J Endocrinol Metab

Journal of Endocrinology and Metabolism, ISSN 1923-2861 print, 1923-287X online, Open Access

Article copyright, the authors; Journal compilation copyright, J Endocrinol Metab and Elmer Press Inc

Journal website http://www.jofem.org

Original Article

Volume 9, Number 4, August 2019, pages 95-107

Primary Hyperparathyroidism: Comparing Cardiovascular Morbidity and Mortality in Patients Treated With Parathyroidectomy Versus Conservative Management

Figure 3. Bar chart showing difference in development of total ASCVD, hypertension, CAD-related events and CVD-related events between the two groups. *Denotes significance. ASCVD: atherosclerotic cardiovascular disease; CAD: coronary artery disease; CVD: cardiovascular disease.

Figure 4. Bar chart showing the difference in development of total ASCVD, hypertension, CAD-related events and CVD-related events between the two age-matched subgroups. *Denotes significance (P < 0.05). ASCVD: atherosclerotic cardiovascular disease; CAD: coronary artery disease; CVD: cardiovascular disease.

**Table 1.** Demographics and Baseline Biochemical Parameters of the Two Groups
	Conservative	Surgical	P value
Statistical tests: Mann-Whitney, unpaired t-test. *Denotes significance (P < 0.05). S.Ca: serum corrected calcium; PTH: parathyroid hormone.
Number of patients	264	72	-
Male/female	59:277	11:61	-
Mean age in years (range)	68.7 (26 - 90)	56.8 (33 - 84)	0.001*
Mean S.Ca at diagnosis (mmol/L)	2.70	2.72	0.498
Mean PTH at diagnosis (pg/mL)	11.4	16.5	0.034*

Table 1. Demographics and Baseline Biochemical Parameters of the Two Groups

Conservative

Surgical

P value

Statistical tests: Mann-Whitney, unpaired t-test. *Denotes significance (P < 0.05). S.Ca: serum corrected calcium; PTH: parathyroid hormone.

Number of patients

264

Male/female

59:277

11:61

Mean age in years (range)

68.7 (26 - 90)

56.8 (33 - 84)

0.001*

Mean S.Ca at diagnosis (mmol/L)

2.70

2.72

0.498

Mean PTH at diagnosis (pg/mL)

11.4

16.5

0.034*

**Table 2.** Primary and Secondary Outcomes
	Conservative	Surgical	P value
Statistical tests: Mann-Whitney, Fisher’s test. *Denotes significance (P < 0.05). Percentage is not stated for total ASCVD as single patient may have had multiple events. CAD: coronary artery disease; ASCVD: atherosclerotic cardiovascular disease.
Mortality	62 (23.4%)	10 (13.8%)	0.104
Hypertension	119 (45.2%)	18 (25%)	0.002*
CAD-related events	45 (17.1%)	2 (2.8%)	0.001*
Cerebrovascular disease	27 (10.2%)	6 (8%)	0.824
Peripheral vascular disease	22 (8%)	6 (8%)	1.00
Total ASCVD events	213	32	0.001*
Duration of clinic follow-up in years (range)	7.25 (0.5 - 23)	2.42 (0.3 - 13)	0.001*

Table 2. Primary and Secondary Outcomes

Conservative

Surgical

P value

Statistical tests: Mann-Whitney, Fisher’s test. *Denotes significance (P < 0.05). Percentage is not stated for total ASCVD as single patient may have had multiple events. CAD: coronary artery disease; ASCVD: atherosclerotic cardiovascular disease.

