Endocr Pathol DOI 10.1007/s12022-013-9294-3
Understanding the Genetic Basis of Parathyroid Carcinoma Anthony J. Gill
# Springer Science+Business Media New York 2014
Abstract Parathyroid carcinoma has always been difficult to diagnose pathologically. In fact, most parathyroid tumors which are classified as carcinoma do not recur after excision, and most parathyroid tumors which actually metastasize or recur repeatedly in the neck are not recognized as malignant at first presentation. In 2002, germline HRPT2 (also known as CDC73) mutation w as reported as the cause of hyperparathyroidism-jaw tumor (HPT-JT) syndrome, an autosomal dominant hereditary tumor syndrome associated with a lifetime risk of parathyroid carcinoma approaching 15 %. Subsequently, bi-allelic inactivation or mutation of HRPT2 has been reported in the majority of parathyroid carcinomas that actually behave in a malignant manner but very rarely in sporadic benign parathyroid disease. Furthermore, germline testing for HRPT2 mutation in patients presenting with parathyroid carcinoma often identifies occult HPT-JT syndrome even in the absence of a family history or other syndromic manifestations. HRPT2 mutation testing is not readily available, and loss of expression of parafibromin (the protein encoded by HRPT2) as determined by immunohistochemistry has been used as a surrogate marker of HRPT2 mutation. Immunohistochemistry for parafibromin can be technically difficult and has been deployed by different investigators with variable enthusiasm and success. However, proponents have found immunohistochemistry for parafibromin useful to definitively confirm a pathological diagnosis of parathyroid A. J. Gill (*) Department of Anatomical Pathology, Royal North Shore Hospital, Pacific Highway, St Leonards, Sydney, NSW 2065, Australia e-mail: [email protected] A. J. Gill Cancer Diagnosis and Pathology Research Group, Kolling Institute of Medial Research, St Leonards, NSW 2065, Australia A. J. Gill University of Sydney, Sydney, NSW, Australia 2006
carcinoma, predict a worse outcome in definite parathyroid carcinomas, triage formal genetic testing for HPT-JT syndrome, and predict the outcome of histologically atypical parathyroid adenomas. Keywords Parathyroid carcinoma . HRPT2 . CDC73 . Parafibromin Over the last 25 years, much has been learnt about the molecular mechanisms of disease and the genetic basis of many hereditary tumor syndromes. The molecular and hereditary basis of parathyroid carcinoma is no exception. Although progress has been slow, there has been a steady evolution in the understanding of this rare endocrine malignancy—particularly its association with the previously obscure hereditary cancer syndrome known as hyperparathyroidism-jaw tumor (HPT-JT) syndrome. This increased understanding has been paralleled by a significant increase in the reported incidence of parathyroid carcinoma, historically one of the rarest endocrine malignancies [1, 2]. This review concentrates on these developments and highlights both what we have learnt and what remains uncertain. In 1973, Shantz and Castleman described their experience with 70 cases of parathyroid carcinoma and reported that the histological features that distinguish parathyroid carcinoma from adenoma were a trabecular growth pattern, mitotic activity, broad fibrous bands, capsular penetration, and vascular space invasion [3]. Although the abnormal architectural and cytological features that they reported are still considered important clues to the diagnosis of carcinoma, it is now clear that the presence of an unequivocally invasive growth pattern is the most important predictor of subsequent malignant behavior of parathyroid tumors. Therefore, the WHO 2004 criteria now recommend that the pathological diagnosis of parathyroid carcinoma be restricted to lesions, which show
Endocr Pathol
unequivocal invasive growth as evidence by perineural invasion, vascular space invasion or full thickness capsular penetration with growth into adjacent tissues or metastasis [4]. Despite these refined criteria, parathyroid carcinoma remains difficult to diagnose with any certainty histologically, and perhaps, the key to understanding the evolution in the molecular and surgical pathology of parathyroid carcinoma over the last 25 years is to first acknowledge that physicians, surgeons, and pathologists have always had difficulty in distinguishing parathyroid carcinoma from benign disease in the absence of metastasis. For example, in one large multi-institutional study of 40 parathyroid carcinomas defined on the basis of biologically malignant behavior (i.e., recurrence and metastasis) rather than by pathological criteria alone, less than half were correctly diagnosed as malignant at first presentation based on the combined clinical and pathological findings [5]. In another review of 27 cases initially considered to be suspicious or malignant parathyroid tumors, 16 (59 % of the cases) were reclassified as benign on blinded pathological review and none of these recurred [6]. The same study reported that only four of the 11 patients given a definitive review diagnosis of carcinoma went on to behave in a malignant manner, and the remaining seven carcinoma patients (including four who had simple rather than en bloc resection) were alive with no evidence of disease at a median follow-up of 65 months. Even in Shantz and Castleman’s series, as many as 58 % of patients were alive without disease at the last follow-up [3]. The subjectivity and inadequacy of the pathological diagnosis of parathyroid carcinoma is further complicated by the geographic and institutional variation in reported incidence. In most series, carcinoma accounts for considerably less than 1 % of all patients with primary hyperparathyroidism [5–7], but this figure is as high as 5 % in some Japanese and Italian reports [8–11]. Of course, specific environmental or genetic factors may account for this difference, but it is intriguing to postulate that different pathological thresholds for diagnosis may also play a role. Although it is confounded by the fact that surgery, particularly radical or en bloc excision, has the capacity to interrupt the natural history of parathyroid carcinoma, that late recurrence is not uncommon, and that different institutions may use subtly different pathological criteria, it appears that perhaps as many as 50 % of parathyroid carcinomas which actually behave in a malignant manner will initially be considered benign using conventional clinicopathological criteria and perhaps as few as 15 % of all cases diagnosed as carcinoma prospectively (36 % if rigid histopathological criteria for malignancy are applied by a specialist endocrine pathologist) will go on to recur or behave in a malignant manner after simple excision. Molecular analysis holds promise to improve the diagnosis of parathyroid carcinoma, and the first steps toward identifying the molecular pathology of parathyroid carcinoma were
