Gross pathology specimen of a parathyroid carcinoma. The tumor was pushing and displacing the thyroid gland but not invading it. The right thyroid lobe was removed regardless due to a suspicious thyroid nodule which on final pathology was a papillary thyroid carcinoma. The recurrent laryngeal nerve was then seen draped over the tumor and the intervening tissues were dissected to the capsule to free the nerve and preserve it. The surrounding tissues were dissected en bloc
Without evidence of metastatic disease or unequivocal invasion of adjacent structures, the histologic diagnosis of parathyroid carcinomas can be very challenging. The classic histopathology features as described by Schantz and Castleman  are listed in Table 31.1 and demonstrated in Figs. 31.2, 31.3, 31.4, and 31.5. Also listed in Table 31.1 are their frequency as originally reported by Schantz and Castleman. None of the criteria are invariably found in every case and there are some nuances that can make their interpretation difficult. Schantz and Castleman felt that the presence of mitosis within parenchymal cells is the single most valuable criterion; however, this must be clearly distinguished from mitosis within endothelial cells. Similarly fibrosis must be distinguished from scarring from previous surgery. They felt that cellular atypia is not a useful distinguishing feature and more likely associated with adenoma. On the contrary, other authors have found mitosis not as helpful and nuclear atypia associated with carcinoma .
Incidence of Shantz and Castleman histopathologic criteria in parathyroid carcinoma
Trabecular or rosette-like cellular architecture
Thick fibrous bands
Parathyroid carcinoma with pleomorphic nuclei, macronucleoli, mitoses, and thick fibrotic bands between nests and trabeculae of tumor. This tumor weighed 73.35 g and measured 9.6 cm × 5.3 cm × 4.4 cm. It invaded through the capsule and into fibroadipose tissue and extended to the resection margin. It showed necrosis, large acellular fibrous bands, vascular space invasion, and perineural invasion (H&E, 200× magnification)
Invasion into adjacent fibroadipose tissue. (H&E, 20× magnification)
Perineural invasion. (H&E, 200× magnification)
Vascular space invasion. (H&E, 100× magnification)
The challenge pathologists have faced with making this diagnosis is exemplified by a 1992 study where tumors were reclassified based on future malignant behavior such as recurrence and metastasis and less than half had been originally correctly diagnosed based on clinical and pathologic criteria alone . These difficulties led the World Health Organization in 2004 to recommend that the diagnosis of parathyroid carcinoma be reserved for cases where there is unequivocal evidence of invasive growth or metastasis . Tumors which have all of the classic features of parathyroid carcinoma but lack evidence of vascular, perineural, or full-thickness capsular invasion or metastasis should be classified as atypical parathyroid adenomas. This has been supported by a study in 2007 of 27 previously diagnosed parathyroid carcinoma patients which were reclassified based on the WHO criteria . 59 % of the patients were reclassified as benign and none of these patients recurred at a median follow-up of 91 months.
There is increasing evidence that immunohistochemistry techniques can substantially aid in the diagnosis of parathyroid carcinoma and may be considerably more accurate than histopathology using hematoxylin and eosin staining alone [13, 14]. As discussed earlier, the mitotic activity of parathyroid carcinomas compared to adenomas is controversial but generally thought to be higher. Similarly, analysis of cellular proliferation fractions using Ki-67 antibody labeling has generally shown higher values in carcinomas than adenomas but there are equivocal cases [15, 16].
Much more promising are targets within the cyclin D1-CDK4/6 pathway of retinoblastoma protein (Rb) phosphorylation which allows for cell cycle progression, (Fig. 31.6). This is now known to be a key mechanism in parathyroid cancer pathogenesis . Parafibromin, the protein product of HRPT2, a tumor suppressor gene mutated in hyperparathyroidism-jaw tumor syndrome, is a strong inhibitor of cyclin D1 amongst other functions which can potentially drive tumorigenesis [15, 17]. When the gene is mutated, it leads to Cyclin D1 overexpression which drives the pathway forward and leads to cell cycle progression. Partial or complete loss of immunohistochemical staining for parafibromin is highly specific (99 %) and sensitive (96 %) for parathyroid carcinoma . However, this is dependent on how partial loss of immunohistochemical staining is interpreted which adds subjectivity to the test . Cyclin-dependent kinase inhibitor 1B (Cdkn1b or p27Kip1) is another inhibitor of the cyclin D1-CDK4 complex . Absence of p27 expression has been shown to have a sensitivity of 83 % and specificity of 100 % for parathyroid carcinoma in one study . In the same study, absence of immunohistochemical staining for Rb had a sensitivity of 94 % but specificity of only 65 % for parathyroid carcinoma.