Mortality

62 (23.4%)

10 (13.8%)

0.104

Hypertension

119 (45.2%)

18 (25%)

0.002*

CAD-related events

45 (17.1%)

2 (2.8%)

0.001*

Cerebrovascular disease

27 (10.2%)

6 (8%)

0.824

Peripheral vascular disease

22 (8%)

6 (8%)

1.00

Total ASCVD events

213

0.001*

Duration of clinic follow-up in years (range)

7.25 (0.5 - 23)

2.42 (0.3 - 13)

0.001*

**Table 3.** Logistic Regression Analysis
	B (coefficient)	SE	Sig.	OR (Exp(b))	95% CI for OR
*Denotes significance (P < 0.05). CI: confidence interval; OR: odds ratio.
Age	0.033	0.010	0.001*	1.033	1.014	1.053
Surgery	-0.729	0.330	0.027*	0.482	0.253	0.921
Constant	-2.816	0.747	0.000	0.060

Table 3. Logistic Regression Analysis

B (coefficient)

Sig.

OR (Exp(b))

95% CI for OR

Lower

Upper

*Denotes significance (P < 0.05). CI: confidence interval; OR: odds ratio.

Age

0.033

0.010

0.001*

1.033

1.014

1.053

Surgery

-0.729

0.330

0.027*

0.482

0.253

0.921

Constant

-2.816

0.747

0.000

0.060

**Table 4.** Demographics and Baseline Biochemical Parameters of the Two Groups
	Conservative (age-matched subgroup)	Surgical	P value
Statistical tests: Mann-Whitney, unpaired t-test. *Denotes significance (P < 0.05). S.Ca: serum corrected calcium; PTH: parathyroid hormone.
Number of patients	72	72	-
Male/female	11:61	11:61	-
Mean age at diagnosis	59.8 (26 - 87)	56.8 (33 - 84)	0.139
Mean age now/age at death (years)	66.3 (30 - 96)	65.9 (40 - 95)	0.865
Mean S.Ca at diagnosis (mmol/L)	2.67	2.73	0.221
Mean PTH at diagnosis (mmol/L)	10.07	15.1	0.018*

Table 4. Demographics and Baseline Biochemical Parameters of the Two Groups

Conservative (age-matched subgroup)

Surgical

P value

Statistical tests: Mann-Whitney, unpaired t-test. *Denotes significance (P < 0.05). S.Ca: serum corrected calcium; PTH: parathyroid hormone.

Number of patients

Male/female

11:61

Mean age at diagnosis

59.8 (26 - 87)

56.8 (33 - 84)

0.139

Mean age now/age at death (years)

66.3 (30 - 96)

65.9 (40 - 95)

0.865

Mean S.Ca at diagnosis (mmol/L)

2.67

2.73

0.221

Mean PTH at diagnosis (mmol/L)

10.07

15.1

0.018*

**Table 5.** Primary and Secondary Outcome Data in the Two Subgroups
	Conservative (age-matched subgroup)	Surgical	P value
Statistical tests: Chi-squared, Fisher’s test. *Denotes significance (P < 0.05). Percentage is not stated for total ASCVD as single patient may have had multiple events. CAD: coronary artery disease; ASCVD: atherosclerotic cardiovascular disease.
Mortality	6 (8.3%)	10 (13.8%)	0.427
Hypertension	31 (43%)	18 (25%)	0.022*
CAD-related events	14 (19%)	2 (2.8%)	0.002*
Cerebrovascular disease	3 (4%)	6 (8%)	0.299
Peripheral vascular disease	5 (7%)	6 (8%)	0.089
Total ASCVD	53	32	0.001*

Table 5. Primary and Secondary Outcome Data in the Two Subgroups

Conservative (age-matched subgroup)

Surgical

P value

Statistical tests: Chi-squared, Fisher’s test. *Denotes significance (P < 0.05). Percentage is not stated for total ASCVD as single patient may have had multiple events. CAD: coronary artery disease; ASCVD: atherosclerotic cardiovascular disease.

