Use of Intraoperative Parathyroid Hormone in Minimally Invasive Parathyroidectomy for Primary Hyperparathyroidism: A Systematic Review and Meta-analysis

Key Points

Question  Is use of intraoperative parathyroid hormone associated with higher cure rates in minimally invasive parathyroidectomy for primary hyperparathyroidism?

Findings  In this systematic review and meta-analysis of 12 studies involving 2290 patients with primary hyperparathyroidism, use of intraoperative parathyroid hormone in minimally invasive parathyroidectomy was associated with increased rates of cure and reduced need for reoperation.

Meaning  These study findings suggest that there is a reduction in surgical failure and need for reoperation in patients undergoing minimally invasive parathyroidectomy for primary hyperparathyroidism when intraoperative parathyroid hormone is used.

Importance  Intraoperative parathyroid hormone (ioPTH) is a surgical adjunct that has been increasingly used during minimally invasive parathyroidectomy (MIP). Despite its growing popularity, to our knowledge a meta-analysis comparing MIP with ioPTH vs MIP without ioPTH has not yet been conducted.

Objective  To evaluate the safety and efficacy of MIP with ioPTH for treatment of primary hyperparathyroidism.

Data Sources  A systematic search of the databases PubMed, Embase, Scopus, Web of Science, and Cochrane Collaboration was performed to identify studies that compared MIP with and without ioPTH. Data were analyzed between August and September 2019.

Study Selection  Inclusion criteria consisted of randomized clinical trials and observational studies with a retrospective/prospective design, comparing MIP using ioPTH vs MIP not using ioPTH for treatment of primary hyperparathyroidism. Eligible studies had to present odds ratio (OR), risk ratio, or hazard ratio estimates (with 95% CI), standard errors, or number of events necessary to calculate these for the outcome of interest rate. Studies involving patients with secondary or tertiary hyperparathyroidism or those with multiple endocrine neoplasia syndrome were excluded.

Data Extraction  Two reviewers independently reviewed the literature according to Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guidelines. Dichotomous variables were pooled as ORs while continuous variables were compared using weighted mean differences. Quality assessment was performed using the Newcastle-Ottawa Scale.

Main Outcomes and Measures  The primary outcome was rate of cure. Secondary outcomes included need for reoperation, need for bilateral neck exploration, morbidity, and length of surgery.

Results  A total of 12 studies, involving 2290 patients with primary hyperparathyroidism, were eligible for inclusion. The median (SD) age of participants was 60.1 (11.8) years and 77.3% of participants were women. The median Newcastle-Ottawa score was 7. Patients who underwent MIP with ioPTH had higher cure rates (OR, 3.88; 95% CI, 2.12-7.10; P < .001). There was a greater need for reoperation in the group of patients who had surgery without ioPTH (OR, 0.40; 95% CI, 0.19-0.86; P = .02). There was a trend toward longer operating times/increased duration of surgery in the ioPTH group; however, this did not reach statistical significance (weighted mean difference, 21.62 minutes; 95% CI, −0.93 to 44.17 minutes; P = .06). The use of ioPTH was associated with higher rates of bilateral neck exploration (OR, 3.55; 95% CI, 1.27-9.92; P = .02).

Conclusions and Relevance  Use of ioPTH is associated with higher cure rates for patients with primary hyperparathyroidism undergoing MIP. Minimally invasive parathyroidectomy performed without ioPTH is associated with less conversion to bilateral neck exploration at initial surgery but with lower cure rates and an increased risk for reoperation.

Trial Registration  PROSPERO identifier: CRD42020148588

Primary hyperparathyroidism (PHP) is a common condition with an estimated incidence of 28 out of 100 000 individuals in the general population.¹ The nature of the disease in 70% to 95% of cases is an aberrant adenoma oversecreting parathyroid hormone (PTH); however, the cause also includes multigland hyperplasia (up to 15%) and carcinoma (1%).² Surgical excision remains the only means of cure and is indicated for symptomatic patients and asymptomatic patients with a serum calcium more than 0.25 mmol/L above the normal upper limit.^3,4

