological characteristics of parathyroid nodules with postoperative histology in a large series of patients in order to see whether or not the FNAC report had an impact on the clinical handling, and if the morphological aspects observed on FNACs were correlated to specific histological attributes. Of course, if parathyroid lesions could be recognized cyto-morphologically, it would be of direct clinical value, to avoid falsely diagnosing a parathyroid lesion as thyroid-derived.

2 MATERIAL AND METHODS

The pathology database at the Karolinska University Hospital, Stockholm, Sweden was searched retrospectively for parathyroid lesions diagnosed by cytology. For this purpose, we used the SNOMED nomenclature for “parathyroid” (T97xxx) and investigated the cytology records between January 1, 2000 and May 20, 2020. In total, 37 cytology reports were found, which should be compared to the >1600 thyroid-related FNACs performed annually at our institution. The diagnoses, biochemical tests, ultrasonography-guided FNAC smears and available histological tissue sections (if surgical resection was performed) were retrieved. All FNA smears had been air-dried and stained with May Grunwald-Giemsa (MGG). Unstained smears were also previously acquired for the purpose of PTH ICC which had been performed for the majority of cases. The PTH ICC was conducted in a clinically accredited laboratory setting using the Novocastra anti-PTH antibody (clone NCL-PTH-488, Leica Biosystems, Newcastle, UK) and an automated Ventana Benchmark ULTRA system (Roche, Basel, Switzerland). Additional clinical information, including patient follow-up, was obtained through accessing the medical files of each patient. Ethical approval was granted by the local ethics committee of Karolinska Institutet.

All cytological smears were reviewed by two of the authors (S.S., M.H.), and evaluated in terms of the following predefined parameters; architectural patterns, nuclear features, cytoplasmic appearance (granularity, vacuolation), presence of macrophages, mitoses and colloid. Histological tissue sections for all parathyroids that were subsequently surgically resected were reviewed by three authors (S.S., M.H., C.C.J.) for proportions of cell types, proportions of growth patterns as well as signs of biopsy artifacts. Comparisons were then made between smears and histology.

3 RESULTS 3.1 Clinical aspects of the study cohort

At our institution, the parathyroid lesions diagnosed by FNAC have increased over the past years, with seven cases diagnosed in 2000–2010 compared with 29 cases between 2011 and 2020. Out of 37 FNACs from 36 patients in which parathyroid tissue was acknowledged in the cytology report, two cases were not found in our archives and hence excluded. Of the remaining 35 cases, two cases were excluded based on PTH biochemical estimations under the threshold value. Moreover, an additional five cases contained too few cells for evaluation. In total, 28 FNACs of bona fide parathyroid tissue were assessed cyto-morphologically. Seven of these cases did not proceed to surgical intervention and were therefore not included in the comparative analyses with histology, and for one case the histological sections were not available for evaluation, leaving 20 cases in which comparison with histopathologic characteristics of the subsequently excised tumor could be made.

The clinical parameters of this cohort are summarized in Table 1. In detail, out of the 28 patients investigated, the female: male ratio was 3:1, and the median age at the time of cytology was 55 years (range 16–76). Twenty-one patients (75%) had a pre-FNAC clinical diagnosis of pHPT. Of these, the FNAC investigation was initiated with the aim to verify the assumed location of a parathyroid lesion in 17 patients (81%), of which many exhibited a history of inconclusive tumor localization using radiology and/or negative findings using surgical exploration. Therefore, the reason for performing the FNAC was primarily not to establish a preoperative parathyroid tumor diagnosis, but rather to verify the location in cases in which standard clinical work-up was inconclusive. Cases with an established PHPT and initial negative findings using imaging and/or surgical exploration were followed biochemically and re-investigated using scintigraphy, and if the results were inconclusive, the patients were then assessed with ultrasonography-guided FNAC in order to try to establish the exact location in order to avoid excessive explorative surgery. The majority of patients (15/28; 54%) was somewhat unconventional in its clinical presentation, as patients either exhibited synchronous thyroid nodules upon ultrasonographic investigations (n = 4), presented with an intrathyroidal nodule (n = 4) or had no previous diagnosis of PHPT (n = 7). In these cases, the FNAC was therefore needed to pinpoint the parathyroid origin in order to plan the correct surgical procedure. An additional 12 patients (12/28; 43%) had previous neck surgery (either parathyroid surgery or surgery for other causes), and the FNAC was ordered to verify the parathyroid origin in order to plan the re-operation carefully. When looking into these latter cases more carefully, it appears that the preoperative ultrasonographic investigation in many cases was uncertain, and the surgeons specifically ordered another round of ultrasound with FNAC verification before attempting another round of surgery. In these instances, the risk of biopsy-associated changes was weighted against the potential morbidity of additional neck surgery in a scar tissue area (recurrent laryngeal nerve damage and/or hypoparathyroidism).

