Abstract
Background:
Primary hyperparathyroidism (PHPT) caused by double adenoma may carry a higher risk of failure to cure. We compared outcomes in single adenoma (SA), double adenoma (DA) and four-gland hyperplasia (HP).
Methods:
Patients undergoing initial parathyroidectomy for PHPT were categorized by diagnosis. The primary outcome was persistent/recurrent disease postoperatively.
Results:
Of 3,408 patients, 81.3% had SA, 9.5% had DA, and 9.3% had HP. Rates of persistence/recurrence were 2.9%, 5.3%, and 4.5% in SA, DA, and HP, respectively (p=0.281). Patients with persistence/recurrence had higher preoperative calcium (11.0 vs 10.7 mg/dl, p=0.028) and PTH (96 vs 77 pg/ml, p=0.015), and lower rates of IOPTH normalization (77% vs 96% p<0.001). On multivariable regression, DA was associated with increased risk of persistent/recurrent disease (OR 3.0, p=0.017).
Conclusions:
Most patients with DA are cured with removal of two glands, but approximately 5% experience disease persistence/recurrence. Low-normal final IOPTH was associated with lower risk of persistent/recurrent disease.
Keywords: primary hyperparathyroidism, double adenoma, hypercalcemia, parathyroid hyperplasia
Introduction
Primary hyperparathyroidism (PHPT) is the most common cause of hypercalcemia, affecting approximately 1% of adults in the general population (1). Parathyroidectomy remains the only definitive cure, and curative surgery is the most cost-effective treatment (2). Surgery is highly successful, with cure rates ranging from 94–99% (3, 4). However, one large population-based study found more than 7% of patients who underwent parathyroidectomy for PHPT experienced disease persistence and more than 5% eventually experienced disease recurrence (5). Although relatively uncommon, it is critical to understand the etiology of operative failure when it does occur, as reoperation for hyperparathyroidism carries increased risk of procedure-specific complications and 30 day readmission (6).
While most cases of PHPT are caused by a single benign parathyroid adenoma, multi-gland disease (MGD) is present in up to 15% of patients. Double adenomas (DA) have been reported in up to 12% of patients, with four-gland hyperplasia (HP) accounting for the remainder of MGD (7, 8). MGD is associated with increased risk of both intraoperative failure and persistent hyperparathyroidism and has been identified in almost 90% of patients with recurrent disease (3, 4, 9–11). Historically, DA has been considered to be a distinct diagnosis based on the observation of cure following removal of two parathyroid glands (7). Among patients with DA, reported rates of persistent and recurrent disease are 4% and 7%, respectively (12). Due to the risk of persistent/recurrent disease in DA, some authors suggest that DA may represent asymmetric or asynchronous hyperplasia (12). Distinguishing between DA and asymmetric hyperplasia has important implications for patient care. While patients with true DA can be cured after removal of two abnormally functioning glands, those with asymmetric hyperplasia remain at elevated risk of persistence/recurrence. Because of uncertainty surrounding DA and asymmetric hyperplasia as separate entities, patients continue to require careful intraoperative observation and close follow-up with serum calcium levels, as they may be more likely to require reoperation (12). An improved understanding of the potentially distinct entities of DA and asymmetric hyperplasia may better inform the risk of curative failure and need for reoperation. However, data on long-term follow up of patients with DA is limited.
Therefore, the goals of the current study are to characterize long-term outcomes for patients with DA and to describe risk factors for disease persistence/recurrence.
Materials and Methods
Study Population:
We performed a retrospective cohort study of patients undergoing initial parathyroidectomy for PHPT at our institution from 1997–2018. A total of 3,779 subjects were screened for inclusion. Patients with re-operative parathyroidectomy (n=258), inconclusive final diagnosis (n=76), parathyroid cancer (n=2) or missing diagnosis (n=35) on surgical pathology were excluded. The final cohort consisted of 3,408 patients. This study was approved by the Institutional Review Board of the University of Pennsylvania (protocol #828743).
Variables:
Patient data was collected prospectively from the electronic medical record and stored in quality assurance databases. Demographic and clinical variables included age, gender, self-reported race, preoperative biochemical and imaging data, intraoperative biochemical data, and surgical pathology. Final diagnosis of parathyroid disease was determined by a combination of intraoperative PTH monitoring and surgical pathology as either single adenoma (SA), double adenoma (DA) or four-gland hyperplasia (HP).