The traditional surgical approach for patients with PHP has been bilateral neck exploration (BNE) with histopathologic confirmation of pathologic gland excision via frozen section.⁵ This has a reported success rate exceeding 95%.^6-10 With the advent of specialist imaging, including ultrasonography and technetium-99m sestamibi methoxyisobutylisonitrile (MIBI) scan, and intraoperative adjuncts, including intraoperative PTH (ioPTH), a targeted approach has been advocated in cases of single-gland disease.^8,9 The reported advantages of minimally invasive parathyroidectomy (MIP) include reduced length of hospital stay, reduced incidence of postoperative hypocalcemia, lower rates of recurrent laryngeal injury, and equivalent cure rates.^5,7

The use of ioPTH was first reported by Nussbaum et al¹¹ in 1988. It is based on the principle that the half-life of PTH is 3 minutes, and thus results can be obtained while surgery is ongoing.¹² An adequate reduction in PTH measured intraoperatively provides the surgeon with the confidence to terminate the procedure on successful excision of the pathologic gland.¹² The main proposed advantage of ioPTH is an improved cure rate in patients undergoing MIP.^13,14 Advocates of ioPTH maintain that regardless of the advances made in imaging modalities, ioPTH assists the surgeon in confirming that the pathologic gland has been appropriately excised and is “essential to avoid a surgical failure,”³ and that ioPTH should be used routinely by all surgeons when performing surgery for PHP.⁷ Other proposed benefits of ioPTH use include a reduced length of hospital stay and consequent economic benefits.⁵

Despite its increasing and widespread use, controversy regarding the utility of ioPTH remains.^13,15 Some authors maintain that the benefits of ioPTH are marginal in cases where a pathologic gland has been localized by concordant dual imaging preoperatively.^5,15-17 It is argued that in these cases, the use of ioPTH unnecessarily uses hospital resources and equipment for an intraoperative adjunct that is unlikely to save operative time or diagnostic yield.¹⁶ Moreover, some authors contend that the interpretation of ioPTH results remains subjective, as there is a lack of uniformity in relation to which algorithm is optimal to analyze assay results.¹⁷ There is also concern regarding potential false-negative and false-positive results in the analysis of ioPTH that may lead to unnecessary exploration or premature termination of surgery when the pathologic gland(s) has not yet been excised.^15,17,18 Given this uncertainty, the aim of the present study was to perform a systematic review and meta-analysis comparing the safety and efficacy of MIP with ioPTH vs without ioPTH for treatment of PHP.

The population examined was patients with PHP undergoing parathyroidectomy. Parathyroidectomy with ioPTH was compared with parathyroidectomy without ioPTH to determine cure rate, defined as postoperative normocalcemia (calcium <2.6 mmol/L).

The dichotomous outcomes of interest included conversion to BNE, need for reoperation, recurrent laryngeal nerve injury (RLNI), postoperative hematoma, and hypocalcemia. The continuous outcomes of interest included age, operative time, length of hospital stay, and postoperative PTH and calcium levels.

Following Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guidelines, a systematic search for relevant publications was performed using PubMed, Embase, Scopus, Web of Science, and the Cochrane Collaboration databases (Figure 1). Each of the relevant publication’s reference sections and Google Scholar were also screened for other applicable publications. The search of the databases was performed by combining the following search terms using the Boolean AND/OR operators: surgery, parathyroidectomy, primary hyperparathyroidism, intraoperative parathyroid hormone, intraoperative PTH. Studies were included if they fulfilled the inclusion criteria and reported on primary and/or secondary outcomes. The search was conducted between August 5, 2019, and September 22, 2019.

Study inclusion criteria were as follows:

1. All randomized clinical trials (RCTs) or observational studies with a retrospective or prospective design.

2. Evaluated parathyroidectomy with vs without ioPTH for PHP.

3. Presented odds ratio (OR), relative risk (RR), or hazard ratio (HR) estimates (with 95% CI), SEs, or number of events necessary to calculate these for the outcome of interest rate.

4. Included at least 5 patients treated with primary hyperparathyroidism (PHP) in each arm.

5. Reported on a previously unreported group.

6. The most recently published or largest study was used, when multiple publications from the same study or institution were available.