made in the 1970 when it was recognized that parathyroid carcinoma demonstrates a strong familial tendency. For example, in 1972, Frayha et al. [12] reported parathyroid carcinoma in siblings each with an adenoma and carcinoma; in 1974, Malette et al. [13] reported parathyroid carcinoma in the setting of familial hyperparathyroidism; in 1975, Leborgne et al. [14] reported parathyroid carcinoma in siblings, and in 1977, Dinnen et al. [15] reported parathyroid carcinoma in two separate kindreds with familial hyperparathyroidism one of which also had jaw tumors. In fact, careful reading of many subsequent large series of parathyroid carcinoma almost inevitably reveals a passing comment that some of the cases were familial or associated with familial hyperparathyroidism, although these findings were not always emphasized [16]. This familial syndrome of hyperparathyroidism and parathyroid carcinoma came to be known as hyperparathyroidismjaw tumor (HPT-JT) syndrome [OMIM#145001]. Given the rarity of parathyroid carcinoma in the general community, the striking lifetime risk of parathyroid carcinoma in HPT-JT that was reported to be as high as 15 % [4] suggested a strong association between the molecular pathology of this syndrome and parathyroid carcinoma. In 2002, Carpten and colleagues [17] studied 14 families with HPT-JT and found inactivating mutations at 1q25-32 in a gene initially named HRPT2 for ‘Hyperparathyroidism 2’ but subsequently renamed CDC73. HRPT2 contains 17 exons and encodes a 531 amino acid protein termed parafibromin because of its relationship to parathyroid disease and fibro-osseous lesions and was postulated as a tumor suppressor gene. It is not surprising that following the discovery of HRPT2 mutation as the cause of HPT-JT syndrome, investigators began to focus on the role of HRPT2 mutation in parathyroid carcinoma. In 2003, Howell et al. [18] found HRPT2 mutations in four of five apparently sporadic carcinomas but in none of 25 adenomas, 11 secondary hyperplasia, or six tertiary hyperplasias. In 2003 Shattuck et al. [19] also examined 15 patients with apparently sporadic parathyroid carcinomas and found that ten (67 %) of the carcinomas harbored pathogenic HRPT2 mutations. Furthermore, even in the absence of a family history, three of these ten patients (30 %) also had germline mutation of HRPT2. That is, they had unrecognized HPT-JT syndrome. In 2004, Cetani et al. [20] reported seven patients with parathyroid carcinoma, six of which were associated with HRPT2 mutation. Again, two of the six HRPT2 mutated parathyroid carcinomas were associated with germline HRPT2 mutation. Krebs et al. [21] went on to report no HRPT2 mutations in 60 consecutive apparently sporadic benign adenomas, and their meta-analysis suggested that the incidence of HRPT2 mutation in unselected parathyroid adenomas was less than 0.8 %. The combined results of these studies indicated that HRPT2 mutation occurs commonly (77 %) of parathyroid carcinomas and rarely (less than 1 %) in unselected benign disease. Importantly, because of the difficulties in diagnosing
Endocr Pathol
parathyroid carcinoma confidently pathologically, all three major studies investigating the incidence of HRPT2 mutation in parathyroid carcinoma required biological evidence of malignant behavior (that is recurrence or metastasis) as inclusion criteria. Therefore HRPT2 mutation occurs not in 77 % of cases pathologists have diagnosed as carcinoma but in 77 % of cases which have been diagnosed as carcinoma by pathologists and have also behaved in a malignant manner. Similarly, given that the estimated incidence of parathyroid carcinoma is less than 1 % and that the diagnosis is frequently not made until recurrence, the finding of HRPT2 mutation in less than 1 % of unselected parathyroids strongly supports HRPT2 mutation as being highly specific for parathyroid carcinoma. That is, because parathyroid carcinoma is so difficult to diagnose pathologically, the strength of these three studies lies in the fact that they used malignant behavior as the gold standard with which to correlate the significance of HRPT2 mutation, not just histopathological criteria. Despite the presence of mutational hotspots in exons 1, 2, and 7 of HRPT2 where approximately 80 % of pathogenic mutations occur [22], sequencing this gene is expensive and not readily available. Furthermore, perhaps as many as 35 % of HPT-JT syndrome associated HRPT2 mutations are large deletion mutations that require gross deletion analysis to identify [23]. Therefore despite its potential, HRPT2 mutation testing is difficult to deploy in the diagnostic surgical pathology laboratory. In 2004, Tan and colleagues [24] developed a mouse monoclonal antibody directed against parafibromin. Because HRPT2 mutations are commonly of a frameshift or truncating nature, they postulated that completely absent nuclear staining for parafibromin could be used to confirm HRPT2 mutation and reported that loss of nuclear staining occurs in 96 % of parathyroid carcinomas, in 89 % of HPT-JT associated ‘adenomas’ and in less than 1 % of benign parathyroid tumors. When we examined parafibromin expression in independent cohorts [25, 26], we found completely negative parafibromin staining in eight of 11 apparently sporadic parathyroid carcinomas (defined as not only meeting the WHO criteria for malignancy but also behaving in a malignant manner) and three of four HPT-JT related parathyroid tumors but positive staining in 100 unselected adenomas and 56 of 57 ‘giant adenomas’ (defined as adenomas greater than 2 g). The one giant adenoma that showed negative staining for parafibromin was initially classified as ‘atypical adenoma’—discussed in detail later. Although we and the others have found immunohistochemistry for parafibromin to be useful in both confirming a definitive diagnosis of malignancy and in triaging patients for germline mutation testing for HPT-JT [24, 25, 27–35], others have found the antibody difficult to deploy [36–38]. How can this discrepancy be explained?
There are few commercially available antibodies to parafibromin, and our experience has been that all can be technically difficult to deploy in the routine diagnostic laboratory—in fact, many laboratories have been unable to achieve satisfactory staining, and we have found that staining often needs to be repeated in individual cases using different protocols to achieve satisfactory results. This has led to markedly different staining patterns in different laboratories and also to the evolution of different criteria for interpreting the stain. We require completely absent nuclear staining for parafibromin in the presence of an internal positive control (that is positive staining in endothelial and stromal cells) before staining is considered negative [25–27, 30] (illustrated in Fig. 1), and this approach has been supported by others [35]. This approach is similar to that used to assess microsatellite instability in colorectal adenocarcinoma using mismatch repair immunohistochemistry where early antibodies were equally capricious. However, others have interpreted any loss of staining (i.e., focal or partial loss) as being negative [28, 31]. Furthermore, the difficulty in achieving a pathological diagnosis of carcinoma has resulted in varying reports on the sensitivity and specificity of loss of parafibromin staining for malignancy. Given that the description of the significance of HRPT2 mutation was based on tumors that did not just fulfil histological criteria for malignancy but also went on to behave in a malignant manner [18–20], it is not surprising that different groups report different sensitivities and specificities depending on the gold standard to which parafibromin staining is compared. That is, groups that use biological evidence of malignant behavior as the gold standard for the diagnosis of parathyroid carcinoma tend to report a high rate of negative staining for parafibromin or mutation in HRPT2 in carcinomas, whereas groups that use pathological features of malignancy without the requirement for malignant behavior tend to report a lower rate [18–20, 27]. Whether the difference is due to ‘parathyroid carcinomas’ being cured by surgery, false positive histology (i.e., groups reporting a low incidence of HRPT2 mutation in parathyroid carcinoma ‘overdiagnosing’ cancer in atypical glands), technical difficulties with immunohistochemistry or poor sensitivity of parafibromin is impossible to determine. Suffice it to say that some investigators have found parafibromin staining very useful and others not so useful. Of course, for the practicing surgical pathologist, the main issue is whether parafibromin immunohistochemistry or HRPT2 mutation analysis can predict the likelihood of recurrence or metastasis of histologically malignant parathyroid tumors or the likelihood of malignant behavior of atypical parathyroid adenomas. That is, are carcinomas which show histological features of malignancy and abnormal parafibromin staining more likely to behave in a malignant manner? And can the risk of malignant behavior of atypical parathyroid adenomas be predicted by abnormal staining for parafibromin?