Diagram of parafibromin’s mechanisms of action in parathyroid tumorigenesis. Parafibromin is the protein product of the hyperparathyroidism-2 gene (HRPT2) otherwise known as cell division cycle 73 (CDC73) . It is known to be mutated in over 95 % of parathyroid carcinomas . When mutated, it fails to inhibit the cyclin D1-cyclin dependent kinase 4/6 (CD1-CDK4/6) complex from phosphorylating the retinoblastoma (Rb) protein, which allows the cell to progress from growth 1 (G1) phase to synthesis (S) phase in the cell cycle . It is also known to affect the wingless type (Wnt) signaling pathway which regulates multiple cellular process including cell fate specification, cell proliferation, and cell migration . Loss of the Wnt pathway tumor suppressor protein adenomatous polyposis coli (APC) has been shown to have very high specificity for parathyroid carcinoma . It also facilitates 3′ mRNA processing, known as polyadenylation, which protects mRNA from enzymatic degradation . This is one mechanism by which it affects p53 as well as possibly a direct interaction with p53 mRNA . It is also a component of the polymerase associated factor (PAF) protein complex which associates the RNA polymerase II subunit with histone methyltransferase complex and has multiple effects on gene transcription and potentially tumorigenesis . P27 is also known to inhibit CD1-CDK4 and is frequently absent on immunohistochemistry of parathyroid carcinoma [16, 19]
There are several other markers from alternative pathways that have been targeted by immunohistochemistry techniques to aid in diagnosis. Ubiquitin carboxy-terminal esterase L1 (UCHL1) , also known as protein gene product 9.5, is abundantly present in all neurons and is also highly specific to neuroendocrine cells . Strong staining has been shown to have a sensitivity of 78 % and specificity of 100 % for parathyroid carcinoma . Positive staining for galectin-3, a lectin expressed in several malignant tumors, has been shown to have a sensitivity of 92 % and specificity of 97 % for parathyroid carcinoma . Loss of the adenomatous polyposis coli (APC) gene, part of the Wingless type (Wnt) pathway which is functionally linked to parafibromin, has also been implicated in parathyroid carcinomas but not typical adenomas . It has been shown to have a sensitivity of 75 % and specificity of 100 % for parathyroid carcinoma . Although, it is also frequently absent in atypical adenomas, it is present even in HRPT2 adenomas with absent parafibromin expression . All of these markers have considerably improved pathologists ability to make the correct diagnosis.
While most parathyroid adenomas are found incidentally and patients only endorse mild neurocognitive symptoms, patients with parathyroid carcinoma usually present with significant signs and symptoms of hypercalcemia and at times hypercalcemic crisis [3, 8, 25–27]. These signs and symptoms are listed in Tables 31.2 and 31.3 along with their frequency.
Incidence of symptoms at presentation
Incidence of signs and associated conditions at presentation
Palpable neck mass
Hypercalcemic crisis , also termed parathyrotoxicosis , is a life-threatening condition associated with critically elevated calcium levels >16 mg/dl and characterized by acute renal failure and severe neurologic manifestations of profound weakness, somnolence, and even coma . About 10 % of patients present in this manner  and they require urgent treatment of their hypercalcemia as is discussed later.
A palpable neck mass associated with symptomatic primary hyperparathyroidism is the classic, pathognomonic presentation for parathyroid carcinoma but only occurs in about a third of patients in modern series . Nevertheless, very rarely do patients present without any signs or complications of hypercalcemia as listed in Table 31.3 [3, 27]. These findings in a patient with significant hypercalcemia should alert the physician to the possibility of parathyroid carcinoma as they proceed with further workup. As will be discussed later in this chapter, the surgical approach to parathyroid carcinoma is fundamentally different from parathyroid adenoma. The minority of patients in whom there is high clinical suspicion for parathyroid carcinoma preoperatively, undergo en bloc surgical resections and have better outcomes than when the diagnosis is made on final pathology .
Less than 10 % of patients have nonfunctional parathyroid carcinomas and they often present late from mass affect unless found incidentally . Symptoms may include hoarseness, dysphagia, or even dyspnea and indicate advanced invasive disease.
Patients with parathyroid carcinoma present with mean calcium levels around 13.5–14 mg/dl (about 3.5 mmol/l) but can range from 8.8 to 24 mg/dl (2.2–6 mmol/l) [25, 27, 30]. About a third of patients present with severe hypercalcemia (>13.5 mg/dl), a third moderate hypercalcemia (12–13.5 mg/dl), and a third mild hypercalcemia (<12 mg/dl) . Mean PTH levels from Talat’s and Shulte’s review of the literature (205 patients) were on average 8.7 times the upper limit of normal (566 pg/ml, 60 pmol/l) with a range from 1 to 71.6 times the upper limit of normal (65–4660 pg/ml, 6.9–494 pmol/l) . A PTH level 10 times the upper limit of normal has a positive predictive value of 84 % for the diagnosis of parathyroid carcinoma .