Study exclusion criteria were as follows:

1. Any publication that did not meet the above inclusion criteria.

2. Studies that included patients treated nonoperatively (ie, via medical management).

3. Patients with secondary or tertiary hyperparathyroidism, multiple endocrine neoplasia syndromes, or familial hyperparathyroidism.

4. Studies where the full text was not available in English.

Two reviewers (A.J.Q. and S.G.) independently reviewed the literature according to the above predefined strategy and criteria. Each reviewer extracted the following data variables: title and reference details (first author, journal, year, country), study population characteristics (number in study; number treated by each approach; gender and age; preoperative neurological, bone, urological, or other symptoms/sequelae of PHP), preoperative imaging (neck ultrasonography, computed tomography [CT], MIBI, single-photon emission CT [SPECT]), disease characteristics (disease location and histopathological type, multiple gland disease, resected parathyroid gland weight), type and approach of surgical intervention (focused parathyroidectomy [MIP], unilateral neck exploration, or BNE with or without image guidance, and whether combined thyroidectomy was performed), and outcome data (mean drop of plasma ioPTH level at 10 minutes, postoperative hypocalcemia, total serum calcium after surgery, plasma PTH level after surgery, long-term and short-term success rates, and persistent or recurrent disease), including reoperative outcomes in the case of persistent or recurrent disease (repeated biochemical evaluation, results of new localization tests). All data were recorded independently by both literature reviewers in separate databases and were compared at the end of the reviewing process to limit selection bias. The database was also reviewed by a third person (É.J.R.). Duplicates were removed and any disparities were clarified.

Descriptive statistics were used to report the characteristics of all eligible trials, describing the types of surgical modalities, total patient numbers, mean age, and median length of stay. Statistical analysis was performed using Review Manager, version 5.3 (Cochrane Collaboration, 2014). Analysis of dichotomous variables was performed using ORs, reported with 95% CIs, while weighted mean differences were calculated for continuous variables.

First, pairwise meta-analysis was used to compare the reintervention and complication rates between the approaches, with an a priori decision to use a random-effects model being made owing to the assumption that there is considerable variation between studies. The test for heterogeneity was investigated by visual inspection of the forest plots and I² statistic, which provides the percentage of variability attributed to heterogeneity rather than sampling error. The OR was considered to be statistically significant at the P < .05 level if the 95% CI did not include the value of 1. Sensitivity analyses were carried out where appropriate.

The quality of the studies included in this systematic review was assessed by the Newcastle-Ottawa Scale (NOS) for observational studies and Grading of Recommendations Assessment, Development, and Evaluation criteria for RCTs, to evaluate patient selection methods, comparability, and assessments of outcomes. The quality score rating was determined for each publication, with 8 or more stars representing observational studies of higher quality.

The initial literature search yielded a total of 144 studies for full-text review, of which 12 met the inclusion criteria.^{3,5,7,10,13,14,16-21} A total of 2290 patients with primary hyperparathyroidism (ioPTH, 1148; without ioPTH, 1132) were eligible for inclusion from these studies. A total of 132 articles were excluded as they did not fulfill the requisite criteria. On full text review, 87 studies were excluded as they did not provide a control group for comparison, 34 studies did not set out to perform a MIP or unilateral neck exploration, 8 were available in abstract form only, and 3 studies were not available in the English language.

A total of 10 out of 12 included studies were retrospective observational studies, and 2 studies had a prospective design (Table 1^{3,5,7,10,13,14,16-21}). There were no RCTs eligible for inclusion. The earliest published study was from 2002 and the latest published in 2018. The study period ranged from 1990 through 2017. The included studies were performed in the UK, Poland, Italy, Romania, Israel, Korea, Japan, Canada, and the United States (Table 1). A total of 10 of the 12 studies commented on long-term postoperative follow-up, and the mean length of follow up was 12.66 months.^{3,5,10,13,14,16,18-21}

The NOS scores for the included studies are summarized in Table 1. The NOS score ranged from 6 to 8, with 8 studies having what was considered a high rating (>7). There were no RCTs included as part of the analysis and thus no need to use GRADE criteria.