Definitions & Outcomes:
The primary study outcome was persistence or recurrence of primary hyperparathyroidism following initial parathyroidectomy in patients with long-term follow-up (≥ 6 months). Persistence was defined as failure of serum calcium levels to normalize within 6 months following initial parathyroidectomy. Recurrence was defined as elevated serum calcium levels after 6 months in patients with previous normocalcemia following parathyroidectomy. Secondary outcomes included intraoperative success and percent decrease in intraoperative parathyroid hormone (PTH) level. Intraoperative success was defined as a decrease of ≥50% intraoperative parathyroid hormone levels (IOPTH), similar to the established Miami criteria but with a draw time at 15 minutes post excision (13, 14). Final diagnosis incorporated both histopathologic and IOPTH criteria. Single adenoma was defined as one histologically abnormal gland with an accompanying appropriate decrease in IOPTH after excision. Double adenoma was defined as two histologically abnormal glands with an accompanying appropriate IOPTH decrease after excision or two histologically abnormal glands with two histologically normal biopsies of the remaining two glands and appropriate decrease in IOPTH. Four-gland hyperplasia was defined as four histologically abnormal glands. Percent decrease in IOPTH was calculated from recorded intraoperative values. Our institutional practice is to strive to achieve greater than 50% decrease in IOPTH as well as IOPTH into the low normal range (<40 pg/ml). Surgical approach was documented as bilateral neck exploration or minimally invasive parathyroidectomy (unilateral neck exploration). Gland size was defined by measurement of the dominant abnormal gland and was recorded as both mass (milligrams) and size by greatest dimension (centimeters). Concordant imaging was defined as imaged abnormal gland(s) identified on the same side.
Intraoperative Parathyroid Hormone Monitoring (IOPTH):
Intraoperative serum parathyroid hormone (PTH) levels were obtained by peripheral venipuncture prior to incision and 15 minutes after excision of abnormal parathyroid gland(s). Testing was performed on an Immulite® 2000 Immunoassay System (Siemens Healthcare, Erlangen, Germany) with chemiluminescent immunoassay detection of intact PTH.
Statistics:
Descriptive statistics were performed. Comparisons between two groups utilized the chi-squared test or the Fisher’s exact test for comparison of categorical variables and the Wilcoxon rank-sum test for comparison of continuous variables. Comparisons of three or more groups utilized analysis of variance (ANOVA) or Kruskal-Wallis for parametric and non-parametric variables, respectively. Continuous variables were reported as mean with standard deviation (SD) or median with interquartile interval (IQI), as appropriate. Subjects were stratified by final IOPTH levels as follows: normal (<65 pg/ml), high-normal (40–65 pg/ml), and low-normal (<40 pg/ml). Univariate logistic regression was performed to determine the association between persistent/recurrent disease and clinical covariates. Variables meeting nominal significance (p≤0.2) on univariate analysis were incorporated into a backward stepwise multivariable logistic regression model to identify covariates independently associated with persistence/recurrence. The backwards stepwise regression incorporated a single intraoperative PTH variable in the final model to eliminate confounding variables. Statistical analysis was performed using STATA 15.1 (Stata Corporation, College Station, TX). P-values are two-tailed unless otherwise indicated. A p-value ≤0.05 was considered significant.
Results
Study Cohort
The median age of patients undergoing initial parathyroidectomy for primary hyperparathyroidism was 60 years (IQI 52, 68). The majority of patients in the cohort were female gender (77.2%) and self-identified their race as White (79.4%). Incidence of family history of hyperparathyroidism or personal history of Multiple Endocrine Neoplasia was low (1.31% and 0.29%, respectively). On biochemical evaluation, median preoperative serum calcium was 10.8 mg/dl (IQI 10.2, 11.2), PTH was 88 pg/ml (IQI 56, 124.1) and creatinine was 0.82 mg/dl (IQI 0.7, 1).
Patients by Diagnosis
The majority of patients (81.3%) undergoing initial parathyroidectomy for primary hyperparathyroidism were found to have SA on final pathology, while 9.5% had DA and the remaining 9.3% had HP (Figure 1). Patient characteristics and preoperative variables were analyzed by final diagnosis (Table 1). Age varied across groups, with a median age of 59, 62, and 60.5 years for patients with SA, DA, and HP, respectively (p=0.003). The majority of patients in all groups were female (p=0.153) and a majority self-identified their race as White (p=0.437). Patients with multigland disease (either DA or HP) were more likely to have a family history of PHPT; family history was most common in patients with HP (SA=0.8%, DA=2.7%, HP=4.7%; p<0.001). Similarly, a known diagnosis of Multiple Endocrine Neoplasia was most common in patients with HP (SA=0.1%, DA=0%, HP=2.2%; p<0.001).
Figure 1.
Distribution of final diagnosis on surgical pathology in 3,408 patients presenting for initial parathyroidectomy for PHPT
Table 1.