Endocr Pathol
Fig. 1 H&E of a parathyroid carcinoma with parafibromin staining (inset). We emphasize that for a parathyroid tumor to be considered negative for parafibromin, there should be a strong positive internal controls - in this case endothelial cells and occasional lymphocytes
We recently reviewed our experience of immunohistochemistry for parafibromin in the differential diagnosis of atypical adenomas and in predicting the outcome of tumors which fulfilled the WHO 2004 criteria of malignancy [39]. We subdivided 81 parathyroid tumors that we encountered over a 14-year period into four groups: group A, parathyroid carcinomas, defined by the presence of WHO 2004 criteria of malignancy positive which also showed negative staining for parafibromin; group B, parathyroid carcinomas, which were also of course WHO criteria positive for malignancy but parafibromin positive; group C; atypical adenomas, which were WHO criteria for malignancy negative and parafibromin negative; and group D, atypical adenomas, which were WHO criteria for malignancy negative and parafibromin positive. Across the four groups from A to D, the rates of both death from disease and recurrence varied significantly - the risk of death in group A was 15 %; group B, 7 %; group, C nil; and group D, nil and the risk of recurrence varied in a similar pattern (group A, 38 %; group B, 36 %; group C, 20 %; and group D, nil). Five-year disease-free survival for groups A to D were 55, 80, 78, and 100 % respectively, and the risk of tumor recurrence was significantly associated with negative staining for parafibromin (p=0.048). Our study provided support for the concept that ‘atypical parathyroid adenomas’ which show a normal pattern of parafibromin staining either do not recur or do so extremely rarely (sufficiently rarely for them to be discharged from follow-up), whereas atypical parathyroid adenomas that show negative staining for parafibromin have a significant recurrence risk of 20 % (sufficient for them to be considered tumors of uncertain malignant potential and warrant ongoing followup). Furthermore, our data suggest that even in unequivocally histologically malignant parathyroid carcinomas, in addition to triaging genetic testing for HPT-JT syndrome, loss of
staining for parafibromin can be used to predict an increased risk of death or recurrence. A recent similar study performed by Cetani et al. only focusing on parafibromin as a prognostic indicator in parathyroid carcinoma rather than also investigating atypical adenomas also found that negative staining for parafibromin indicated a higher risk of recurrence and a decreased 5-year survival of 59 % (p=0.107) and very significantly decreased 10-year survival of 23 %, (p=0.0026) [40]. In 2014, what is the current understanding of parathyroid carcinoma, its relationship to HRPT2 mutation and HPT-JT syndrome and the role of parafibromin immunohistochemistry? It is now clearly established that HPT-JT syndrome in which there is a very significantly lifetime risk of parathyroid carcinoma is due to germline mutation of HRPT2. HPT-JT syndrome demonstrates a low penetrance and the absence of a family history of hyperparathyroidism in no way excludes germline HRPT2 mutation. Somatic inactivating mutations of HRPT2 are found in approximately 70 % of parathyroid carcinomas (when defined by malignant behavior but significantly less when defined by pathological criteria alone) but in less than 1 % of unselected parathyroid tumors indicating that the presence of HRPT2 mutation strongly supports a diagnosis of malignancy. In fact, there are even isolated case reports of HRPT2 mutated but histologically benign parathyroid tumors (confirmed by multiple endocrine pathologists) which have gone on to metastasize [41]. All patients with parathyroid carcinoma should be considered for germline testing for HRPT2 because up to 20 % of them will have unrecognized HPT-JT even in the absence of a family history [19, 20]. Immunohistochemistry for parafibromin can be used as a marker of HRPT2 mutation, but experience is mixed with some groups reporting high sensitivity and specificity and other groups finding the antibody difficult to deploy or the results of immunohistochemistry disappointing. Groups that have found parafibromin immunohistochemistry useful have employed it to triage patients with parathyroid carcinoma for formal genetic testing for HPT-JT syndrome. Furthermore, negative staining for parafibromin is beginning to be validated both as a biomarker of poor prognosis in definite parathyroid carcinomas as well as a marker of increased risk of recurrence (perhaps warranting the diagnosis of low grade carcinoma or at least ‘tumor of low or uncertain malignant potential’) in histologically atypical parathyroid tumors which do not quite fulfill conventional histological criteria for carcinoma [39, 40]. Conflict of Interest The author reports no conflict of interest.