The median (SD) age of participants was 60.1 (11.8) and 77.3% of participants were women. Out of 10 studies that reported preoperative imaging, 87% of patients (n = 1353) underwent ultrasonography, 93% of patients (n = 1634) underwent MIBI scan, and only 1 study mentioned using SPECT as part of the preoperative investigation (n = 138). A total of 2 studies^13,18 mentioned only using MIBI scan preoperatively while the remaining studies made reference to dual preoperative imaging. A total of 6 studies commented on concordant imaging preoperatively between the group with ioPTH (n = 717) and the group without ioPTH (n = 674).^{3,5,7,10,16,17} Imaging was concordant if it identified an anomaly in the same location on 2 separate preoperative imaging modalities, which was identified in 92.7% of the patients in the group with ioPTH (n = 665) and 92% of patients in the group without ioPTH (n = 621) (Table 2). The mean (SD) preoperative calcium was 2.88 (0.17) with a mean (SD) preoperative PTH of 203.9 (204.4) (Table 2). A total of 11 studies reported postoperative histologic findings. Where histology was reported, a parathyroid adenoma was confirmed in 92.2% of cases (n = 1993) with a second adenoma identified in 2.2% of cases (n = 47). Four-gland hyperplasia was responsible in 6.9% of cases (n = 149) and only 1 study reported parathyroid carcinoma (n = 2).²⁰ Thielmann and Kerr¹⁹ did not present postoperative histologic findings but defined cure biochemically by normocalcemia postoperatively.

Cure, when defined by studies, was either defined as normocalcemia or defined as calcium less than 2.6 mmol/L. Some studies also stipulated that normocalcemia must persist for 6 months postoperatively.^3,5,16,20 A total of 11 of 12 studies compared rate of cure between the group with ioPTH and the group without ioPTH. The study by Khan et al¹⁴ focused on how ioPTH influenced surgical approach and need for BNE. It did not present cure rates between the 2 groups and thus this article was not included in analysis of the primary outcome but was used for secondary outcomes.

Patients who underwent MIP with ioPTH had a statistically higher likelihood of cure compared with those who underwent MIP without ioPTH (cure OR, 3.88; 95% CI, 2.12-7.10; P < .001) (Figure 2^{3,5,7,10,13,14,16-21}). Patients in the group treated with ioPTH had a cure rate of 98% (n = 1032), while those in the group not treated with ioPTH had a cure rate of 94.8% (n = 1020). There was no significant heterogeneity between the trials (I² = 0%, heterogeneity P = .61).

A total of 10 of the 12 studies reported on the need for reoperation owing to persistent or recurrent disease. In the studies where these were defined, disease persistence was defined as hypercalcemia (calcium >2.6 mmol/L) within 6 months of surgery and disease recurrence was defined as hypercalcemia developing more than 6 months after surgery (preceded by an initial period of normocalcemia). The rate of reoperation in the ioPTH group was 0.90% (n = 8), which was significantly lower than that of patients in the group without ioPTH, in whom the reoperation rate was 3.25% (n = 28) (OR, 0.40; 95% CI, 0.19-0.86; P = .02). A total of 10 studies reported on disease persistence; the rate of persistence was lower in the ioPTH group (0.98%; n = 9) compared with the surgery without ioPTH group (5.65%; n = 55) (OR, 0.24; 95% CI, 0.12-0.45; P < .001). Not all patients with persistent disease underwent further operative management during the study period. A total of 5 studies commented on recurrence rates after initial operation; there was an overall recurrence rate of 0.61% (n = 7) out of 1151 patients, with no statistically significant difference between the 2 groups (OR, 0.67; 95% CI, 0.18-2.49; P = .54).

A total of 11 of the studies commented on the rate of conversion to BNE, which was increased when ioPTH was used, with 12.8% of patients in the ioPTH group undergoing BNE (n = 130), compared with 10.9% of patients in the group without ioPTH (n = 113) (OR, 3.55; 95% CI, 1.27-9.92; P = .02). In all studies, the planned surgical approach was a minimally invasive parathyroidectomy or unilateral neck exploration. The decision to convert to BNE was left to the surgeon’s discretion or based on ioPTH results.