Characteristics of patients presenting for initial parathyroidectomy for PHPT with final diagnosis of single adenoma (SA), double adenoma (DA), or hyperplasia (HP)
| SA (n=2769) | DA (n=323) | HP (n=316) | p-value | |
|---|---|---|---|---|
| Age, years (median, IQR) | 59 (52, 68) | 62 (54, 70) | 60.5 (52, 69) | 0.003 |
| Gender, female (n, %) | 2120 (76.6) | 259 (80.2) | 253 (80.1) | 0.153 |
| Race, white (n, %) | 1760 (79.1) | 194 (81.5) | 181 (80.1) | 0.437 |
| Family history (n, %) | 19 (0.8) | 8 (2.7) | 13 (4.7) | 0.000 |
| Multiple Endocrine Neoplasia (n, %) | 3 (0.1) | 0 (0) | 7 (2.2) | 0.000 |
| Preoperative Biochemical Evaluation | ||||
| Calcium, mg/dl (median, IQR) | 10.8 (10.5, 11.2) | 10.7 (10.4, 11.2) | 10.6 (10.3, 11) | <0.001 |
| PTH, pg/ml (median, IQR) | 89.6 (59, 125) | 80.92 (45, 128) | 79 (39, 115) | <0.001 |
| Creatinine, mg/dl (median, IQR) | 0.81 (0.7, 1) | 0.84 (0.72, 1.0) | 0.87 (0.73, 1.07) | 0.010 |
| Preoperative Imaging | ||||
| Sestamibi alone (n, %) | 479 (17.3) | 24 (7.4) | 34 (10.8) | 0.000 |
| Ultrasound alone (n, %) | 90 (3.3) | 10 (3.1) | 10 (3.2) | 0.987 |
| Sestamibi & ultrasound (n, %) | 2035 (73.5) | 276 (85.5) | 258 (81.7) | 0.000 |
| Concordant Sestamibi & ultrasound (n, %) | 1135 (55.9) | 141 (51.1) | 142 (55.0) | 0.317 |
IQR – Interquartile range; PTH – Parathyroid hormone
Statistically significant results in bold
On preoperative labs, there were small but statistically significant differences in median preoperative calcium (SA=10.8, DA=10.7, HP=10.6 mg/dl; p<0.001). Median preoperative PTH was highest in patients with SA (SA=89.6, DA=80.9, HP=79 pg/ml; p<0.001). Median preoperative creatinine varied across groups and was highest in patients with HP (SA=0.81, DA=0.84, HP=0.87 mg/dl; p=0.010). For preoperative imaging, patients with SA were most likely to have had only sestamibi imaging performed (SA=17.3%, DA=7.4%, HP=10.8%; p<0.001); patients with multigland disease (MGD) were more likely to have both sestamibi and ultrasound performed (SA=73.5%, DA=85.5%, HP=81.7%; p=<0.001). Among patients with dual imaging, concordance between sestamibi and ultrasound findings ranged from 51% to 56% (p=0.317).
Patients with DA had a significantly greater median length of follow-up (SA=6, DA=11, HP=6 months; p<0.001) and were more likely to have long-term follow up (SA=58.9%, DA=76.8%, HP=58.8%; p<0.001). Among patients with ≥ 6 months of follow up, length of follow-up differed across groups (SA=15, DA=26.5, HP=12 months; p=0.005).
Operative Findings and Surgical Outcomes
Surgical characteristics were compared across final diagnosis groups (Table 2). There was a small but statistically significant difference in median gland dimensions in DA versus SA or HP (SA=1.5, DA=1.6, HP=1.5 cm; p<0.001). Dominant gland mass was inversely correlated with the number of glands involved; median gland mass was greatest in patients with SA, followed by DA, and then HP (SA=445, DA=401, HP=253 mg; p<0.001).
Table 2.