References 1. Lee PK, Jarosek SL, Virnig BA, et al. Trends in the incidence and treatment of parathyroid cancer in the United States. Cancer 109: 1736-1741, 2007
Endocr Pathol 2. Brown S, O'Neill C, Suliburk J, et al. Parathyroid carcinoma: increasing incidence and changing presentation. ANZ J Surg 81:528-32, 2011 3. Schantz A, Castleman B Cancer 31:600-605, 1973 4. DeLellis RA, Lloyd RV, Heitz PU, et al. World Health organization classification of tumours Pathology and Genetics: Tumours of Endocrine Organs IARC Press Lyon 2004 5. Sandelin K, Tullgreen O, Farnebo LO. Clinical course of metastatic parathyroid carcinoma World J Surg 18:594-598, 1994 6. Ippolito G, Palazzo FF, Sebag F, et al. Intraoperative diagnosis and treatment of parathyroid cancer and atypical parathyroid adenoma British Journal of Surgery 94:566-570, 2007 7. Shane E Clinical review 122: parathyroid carcinoma J Clin Endocrinol Metabl 86:485-493, 2001 8. Obara T, Fujimoto Y Diagnosis and treatment of patients with parathyroid carcinoma. An update and review. World J Surg 15:738-744, 1991 9. Ishida T, Yokoe T, Izuo M Nationwide survey of parathyroid operations in Japan (1980-1989): Endocrine Surgery (Tokyo) 8:37, 1991 10. Favia G, Lumachi F, Polistina F et al. Parathyroid carcinoma: sixteen new cases and suggestions for correct management World J Surg 22: 1225-1230, 1998 11. Fujimoto Y, Obara T, Ito Y, et al. Localization and surgical resection of metastatic parathyroid carcinoma World J Surg 10:539, 1986 12. Frayha RA, Nassar VH, Dagher F, Salti IS. Familial parathyroid carcinoma J Med Liban 25:299-309, 1972 13. Mallette LE, Bilezikian JP, Ketcham AS, Aurbach GD. Parathyroid carcinoma in familial hyperparathyroidism. Am J Med 57:642-8, 1974 14. Leborgne J, Le Neel JC, Buzelin F, Malvy P Familial cancer of the parathyroid glands. Importance of angiography in the diagnosis of regional recurrences. Considerations on 2 cases. J Chir (Paris) 109: 315-26, 1975 15. Dinnen JS, Greenwoood RH, Jones JH, et al. Parathyroid carcinoma in familial hyperparathyroidism. J Clin Pathol 30:966-75, 1977 16. Smith JF, Coombs RGH Histological diagnosis of carcinoma of the parathyroid glandJ Clin Pathol 37:1370-137, 1984 17. Carpten JD, Robbins Cm, Villablanca A HRPT2, encoding parafibromin, is mutated in hyperparathyroidism-jaw tumor syndrome Nature Genetics 32:676-680, 2002 18. Howell VM, Haven CJ, Kahnoski K et al. HRPT2mutations are associated with malignancy in sporadic parathyroid tumours J Med Genet 40:657-663, 2003 19. Shattuck T, Stiina V, Obara T et al. Somatic and Germ-line mutations of the HRPT2 gene in sporadic parathyroid carcinoma New England Journal of Medicine 349:1722-1729, 2003 20. Cetani F, Pardi E, Borsari S, et al. Genetic Analyses of HRPT2 gene in primary hyperparathyroidism: Germline and Somatic Mutations in Familial and Sporadic Parathyroid Tumours Journal of Clinical Endocrinology and Metabolism 89:5583-5591, 2004 21. Kresbs L , Shattuck TM, Arnold A HRPT mutation analysis of typical sporadic parathyroid adenomas Journal of Clinical Endocrinology and Metabolism 90:5015-5017, 2005 22. Marsh DJ, Hahn MA, Howell VM, Gill AJ Molecular diagnosis of primary hyperparathyroidism in familial cancer syndromes Expert Opinion in Medical Diagnostics 1:377-92, 2007 23. Bricaire L, Odou MF, Cardot-Bauters C, et al. Frequent large germline HRPT2 deletions in a French National cohort of patients with primary hyperparathyroidism. J Clin Endocrinol Metab. 98: E403-8, 2013
24. Tan M-H, Morrison C, Wang P, et al. Loss of parafibromin immunoreactivity is a distinguishing feature of parathyroid carcinoma Clin Cancer Res 10:6629-6637, 2004 25. Gill AJ, Clarkson A, Gimm O, et al. Loss of nuclear expression of parafibromin distinguishes parathyroid carcinomas and Hyperparathyroidsim-Jaw Tumor (HPT-JT) syndrome related adenomas from sporadic parathyroid adenomas and hyperplasias American Journal of Surgical Pathology 30:1140-1149, 2006 26. Meyer-Rochow GY, Alvarado R, Sywak MS, et al. Intraoperative diagnosis and treatment of parathyroid cancer and atypical parathyroid adenoma British Journal of Surgery 94:1044, 2007 27. Howell VM, Gill AJ, Clarkson A, et al. Accuracy of combined protein gene product 9.5 and parafibromin markers for immunohistochemical diagnosis of parathyroid carcinoma Journal of Clinical Endocrinology and Metabolism 94:434-441, 2009 28. Wang O, Wang C, Nie M, et al. Novel HRPT2/CDC73 gene mutations and loss of expression of parafibromin in Chinese patients with clinically sporadic parathyroid carcinomas. PLoS One. 7(9):e45567, 2012 29. Wang O, Wang CY, Shi J, et al. Expression of Ki-67, galectin-3, fragile histidine triad, and parafibromin in malignant and benign parathyroid tumors. Chin Med J (Engl) 125:2895-901, 2012 30. Lim S, Elston MS, Gill AJ, et al. Metastatic parathyroid carcinoma initially misdiagnosed as parathyroid adenoma: the role of parafibromin in increasing diagnostic accuracy. Intern Med J 41: 695-9, 2011 31. Kim HK, Oh YL, Kim SH, et al. Parafibromin immunohistochemical staining to differentiate parathyroid carcinoma from parathyroid adenoma. Head Neck 34:201-6, 2012 32. Juhlin CC, Nilsson IL, Johansson K, et al. Parafibromin and APC as screening markers for malignant potential in atypical parathyroid adenomas. Endocr Pathol 21:166-77, 2010 33. Juhlin CC, Villablanca A, Sandelin K, et al. Parafibromin immunoreactivity: its use as an additional diagnostic marker for parathyroid tumor classification. Endocr Relat Cancer. 14:501-12, 2007 34. Cetani F, Ambrogini E, Viacava P, et al. Should parafibromin staining replace HRTP2 gene analysis as an additional tool for histologic diagnosis of parathyroid carcinoma? Eur J Endocrinol 156:547-54, 2007 35. Witteveen JE, Hamdy NA, Dekkers OM, et al. Downregulation of CASR expression and global loss of parafibromin staining are strong negative determinants of prognosis in parathyroid carcinoma. Mod Pathol. 24:688-97, 2011 36. Cabané T, Patricio, Gac E, et al. Detección inmunohistoquímica de parafibromina en patología de paratiroides. Revista chilena de cirugía, 65:20-24, 2013 37. Guarnieri V, Battista C, Muscarella LA et al. CDC73 mutations and parafibromin immunohistochemistry in parathyroid tumors: clinical correlations in a single-centre patient cohort Cellular Oncology 35: 411-422, 2012 38. DeLellis RA, Mazzaglia P, Mangray S. Primary hyperparathyroidism: a current perspective. Arch Pathol Lab Med. 132:1251-62, 2008 39. Krujiff S, Sidhu SB, Sywak MS, et al. Negative parafibromin staining predicts malignant behaviour in atypical parathyroid adenomas Annals of Surgical Oncology Oct 1 [epub ahead of print] 40. Cetani F, Banti C, Pardi E, et al. CDC73 mutational status and loss of parafibromin in the outcome of parathyroid cancer. Endocr Connect. 28:186-95, 2013 41. M.S. Sarquis, L.G. Silveira, F.J. Pimenta, et al., Familial hyperparathyroidism: surgical outcome after 30 years of follow-up in three families with germline HRPT2 mutations, Surgery 143:630–640, 2008
Understanding the Genetic Basis of Parathyroid Carcinoma Anthony J. Gill