Mortality

6 (8.3%)

10 (13.8%)

0.427

Hypertension

31 (43%)

18 (25%)

0.022*

CAD-related events

14 (19%)

2 (2.8%)

0.002*

Cerebrovascular disease

3 (4%)

6 (8%)

0.299

Peripheral vascular disease

5 (7%)

6 (8%)

0.089

Total ASCVD

0.001*

**Table 6.** Logistic Regression Analysis
	B (coefficient)	SE	Sig.	OR (Exp (b))	95% CI for OR
CI: confidence interval; OR: odds ratio.
Surgery	-1.249	0.357	0.001*	0.287	0.142	0.578
Constant	-0.223	0.237	0.347	1.250

Table 6. Logistic Regression Analysis

B (coefficient)

Sig.

OR (Exp (b))

95% CI for OR

Lower

Upper

CI: confidence interval; OR: odds ratio.

Surgery

-1.249

0.357

0.001*

0.287

0.142

0.578

Constant

-0.223

0.237

0.347

1.250

**Table 7.** Literature Review of Studies Covering the Association of PHPT With Hypertension and Left Ventricular Hypertrophy
Study authors (reference)	Design	Number of participants	Results	Comments
ARR: aldosterone to renin ratio; BP: blood pressure; CKD: chronic kidney disease; EPATH: “Eplerenone in Primary Hyperparathyroidism” trial; HTN: hypertension; LVH: left ventricular hypertrophy; PTH: parathyroid hormone; MESA: Multi-Ethnic Study of Atherosclerosis; PHPT: primary hyperparathyroidism; SBP: systolic blood pressure.
Brown et al, 2015 [44]	Cross-sectional analysis of MESA participants without PHPT or CKD	5,668	Higher serum aldosterone was associated with higher serum PTH. Higher PTH levels were observed in patients with HTN. The use of RAAS inhibitors was associated with lower PTH levels.	Large diverse population. Association shown, causation not clear. Led onto EPATH study.
Heyliger et al, 2009 [28]	Retrospective review of patients undergoing parathyroidectomy for PHPT or thyroidectomy for non-toxic thyroid disease	464	Significant decrease in BP (< 10 mm Hg) after parathyroidectomy in patients with preoperative hypertension.	Large sample size. Retrospective design. Single BP measurements taken pre- and post-operatively. Patients who had preoperative alterations in BP medications were excluded.
Nainby-Luxmoore et al, 1982 [25]	Case control study; patients with PHPT undergoing parathyroidectomy and matched surgical controls	248	Mean BP in patients with PHPT was higher than matched controls. This difference was no longer observed at discharge.	Difference persisted when controlled for renal function.
Stefenelli et al, 1997 [29]	Prospective study; baseline LV mass pre- and post- elective parathyroidectomy on patients with PHPT	119	Regression of LVH after 41 months of parathyroidectomy. LVH reduction was independent of changes in BP.	LV mass measured using echocardiography; potential for operator variability in measurements.
Verheyen et al, 2016 [27]	Cross-sectional analysis of the EPATH trial, studied association between PTH levels and nocturnal hypertension	136	PTH levels did not correlate with nocturnal hypertension. Significant relationship between ARR and nocturnal SBP in patients with high PTH.	Small sample size, single-center study. Daytime SBP readings not assessed.
Yao et al, 2016 [26]	Prospective observational study on health population, median follow-up of 6 years, monitored for development of HTN	7,504	Raised PTH levels were associated with an increased risk of development of hypertension.	Large sample size. Participants with pre-existing ASCVD at baseline excluded. PTH only measured on one occasion.

Table 7. Literature Review of Studies Covering the Association of PHPT With Hypertension and Left Ventricular Hypertrophy

Study authors (reference)

Design

Number of participants

Results

Comments

ARR: aldosterone to renin ratio; BP: blood pressure; CKD: chronic kidney disease; EPATH: “Eplerenone in Primary Hyperparathyroidism” trial; HTN: hypertension; LVH: left ventricular hypertrophy; PTH: parathyroid hormone; MESA: Multi-Ethnic Study of Atherosclerosis; PHPT: primary hyperparathyroidism; SBP: systolic blood pressure.

Brown et al, 2015 [44]

Cross-sectional analysis of MESA participants without PHPT or CKD

5,668

Higher serum aldosterone was associated with higher serum PTH. Higher PTH levels were observed in patients with HTN. The use of RAAS inhibitors was associated with lower PTH levels.