Surgical findings in patients with final diagnosis of single adenoma (SA), double adenoma (DA), or hyperplasia (HP)
| SA (n=2769) | DA (n=323) | HP (n=316) | p-value | |
|---|---|---|---|---|
| Dominant Gland Characteristics | ||||
| Size, cm (median, IQR) | 1.5 (1.2, 2) | 1.6 (1.2, 1.8) | 1.5 (1.1, 1.8) | <0.001 |
| Mass, mg (median, IQR) | 445 (235, 899) | 401 (216, 785) | 253 (143, 530) | <0.001 |
| Intraoperative Findings | ||||
| Cure by modified Miami Criteria (n, %) | 2569 (98.7) | 301 (95.25) | 285 (92.5) | 0.000 |
| IOPTH decrease, % (median, IQR) | 85.3 (77.3, 91.0) | 84.8 (75.4, 90.4) | 82.7 (68.5, 89.7) | <0.001 |
| IOPTH final, pg/ml (median, IQR) | 21.6 (14.1, 32.0) | 21.2 (14.1, 33.8) | 24.4 (15.0, 39.5) | 0.008 |
| IOPTH normalized (n, %) | 2538 (97.4) | 307 (96.5) | 284 (91.9) | 0.000 |
| IOPTH to high normal (n, %) | 307 (12.1) | 38 (12.4) | 51 (18.0) | 0.019 |
| IOPTH to low normal (n, %) | 2231 (85.6) | 269 (84.6) | 233 (75.4) | 0.000 |
IQR – Interquartile range; IOPTH – Intra-operative Parathyroid hormone
Statistically significant results in bold
Anatomical distribution of pathological glands in patients with DA was characterized. The six possible distributions for DA were bilateral superior, bilateral inferior, unilateral right, unilateral left, asymmetric, and ectopic/supernumerary. The most common distribution was bilateral superior (39%) followed by asymmetric (26%), with the other distributions accounting for the remaining minority of cases (unilateral left=14%, unilateral right=10%, bilateral inferior=9%, and ectopic supernumerary=2%).
Rates of intraoperative success by modified Miami criteria (SA=98.7%, DA=95.3%, HP=92.5%; p<0.001) and median decrease in IOPTH (SA=85.3%, DA 84.8%, HP=82.7%; p=<0.001) varied across final diagnosis groups (Table 2). Median final IOPTH differed across groups (SA=21.6, DA=21.2, HP=24.2 pg/ml; p=0.008), although all groups demonstrated a median final IOPTH in the normal range. However, while IOPTH normalized in most patients, there was significant variation in the percent of patients who achieved normal IOPTH by diagnosis (SA=97.4%, DA=96.5%, HP=91.9%; p<0.001). Patients with MGD were less likely to have final IOPTH that reached the low-normal range; this was especially true for HP (SA=85.6%, DA=84.6%, HP=75.4%; p<0.001). Patients with DA had significantly longer median follow-up (p=0.005) and rates of long term follow up (p<0.001)
Persistent and Recurrent Hyperparathyroidism
The primary outcome of persistent/recurrent disease after initial parathyroidectomy was compared across diagnosis groups (Figure 2). There was no statistically significant difference in primary outcome by diagnosis in patients with ≥ 6 months of follow-up (SA=2.9%, DA=5.3%, HP=4.5%; p=0.281).
Figure 2.
Rates of persistent/recurrent PHPT after initial parathyroidectomy, by final diagnosis. There was no statistically significant difference in persistent/recurrent disease between single adenoma, double adenoma, and hyperplasia on long-term follow up (p=0.281)
We compared clinical characteristics between patients with persistence/recurrence and patients with long-term cure (Table 3). There were no differences in age, gender, race or family history of PHPT between patients with persistent/recurrent disease and patients cured with initial parathyroidectomy, although patients with persistent/recurrent disease were more likely to have a diagnosis of multiple endocrine neoplasia (p=0.042). Median preoperative calcium (11.0 vs 10.7 mg/dl, p =0.028), PTH (96.0 vs 76.5 pg/ml, p=0.015), and creatinine (1.03 vs 0.82 mg/dl, p =0.002) were higher in patients with persistence/recurrence versus patients with no evidence of disease. There was no difference in frequency of operative approaches between groups (p=0.698). There were no differences in median gland size on surgical pathology between groups (size: p=0.678; mass: p=0.615).
Table 3.
Clinical characteristics of patients with long-term follow-up, by cure status
| Cured (n=1011) | Not Cured n=36) | |
|---|---|---|
| Age, years (median, IQR) | 61 (53, 69) | 61.5 (50, 71) |
| Gender, female (n, %) | 788 (77.9) | 28 (77.8) |
| Race, white (n, %) | 507 (77.3) | 22 (81.5) |
| Family history (n, %) | 9 (0.98) | 0 (0.0) |
| Multiple Endocrine Neoplasia (n, %) | 4 (0.4) | 1 (2.78) |
| Preoperative Biochemical Evaluation | ||
| Calcium, mg/dl (median, IQR) | 10.7 (10.4, 11.1) | 11 (10.4, 11.6) |
| PTH, pg/ml (median, IQR) | 76.5 (14.6, 117) | 96 (66, 177.7) |
| Creatinine, mg/dl (median, IQR) | 0.82 (0.7, 1.01) | 1.03 (0.82, 1.6) |
| Preoperative Imaging (n, %) | ||
| Sestamibi alone | 160 (15.8) | 7 (19.4) |
| Ultrasound alone | 48 (4.8) | 2 (5.6) |
| Sestamibi & ultrasound | 744 (73.6) | 21 (58.3) |
| Concordant Sestamibi & ultrasound | 372 (47.6) | 13 (59.1) |
IQR – Interquartile range; PTH – Parathyroid hormone
Regarding secondary outcomes, patients with persistent/recurrent disease had significantly lower rates of intraoperative success by modified Miami criteria (90.3 vs 97.3%, p=0.025) and trended towards a smaller percent drop in IOPTH (78.9 vs 84.3%, p=0.051). Patients with disease persistence/recurrence had an almost 90% higher median final IOPTH (42.3 vs 22.6 pg/ml, p<0.001) and were less likely to have a final IOPTH fall within the normal range (77.4% vs 96.3%, p<0.001). Similarly, patients with persistent/recurrent disease were less than half as likely to have a final IOPTH in the low-normal (<40 pg/ml) range (36.1 vs. 78.3%, p<0.001).
