Parathyroid Surgery: Separating Promise from Reality

We set out to determine the accuracy in predicting the success of biochemical and localizing studies for use in a minimally invasive parathyroidectomy. Preoperative sestamibi scans, intraoperative γ-probe examinations, and intraoperative PTH (IOPTH) monitoring were performed on a prospective cohort of patients.

Seventy-one patients were included in the study. Of the 59 patients (83%) with primary HPT, adenoma localization by sestamibi scanning was correct in 95% with solitary adenomas, but was correct in only 25% of the 14 patients with multiple adenomas. In patients with secondary and tertiary disease, sestamibi scanning incorrectly identified a single hot spot in 64% of cases. In no case of hyperplasia was the probe useful in locating other glands after a single gland was removed. IOPTH was accurate in 78% of patients with primary disease and in only 45% of patients with nonprimary disease.

A minimal approach can be considered in a select group of patients that does not have familial primary HPT, secondary or tertiary disease, coexisting thyroid pathology, or an equivocal sestamibi scan. Only patients with a positive single hot spot on sestamibi scan can be considered candidates. Using this criteria only 64% of all patients with hyperparathyroidism are candidates for a minimally invasive approach. The combination of a solitary hot spot on sestamibi scan and a fall in IOPTH allows the surgeon to make the correct decision regarding the need to convert to a bilateral approach in 93% of these selected patients.

CURRENTLY THERE is considerable discussion about whether focal, unilateral, or bilateral neck exploration should be used in patients with primary sporadic hyperparathyroidism (HPT) because of improved preoperative localization tests and intraoperative PTH (IOPTH) assay. Making sense of applying these new technologies is difficult and poses the question: is limited parathyroid surgery better surgery? Most clinicians deduce that limited surgery is better only if the success rate is equivalent to the “tried and true” gold standard, bilateral exploration. Currently the success rate of parathyroidectomy is greater than 95% in centers experienced with this operation and serious complications are rare (1). Although these results are good, we must continue to improve these results or obtain similar results in a less invasive manner.

No completely reliable means has yet been developed for distinguishing normal from abnormal parathyroid glands, but complementary innovations in biochemical and imaging techniques have altered the surgical approach by many surgeons. A combination of preoperative localization of parathyroid tumors with scintigraphy and IOPTH monitoring was first suggested by Irvin et al. in the mid 1990s (2). The promise of both technologies has also been recommended by others to significantly reduce the extent of surgery and to shorten the hospital stay (3). To help separate the promise from reality, we prospectively studied a relatively large parathyroid patient population operated on during a 9-month period to assess the utility of independent and combined imaging and biochemical techniques.

Subjects and Methods

A prospective cohort of patients requiring parathyroidectomy was established at the University of California-San Francisco/Mount Zion Medical Center. The target population was all patients with documented hyperparathyroidism requiring surgical intervention. The accessible population was patients referred to a single endocrine surgeon during the months of October 1998 through June 1999. All patients were scheduled to have a standard bilateral exploration under general anesthesia with intraoperative visualization of all parathyroid glands. Frozen section was performed on abnormal-appearing glands. A 4- to 5-cm Kocher transverse cervical incision positioned 1 cm caudal to the cricoid cartilage was used to position the incision directly over the midportion of the thyroid gland. In addition, preoperative imaging with sestamibi scintigraphy, intraoperative γ-probe examination, and intraoperative measurement of PTH by quick assay were performed on all patients. These modalities did not influence the action of the surgeons at the time of surgery. Informed consent was obtained and conducted in accordance with the ethical standards of the committee on human experimentation at University of California-San Francisco. Success of surgery was defined as documented cure of hypercalcemia after the operation. Near the end of the study period we used a minimally invasive technique in two patients through a 2.5-cm focal incision positioned more laterally in the neck.

Imaging

Preoperative sestamibi scintigraphy using 15–18 mCi Tc-99 was planned for all patients studied. Studies were scheduled for the week before operation or on the morning of operation. A single nuclear medicine physician interpreted all scans preoperatively. Sestamibi was injected iv, followed by pinhole and planar projection images, immediately and selectively at 2 h after injection. All findings were recorded.

