World Journal of Surgery

, Volume 35, Issue 1, pp 147–153

Parathyroid Carcinoma Encountered After Minimally Invasive Focused Parathyroidectomy may not Require Further Radical Surgery

Authors

  • Christine J. O’Neill
    • University of Sydney Endocrine Surgical Unit
  • Conan Chan
    • University of Sydney Endocrine Surgical Unit
  • James Symons
    • University of Sydney Endocrine Surgical Unit
  • Diana L. Learoyd
    • Department of EndocrinologyRoyal North Shore Hospital
  • Stan B. Sidhu
    • University of Sydney Endocrine Surgical Unit
  • Leigh W. Delbridge
    • University of Sydney Endocrine Surgical Unit
  • Anthony Gill
    • Department of Anatomical PathologyRoyal North Shore Hospital
    • University of Sydney
    • University of Sydney Endocrine Surgical Unit
Article

DOI: 10.1007/s00268-010-0826-4

Cite this article as:
O’Neill, C.J., Chan, C., Symons, J. et al. World J Surg (2011) 35: 147. doi:10.1007/s00268-010-0826-4

Abstract

Background

Parathyroid carcinoma accounts for <1% of tumors in primary hyperparathyroidism (PHPT). Distinguishing parathyroid malignancy from benign disease is difficult both before and after initial surgery. Despite the improved specificity of a malignant diagnosis with immunohistochemistry for parafibromin and PGP9.5, proven metastatic behavior remains the gold standard of diagnosis. Minimally invasive focused parathyroidectomy (MIP) is widely performed in patients with PHPT and positive localization studies; thus, it is inevitable that some parathyroid caricnomas will be encountered at MIP. We present our experience of this rare entity.

Methods

The present study represents a surgical case series of patients with parathyroid carcinoma encountered after MIP. The clinicopathological features of benign and malignant parathyroid tumors were compared. Multiple regression analysis was undertaken to compare indicators of malignancy.

Results

Between May 1999 and April 2010, a total of 1,292 patients underwent MIP at the University of Sydney Endocrine Surgical Unit, and a histopathological diagnosis of parathyroid carcinoma was made in seven patients (0.5%). Staining for parafibromin and/or PGP9.5 was abnormal in five carcinomas (71%). Despite subsequent unilateral thyroid lobectomy and lymphadenectomy in six patients, no further malignancy was identified in any specimens. Compared to controls, preoperative calcium (p = 0.04) and parathyroid hormone (p = 0.01) were significantly higher in patients with malignancy. The positive predictive value of these parameters for carcinoma was 56 and 75%, respectively.

Conclusions

In patients diagnosed with parathyroid carcinoma after MIP where preoperative imaging had already demonstrated localized disease, revision en bloc surgery did not reveal any residual disease. The benefits of further radical surgery for parathyroid carcinoma after MIP remain controversial.

Introduction

Parathyroid carcinoma is a rare disease representing less than 1% of all cases of primary hyperparathyroidism (PHPT) [1]. Differentiating parathyroid carcinoma from benign PHPT is difficult and in the absence of metastases is rarely achieved preoperatively [2]. Intraoperative features of a firm and adherent mass raise the suspicion of malignancy but are insufficient to distinguish between atypical adenoma and parathyroid carcinoma [3]. The histopathological criteria for malignancy (as defined by the World Health Organisation; WHO) require unequivocal capsular, vascular, or perineural spread or the presence of metastases [4]. However, these features are poor predictors of behavior as they may be absent in cases that recur or metastasize and most cases that display features of malignancy will never recur. In fact, fewer than half of the cases that actually metastasize may be recognized as carcinoma at first presentation [5], and perhaps as few as 15% of cases that are classified as carcinoma (36% if using rigid histological criteria) will ever recur [3].