There was no statistically significant difference in the incidence of MGD, DA, or HP in patients with persistent/recurrent disease compared to patients with no evidence of disease after initial parathyroidectomy (MGD: 36.1 vs 24.6%, p=0.118; DA: 22.2 vs 14.2%, p=0.182; HP: 13.9 vs 10.4, p=0.501).
Predictors of Persistent/Recurrent Disease
Univariate logistic regression analysis was performed to investigate the association of clinical variables with disease persistence/recurrence (Table 4). Among all patients, higher preoperative calcium (OR 1.809, p=0.002), higher preoperative PTH (OR 1.001, p=0.005), higher preoperative creatinine (OR 1.540, p=0.004), a higher final IOPTH (OR 1.014, p<0.001) and a high-normal final intraoperative PTH (OR 5.277, p<0.001) were associated with increased risk of persistence/recurrence. For all patients, intraoperative success by IOPTH (OR 0.262, p=0.036), a greater intraoperative drop in PTH (OR 0.983, p=0.043), a normal final intraoperative PTH (OR 0.131, p<0.001), and a low-normal final intraoperative PTH (OR 0.148, p<0.001) were associated with lower risk of persistent/recurrent disease. Covariates meeting nominal significant (p≤0.20) were incorporated into a multivariable logistic regression model to determine predictors of persistent/recurrent disease. On final multivariable regression, higher preoperative calcium (OR 1.617, p=0.034), higher preoperative serum PTH (OR 1.001, p=0.040), and a final diagnosis of DA (OR 3.007, p=0.017) were independently associated with disease persistence/recurrence. A low-normal final IOPTH (OR 0.174, p <0.001) was independently associated with lower odds of persistent/recurrent disease.
Table 4.
Predictors of disease persistence/recurrence among patients with long-term follow-up, on univariable and multivariable models
| Univariable | Multivariable | |||||
|---|---|---|---|---|---|---|
| OR | 95% CI | p-value | OR | 95% CI | p-value | |
| Preoperative Biochemical Evaluation | ||||||
| Preoperative calcium, mg/dl | 1.809 | (1.240, 2.640) | 0.002 | 1.617 | (1.036, 2.523) | 0.034 |
| Preoperative PTH, pg/ml | 1.001 | (1.000, 1.002) | 0.005 | 1.001 | (1.000, 1.002) | 0.040 |
| Preoperative creatinine, mg/dl | 1.540 | (1.145, 2.071) | 0.004 | -- | -- | |
| Minimally Invasive Approach | 1.164 | (0.540, 2.512) | 0.698 | -- | -- | |
| Dominant Gland Characteristics | ||||||
| Size, cm | 1.080 | (0.742, 1.572) | 0.689 | -- | -- | |
| Gland mass, mg | 1.000 | (1.000, 1.000) | 0.228 | -- | -- | |
| Intraoperative Findings | ||||||
| Cure by modified Miami Criteria | 0.262 | (0.075, 0.918) | 0.036 | -- | -- | |
| IOPTH decrease, % | 0.983 | (0.967, 0.999) | 0.043 | -- | -- | |
| IOPTH final, pg/ml | 1.014 | (1.007, 1.022) | <0.001 | -- | -- | |
| IOPTH normal | 0.131 | (0.053, 0.323) | <0.001 | -- | -- | |
| IOPTH high-normal | 5.277 | (2.314, 12.035) | <0.001 | -- | -- | |
| IOPTH low-normal | 0.148 | (0.071, 0.308) | <0.001 | 0.174 | (0.076, 0.394) | 0.000 |
| Final Diagnosis | ||||||
| Double adenoma | 1.841 | (0.807, 4.196) | 0.147 | 3.007 | (1.219, 7.418) | 0.017 |
| Hyperplasia | 1.578 | (0.587, 4.239) | 0.366 | 0.741 | (0.160, 3.426) | 0.701 |
CI – Confidence interval; PTH – Parathyroid hormone; IOPTH – Intra-operative parathyroid hormone
Statistically significant results in bold
Univariate regression analysis was separately performed on the subgroup of patients with final diagnosis of DA. On single variable logistic modeling of patients with DA, only preoperative creatinine was significantly associated with increased risk of persistence/recurrence (OR 6.654, p=0.014).