For γ-probe examinations, 15–18 mCi Tc-99 were iv injected 2 h before operation. If not previously scanned, the patient was first imaged with the γ-camera. After positioning on the operating room table, an 11-mm hand-held γ-counter (Neoprobe Corp., Dublin, OH) was used to measure radioactivity in four quadrants of the patient’s neck. This probe was chosen because of its availability at our institution. Subcutaneous γ emission examinations with the probe took place from 60–300 min after injection depending on the operative schedule. Longer time periods were observed near the beginning of the study and were due to prolonged patient transport or unexpected operating room delays. Counts were made, and auditory signals were noted. An overall γ measure baseline for the entire area was referred to as background. A greater than 15% increase from background was noted by a built-in increased pitch in the auditory signal of the machine. Counts were also quantified and recorded. No collimator was used. A single surgeon performed all exams. The upper and lower poles of the thyroid gland on each side defined quadrants. Examinations were performed and recorded transcutaneously before 2- to 3-cm subplatysmal flaps were developed and again before thyroid mobilization. Radioactive sites were localized by following increasing counts. Sites with the highest counts were identified and recorded. After parathyroid gland excision, the γ counts were again measured, and additional radioactive sites were identified if present. Counts were also obtained from resected ex vivo glands. Ultrasound examination was not systematically performed, but information was collected if available.

Intraoperative hormone measurement

All patients were to receive IOPTH monitoring with the Quick-IntraOperative Intact PTH assay (Nichols Institute Diagnostics, San Juan Capistrano, CA). A certified clinical laboratory scientist with training in running high complexity tests performed each assay. Pre-abnormal gland excision serum values were measured and compared with values obtained 10 min after abnormal gland excision. A more than 50% decrease in PTH in a blood sample obtained 10 min after removal of one or more parathyroid glands was defined as a marker of adequate abnormal gland excision. For the purposes of the study a less than 50% decrease after adequate gland excision or intraoperative visualization of other enlarged glands, presumed to be abnormal, was described as inaccuracy of the assay.

The predictor variables were 1) accuracy of preoperative localization with sestamibi scintigraphy, 2) accuracy of intraoperative localization with the γ-probe, and 3) accuracy of quick intraoperative PTH assay as a predictor of adequate gland excision. Accuracy was defined as completion of resection had the surgeon depended solely on the above studies. For example, sestamibi was defined as being accurate if it suggested one abnormal gland and at operation that one gland was indeed the correct and only abnormal gland. Sestamibi was defined as being inaccurate if it suggested a single abnormal gland and at operation there was more than one abnormal gland or if the suggested gland was not the correct abnormal gland. Patients with more than one abnormal parathyroid gland were termed as having hyperplasia when all four parathyroid glands were abnormal and as having double adenomas when two abnormal and histologically proven glands were seen at operation and other normal appearing glands were identified and/or patients had normal calcium and PTH levels postoperatively. The dependent measure was the success rate for surgery, which was a percentage calculated by dividing the number of tests that were of value by the total number of patients who received that test. Percentages of success were calculated for each type of test and for each test when paired with IOPTH. We expected that the percentages would be higher for any test when paired with IOPTH compared with the percentage of any test used alone. Correlation coefficients (Pearson r) and contingency table analyses (χ²) were used for statistical analyses.

Results

Demographics

During a 9-month period from October 1998 through June 1999, 71 parathyroidectomies were performed by a single surgeon. The patient sample included 44 women and 27 men. The mean age was 54 yr (range, 33–77). Sixty-three patients underwent bilateral standard neck explorations with intraoperative visualization of all glands. Eight patients received a unilateral focused approach identifying the abnormal gland only; 6 received focused reoperations where abnormal tissue had been excised at the previous operation (persistent disease). These 6 previous parathyroid operations had been performed at other medical centers. Two patients were operated on using a mini-incision focused initial exploration. These 2 operations were performed near the terminal point of the study. Seven patients had concurrent thyroid disease.

Among our entire group, 1 patient had parathyroid cancer and 59 patients had primary HPT. Of these, 45 had a single adenoma, and 14 (23%) had multiple abnormal parathyroid glands. Four patients (5.6%) had MEN1, and all of these had multiple abnormal parathyroid glands.

Four patients (5.6%) had MEN1, two patients (2.8%) had secondary HPT, and 5 (7.0%) had tertiary HPT. Seventy (98.5%) of the operations were performed under a general anesthetic. One (1.5%) was attempted under local sedation, but required conversion to general anesthesia. All patients were cured, defined as a normal calcium level at 3 month follow-up. The mean and median duration of hospitalization for patients with primary HPT were both 1 d, and those for secondary and tertiary HPT were 1.5 and 1 d, respectively. One patient was hospitalized for 3 d because of a postoperative arrhythmia, yielding a complication rate of 1.4%. No patient developed permanent hypoparathyroidism or recurrent laryngeal nerve palsy.