There was a trend toward increased length of operation in the ioPTH group; however, this trend did not reach statistical significance (weighted mean difference, 21.62 minutes; 95% CI, −0.93 to 44.17 minutes; P = .06). The ioPTH group had a mean (SD) length of operation of 75.25 (39.6) minutes while the group without ioPTH had a mean (SD) operating time of 54.01 (20.8) minutes (Figure 3^{3,5,7,10,13,16-18}).

A total of 6 studies reported on postoperative morbidity. Morbidity, as defined in the studies, included postoperative hypocalcemia, hematoma, wound infection, and RLNI. The overall morbidity rate out of 1157 patients was 2.8%. The overall percentage of RLNI was 0.5% and did not differ between groups (RR, 1.48; 95% CI, 0.31-7.02; P = .62). Similarly, there was no significant difference between the 2 groups in the overall rates of morbidity (RR, 0.88; 95% CI, 0.44-1.78; P = .73).

Sensitivity analysis on the primary outcome, the rate of cure after surgery, was performed to evaluate the stability of results. With regard to cure rate, when the 3 largest studies were analyzed, there was a statistically significant increase in the rate of cure for the ioPTH group (3.86; 95% CI, 1.67-8.95; P = .002). When studies with the highest NOS scores were analyzed, it was again found that there was a significantly higher rate of cure associated with the ioPTH group (2.67; 95% CI, 1.17-6.12; P = .02). However, when the most recently published studies were interpreted, it was found that although there was a trend toward higher rate of cure in the ioPTH group, it did not reach statistical significance (2.02; 95% CI, 0.79-5.12; P = .14).

Recent advances in imaging technology and surgical adjuncts have increased the use of the MIP approach for parathyroid surgery. This approach aims to achieve similar or higher success rates with minimal morbidity.⁹ The introduction of ioPTH is deemed by some authors to have been a critical part of this movement.^7,14 Despite its increasing use as a component of the management of a common disorder, controversy regarding the utility of ioPTH for PHP remains with regard to operative success, operative time, cost, and other considerations.^5,15-17,22 To our knowledge, this meta-analysis is the first to compare attempted MIP with and without the use of ioPTH for PHP. The main finding is that use of ioPTH is associated with a higher rate of cure and a lower need for reoperation. This suggests that MIP should not be undertaken without the availability of ioPTH as an adjunct.

Although PHP is a common disorder, it can have serious consequences. The associated hypercalcemia may result in renal complications, skeletal involvement, cognitive impairment, and potentially fatal cardiac events.²³ Prompt management is necessary to avoid sometimes-irreversible morbidity. For this reason, the primary end point when evaluating MIP for PHP should ultimately be cure of disease with return to normocalcemia. Patients who undergo unsuccessful surgery remain persistently hypercalcemic and are at increased risk of morbidity.²⁴ There may be increased patient anxiety regarding the need for reoperation and the potential loss of faith by the patient or the surgeon in the surgical process.²⁵ Consideration should also be given to its economic effect, including reoperation and other health care costs.²⁶

Unsurprisingly, the increased rate of cure in the ioPTH group was associated with a lower requirement for reoperation in the present study, with 0.9% necessitating a second procedure as compared with 3.4% of patients in the without ioPTH cohort. This is somewhat offset by the fact that ioPTH use was associated with a higher conversion to BNE vs those without (12.8% vs 10.9%). This is not in keeping with previous studies, many of which advocate for the use of ioPTH to complement the MIP approach.^3,13,14 However, the reason for BNE in the group with the use of ioPTH is the failure of the intraoperative level to drop sufficiently to satisfy the respective criterion, due to further secreting gland(s).¹⁷ Once this is appreciated, the reason for the increased rate of BNE compared with MIP without this adjunct available is intuitive and the higher cure rate indicates that the conversion to BNE was appropriate. Thus, while the use of ioPTH is associated with higher BNE, this appears to be justified, as the rate of cure is higher and the need for reoperation is lower. This also may have been associated with the trend toward increased length of operative time in the ioPTH group.¹⁹ However, it must be noted that this was not associated with an increased rate of complications at the index surgery. Moreover, the potential pitfalls of revision neck surgery, including a more challenging neck dissection with the potential for increased rates of complications, specifically RLNI, may be avoided.^27-30