Additional subgroup analysis was performed to investigate the impact of normal versus elevated preoperative calcium on our primary outcome. First, we found a significant difference in rates of normocalcemia across diagnosis groups (SA=6.8%, DA=10.2%, HP=18.4%; p<0.001). However, on univariable modeling, normal preoperative calcium was not associated with increased risk of persistence/recurrence (OR 0.984, p=0.975). Finally, to account for possible confounding secondary to lower cure rates in patients with normal preoperative calcium, we performed subgroup multivariable regression analysis limited to patients with elevated preoperative calcium (≥10.1 mg/dl). Multivariable modeling with this subgroup demonstrated that a diagnosis of double adenoma remained associated with an increased risk of persistence/recurrence (OR 2.963; p=0.026), while low-normal IOPTH remained associated with decreased risk (OR 0.176; p<0.001).
Discussion
In this large series, we describe long-term outcomes for patients with DA as well as risk factors for cure failure after parathyroidectomy. We observed a DA incidence of 9.5% in patients undergoing initial parathyroidectomy for PHPT, which is within the range of reported rates (8–10, 15–17). We found no significant difference in the incidence of persistent/recurrent disease in patients with DA versus SA and HP. Published reports conflict regarding the incidence of persistent/recurrent PHPT in patients with DA (12, 17). Mazotas et al reported an increased risk of persistent disease but not recurrent disease (10). In contrast, Alhefdi et al found higher rates of both persistent and recurrent PHPT, while Abboud and colleagues discovered no difference in persistent or recurrent disease (12, 17). Regarding distribution of abnormal glands in DA, we observed predominantly bilateral superior disease followed by asymmetric disease. Published data are mixed regarding the distribution of abnormal glands in DA. While several groups have found that abnormal glands occurred in a disproportionately bilateral superior distribution, others reported disproportionately crossed bilateral/asymmetric adenomas (8, 16, 17). Importantly, we found no association between the anatomic distribution of DA and risk of persistence/recurrence.
We examined clinical risk factors for persistent/recurrent hyperparathyroidism after parathyroidectomy. In our cohort, patients with persistent/recurrent disease had significantly higher preoperative serum calcium, PTH, and creatinine. Not surprisingly, failure to achieve intraoperative success by IOPTH was associated with persistent/recurrent PHPT. Consistent with prior studies, patients with persistent/recurrent disease had a median final IOPTH nearly two times higher than patients who achieved cure (18). In our final multivariable model, higher preoperative calcium and PTH as well as final diagnosis of double adenoma were independently associated with persistent/recurrent disease, while a final IOPTH within the low-normal range was independently associated with decreased risk of persistence/recurrence.
Our findings support the importance of careful IOPTH monitoring in the intraoperative diagnosis and surgical cure of DA. Existing research suggests that a drop of ≥50% is insufficient to demonstrate intraoperative cure in cases of MGD (19, 20). In their study of IOPTH monitoring which included 271 patients with MGD, Richards and colleagues demonstrated that only using the criteria of 50% decrease in IOPTH would lead to failed operation in more than 20% of patients with MGD (21). As a result, some authors advocate the use of more stringent criteria, requiring IOPTH to drop ≥50% and into the normal range, to avoid missing MGD (22, 23). Wharry et al showed that failure to normalize IOPTH increases the risk of cure failure and recommended that patients with final IOPTH in the high-normal range warrant long-term surveillance for recurrence (18). Our findings corroborate that a drop of ≥50% of IOPTH may be an insufficient criteria for cure in DA (19, 21, 23). Furthermore, our data support that a final IOPTH in the low-end of the normal range was associated with significantly lower risk of persistent/recurrent disease.
In this series, approximately 95% of patients with DA achieved long-term eucalcemia after removal of two abnormal glands, supporting DA as a clinical entity distinct from four-gland hyperplasia. This is consistent with research demonstrating that resection of two affected glands can normalize parathyroid hormone and calcium levels (8, 17). However, approximately 5% of DA patients in our series demonstrate persistent/recurrent disease, suggesting that a subset of patients may have had asymmetric or asynchronous hyperplasia misdiagnosed on initial surgery. While we hypothesize that a subset of cases diagnosed as DA represent asymmetric hyperplasia leading to cure failure, current evidence is insufficient to fully characterize these disparate diagnoses using clinical or pathologic information.