# Springer Science+Business Media New York 2014
Abstract Parathyroid carcinoma has always been difficult to diagnose pathologically. In fact, most parathyroid tumors which are classified as carcinoma do not recur after excision, and most parathyroid tumors which actually metastasize or recur repeatedly in the neck are not recognized as malignant at first presentation. In 2002, germline HRPT2 (also known as CDC73) mutation w as reported as the cause of hyperparathyroidism-jaw tumor (HPT-JT) syndrome, an autosomal dominant hereditary tumor syndrome associated with a lifetime risk of parathyroid carcinoma approaching 15 %. Subsequently, bi-allelic inactivation or mutation of HRPT2 has been reported in the majority of parathyroid carcinomas that actually behave in a malignant manner but very rarely in sporadic benign parathyroid disease. Furthermore, germline testing for HRPT2 mutation in patients presenting with parathyroid carcinoma often identifies occult HPT-JT syndrome even in the absence of a family history or other syndromic manifestations. HRPT2 mutation testing is not readily available, and loss of expression of parafibromin (the protein encoded by HRPT2) as determined by immunohistochemistry has been used as a surrogate marker of HRPT2 mutation. Immunohistochemistry for parafibromin can be technically difficult and has been deployed by different investigators with variable enthusiasm and success. However, proponents have found immunohistochemistry for parafibromin useful to definitively confirm a pathological diagnosis of parathyroid A. J. Gill (*) Department of Anatomical Pathology, Royal North Shore Hospital, Pacific Highway, St Leonards, Sydney, NSW 2065, Australia e-mail: [email protected] A. J. Gill Cancer Diagnosis and Pathology Research Group, Kolling Institute of Medial Research, St Leonards, NSW 2065, Australia A. J. Gill University of Sydney, Sydney, NSW, Australia 2006
carcinoma, predict a worse outcome in definite parathyroid carcinomas, triage formal genetic testing for HPT-JT syndrome, and predict the outcome of histologically atypical parathyroid adenomas. Keywords Parathyroid carcinoma . HRPT2 . CDC73 . Parafibromin Over the last 25 years, much has been learnt about the molecular mechanisms of disease and the genetic basis of many hereditary tumor syndromes. The molecular and hereditary basis of parathyroid carcinoma is no exception. Although progress has been slow, there has been a steady evolution in the understanding of this rare endocrine malignancy—particularly its association with the previously obscure hereditary cancer syndrome known as hyperparathyroidism-jaw tumor (HPT-JT) syndrome. This increased understanding has been paralleled by a significant increase in the reported incidence of parathyroid carcinoma, historically one of the rarest endocrine malignancies [1, 2]. This review concentrates on these developments and highlights both what we have learnt and what remains uncertain. In 1973, Shantz and Castleman described their experience with 70 cases of parathyroid carcinoma and reported that the histological features that distinguish parathyroid carcinoma from adenoma were a trabecular growth pattern, mitotic activity, broad fibrous bands, capsular penetration, and vascular space invasion [3]. Although the abnormal architectural and cytological features that they reported are still considered important clues to the diagnosis of carcinoma, it is now clear that the presence of an unequivocally invasive growth pattern is the most important predictor of subsequent malignant behavior of parathyroid tumors. Therefore, the WHO 2004 criteria now recommend that the pathological diagnosis of parathyroid carcinoma be restricted to lesions, which show
Endocr Pathol
unequivocal invasive growth as evidence by perineural invasion, vascular space invasion or full thickness capsular penetration with growth into adjacent tissues or metastasis [4]. Despite these refined criteria, parathyroid carcinoma remains difficult to diagnose with any certainty histologically, and perhaps, the key to understanding the evolution in the molecular and surgical pathology of parathyroid carcinoma over the last 25 years is to first acknowledge that physicians, surgeons, and pathologists have always had difficulty in distinguishing parathyroid carcinoma from benign disease in the absence of metastasis. For example, in one large multi-institutional study of 40 parathyroid carcinomas defined on the basis of biologically malignant behavior (i.e., recurrence and metastasis) rather than by pathological criteria alone, less than half were correctly diagnosed as malignant at first presentation based on the combined clinical and pathological findings [5]. In another review of 27 cases initially considered to be suspicious or malignant parathyroid tumors, 16 (59 % of the cases) were reclassified as benign on blinded pathological review and none of these recurred [6]. The same study reported that only four of the 11 patients given a definitive review diagnosis of carcinoma went on to behave in a malignant manner, and the remaining seven carcinoma patients (including four who had simple rather than en bloc resection) were alive with no evidence of disease at a median follow-up of 65 months. Even in Shantz and Castleman’s series, as many as 58 % of patients were alive without disease at the last follow-up [3]. The subjectivity and inadequacy of the pathological diagnosis of parathyroid carcinoma is further complicated by the geographic and institutional variation in reported incidence. In most series, carcinoma accounts for considerably less than 1 % of all patients with primary hyperparathyroidism [5–7], but this figure is as high as 5 % in some Japanese and Italian reports [8–11]. Of course, specific environmental or genetic factors may account for this difference, but it is intriguing to postulate that different pathological thresholds for diagnosis may also play a role. Although it is confounded by the fact that surgery, particularly radical or en bloc excision, has the capacity to interrupt the natural history of parathyroid carcinoma, that late recurrence is not uncommon, and that different institutions may use subtly different pathological criteria, it appears that perhaps as many as 50 % of parathyroid carcinomas which actually behave in a malignant manner will initially be considered benign using conventional clinicopathological criteria and perhaps as few as 15 % of all cases diagnosed as carcinoma prospectively (36 % if rigid histopathological criteria for malignancy are applied by a specialist endocrine pathologist) will go on to recur or behave in a malignant manner after simple excision. Molecular analysis holds promise to improve the diagnosis of parathyroid carcinoma, and the first steps toward identifying the molecular pathology of parathyroid carcinoma were
made in the 1970 when it was recognized that parathyroid carcinoma demonstrates a strong familial tendency. For example, in 1972, Frayha et al. [12] reported parathyroid carcinoma in siblings each with an adenoma and carcinoma; in 1974, Malette et al. [13] reported parathyroid carcinoma in the setting of familial hyperparathyroidism; in 1975, Leborgne et al. [14] reported parathyroid carcinoma in siblings, and in 1977, Dinnen et al. [15] reported parathyroid carcinoma in two separate kindreds with familial hyperparathyroidism one of which also had jaw tumors. In fact, careful reading of many subsequent large series of parathyroid carcinoma almost inevitably reveals a passing comment that some of the cases were familial or associated with familial hyperparathyroidism, although these findings were not always emphasized [16]. This familial syndrome of hyperparathyroidism and parathyroid carcinoma came to be known as hyperparathyroidismjaw tumor (HPT-JT) syndrome [OMIM#145001]. Given the rarity of parathyroid carcinoma in the general community, the striking lifetime risk of parathyroid carcinoma in HPT-JT that was reported to be as high as 15 % [4] suggested a strong association between the molecular pathology of this syndrome and parathyroid carcinoma. In 2002, Carpten and colleagues [17] studied 14 families with HPT-JT and found inactivating mutations at 1q25-32 in a gene initially named HRPT2 for ‘Hyperparathyroidism 2’ but subsequently renamed CDC73. HRPT2 contains 17 