A number of studies state that ioPTH is not required when imaging is concordant on 2 separate imaging modalities, namely focused ultrasonography scan of the neck and MIBI scan,^17,19,31 owing to the costs and time associated with its use and the likelihood that it will not add anything to an already straightforward procedure.^17,19,31 However, ultrasonography is operator dependent and, although MIBI scan has a sensitivity of 68% to 95% and specificity of 90%, it has a high false negative rate, with rates as high as 20% reported.^32,33 This is primarily due to the small size of parathyroid glands, interference from thyroid nodules, and the oxyphil cell content of the glands.^32,33 These difficulties may be compounded by concomitant thyroid disease (multinodular goiter or autoimmune disease).³⁴ Moreover, half of the included studies reporting radiological data had concordance rates of about 90% or more. This suggests there is a role for ioPTH in MIP for PHP even in the setting of concordant preoperative imaging.^3,20,35

One of the main issues that has been raised regarding the use of ioPTH surrounds interpretation of the biochemical results.^21,36 Miami and Vienna criteria have the highest rate of false positives (0.4%), wherein patients are deemed cured but are in fact persistently hypercalcemic postoperatively, due to a second adenoma or undetected multigland disease. Halle and Rome criteria have a significantly higher false negative rate (35% vs 16.2%), defined as incorrect prediction of incomplete excision.^17,36 Unfortunately no comparison of the relative performance of the available techniques was possible due to scant reporting of the requisite data, and at present there is no recommendation regarding which criterion to use.¹⁷ However, Miami criteria are suggested to be the most balanced with regard to prediction of cure, while Rome criteria are favored if multigland disease is suspected.³⁶ Regardless, it is imperative that the operating surgeon understands that the accuracy of ioPTH is contingent on the criteria used and that careful judgment is used when selecting a sampling strategy.

Another proposed drawback of ioPTH is the perceived increased procedural cost.^18,37 It is argued that ioPTH may be associated with increase in the length of the operation while the surgeon awaits the laboratory result^16,19 and that there is significant expense associated with the infrastructure and the biochemical expertise required to rapidly process ioPTH.¹⁵ However, in this study the difference in mean operative time was nonsignificant and patients in the ioPTH group were less likely to undergo a further procedure; thus, costs related to reoperation and its potential morbidity, as well as the management of persistent hypercalcemia, are avoided. A cost-effectiveness analysis of both approaches is an area warranting evaluation. The inclusion of nonrandomized studies also increased the risk of selection bias. However, the groups are broadly similar and sensitivity analysis was performed to account for this heterogeneity.

A limitation of the review is the heterogeneity in the length of follow-up of the included studies. Many of the studies report follow-up periods of 6 months, which raises the question of lower rates of cure due to unforeseen and undocumented recurrences.^{3,13,14,16,21}

In conclusion, this meta-analysis demonstrates that using ioPTH in patients with PHP undergoing attempted MIP is associated with higher cure rates and a reduced need for reoperation. While patients are more likely to have their MIP transformed to a BNE, this is likely due to the presence of multigland disease or an aberrant adenoma, and necessary to facilitate cure, and there is no associated increased morbidity with ioPTH. For these reasons, the authors advocate that attempted MIP should only be undertaken with the availability of ioPTH as an adjunct.

Accepted for Publication: September 5, 2020.

Published Online: November 19, 2020. doi:10.1001/jamaoto.2020.4021

Corresponding Author: Éanna J. Ryan, MBBCh, BAO, MRCSI, MD, Department of Surgery, Galway University Hospital, Newcastle Road, Galway, H91 YR71, Ireland (eannaryan@rcsi.com).

Author Contributions: Dr Ryan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Quinn, Ryan, James, Boland, Kerin, Lowery.

Acquisition, analysis, or interpretation of data: Quinn, Ryan, Garry, Young, Lowery.

Drafting of the manuscript: Quinn, Ryan, Garry, Kerin.

Critical revision of the manuscript for important intellectual content: Quinn, Ryan, James, Boland, Young, Lowery.

Statistical analysis: Quinn, Ryan, Garry.

Administrative, technical, or material support: Quinn, Ryan, Garry, James, Boland, Young.

Supervision: Ryan, Young, Kerin, Lowery.

Conflict of Interest Disclosures: None reported.