Sestamibi

Sixty-three of our 71 (87.5%) patients had a preoperative sestamibi scan. Eight patients did not have a preoperative scan because of scheduling difficulties and were omitted from our analyses pertaining to sestamibi scans (see Table 1). All of these patients had solitary adenomas. Localization was correct in 47 patients (74.6%). One cystic parathyroid adenoma was not visualized. Parathyroid cancer was visualized, but had no specific defining characteristics. Sestamibi scans were more accurate in patients with primary sporadic HPT than in those with MEN or secondary or tertiary HPT (by Fisher’s exact test: χ² = 5.68; P < 0.05).

Primary disease. Of the 45 patients with a single adenoma, 42 were scanned, and 40 (95.2%) had a positive scan with correct localization (see Fig. 1). In 1 of the 2 incorrect scans the abnormal gland was on the contralateral side of the neck. In the other case the scan was negative. This latter patient had a reoperation for persistent disease. At reoperation no obvious adenoma was identified. This patient was treated by left lobectomy and thymectomy and has remained normocalcemic. As mentioned above, 1 patient with parathyroid cancer had a positive scan.

Of 14 patients with primary sporadic disease and multiple abnormal parathyroid glands, 12 were scanned. In 3 of these 12 patients (25%), all abnormal parathyroid glands were correctly visualized. In 5 of these 12 cases (41.6%) only a single abnormal parathyroid gland was identified, and in 4 (33%) sestamibi scanning was negative. Missed parathyroid glands ranged in size from 0.5–1.8 cm (mean = 1.0 cm), whereas those that were identified ranged in size from 0.6–1.2 cm (mean = 1.1 cm).

Secondary and tertiary HPT and MEN1. Sestamibi scanning suggested multiple gland involvement in 3 of 8 (37.5%) patients scanned (see Table 1). Five of 8 (62.5%) had a single focus, 2 (25%) showed 2 foci, and 1 (12.5%) showed 3 foci. Overall, only 12 of 32 (37.5%) hyperplastic glands in this subset of patients were visualized.

Coexisting thyroid disease. Of six patients who were scanned with coexisting thyroid disease, one (16.6%) had false positive uptake ipsilateral to the thyroid nodule and contralateral to the parathyroid adenoma. The five others (83.4%) had true positive scans, three of which had thyroid nodules on the side of the parathyroid adenoma, one had multinodular goiter, and 1 had diffuse thyroiditis.

γ-Probe

Sixty of the 71 patients had γ-probe examinations. The mean time to examination was 120 min after iv injection. The γ-probe proved to be only as successful as the preoperative sestamibi scan. In 1 patient (1.6%), however, the probe identified an adenoma that was not localized preoperatively by the γ-camera. This localization identified a preoperative false negative and took place 150 min after iv injection. High counts were from abnormal parathyroid tissue in 47 patients (78.3%). In the other 13 patients (21.7%), the focal site was a nonparathyroid source. In all these cases the thyroid gland was the source of the focal increased count. Eighteen patients (30%) had elevated counts from a particular quadrant after excision of all abnormal parathyroid glands (false positive studies). These 18 patients did not have corresponding elevated counts when the hypercellular tissue was in situ.

Primary sporadic disease. Ex vivo gland counts were 30% or more of baseline for all 46 excised parathyroid adenomas (range, 30–133%). For the functioning cystic adenoma and the parathyroid carcinoma, ex vivo gland counts were 50% and 65%, respectively, above baseline. In patients with primary HPT who had double adenomas and a negative preoperative scan, ex vivo γ-probe gland counts were only slightly elevated (30–50%) compared with baseline. In the 3 patients with primary HPT who had multiple abnormal parathyroid glands and a preoperative scan that demonstrated multiple foci, ex vivo γ-probe counts were significantly more elevated (100–133%) compared with baseline.

Secondary and tertiary HPT and MEN disease. In all cases the γ-probe failed to identify remaining adenomas or hyperplastic parathyroid glands from the thyroid bed after the initial abnormal gland removal. In 50% of cases, the excised hyperplastic or adenomatous parathyroid gland had counts less than 30% of baseline. For each patient there was at least a 25% difference in counts between the glands. In patients with a solitary focus on preoperative sestamibi scanning, the hottest gland was in concordance with preoperative MIBI.