exons and encodes a 531 amino acid protein termed parafibromin because of its relationship to parathyroid disease and fibro-osseous lesions and was postulated as a tumor suppressor gene. It is not surprising that following the discovery of HRPT2 mutation as the cause of HPT-JT syndrome, investigators began to focus on the role of HRPT2 mutation in parathyroid carcinoma. In 2003, Howell et al. [18] found HRPT2 mutations in four of five apparently sporadic carcinomas but in none of 25 adenomas, 11 secondary hyperplasia, or six tertiary hyperplasias. In 2003 Shattuck et al. [19] also examined 15 patients with apparently sporadic parathyroid carcinomas and found that ten (67 %) of the carcinomas harbored pathogenic HRPT2 mutations. Furthermore, even in the absence of a family history, three of these ten patients (30 %) also had germline mutation of HRPT2. That is, they had unrecognized HPT-JT syndrome. In 2004, Cetani et al. [20] reported seven patients with parathyroid carcinoma, six of which were associated with HRPT2 mutation. Again, two of the six HRPT2 mutated parathyroid carcinomas were associated with germline HRPT2 mutation. Krebs et al. [21] went on to report no HRPT2 mutations in 60 consecutive apparently sporadic benign adenomas, and their meta-analysis suggested that the incidence of HRPT2 mutation in unselected parathyroid adenomas was less than 0.8 %. The combined results of these studies indicated that HRPT2 mutation occurs commonly (77 %) of parathyroid carcinomas and rarely (less than 1 %) in unselected benign disease. Importantly, because of the difficulties in diagnosing
Endocr Pathol
parathyroid carcinoma confidently pathologically, all three major studies investigating the incidence of HRPT2 mutation in parathyroid carcinoma required biological evidence of malignant behavior (that is recurrence or metastasis) as inclusion criteria. Therefore HRPT2 mutation occurs not in 77 % of cases pathologists have diagnosed as carcinoma but in 77 % of cases which have been diagnosed as carcinoma by pathologists and have also behaved in a malignant manner. Similarly, given that the estimated incidence of parathyroid carcinoma is less than 1 % and that the diagnosis is frequently not made until recurrence, the finding of HRPT2 mutation in less than 1 % of unselected parathyroids strongly supports HRPT2 mutation as being highly specific for parathyroid carcinoma. That is, because parathyroid carcinoma is so difficult to diagnose pathologically, the strength of these three studies lies in the fact that they used malignant behavior as the gold standard with which to correlate the significance of HRPT2 mutation, not just histopathological criteria. Despite the presence of mutational hotspots in exons 1, 2, and 7 of HRPT2 where approximately 80 % of pathogenic mutations occur [22], sequencing this gene is expensive and not readily available. Furthermore, perhaps as many as 35 % of HPT-JT syndrome associated HRPT2 mutations are large deletion mutations that require gross deletion analysis to identify [23]. Therefore despite its potential, HRPT2 mutation testing is difficult to deploy in the diagnostic surgical pathology laboratory. In 2004, Tan and colleagues [24] developed a mouse monoclonal antibody directed against parafibromin. Because HRPT2 mutations are commonly of a frameshift or truncating nature, they postulated that completely absent nuclear staining for parafibromin could be used to confirm HRPT2 mutation and reported that loss of nuclear staining occurs in 96 % of parathyroid carcinomas, in 89 % of HPT-JT associated ‘adenomas’ and in less than 1 % of benign parathyroid tumors. When we examined parafibromin expression in independent cohorts [25, 26], we found completely negative parafibromin staining in eight of 11 apparently sporadic parathyroid carcinomas (defined as not only meeting the WHO criteria for malignancy but also behaving in a malignant manner) and three of four HPT-JT related parathyroid tumors but positive staining in 100 unselected adenomas and 56 of 57 ‘giant adenomas’ (defined as adenomas greater than 2 g). The one giant adenoma that showed negative staining for parafibromin was initially classified as ‘atypical adenoma’—discussed in detail later. Although we and the others have found immunohistochemistry for parafibromin to be useful in both confirming a definitive diagnosis of malignancy and in triaging patients for germline mutation testing for HPT-JT [24, 25, 27–35], others have found the antibody difficult to deploy [36–38]. How can this discrepancy be explained?
There are few commercially available antibodies to parafibromin, and our experience has been that all can be technically difficult to deploy in the routine diagnostic laboratory—in fact, many laboratories have been unable to achieve satisfactory staining, and we have found that staining often needs to be repeated in individual cases using different protocols to achieve satisfactory results. This has led to markedly different staining patterns in different laboratories and also to the evolution of different criteria for interpreting the stain. We require completely absent nuclear staining for parafibromin in the presence of an internal positive control (that is positive staining in endothelial and stromal cells) before staining is considered negative [25–27, 30] (illustrated in Fig. 1), and this approach has been supported by others [35]. This approach is similar to that used to assess microsatellite instability in colorectal adenocarcinoma using mismatch repair immunohistochemistry where early antibodies were equally capricious. However, others have interpreted any loss of staining (i.e., focal or partial loss) as being negative [28, 31]. Furthermore, the difficulty in achieving a pathological diagnosis of carcinoma has resulted in varying reports on the sensitivity and specificity of loss of parafibromin staining for malignancy. Given that the description of the significance of HRPT2 mutation was based on tumors that did not just fulfil histological criteria for malignancy but also went on to behave in a malignant manner [18–20], it is not surprising that different groups report different sensitivities and specificities depending on the gold standard to which parafibromin staining is compared. That is, groups that use biological evidence of malignant behavior as the gold standard for the diagnosis of parathyroid carcinoma tend to report a high rate of negative staining for parafibromin or mutation in HRPT2 in carcinomas, whereas groups that use pathological features of malignancy without the requirement for malignant behavior tend to report a lower rate [18–20, 27]. Whether the difference is due to ‘parathyroid carcinomas’ being cured by surgery, false positive histology (i.e., groups reporting a low incidence of HRPT2 mutation in parathyroid carcinoma ‘overdiagnosing’ cancer in atypical glands), technical difficulties with immunohistochemistry or poor sensitivity of parafibromin is impossible to determine. Suffice it to say that some investigators have found parafibromin staining very useful and others not so useful. Of course, for the practicing surgical pathologist, the main issue is whether parafibromin immunohistochemistry or HRPT2 mutation analysis can predict the likelihood of recurrence or metastasis of histologically malignant parathyroid tumors or the likelihood of malignant behavior of atypical parathyroid adenomas. That is, are carcinomas which show histological features of malignancy and abnormal parafibromin staining more likely to behave in a malignant manner? And can the risk of malignant behavior of atypical parathyroid adenomas be predicted by abnormal staining for parafibromin?