Recently, a better understanding of the molecular pathways involved in carcinogenesis, particularly its association with germline and sporadic HRPT2 mutations in an apparent classic double-hit relationship, has facilitated a more specific diagnosis [68]. The combined results of these studies indicate that HRPT2 mutations occur in approximately 77% of parathyroid carcinomas that actually behave in a malignant manner (rather than just meet the histopathological criteria for malignancy) but in less that 1% of unselected parathyroid tumors [9]. Given that even “histologically benign” HRPT2 mutated tumors (reviewed by multiple experienced endocrine pathologists) have been reported to metastasize [10], and that “benign” HRPT2 mutated tumors display the same gene expression profile as parathyroid carcinomas [11], it is reasonable to classify all HRPT2 mutated tumors as having at least some malignant potential. Immunohistochemistry (IHC), particularly staining for the expression of parafibromin and protein gene product 9.5 (PGP9.5), has been used as a de facto marker of HRPT2 mutation in order to improve the specificity of the histological diagnosis of carcinoma and triage genetic testing [12].

Minimally invasive focused parathyroidectomy (MIP) has become the standard of care in many centers for patients with PHPT where localization studies are positive [13]. This technique has been shown to be safe, with initial cure rates equivalent to bilateral neck exploration [1416]. Given the difficulty in differentiating parathyroid carcinoma from benign PHPT preoperatively, it is inevitable that some parathyroid carcinomas will be encountered at MIP. However, it is likely, given that this sub-group of patients has already been selected out on the basis of imaging studies demonstrating localized disease, that these particular cases will not have widespread local involvement, metastatic local nodal disease, or distant metastases.

The purpose of the present study was twofold: first, to reassess the clinical and biochemical parameters that can be used to distinguish parathyroid carcinoma from benign PHPT and second to review the surgical management and clinical follow-up of patients with parathyroid carcinoma encountered at MIP and thus determine the role of further surgery in this rare group of patients.

Material and methods

Patients who underwent parathyroidectomy within the 11-year period between May 1999 and April 2010 were identified from the prospectively maintained University of Sydney Endocrine Surgery Database. May 1999 was selected because minimally invasive focused parathyroidectomy (via a lateral mini-incision) was introduced to the University of Sydney Endocrine Surgical Unit at that time. Patients with a diagnosis of parathyroid carcinoma were identified and their medical charts reviewed.

The histopathological diagnosis of parathyroid carcinoma in the present study was based on WHO 2004 criteria and required unequivocal capsular, vascular, or perineural invasion or metastasis [4]. In more recent years, parafibromin and PGP9.5 IHC has been performed in conjunction with histology; however, the WHO criteria for malignancy were still the bases of the diagnosis of carcinoma. In three of the patients with carcinoma in the present series, IHC was performed at the time of histopathological examination. For the remaining patients, IHC was performed as part of a study by Howell et al. in 2009, and these results were obtained [12]. The details of additional surgical procedures and clinical follow-up for all patients with parathyroid carcinoma encountered after MIP were obtained.

A control group of 60 patients was derived from consecutive patients who underwent MIP for benign PHPT during a 6-month period between May and October 2007 (inclusive). Patients were excluded from the control group if they had secondary or tertiary hyperparathyroidism or underwent a concomitant thyroid procedure at the time of parathyroidectomy. Clinical, biochemical, and pathological characteristics of the control group were tabulated and compared to those from patients with parathyroid carcinoma encountered at MIP.

Statistical analysis was performed with SPSS software (version 17.0, SPSS Inc, Chicago, IL). Categorical variables were analyzed by Fisher’s exact test. The ability of clinical and biochemical parameters to predict malignancy were assessed using binary logistic regression analysis. The predictive ability was assessed by odds ratios and sensitivity/specificity. The positive and negative predictive values for different parameters were also calculated. Statistical significance was set at p < 0.05.

Results

During the 11-year study period, 2,747 parathyroid operations were performed in the University of Sydney Endocrine Surgical Unit, 1,292 via a minimally invasive focused approach and 1,455 via bilateral four-gland neck exploration. During this time 17 patients (0.6%) had a histopathological diagnosis of parathyroid carcinoma, 10 (0.7%) in the bilateral neck exploration group and 7 (0.5%) encountered at MIP (p = 0.46) (Fig. 1). The majority (14 of 17 tumors, including all seven encountered at MIP) were diagnosed at final histology and had not been suspected to be carcinoma preoperatively. Of the ten patients who underwent initial bilateral neck exploration, three had carcinomas that were suspected preoperatively and six had carcinomas that were suspected at the time of surgery. All of these nine patients underwent an initial cancer resection including ipsilateral thyroid lobectomy and central lymph node dissection. The one remaining patient in this group underwent further surgery on a separate occasion after the histological diagnosis of cancer.
https://static-content.springer.com/image/art%3A10.1007%2Fs00268-010-0826-4/MediaObjects/268_2010_826_Fig1_HTML.gif
Fig. 1