Large diverse population. Association shown, causation not clear. Led onto EPATH study.

Heyliger et al, 2009 [28]

Retrospective review of patients undergoing parathyroidectomy for PHPT or thyroidectomy for non-toxic thyroid disease

464

Significant decrease in BP (< 10 mm Hg) after parathyroidectomy in patients with preoperative hypertension.

Large sample size. Retrospective design. Single BP measurements taken pre- and post-operatively. Patients who had preoperative alterations in BP medications were excluded.

Nainby-Luxmoore et al, 1982 [25]

Case control study; patients with PHPT undergoing parathyroidectomy and matched surgical controls

248

Mean BP in patients with PHPT was higher than matched controls. This difference was no longer observed at discharge.

Difference persisted when controlled for renal function.

Stefenelli et al, 1997 [29]

Prospective study; baseline LV mass pre- and post- elective parathyroidectomy on patients with PHPT

119

Regression of LVH after 41 months of parathyroidectomy. LVH reduction was independent of changes in BP.

LV mass measured using echocardiography; potential for operator variability in measurements.

Verheyen et al, 2016 [27]

Cross-sectional analysis of the EPATH trial, studied association between PTH levels and nocturnal hypertension

136

PTH levels did not correlate with nocturnal hypertension. Significant relationship between ARR and nocturnal SBP in patients with high PTH.

Small sample size, single-center study. Daytime SBP readings not assessed.

Yao et al, 2016 [26]

Prospective observational study on health population, median follow-up of 6 years, monitored for development of HTN

7,504

Raised PTH levels were associated with an increased risk of development of hypertension.

Large sample size. Participants with pre-existing ASCVD at baseline excluded. PTH only measured on one occasion.

**Table 8.** Literature Review of Studies Covering Impact of PHPT on Vascular Endothelium, Cardiac Function and Valvular Calcification
Study authors (reference)	Design	Number of participants	Results	Comments
25(OH)D: 25-hydroxyvitamin D; Aix: augmentation index; AVCA: aortic valve calcification area; CVD: cardiovascular disease; HF: heart failure; HTN: hypertension; IMT: intima media thickness; LVM: left ventricular mass; PHPT: primary hyperparathyroidism; PTH: parathyroid hormone; PWA: pulse wave analysis; PWV: pulse wave velocity; MRI: magnetic resonance imagining.
Bansal et al, 2014 [39]	Cross-sectional analysis of participants of MESA. Healthy cohort without ASCVD.	6,459	Higher PTH associated with higher levels of incident HF (hazard ratio: 1.96). Higher PTH associated with higher LVM (11g mean difference).	Multi-ethnic large population. Cardiac MRI used to quantify LVM. Single baseline measurement of PTH and 25(OH)D.
Dalberg et al, 1996 [40]	Prospective case control study; patients with PHPT undergoing parathyroidectomy and matched controls undergoing thyroidectomy.	67	Higher BP, left atrial diameter and diastolic dysfunction in PHPT patients when compared to controls. Reduced cardiac calcifications post parathyroidectomy.	Prospective design. Small sample size. Use of echocardiography associated with potential of operator bias.
Iwata et al, 2012 [10]	Case-control study; patients with PHPT and matched controls (further analysis of Walker 2009 cohort).	100	AVCA positively associated with PTH level; no difference in peak transaortic pressure gradient.	Control group had more cardiovascular risk factors. Use of echocardiography associated with potential operator bias. AVCA is an indirect marker for ASCVD.
Rosa et al, 2011 [31]	Case-control study; 28 patients with PHPT and hypertension, 16 with PHPT without HTN, 28 essential HTN patients, 18 healthy controls.	90	PWV significantly higher in patients with PHPT and HTN compared with those with essential HTN. Surgery significantly decreased systolic BP and PWV.	Well matched baseline characteristics. Singular BP measurements (not ambulatory).
Rubin et al, 2005 [32]	Case-control study; patients with mild PHPT and matched healthy controls.	173	PHPT was an independent predictor of increased arterial stiffness. PTH levels directly correlated with arterial stiffness (measured by AIx)	Small proportion of cases compared to controls. Surrogate marker of arterial stiffness used.
Schillaci et al, 2011 [34]	Case-control study; patients with mild PHPT and matched healthy controls.	72	Aortic PWV was significantly higher among PHPT patients. Aortic PWV decreased after surgery. Association remained significant when controlled for BP.	Patients with pre-existing hypertension, CV disease excluded. PWV is a surrogate marker.
Smith et al, 2000 [33]	Case-control study; patients with mild PHPT and matched healthy controls.	42	PWA showed augmentation in PHPT group significantly higher than controls. Central aortic systolic pressure also elevated.	PWA indirect measure of vascular stiffness (surrogate marker of clinical disease). Small sample size.
Vestergaard et al, 2003 [41]	Case-control study, patients with PHPT undergoing parathyroidectomy compared to matched controls.	2,695	Increased incidence of acute MI in PHPT patients up to 10 years prior and 1 year post surgery. Increased risk of HTN and HF, which did not resolve by 1 year post surgery.	Large sample size. Historical follow-up.
Walker et al, 2009 [35]	Case-control study; patients with PHPT and matched healthy controls.	1,040	IMT higher in PHPT patients. Odds of abnormal carotid stiffness increased for every 10 pg/mL increase in PTH.	Small proportion of cases compared to controls. Control group was not well matched with respect to cardiovascular risk factors.
Wannamethee et al, 2014 [38]	Prospective observational study of healthy cohort (British Regional Heart Study); 13-year follow-up.	3,731	Raised PTH at baseline was associated with increased CV risk factors. Higher risk of incident HF with elevated PTH levels (hazard ratio: 1.66) independent of HTN, renal dysfunction, calcium and vitamin D levels.	Large sample size. Study based on older, predominately white, male population. Doctor diagnosed HF, echo not available in all patients.