Imaging plays a fundamentally different but equally important role in parathyroid carcinoma as compared to benign parathyroid disease. The goal is not only to identify which side the tumor is on, which is usually readily apparent, but more so the presence and extent of local invasion and any associated regional and distant metastases. In the vast majority of cases, the diagnosis of malignancy is uncertain or not even suspected and imaging can provide clues that point towards an invasive process. There are a variety of functional and anatomic imaging modalities that can be used for suspected cases of parathyroid carcinoma and a thorough workup should include both types.
The functional imaging modality most frequently used in benign parathyroid disease (and often the only modality) is technetium-99 m sestamibi scanning with or without single-photon emission tomography (SPECT). This is also useful in malignant disease not only to identify the primary lesion but also any regional or distant metastasis . Positron emission tomography (PET) can also be used to characterize the primary and identify regional and distant metastasis and can be complimentary [32, 33]. There have been reports of metastatic lesions that were nondetectable on technetium-99m sestamibi scanning but localized with PET , and vice-versa .
Anatomic imaging has a more critical role in surgical planning in parathyroid cancer. Cervical ultrasound is often the first modality employed especially in patients that present with a neck mass. Ultrasound has a very high accuracy at detecting parathyroid glands the size of carcinomas and certain sonographic features can be very suggestive of malignancy [27, 35]. These features include a large hypoechoic gland, lobulated appearance, indistinct margins, calcification, abnormal vascularity, and a thick capsule [36, 37]. Figure 31.7 shows an example with some of these features. Ultrasound can also identify potentially involved structures; however, this is probably best demonstrated with computed tomography (CT) or magnetic resonance imaging (MRI), one of which is essential for adequate preoperative preparation. Figure 31.8 shows the CT of the same patient as in Fig. 31.7. These modalities can help the surgeon formulate a surgical plan in advance should invasion of critical structures such as the esophagus or trachea be seen and thereby help him or her achieve an en-bloc resection .
Preoperative axial (a) and sagittal (b) ultrasonography images of the parathyroid carcinoma from Fig. 31.1. Note the lobulated, indistinct margins and thick capsule
Needle biopsy of a primary parathyroid lesion suspicious for parathyroid carcinoma is not recommended primarily due to high nondiagnosis rate . There have also been two case reports of potential tumor seeding along FNA tracks [39, 40]. In cases of recurrence, needle biopsy of suspicious lesions can help localize and confirm the site of recurrence .
A comprehensive head and neck examination is vital in the workup of parathyroid carcinoma as a high percentage of patients will present with a palpable neck mass. A mass that feels fixed to adjacent tissues provides important information that will aid in surgical planning. In addition, flexible endoscopic examination of the vocal folds should be strongly considered as this can alert the surgeon to recurrent laryngeal nerve involvement which can also greatly facilitate decision making at the time of surgery.
The differential diagnosis for patients presenting with primary hyperparathyroidism and findings suggestive of malignancy such as a neck mass includes the very unusual situation of parathyroid adenoma or hyperplasia with a concomitant head and neck malignancy. Malignancies that present with central neck metastasis, most commonly papillary thyroid carcinoma and squamous cell carcinomas of the laryngopharyngeal tract could conceivably be mistaken for a parathyroid carcinoma. This should be distinguishable by imaging characteristics, endoscopic examination, and lower calcium and PTH levels.
A more common situation is a patient who was thought to have a parathyroid adenoma or an atypical parathyroid adenoma but presents several years later with multiple functioning parathyroid nodules in the superficial and deep soft tissues of the neck, termed parathyromatosis . This may be very difficult to distinguish from locoregional recurrence of a mistaken parathyroid carcinoma. Parathyromatosis is usually seen after surgery for parathyroid adenoma and is thought to be the result of tumor spillage and reimplantation. It usually has a characteristic distribution in the previous surgical bed and along the incision; however, at times it can appear to extend outside the previous surgical field. There have also been reports of de novo parathyromatosis in patients who have not undergone surgery . This is postulated to be the result of an overgrowth of embryologic rests of parathyroid tissue.
As previously discussed, atypical parathyroid adenomas are parathyroid tumors that show the same characteristics of parathyroid cancer but not unequivocal evidence of invasion. When distant metastasis occur in this situation, the diagnosis of carcinoma becomes clear.
Despite the difficulties arising from all of the entities sharing overlapping histopathologic features, they can often be differentiated based on clinical presentation. Parathyroid carcinoma usually presents with much higher calcium levels and associated end organ damage . Nevertheless, there are no absolute clinical or histopathologic findings other than invasion. With identification of the HRPT2 mutation and its almost exclusive presence in parathyroid carcinoma, these lesions may become easier to characterize . There have been several examples of incorrectly diagnosed atypical parathyroid adenomas that have later metastasized that would have been correctly identified with HRPT2 mutation testing .