Parathyroid surgery for primary hyperparathyroidism from May 1999 to June 2010 at the University of Sydney Endocrine Surgical Unit

The characteristics of the seven patients with parathyroid carcinoma encountered at MIP are summarized in Table 1. Results of parafibromin and PGP9.5 staining are also reported, with abnormal staining patterns (parafibromin negative or PGP9.5 positive) confirming the diagnosis of malignancy in five of the seven patients. As has been the routine practice in this unit, six of the seven patients in this series underwent unilateral thyroid lobectomy and central lymph node dissection within a few weeks after MIP. The patient who did not undergo further surgery had a number of medical co-morbidities, and therefore further surgical intervention was deferred. All patients remain normocalcemic at a median of 44 months follow-up (range: 6–78 months). Two patients (cases 2 and 3) have persistently mildly elevated parathyroid hormone (PTH) levels (<1.5× upper limit of normal) in the presence of normal vitamin D metabolism and remain clinically well. Both of these patients continue to be followed with serial calcium and PTH levels without surgical intervention.
Table 1

Clinical, biochemical, and pathological characteristics of seven patients with parathyroid carcinoma encountered at MIP

Patient no.

1

2

3

4

5

6

7

Sex/age (years)

M/39

M/74

F/60

M/70

M/38

M/61

M/59

Preoperative cancer (mmol/l)

4.30

3.23

3.17

2.62

2.87

3.53

3.16

PTH (pmol/l)

173.0

42.4

49.0

27.9

15.7

43.9

100

Weight (mg)

1,630

4,400

3,760

2,500

1,880

1,348

1,098

Max diameter (mm)

21

40

30

35

24

20

25

Parafibromin

Positive

Positive

Positive

Positive

Negative

Positive

Negative

PGP9.5

Negative

Positive

Positive

Negative

Positive

Positive

Positive

Further surgerya

Yes

No

Yes

Yes

Yes

Yes

Yes

Clinical outcomeb

Disease-free

↑PTHb

↑PTHb

Disease-free

Disease-free

Disease-free

Disease-free

Months follow-up

66

55

29

33

17

25

6

aFurther intervention in all five cases where this occurred included thyroid lobectomy and central lymph node dissection on the side of the parathyroid cancer. In no case was residual malignancy identified

bIn both cases of high PTH, this is only mildly elevated (<1.5× upper limit of normal) with normal serum calcium

The clinical, biochemical, and histopathological features of the seven patients with parathyroid carcinoma encountered at MIP were compared to a control group of 60 patients with benign PHPT who were treated by MIP (Table 2). There was no significant difference in age between the two groups. Although after univariate analysis, gender and tumor weight were associated with malignancy, this association was no longer significant following multivariate analysis. Preoperative calcium, PTH, and maximal tumor diameter were significantly associated with malignancy after both univariate and multivariate analysis, with preoperative PTH being the most predictive (p = 0.01). The adjusted odds ratios for serum calcium and PTH were 1.52 and 1.33, respectively (Table 3).
Table 2

Comparison of clinical, biochemical, and pathological characteristics between cases of parathyroid cancer and controls with benign primary hyperparathyroidism

Variable

Cancer (n = 7)

Benign (n = 60)

Univariate analysis

Multivariate analysisa

Age, years (mean)

57.3

59.6

p = 0.68

p = 0.37

Male gender

6 (86%)

12 (20%)

p = 0.005

p = 0.07

Preoperative mean Ca (mmol/l)

3.27

2.76

p = 0.005

p = 0.04

PTH (pmol/l)

64.6

13.4

p = 0.002

p = 0.01

Mean weight (mg)

2,374

819

p = 0.003

p = 0.05

Mean diameter (mm)

28

19

p = 0.04

p = 0.04

aIn multivariate analysis not all covariates have been used due to collinearity

Ca cancer, PTH parathyroid hormone

Table 3

Adjusted odds ratio (OR) estimate of parathyroid cancer compared with benign primary hyperparathyroidism