Discriminating between single and multiple gland disease has posed a challenge on the molecular level. Early work demonstrated a prevalence of monoclonal derangements in SA but an absence of monoclonality in HP (24). More recent work by Olson and colleagues supports a polyclonal etiology of PHPT, especially in cases of MGD (25). These findings suggest a difference in molecular pathology in single parathyroid adenomas (monoclonal) versus four gland hyperplasia (polyclonal). However, evidence in DA is limited. Understanding the mechanisms of disease in DA may allow a more detailed characterization of abnormal glands and a more accurate understanding of persistence and recurrence after parathyroidectomy. Specifically, this may help explain the subset of patients diagnosed with DA on surgical pathology who go on to experience persistent/recurrent disease, consistent with asymmetric hyperplasia. However, current criteria for the diagnosis of DA is insufficient to separate true DA from asymmetric hyperplasia.
Our study has several limitations. First, as a retrospective study, it carries the risk of bias and confounding. However, data was collected prospectively in order to mitigate these inherent risks. Second, our sample is limited to patients presenting for initial parathyroidectomy at a single tertiary care center and therefore results may not be generalizable to all patients with PHPT. Third, our primary outcome included persistence or recurrence of PHPT as determined by serum calcium levels, but there was significantly longer follow up in patients with DA versus SA or HP. In this retrospective study, we were unable to identify a clear explanation for the variation in follow-up time across diagnosis groups. This potentially underestimates the recurrence of hypercalcemia occurring after patients with SA or HP were lost to follow-up and potentially biases our findings towards higher rates of recurrence in double adenoma. However, we found no significant increase in persistent/recurrent disease in DA. Finally, it is probable that our study is underpowered to detect differences in long term cure failure, because of the low rates of persistent/recurrent disease. Nevertheless, DA was independently associated with increased odds of persistent/recurrent disease in our final multivariable model.
Conclusions
This large retrospective study confirms the existence of DA as a distinct pathological entity and provides guidance for assessing the risk of persistence/recurrence following initial parathyroidectomy for PHPT. Our findings support the careful use of IOPTH to decrease the risk of persistent/recurrent disease in patients with a diagnosis of DA. Future work may provide a more detailed characterization of pathology in DA to better characterize the molecular pathogenesis.
Highlights.
Double adenoma occurs in almost 10% of patients undergoing initial surgery for PHPT
Rates of recurrent/persistent disease are similar in DA and four-gland hyperplasia
In a multivariate model, DA is associated with higher risk of failure to cure
Low-normal IOPTH is associated with decreased risk of persistence/recurrence
Careful use of IOPTH may minimize the risk of failure to cure in DA
Acknowledgements
The authors gratefully acknowledge the contributions of Lindsay Kuo, MD and Jonathan Sataloff, MD to data acquisition.
Disclosures/Funding: The authors have no disclosures. HW received support from the National Center for Advancing Translational Sciences of the National Institutes of Health, Award Number KL2-TR001879.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Mohebati A, Shaha AR. Imaging techniques in parathyroid surgery for primary hyperparathyroidism. Am J Otolaryngol. July;33(4):457–68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Zanocco K, Angelos P, Sturgeon C. Cost-effectiveness analysis of parathyroidectomy for asymptomatic primary hyperparathyroidism. Surgery. 2006December;140(6):874–81; discussion 81–2. [DOI] [PubMed] [Google Scholar]
- 3.Lew JI, Solorzano CC. Surgical management of primary hyperparathyroidism: state of the art. Surg Clin North Am. 2009October;89(5):1205–25. [DOI] [PubMed] [Google Scholar]
- 4.Wilhelm SM, Wang TS, Ruan DT, et al. The American Association of Endocrine Surgeons Guidelines for Definitive Management of Primary Hyperparathyroidism. JAMA Surg. 201610;151(10):959–68. [DOI] [PubMed] [Google Scholar]