Table 8. Literature Review of Studies Covering Impact of PHPT on Vascular Endothelium, Cardiac Function and Valvular Calcification

Study authors (reference)

Design

Number of participants

Results

Comments

25(OH)D: 25-hydroxyvitamin D; Aix: augmentation index; AVCA: aortic valve calcification area; CVD: cardiovascular disease; HF: heart failure; HTN: hypertension; IMT: intima media thickness; LVM: left ventricular mass; PHPT: primary hyperparathyroidism; PTH: parathyroid hormone; PWA: pulse wave analysis; PWV: pulse wave velocity; MRI: magnetic resonance imagining.

Bansal et al, 2014 [39]

Cross-sectional analysis of participants of MESA. Healthy cohort without ASCVD.

6,459

Higher PTH associated with higher levels of incident HF (hazard ratio: 1.96). Higher PTH associated with higher LVM (11g mean difference).

Multi-ethnic large population. Cardiac MRI used to quantify LVM. Single baseline measurement of PTH and 25(OH)D.

Dalberg et al, 1996 [40]

Prospective case control study; patients with PHPT undergoing parathyroidectomy and matched controls undergoing thyroidectomy.

Higher BP, left atrial diameter and diastolic dysfunction in PHPT patients when compared to controls. Reduced cardiac calcifications post parathyroidectomy.

Prospective design. Small sample size. Use of echocardiography associated with potential of operator bias.

Iwata et al, 2012 [10]

Case-control study; patients with PHPT and matched controls (further analysis of Walker 2009 cohort).

100

AVCA positively associated with PTH level; no difference in peak transaortic pressure gradient.

Control group had more cardiovascular risk factors. Use of echocardiography associated with potential operator bias. AVCA is an indirect marker for ASCVD.

Rosa et al, 2011 [31]

Case-control study; 28 patients with PHPT and hypertension, 16 with PHPT without HTN, 28 essential HTN patients, 18 healthy controls.