There is no widely accepted staging system for parathyroid carcinoma. Shaha and Shah proposed a staging system in an editorial in 1999 following the TNM format (Table 31.4) and based on findings from a National Cancer Data Base (NCDB) study by Hundahl et al. [4, 42]. This study of 286 cases was the largest reported cohort of parathyroid cancer patients at the time and the first registry type study for parathyroid cancer. The staging system used a 3 cm cutoff to separate T1 and T2 patients based on the mean size of tumors found in the NCDB study as well as lymph node status to define stage III disease. However, these criteria were used despite a clear finding in the NCDB study against them acting as significant prognostic factors.
Shaha and Shah staging system
Primary tumor <3 cm
Primary tumor >3 cm
Primary tumor of any size with invasion of the surrounding soft tissues (i.e. thyroid gland, strap muscles, etc)
Massive central compartment disease, invading the trachea or the esophagus, or recurrent parathyroid carcinoma
No regional lymph node metastasis
Regional lymph node metastasis
No evidence of distant metastasis
Evidence of distant metastasis
Any T, N1, M0
Any T, Any N, M1
More recently, Talat and Shulte proposed a different TNM type staging system which excluded size as a criterion and placed more emphasis on vascular invasion (Shulte a, Table 31.5) . They also proposed a simpler low/high risk stratification system (Shulte b, Table 31.5) which essentially defined Shulte a stage I disease as low risk and stages II to IV as high risk. Both of these systems were then evaluated in a systematic review of the literature that included 706 patients which found them to be significantly related to risk of recurrence and death . They further verified these two staging systems in a multi-institutional study of 82 patients which confirmed significant differences in recurrence and survival between low-risk and high-risk groups, as well as between stage I and II/III disease . There were no significant differences seen between stage II and III disease, only 2 recurrences in the low risk group and no deaths in this study.
Talat and Shulte staging system
Evidence of capsular invasion
Invasion of surrounding soft tissues, excluding the vital organs (trachea, larynx, and esophagus)
Evidence of vascular invasion
Invasion of vital organs (hypopharynx, trachea, esophagus, larynx, recurrent laryngeal nerve, carotid artery)
No regional lymph node metastasis
Regional lymph node metastasis
No evidence of distant metastasis
Evidence of distant metastasis
T1N0M0 or T2N0M0
T4N0M0 or Any T, N1, M0
Any T, Any N, M1
The Shaha and Shulte staging systems were recently further compared by the Spanish Parathyroid Carcinoma Study Group (SPCSG) , a multi-institutional cancer registry in Spain . This prospectively collected registry study of 62 patients again demonstrated significant differences in recurrence and survival only between stage I and III patients in both staging systems.
The recurrence and survival data for the Shaha and Shulte staging systems based on Talat and Schulte  systematic review is listed in Table 31.6 . As is clearly evident, we do not have a reliable multistage TNM type staging system and can differentiate between high and low risk disease.
Recurrence and survival by stage
5 Year recurrence
5 Year survival
The link between parathyroid cancer and hyperparathyroidism-jaw tumor syndrome HPT-JT is now clearly established . Fifteen percent of patients with HPT-JT will develop parathyroid cancer in their lifetime . This realization led to the discovery of the HRPT2 tumor suppressor gene (now officially named CDC73) in 2002 by Carpten and colleagues in 14 families with HPT-JT . The gene codes for the parafibromin protein, so named for its relationship to parathyroid disease and fibro-osseous lesions of the jaw. The HRPT2 mutation is highly specific for parathyroid carcinoma, occurring in 77 % of tumors with confirmed malignant behavior and is rarely seen in adenomas and hyperplasia . In fact, loss of parafibromin expression correlates with recurrence in tumors that contain some histopathologic features of parathyroid carcinoma but lack unequivocal evidence of invasion necessary to meet WHO criteria for malignancy .
The incidence of a germline HRPT2 mutation in apparent sporadic cases of parathyroid carcinoma is about 30 % . Therefore, about a third of patients with parathyroid carcinoma have unrecognized HPT-JT despite no family history of the disorder. This has led some authors to recommend genetic testing for all patients with parathyroid carcinoma . Unfortunately, sequencing the HRPT2 is currently expensive and many of the mutations are large deletion mutations that require gross deletion analysis to detect . As discussed earlier, there is a monoclonal antibody to parafibromin used for immunohistochemistry which can confirm malignancy and triage patients for germline mutation testing. However, it requires a great deal of technical expertise to employ successfully and there are mixed results in the literature . Hopefully, 2nd and 3rd generation sequencing will make identification of the mutation more accessible.