Variable

OR

95% CI

p value

Preoperative PTH

1.33

1.05–1.55

0.01

Preoperative Caa

1.52

1.02–2.29

0.04

Tumor weight

1.00

1.00–1.01

0.05

Tumor diameter

1.13

1.01–1.26

0.04

aCalcium was transformed to calculate OR for an increase of 0.1 mmol/l

CI confidence interval

The sensitivity, specificity, and positive and negative predictive values of each of preoperative serum calcium, PTH, and tumor weight are presented in Table 4. Although cut-points of Ca 3.10 mmol/l, PTH 27.6 pmol/l (four times the upper limit of normal), and tumor weight of 1,500 mg have high negative predictive values (93–98%), the positive predictive value of these cut-points for calcium, PTH, and tumor weight were 56, 75, and 50%, respectively.
Table 4

Ability of preoperative serum calcium, PTH, and tumor weight to predict parathyroid cancer

Variable

Cut-off

Sensitivity, %

Specificity, %

PPV, %

NPV, %

Ca (mmol/l)

3.10

71

93

56

97

PTH (pmol/l)

27.6a

86

95

75

98

Weight (mg)

1,500

71

92

50

93

aThis measure is 4 × the upper limit of normal

PPV positive predictive value; NPV negative predictive value

Discussion

To our knowledge, this article documents the first series of parathyroid carcinoma encountered after MIP and raises a question about the need for further radical surgery. In the present study, the diagnosis of malignancy was based on the WHO 2004 criteria that require capsular, vascular, or perineural invasion or metastasis [4]. As these criteria are unreliable predictors of behavior, they have been supplemented with IHC for parafibromin and PGP9.5, which we believe increase the specificity of a diagnosis of carcinoma [12, 17]. Because not all parathyroid carcinomas are HRPT2 mutated, this ancillary testing may not increase the sensitivity of the diagnosis and thus histological criteria of malignancy remain central to diagnosis.

Previous series of parathyroid carcinomas aimed at improved preoperative prediction of malignancy concur with the results of this manuscript that patients with parathyroid carcinoma have significantly higher preoperative calcium and PTH levels [2, 1822]. The negative predictive value of calcium levels below 3.10 mmol/l and/or PTH levels below 27.6 pmol/l are high (97 and 98%, respectively). However, for the positive predictive value of calcium and PTH levels above these cut-points are lower (56 and 75%, respectively). These results agree with previously published data that the clinical utility of these biochemical parameters is limited in the prediction of parathyroid carcinoma [2, 22]. Thus, the decision to offer minimally invasive focused parathyroidectomy or conventional bilateral neck exploration will be based on a combination of these parameters and on the surgeon’s suspicion of malignancy. Consequently, in centers where MIP is offered to most patients with positive localization studies it is likely that some parathyroid carcinomas will be diagnosed after MIP.

In our unit the criteria used to offer a minimally invasive focused approach is confident concordant localization to a single site with sestamibi and ultrasound. Therefore it is also very likely that this subgroup of parathyroid carcinomas, having already been selected out on the basis of imaging studies that demonstrated disease localized to a single site, will be the very tumors that will not have widespread local involvement, metastatic local nodal disease, or distant metastases.

In the present study we have attempted to address the question of whether further surgery is required in patients diagnosed with parathyroid carcinoma following MIP. It has been the practice of this unit to perform thyroid lobectomy and central nodal dissection on the side of the malignancy ideally at the time of initial surgery or where diagnosis is not made until final pathology is available, as a secondary procedure. This practice was based on previously published case series of patients suggesting an inferior outcome in those who did not undergo en bloc resection [2227]. However, of the six patients in this series who underwent further surgery, no residual malignancy was identified at histopathological examination in any specimen. Similar results were recently reported by Enomoto et al. where in a series of 11 patients with parathyroid carcinoma, eight underwent prophylactic nodal dissection. No malignancy was identified in the specimens of these eight patients [28].

These data of histopathological specimens suggest that the therapeutic benefit of prophylactic nodal dissection should be questioned. However, the follow-up data of patients comparing unilateral clearance with parathyroidectomy alone in parathyroid carcinoma is scant. In this series, at a median follow-up of 44 months (range: 6–78 months), all patients remain normocalcemic. Two patients have a mildly elevated PTH (<1.5×, the upper limit of the normal range) with no clinical manifestations. It is too early to know if these mild elevations of PTH represent early recurrence, as an elevated PTH in the presence of normocalcemia is found in 9–27% of patients when measured at least 6 months after successful parathyroidectomy for benign PHPT [29]. These preliminary results will be followed and it is intended that long-term data will be subsequently reported.