PWV significantly higher in patients with PHPT and HTN compared with those with essential HTN. Surgery significantly decreased systolic BP and PWV.

Well matched baseline characteristics. Singular BP measurements (not ambulatory).

Rubin et al, 2005 [32]

Case-control study; patients with mild PHPT and matched healthy controls.

173

PHPT was an independent predictor of increased arterial stiffness. PTH levels directly correlated with arterial stiffness (measured by AIx)

Small proportion of cases compared to controls. Surrogate marker of arterial stiffness used.

Schillaci et al, 2011 [34]

Case-control study; patients with mild PHPT and matched healthy controls.

Aortic PWV was significantly higher among PHPT patients. Aortic PWV decreased after surgery. Association remained significant when controlled for BP.

Patients with pre-existing hypertension, CV disease excluded. PWV is a surrogate marker.

Smith et al, 2000 [33]

Case-control study; patients with mild PHPT and matched healthy controls.

PWA showed augmentation in PHPT group significantly higher than controls. Central aortic systolic pressure also elevated.

PWA indirect measure of vascular stiffness (surrogate marker of clinical disease). Small sample size.

Vestergaard et al, 2003 [41]

Case-control study, patients with PHPT undergoing parathyroidectomy compared to matched controls.

2,695

Increased incidence of acute MI in PHPT patients up to 10 years prior and 1 year post surgery. Increased risk of HTN and HF, which did not resolve by 1 year post surgery.

Large sample size. Historical follow-up.

Walker et al, 2009 [35]

Case-control study; patients with PHPT and matched healthy controls.

1,040

IMT higher in PHPT patients. Odds of abnormal carotid stiffness increased for every 10 pg/mL increase in PTH.

Small proportion of cases compared to controls. Control group was not well matched with respect to cardiovascular risk factors.

Wannamethee et al, 2014 [38]

Prospective observational study of healthy cohort (British Regional Heart Study); 13-year follow-up.

3,731

Raised PTH at baseline was associated with increased CV risk factors. Higher risk of incident HF with elevated PTH levels (hazard ratio: 1.66) independent of HTN, renal dysfunction, calcium and vitamin D levels.

Large sample size. Study based on older, predominately white, male population. Doctor diagnosed HF, echo not available in all patients.

**Table 9.** Literature Review of Studies Covering Mortality and PHPT
Study authors (reference)	Design	Number of participants	Results	Comments
CI: confidence interval; CVD: cardiovascular disease; PHPT: primary hyperparathyroidism; RR: relative risk.
Hedback et al, 1991 [45]	Retrospective observational study on patients with PHPT who had undergone parathyroidectomy; 13-year follow-up.	896	Increased risk of death in patients who had undergone parathyroidectomy when compared to the general population. This effect was more marked in those operated on in earlier calendar years, and in older patients.	Only included patients who underwent surgery. Historical study, therefore, caution should be applied if extrapolating these results to current surgical practice.
Lundgren et al, 2001 [9]	Observational case-control study; patients with hypercalcemia and matched normocalcemic controls.	516	Hypercalcemia was independently associated with increased CV mortality. There were also higher PTH levels in cases versus controls.	Large sample size, follow-up over 20 years. Association shown but cannot comment on causality.
Nilsson et al, 2002 [44]	Retrospective cohort study; patients with PHPT who had undergone parathyroidectomy.	10,995	In earlier calendar years increased risk of death following surgery. In those operated in later calendar years mortality did not differ from normal population. Principle causes of mortality included cardiovascular disease.	Large sample size. Authors hypothesise difference in mortality may be due to increased severity of disease when operated (as routine calcium screening introduced in middle of study period).
Ogard et al, 2004 [46]	Retrospective cohort study, patient with PHPT who were managed either conservatively or surgically.	1,578	Patients with a diagnosis of PHPT had a significantly increased risk of premature death compared to general population. Lower risk of death from all causes in women who underwent surgery. Significantly increased mortality for CVD for women treated conservatively.	Patients identified via national register. Increased risk of cancer in PHPT patients.
Palmer et al, 1987 [42]	Observational study of patients with PHPT who had undergone parathyroidectomy; 22-year follow-up	441	Higher mortality postoperatively than general population.	Long follow-up period. Historical study, therefore, caution should be applied if extrapolating these results to current surgical practice.
Uden et al, 1990 [43]	Multi-center retrospective case control study, long-term follow-up of patients with PHPT who had undergone parathyroidectomy and matched controls.	282	Increased mortality in patients who had undergone parathyroidectomy (survival reduced by 18%). Only 1% operative mortality.	Multi-center study, large sample size. Historical study, therefore, caution should be applied if extrapolating these results to current surgical practice.
Wermers et al, 1998 [11]	Retrospective cohort study; patients with confirmed PHPT or hypercalcemia of unknown cause.	435	29% underwent parathyroidectomy. Patients with higher serum calcium levels had an increased risk of death. Conservative patients had a non-significant trend towards decreased survival.	Large sample group. Long follow-up period. Majority of patients were white.