- 5.Yeh MW, Wiseman JE, Chu SD, et al. Population-level predictors of persistent hyperparathyroidism. Surgery. December;150(6):1113–9. [DOI] [PubMed] [Google Scholar]
- 6.Kuo LE, Wachtel H, Fraker D, Kelz R. Reoperative parathyroidectomy: who is at risk and what is the risk? J Surg Res. 2014October;191(2):256–61. [DOI] [PubMed] [Google Scholar]
- 7.Tezelman S, Shen W, Shaver JK, et al. Double parathyroid adenomas. Clinical and biochemical characteristics before and after parathyroidectomy. Ann Surg. 1993September;218(3):300–7; discussion 7–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Bergson EJ, Heller KS. The clinical significance and anatomic distribution of parathyroid double adenomas. J Am Coll Surg. 2004February;198(2):185–9. [DOI] [PubMed] [Google Scholar]
- 9.Wachtel H, Cerullo I, Bartlett EK, et al. What Can We Learn from Intraoperative Parathyroid Hormone Levels that Do Not Drop Appropriately? Ann Surg Oncol. 2015;22(6):1781–8. [DOI] [PubMed] [Google Scholar]
- 10.Mazotas IG, Yen TWF, Doffek K, et al. Persistent/Recurrent Primary Hyperparathyroidism: Does the Number of Abnormal Glands Play a Role? J Surg Res. 2020February;246:335–41. [DOI] [PubMed] [Google Scholar]
- 11.Gough I Reoperative parathyroid surgery: the importance of ectopic location and multigland disease. ANZ J Surg. 2006December;76(12):1048–50. [DOI] [PubMed] [Google Scholar]
- 12.Alhefdhi A, Schneider DF, Sippel R, Chen H. Recurrent and persistence primary hyperparathyroidism occurs more frequently in patients with double adenomas. J Surg Res. 2014July;190(1):198–202. [DOI] [PubMed] [Google Scholar]
- 13.Irvin GL 3rd, Dembrow VD, Prudhomme DL. Clinical usefulness of an intraoperative “quick parathyroid hormone” assay. Surgery. 1993December;114(6):1019–22; discussion 22–3. [PubMed] [Google Scholar]
- 14.Carneiro DM, Solorzano CC, Nader MC, et al. Comparison of intraoperative iPTH assay (QPTH) criteria in guiding parathyroidectomy: which criterion is the most accurate? Surgery. 2003December;134(6):973–9; discussion 9–81. [DOI] [PubMed] [Google Scholar]
- 15.Ruda JM, Hollenbeak CS, Stack BC Jr. A systematic review of the diagnosis and treatment of primary hyperparathyroidism from 1995 to 2003. Otolaryngol Head Neck Surg. 2005March;132(3):359–72. [DOI] [PubMed] [Google Scholar]
- 16.Milas M, Wagner K, Easley KA, et al. Double adenomas revisited: nonuniform distribution favors enlarged superior parathyroids (fourth pouch disease). Surgery. 2003December;134(6):995–1003; discussion - 4. [DOI] [PubMed] [Google Scholar]
- 17.Abboud B, Sleilaty G, Helou E, et al. Existence and anatomic distribution of double parathyroid adenoma. Laryngoscope. 2005June;115(6):1128–31. [DOI] [PubMed] [Google Scholar]
- 18.Wharry LI, Yip L, Armstrong MJ, et al. The final intraoperative parathyroid hormone level: how low should it go? World J Surg. 2014March;38(3):558–63. [DOI] [PubMed] [Google Scholar]
- 19.Karakousis GC, Han D, Kelz RR, et al. Interpretation of intra-operative PTH changes in patients with multi-glandular primary hyperparathyroidism (pHPT). Surgery. 2007December;142(6):845–50; discussion 50.e1–2. [DOI] [PubMed] [Google Scholar]
- 20.Gauger PG, Agarwal G, England BG, et al. Intraoperative parathyroid hormone monitoring fails to detect double parathyroid adenomas: a 2-institution experience. Surgery. 2001December;130(6):1005–10. [DOI] [PubMed] [Google Scholar]
- 21.Richards ML, Thompson GB, Farley DR, Grant CS. An optimal algorithm for intraoperative parathyroid hormone monitoring. Arch Surg. 2011March;146(3):280–5. [DOI] [PubMed] [Google Scholar]
- 22.Karakousis GC, Han D, Kelz RR, et al. Interpretation of intra-operative PTH changes in patients with multi-glandular primary hyperparathyroidism (pHPT). Surgery. 2007December;142(6):845–50; discussion 50 e1–2. [DOI] [PubMed] [Google Scholar]
- 23.Hughes DT, Miller BS, Doherty GM, Gauger PG. Intraoperative parathyroid hormone monitoring in patients with recognized multiglandular primary hyperparathyroidism. World J Surg. 2011February;35(2):336–41. [DOI] [PubMed] [Google Scholar]
- 24.Arnold A, Staunton CE, Kim HG, et al. Monoclonality and abnormal parathyroid hormone genes in parathyroid adenomas. N Engl J Med. 1988March17;318(11):658–62. [DOI] [PubMed] [Google Scholar]
- 25.Shi Y, Azimzadeh P, Jamingal S, et al. Polyclonal origin of parathyroid tumors is common and is associated with multiple gland disease in primary hyperparathyroidism. Surgery. 2018January;163(1):9–14. [DOI] [PMC free article] [PubMed] [Google Scholar]