Endocr Pathol
Fig. 1 H&E of a parathyroid carcinoma with parafibromin staining (inset). We emphasize that for a parathyroid tumor to be considered negative for parafibromin, there should be a strong positive internal controls - in this case endothelial cells and occasional lymphocytes
We recently reviewed our experience of immunohistochemistry for parafibromin in the differential diagnosis of atypical adenomas and in predicting the outcome of tumors which fulfilled the WHO 2004 criteria of malignancy [39]. We subdivided 81 parathyroid tumors that we encountered over a 14-year period into four groups: group A, parathyroid carcinomas, defined by the presence of WHO 2004 criteria of malignancy positive which also showed negative staining for parafibromin; group B, parathyroid carcinomas, which were also of course WHO criteria positive for malignancy but parafibromin positive; group C; atypical adenomas, which were WHO criteria for malignancy negative and parafibromin negative; and group D, atypical adenomas, which were WHO criteria for malignancy negative and parafibromin positive. Across the four groups from A to D, the rates of both death from disease and recurrence varied significantly - the risk of death in group A was 15 %; group B, 7 %; group, C nil; and group D, nil and the risk of recurrence varied in a similar pattern (group A, 38 %; group B, 36 %; group C, 20 %; and group D, nil). Five-year disease-free survival for groups A to D were 55, 80, 78, and 100 % respectively, and the risk of tumor recurrence was significantly associated with negative staining for parafibromin (p=0.048). Our study provided support for the concept that ‘atypical parathyroid adenomas’ which show a normal pattern of parafibromin staining either do not recur or do so extremely rarely (sufficiently rarely for them to be discharged from follow-up), whereas atypical parathyroid adenomas that show negative staining for parafibromin have a significant recurrence risk of 20 % (sufficient for them to be considered tumors of uncertain malignant potential and warrant ongoing followup). Furthermore, our data suggest that even in unequivocally histologically malignant parathyroid carcinomas, in addition to triaging genetic testing for HPT-JT syndrome, loss of
staining for parafibromin can be used to predict an increased risk of death or recurrence. A recent similar study performed by Cetani et al. only focusing on parafibromin as a prognostic indicator in parathyroid carcinoma rather than also investigating atypical adenomas also found that negative staining for parafibromin indicated a higher risk of recurrence and a decreased 5-year survival of 59 % (p=0.107) and very significantly decreased 10-year survival of 23 %, (p=0.0026) [40]. In 2014, what is the current understanding of parathyroid carcinoma, its relationship to HRPT2 mutation and HPT-JT syndrome and the role of parafibromin immunohistochemistry? It is now clearly established that HPT-JT syndrome in which there is a very significantly lifetime risk of parathyroid carcinoma is due to germline mutation of HRPT2. HPT-JT syndrome demonstrates a low penetrance and the absence of a family history of hyperparathyroidism in no way excludes germline HRPT2 mutation. Somatic inactivating mutations of HRPT2 are found in approximately 70 % of parathyroid carcinomas (when defined by malignant behavior but significantly less when defined by pathological criteria alone) but in less than 1 % of unselected parathyroid tumors indicating that the presence of HRPT2 mutation strongly supports a diagnosis of malignancy. In fact, there are even isolated case reports of HRPT2 mutated but histologically benign parathyroid tumors (confirmed by multiple endocrine pathologists) which have gone on to metastasize [41]. All patients with parathyroid carcinoma should be considered for germline testing for HRPT2 because up to 20 % of them will have unrecognized HPT-JT even in the absence of a family history [19, 20]. Immunohistochemistry for parafibromin can be used as a marker of HRPT2 mutation, but experience is mixed with some groups reporting high sensitivity and specificity and other groups finding the antibody difficult to deploy or the results of immunohistochemistry disappointing. Groups that have found parafibromin immunohistochemistry useful have employed it to triage patients with parathyroid carcinoma for formal genetic testing for HPT-JT syndrome. Furthermore, negative staining for parafibromin is beginning to be validated both as a biomarker of poor prognosis in definite parathyroid carcinomas as well as a marker of increased risk of recurrence (perhaps warranting the diagnosis of low grade carcinoma or at least ‘tumor of low or uncertain malignant potential’) in histologically atypical parathyroid tumors which do not quite fulfill conventional histological criteria for carcinoma [39, 40]. Conflict of Interest The author reports no conflict of interest.