Table 9. Literature Review of Studies Covering Mortality and PHPT

Study authors (reference)

Design

Number of participants

Results

Comments

CI: confidence interval; CVD: cardiovascular disease; PHPT: primary hyperparathyroidism; RR: relative risk.

Hedback et al, 1991 [45]

Retrospective observational study on patients with PHPT who had undergone parathyroidectomy; 13-year follow-up.

896

Increased risk of death in patients who had undergone parathyroidectomy when compared to the general population. This effect was more marked in those operated on in earlier calendar years, and in older patients.

Only included patients who underwent surgery. Historical study, therefore, caution should be applied if extrapolating these results to current surgical practice.

Lundgren et al, 2001 [9]

Observational case-control study; patients with hypercalcemia and matched normocalcemic controls.

516

Hypercalcemia was independently associated with increased CV mortality. There were also higher PTH levels in cases versus controls.

Large sample size, follow-up over 20 years. Association shown but cannot comment on causality.

Nilsson et al, 2002 [44]

Retrospective cohort study; patients with PHPT who had undergone parathyroidectomy.

10,995

In earlier calendar years increased risk of death following surgery. In those operated in later calendar years mortality did not differ from normal population. Principle causes of mortality included cardiovascular disease.

Large sample size. Authors hypothesise difference in mortality may be due to increased severity of disease when operated (as routine calcium screening introduced in middle of study period).

Ogard et al, 2004 [46]

Retrospective cohort study, patient with PHPT who were managed either conservatively or surgically.

1,578

Patients with a diagnosis of PHPT had a significantly increased risk of premature death compared to general population. Lower risk of death from all causes in women who underwent surgery. Significantly increased mortality for CVD for women treated conservatively.

Patients identified via national register. Increased risk of cancer in PHPT patients.

Palmer et al, 1987 [42]

Observational study of patients with PHPT who had undergone parathyroidectomy; 22-year follow-up

441

Higher mortality postoperatively than general population.

Long follow-up period. Historical study, therefore, caution should be applied if extrapolating these results to current surgical practice.

Uden et al, 1990 [43]

Multi-center retrospective case control study, long-term follow-up of patients with PHPT who had undergone parathyroidectomy and matched controls.

282

Increased mortality in patients who had undergone parathyroidectomy (survival reduced by 18%). Only 1% operative mortality.

Multi-center study, large sample size. Historical study, therefore, caution should be applied if extrapolating these results to current surgical practice.

Wermers et al, 1998 [11]

Retrospective cohort study; patients with confirmed PHPT or hypercalcemia of unknown cause.

435

29% underwent parathyroidectomy. Patients with higher serum calcium levels had an increased risk of death. Conservative patients had a non-significant trend towards decreased survival.

Large sample group. Long follow-up period. Majority of patients were white.