The evidence base in parathyroid carcinoma remains poor due to the relative rarity of the disease, its indolent nature and, perhaps most importantly, the difficulties of obtaining an accurate diagnosis even after excision. Thus studies are largely limited to case series. Given that the reported incidence of parathyroid carcinoma varies from less that 1–5% of all cases of primary hyperparathyroidism despite little apparent variation in mortality from parathyroid carcinoma, it may even be that the tendency of a unit to “overdiagnose” parathyroid carcinoma may be at least as important in predicting long-term survival as any surgical or biological factors [4, 22, 3033].

As the pathological criteria for the diagnosis of parathyroid carcinoma have been historically inaccurate, the true natural history, particularly of localized parathyroid carcinoma, remains poorly defined. Recurrence is heralded by rising serum calcium and PTH and is most commonly due to either locoregional recurrence or pulmonary metastases [34, 35]. Surgical resection of recurrent or metastatic disease remains the mainstay of treatment and has been associated with improved survival [5, 36, 37]. Little benefit has been shown with conventional chemotherapy [5, 23, 38]. Traditionally, external beam radiotherapy was thought to be of little benefit; however, some small series suggest a benefit when used in either the adjuvant setting or for the treatment of locoregional recurrent disease [39, 40]. More recently, the availability of cinacalcet has revolutionized the management of patients who would otherwise have intractable hypercalcemia [41, 42] Cinacalcet hydrochloride increases the sensitivity of the calcium-sensing receptor to extracellular calcium. It reduces PTH while simultaneously lowering calcium phosphate product.

This small series reports seven patients with parathyroid carcinoma diagnosed after MIP in a unit with a relatively low incidence of parathyroid carcinoma (0.6%). Because in many centers MIP has become the standard of care for patients with well-localized PHPT, and because the ability of preoperative parameters to predict malignancy remains poor, it is inevitable that some parathyroid carcinomas will be diagnosed following MIP. It is likely that parathyroid carcinomas diagnosed in this manner are a highly selected subset, with no strong preoperative evidence of malignancy and without locoregional invasion. Thus in such a subgroup, where the diagnosis of parathyroid carcinoma is made postoperatively, we believe that the need for reoperation with unilateral thyroid lobectomy and central neck dissection should be questioned.

Given that conventional histology is inaccurate, and that HRPT2 mutation has been reported to be associated with parathyroid carcinomas that recur rather than simply meeting histological criteria for malignancy [68], it is not surprising that evidence of HRPT2 mutation predicts a worse prognosis after surgery. For example, in the series presented by Enomoto et al. in 2010, HRPT2 mutation was found in two of four cases that recurred, compared with zero of two that did not [28]. In our series two of two cases with mildly elevated PTH showed an abnormal pattern of staining for markers of HRPT2 versus three of the five with normal PTH in follow-up. Although in the present series the numbers of cases are too small to allow definitive conclusions, one approach may be to reserve radical reoperation to patients with evidence of HRPT2 mutation (parafibromin positive, PGP9.5 negative pattern of staining). For patients without evidence of HRPT2 mutation, further radical surgery may not be mandatory if there is no macroscopic impression of residual disease.

Conclusions

The ability to distinguish parathyroid carcinoma from benign primary hyperparathyroidism remains difficult, and therefore some cases will be diagnosed following minimally invasive focused parathyroidectomy. In patients with a postoperative diagnosis of malignancy the benefits of en bloc resection are questionable. Negative parafibromin and positive PGP9.5 staining may be associated with more aggressive disease and could be used as a marker for selection of patients for further radical surgery with thyroid lobectomy and lymph node clearance.

Acknowledgments

Dr. C. J. O’Neill is supported by a grant from the New South Wales Cancer Institute. Stan B. Sidhu is a New South Wales Cancer Institute Fellow. An earlier form of this article was presented as a poster presentation at the Society of Surgical Oncology, St. Louis, MO, March 2010.

Copyright information

© Société Internationale de Chirurgie 2010