References 1. Lee PK, Jarosek SL, Virnig BA, et al. Trends in the incidence and treatment of parathyroid cancer in the United States. Cancer 109: 1736-1741, 2007
Endocr Pathol 2. Brown S, O'Neill C, Suliburk J, et al. Parathyroid carcinoma: increasing incidence and changing presentation. ANZ J Surg 81:528-32, 2011 3. Schantz A, Castleman B Cancer 31:600-605, 1973 4. DeLellis RA, Lloyd RV, Heitz PU, et al. World Health organization classification of tumours Pathology and Genetics: Tumours of Endocrine Organs IARC Press Lyon 2004 5. Sandelin K, Tullgreen O, Farnebo LO. Clinical course of metastatic parathyroid carcinoma World J Surg 18:594-598, 1994 6. Ippolito G, Palazzo FF, Sebag F, et al. Intraoperative diagnosis and treatment of parathyroid cancer and atypical parathyroid adenoma British Journal of Surgery 94:566-570, 2007 7. Shane E Clinical review 122: parathyroid carcinoma J Clin Endocrinol Metabl 86:485-493, 2001 8. Obara T, Fujimoto Y Diagnosis and treatment of patients with parathyroid carcinoma. An update and review. World J Surg 15:738-744, 1991 9. Ishida T, Yokoe T, Izuo M Nationwide survey of parathyroid operations in Japan (1980-1989): Endocrine Surgery (Tokyo) 8:37, 1991 10. Favia G, Lumachi F, Polistina F et al. Parathyroid carcinoma: sixteen new cases and suggestions for correct management World J Surg 22: 1225-1230, 1998 11. Fujimoto Y, Obara T, Ito Y, et al. Localization and surgical resection of metastatic parathyroid carcinoma World J Surg 10:539, 1986 12. Frayha RA, Nassar VH, Dagher F, Salti IS. Familial parathyroid carcinoma J Med Liban 25:299-309, 1972 13. Mallette LE, Bilezikian JP, Ketcham AS, Aurbach GD. Parathyroid carcinoma in familial hyperparathyroidism. Am J Med 57:642-8, 1974 14. Leborgne J, Le Neel JC, Buzelin F, Malvy P Familial cancer of the parathyroid glands. Importance of angiography in the diagnosis of regional recurrences. Considerations on 2 cases. J Chir (Paris) 109: 315-26, 1975 15. Dinnen JS, Greenwoood RH, Jones JH, et al. Parathyroid carcinoma in familial hyperparathyroidism. J Clin Pathol 30:966-75, 1977 16. Smith JF, Coombs RGH Histological diagnosis of carcinoma of the parathyroid glandJ Clin Pathol 37:1370-137, 1984 17. Carpten JD, Robbins Cm, Villablanca A HRPT2, encoding parafibromin, is mutated in hyperparathyroidism-jaw tumor syndrome Nature Genetics 32:676-680, 2002 18. Howell VM, Haven CJ, Kahnoski K et al. HRPT2mutations are associated with malignancy in sporadic parathyroid tumours J Med Genet 40:657-663, 2003 19. Shattuck T, Stiina V, Obara T et al. Somatic and Germ-line mutations of the HRPT2 gene in sporadic parathyroid carcinoma New England Journal of Medicine 349:1722-1729, 2003 20. Cetani F, Pardi E, Borsari S, et al. Genetic Analyses of HRPT2 gene in primary hyperparathyroidism: Germline and Somatic Mutations in Familial and Sporadic Parathyroid Tumours Journal of Clinical Endocrinology and Metabolism 89:5583-5591, 2004 21. Kresbs L , Shattuck TM, Arnold A HRPT mutation analysis of typical sporadic parathyroid adenomas Journal of Clinical Endocrinology and Metabolism 90:5015-5017, 2005 22. Marsh DJ, Hahn MA, Howell VM, Gill AJ Molecular diagnosis of primary hyperparathyroidism in familial cancer syndromes Expert Opinion in Medical Diagnostics 1:377-92, 2007 23. Bricaire L, Odou MF, Cardot-Bauters C, et al. Frequent large germline HRPT2 deletions in a French National cohort of patients with primary hyperparathyroidism. J Clin Endocrinol Metab. 98: E403-8, 2013
24. Tan M-H, Morrison C, Wang P, et al. Loss of parafibromin immunoreactivity is a distinguishing feature of parathyroid carcinoma Clin Cancer Res 10:6629-6637, 2004 25. Gill AJ, Clarkson A, Gimm O, et al. Loss of nuclear expression of parafibromin distinguishes parathyroid carcinomas and Hyperparathyroidsim-Jaw Tumor (HPT-JT) syndrome related adenomas from sporadic parathyroid adenomas and hyperplasias American Journal of Surgical Pathology 30:1140-1149, 2006 26. Meyer-Rochow GY, Alvarado R, Sywak MS, et al. Intraoperative diagnosis and treatment of parathyroid cancer and atypical parathyroid adenoma British Journal of Surgery 94:1044, 2007 27. Howell VM, Gill AJ, Clarkson A, et al. Accuracy of combined protein gene product 9.5 and parafibromin markers for immunohistochemical diagnosis of parathyroid carcinoma Journal of Clinical Endocrinology and Metabolism 94:434-441, 2009 28. Wang O, Wang C, Nie M, et al. Novel HRPT2/CDC73 gene mutations and loss of expression of parafibromin in Chinese patients with clinically sporadic parathyroid carcinomas. PLoS One. 7(9):e45567, 2012 29. Wang O, Wang CY, Shi J, et al. Expression of Ki-67, galectin-3, fragile histidine triad, and parafibromin in malignant and benign parathyroid tumors. Chin Med J (Engl) 125:2895-901, 2012 30. Lim S, Elston MS, Gill AJ, et al. Metastatic parathyroid carcinoma initially misdiagnosed as parathyroid adenoma: the role of parafibromin in increasing diagnostic accuracy. Intern Med J 41: 695-9, 2011 31. Kim HK, Oh YL, Kim SH, et al. Parafibromin immunohistochemical staining to differentiate parathyroid carcinoma from parathyroid adenoma. Head Neck 34:201-6, 2012 32. Juhlin CC, Nilsson IL, Johansson K, et al. Parafibromin and APC as screening markers for malignant potential in atypical parathyroid adenomas. Endocr Pathol 21:166-77, 2010 33. Juhlin CC, Villablanca A, Sandelin K, et al. Parafibromin immunoreactivity: its use as an additional diagnostic marker for parathyroid tumor classification. Endocr Relat Cancer. 14:501-12, 2007 34. Cetani F, Ambrogini E, Viacava P, et al. Should parafibromin staining replace HRTP2 gene analysis as an additional tool for histologic diagnosis of parathyroid carcinoma? Eur J Endocrinol 156:547-54, 2007 35. Witteveen JE, Hamdy NA, Dekkers OM, et al. Downregulation of CASR expression and global loss of parafibromin staining are strong negative determinants of prognosis in parathyroid carcinoma. Mod Pathol. 24:688-97, 2011 36. Cabané T, Patricio, Gac E, et al. Detección inmunohistoquímica de parafibromina en patología de paratiroides. Revista chilena de cirugía, 65:20-24, 2013 37. Guarnieri V, Battista C, Muscarella LA et al. CDC73 mutations and parafibromin immunohistochemistry in parathyroid tumors: clinical correlations in a single-centre patient cohort Cellular Oncology 35: 411-422, 2012 38. DeLellis RA, Mazzaglia P, Mangray S. Primary hyperparathyroidism: a current perspective. Arch Pathol Lab Med. 132:1251-62, 2008 39. Krujiff S, Sidhu SB, Sywak MS, et al. Negative parafibromin staining predicts malignant behaviour in atypical parathyroid adenomas Annals of Surgical Oncology Oct 1 [epub ahead of print] 40. Cetani F, Banti C, Pardi E, et al. CDC73 mutational status and loss of parafibromin in the outcome of parathyroid cancer. Endocr Connect. 28:186-95, 2013 41. M.S. Sarquis, L.G. Silveira, F.J. Pimenta, et al., Familial hyperparathyroidism: surgical outcome after 30 years of follow-up in three families with germline HRPT2 mutations, Surgery 143:630–640, 2008