TABLE 1. Clinical characteristics and outcomes of the study cohort Case no. Age Sex Pre-FNAC pHPT Indication for FNAC Histopathological diagnosis Did FNAC affect the choice of treatment?* 1 76 F Yes (mild) Persistent pHPT after resection of an adenoma No surgery No, no surgery performed because of mild pHPT only 2 74 M No Other radiological indication for FNAC Parathyroid adenoma Yes, leading to the diagnosis of pHPT 3 22 M Yes pHPT and previous negative surgical exploration Parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 4 16 M Yes Persistent pHPT after resection of an adenoma Parathyroid carcinoma Yes, additional adenoma localized leading to parathyroidectomy 5 70 F Yes pHPT and previous surgery in the neck Parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 6 63 F No Other radiological indication for FNAC No surgery Yes, no surgery needed because no malignancy was detected 7 33 F Yes pHPT and a concomitant thyroid nodule Parathyroid adenoma Yes, adenoma localized, no concomitant thyroid neoplasia 8 46 F Yes pHPT and a concomitant thyroid nodule Parathyroid adenoma Yes, adenoma localized, no concomitant thyroid neoplasia 9 52 M Yes pHPT and previous negative exploration Parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 10 65 F No Other radiological indication for FNAC No surgery Yes, no surgery needed because no malignancy was detected 11 26 M No Other radiological indication for FNAC No surgery Yes, no surgery needed because no malignancy was detected 12 43 F Yes pHPT and concomitant thyroid nodule Parathyroid adenoma Yes, adenoma localized, no concomitant thyroid neoplasia 13 49 F No Other radiological indication for FNAC No surgery Yes, no surgery needed because no malignancy was detected 14 65 M No Other radiological indication for FNAC No surgery Yes, no surgery needed because no malignancy was detected 15 64 M No Other radiological indication for FNAC Secondary hyperplasia Yes, leading to the diagnosis of pHPT 16 75 F Yes pHPT and previous surgery in the neck Parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 17 49 F Yes pHPT and previous negative surgical exploration Parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 18 37 F Yes Recurrent pHPT after resection of an adenoma Parathyroid carcinoma Yes, additional tumor localized leading to parathyroidectomy 19 66 F Yes Persistent pHPT after resection of an adenoma Parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 20 53 F Yes pHPT with suspicious intrathyreoidal adenoma No surgery Yes, adenoma localized, surgery not performed due to comorbidity 21 60 F Yes Persistent pHPT after resection of an adenoma Atypical parathyroid tumor Yes, adenoma localized leading to parathyroidectomy 22 61 F Yes pHPT and a concomitant thyroid nodule Parathyroid adenoma Yes, concomitant parathyroid and thyroid lesions localized. 23 74 F Yes pHPT, localization of adenoma Parathyroid adenoma Yes, adenoma localized, no concomitant thyroid neoplasia 24 59 F Yes Recurrent pHPT after resection of an adenoma Parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 25 62 F Yes pHPT with suspicion of intrathyreoidal adenoma Parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 26 50 F Yes pHPT with suspicion of intrathyreoidal adenoma Atypical parathyroid tumor Yes, adenoma localized leading to parathyroidectomy 27 68 F Yes pHPT and previous negative surgical exploration Intrathyroidal parathyroid adenoma Yes, adenoma localized leading to parathyroidectomy 28 49 F Yes Persistent pHPT after resection of an adenoma Histological sections not available Yes, concomitant parathyroid and thyroid lesions localized. Abbreviations: FNAC, fine-needle aspiration cytology; pHPT, primary hyperparathyroidism.

For 21 FNAC cases (75%), the parathyroid nature of the cells could be anticipated based upon previous clinical information, but for seven cases (25%), the parathyroid origin was not expected. A total of 13 cases (46%) were verified by laboratory estimation of PTH from cytology aspirates, and in 18 patients (64%), a PTH immunocytochemical staining was performed (Figure 1). In eight patients (29%), the diagnosis was verified by both immunocytochemistry and laboratory estimation of PTH. In two patients (7%), no ancillary methods were used, thus the diagnosis was based on morphology alone. For 27 patients (96%), the result of the FNAC investigation affected clinical decisions concerning treatment, defined as a change of planned procedures (Table 1). More specifically, 20 pHPT patients exhibited parathyroid tumors without a clear localization that were successfully identified using FNAC, thereby guiding the surgical approach from extensive explorations to focused surgery. Seven patients without any suspicion of pHPT were investigated for lesions suspicious for neck malignancies, and the subsequent FNAC reports indicating parathyroid tissue thus saved the patients from unnecessary surgery, while in two patients the FNAC and subsequent biochemical evaluation diagnosed them with pHPT. For the remaining patient, the decision to ignore the FNAC result was based on clinical considerations with mild hypercalcemia and mild clinical symptoms. Twenty-one patients (75%) were subsequently operated, and the final diagnoses were parathyroid adenoma (n = 15, 71%), atypical parathyroid tumor (n = 2, 10%), parathyroid carcinoma (n = 2, 10%) and a single case (5%) of secondary hyperplasia. The final histopathological diagnosis for the remaining patient was not available.

Cytological features of parathyroid adenomas. (A) May Grunwald-Giemsa (MGG) stain showing a complex, branched, three-dimensional epithelial group with partly centrally located capillaries (×200 magnification). (B) MGG stain showing a mixture of architectural patterns, including tight clusters, loose aggregates and some dispersed cells in the background (×100 magnification). A combination of different patterns was the most frequent feature in our series. (C) MGG stain showing small group of parathyroid tumor cells arranged in microacinar formations (×400 magnification). (D) Parathyroid hormone (PTH) immunocytochemistry was used for most cases, proving the parathyroid origin of the cells via diffuse cytoplasmic staining (×400 magnification)

3.2 Cytological hallmarks of parathyroid lesions

The cyto-morphological aspects of the parathyroid FNACs are listed in Table 2, and illustrated in Figures 1-3. A few common denominators were found among parathyroid FNACs. In terms of cellular characteristics, the smears were generally monomorphic to slightly pleomorphic. Cells were commonly small, ranging from sizes slightly larger than an erythrocyte up to twice the size of an erythrocyte in the majority of cases (24/28 cases, 86%). In most smears, the cells showed a low to moderate nuclear-to-cytoplasm ratio (seen in 23/28 cases, 82%), a round to oval (seen in 25/28 cases, 89%) decentralized nucleus (seen in 21/28 cases, 75%) with stippled chromatin (seen in 25/28 cases, 89%). Cytoplasmic granularity was seen in 14 cases (50%) and intracytoplasmic vacuolization was noted in seven cases (25%; Figure 3). The background characteristics were generally without macrophages (only seen in four cases, 14%) and colloid (only seen in four cases, 14%). No mitotic activity was identified in any smear. Outliers in many parameters were two cases (7%) in which the corresponding histological sections were diagnosed as parathyroid carcinoma. Those smears showed size-varied cells (from twice up to four times the size of an erythrocyte) with round to irregular nuclei and prominent nucleoli, a feature not seen in the adenoma pool (Figures 2 and 3).

TABLE 2. Histological and cyto-morphological attributes of the study cohort Case no. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Diagnosis ND PA PA PC PA ND PA PA PA ND ND PA ND ND SH PA PA PC PA ND AT PA PA PA PA AT PA ND Histology Growth patterns (%) Microacinary 30 20 40 100 60 30 10 Solid 70 60 100 80 50 60 100 100 100 100 100 50 80 100 40 80 70 90 Trabecular 40 50 20 20 Other (cystic/sinusoidal) 50 100 Cell proportions (%) Conventional chief cells 80 100 80 80 20 10 95 80 100 100 70 30 90 20 Clear cells 20 20 30 10 20 60 20 30 40 10 70 Oxyphilic cells 20 20 100 100 60 60 90 5 100 40 80 30 10 Cytology Architectural patterns Dispersed