IOPTH

IOPTH monitoring was used in all 71 patients (see Table 2). There was a significant difference in IOPTH values for primary and nonprimary disease (χ² = 4.96; P < 0.05). There was no significant correlation between the preexcision PTH value and the percent decrease at 10 min or between the preexcision PTH and postoperative PTH values. This was true across all diagnoses (r = 0.19) and when analyses were conducted by type of disease (primary HPT vs. nonprimary HPT, r = 0.18).

Primary disease. In primary sporadic HPT, IOPTH was accurate in 46 of 59 (78%) of patients. In 13 of 59 patients (22%), IOPTH monitoring was not accurate, in that the PTH level did not decrease by 50% from the preexcision value despite all abnormal tissue being removed. PTH did not decrease 50% or more in any of these patients when additional abnormal parathyroid glands were still in the neck. IOPTH fell by more than 50% in the one patient with cancer, suggesting that there were no other abnormal parathyroid glands or metastatic disease.

Secondary and tertiary HPT and MEN1 disease. IOPTH was accurate in 5 of 11 patients (45%). In these cases IOPTH dropped more than 50% after subtotal parathyroidectomy. In 3 of these cases blood specimens were not obtained between each gland excision, and we were unable to delineate whether there was an incremental drop after the removal of each hypercellular gland. In the other 2 cases, IOPTH dropped less than 50% when only 1 gland was removed and more than 50% when all glands were removed. In 6 of the other 11 cases (54.6%), IOPTH gave misleading information. For example, in 2 patients with tertiary disease, the IOPTH dropped more than 50% when hyperplastic glands remained in situ.

Combination of studies. Of the 59 patients with primary HPT, 45 had single gland disease, and 14 had multiple gland disease. Forty-two of the former were scanned, as were 12 of the latter. A single hot spot, suggestive of 1 gland, was noted in 41 of the patients who actually had single gland disease and in 5 patients who actually had multiple gland primary hyperparathyroidism. These 46 patients would have been identified as candidates for a minimally invasive approach. This same subset of patients all had IOPTH monitoring. Thirty-seven of the 40 subjects with a single gland actually had a greater than 50% drop in hormone level and would have undergone a successful minimally invasive parathyroidectomy. Three patients with solitary adenomas did not experience a drop of greater than 50% in IOPTH monitoring after solitary gland removal, and this would have wrongly prompted the surgeon to repeat the PTH tests and/or abort the minimally invasive approach and convert to the standard bilateral cervical exploration. This would not have been considered a successful minimally invasive approach. One other patient with single gland disease did not have a drop in hormone level after a search on the side of the neck suggested by sestamibi, as only a thyroid nodule was noted on this side. The actual adenoma was found on the contralateral side of the neck. This could be considered a successful combination of studies, as IOPTH would have not been expected to accurately drop because residual disease remained.

All five patients with primary hyperparathyroidism and multiple gland disease that was not predicted by preoperative sestamibi scan failed to have a drop greater than 50% in the intraoperative PTH monitoring. These would have been considered a successful combination of studies, as the failure to decrease by 50% suggested further evaluation of the neck before wound closure.

Combining these studies, 46 patients would have been eligible for a minimal approach based on sestamibi. Of these, 43 patients had appropriate IOPTH studies that suggested to the surgeon the correct time to stop the dissection. Using these combined studies, 43 of the 46 (93%) patients would have been expected to leave the operating room cured by having had the least invasive procedure possible.

Discussion

For patients with primary sporadic HPT, if we had needed a focal approach using a combination of sestamibi scanning and then using IOPTH, we would have achieved a success rate of surgery of 93%. Nonprimary and familial disease relying on sestamibi and IOPTH would only predict a 62.5% success rate. The γ-probe was as effective for identifying abnormal parathyroid glands as preoperative sestamibi scanning, but added little to the preoperative radionucleide study results and increased the operation time. We therefore no longer use the γ-probe.

Imaging

Sestamibi scanning proved valuable as a preoperative localizing tool in patients with primary HPT due to a solitary adenoma without coexisting clinical thyroid pathology. The patients with positive sestamibi scan demonstrating one abnormal parathyroid with sporadic HPT are usually the easiest to treat successfully. As has previously been reported by us and other investigators (4, 5), coexisting thyroid nodules mimic parathyroid adenomas and may direct the surgeon to an incorrect site. Our data suggest that sestamibi scanning, although useful in many patients, is of little help in patients with multiglandular disease. The major reason to use preoperative sestamibi scanning in such patients would be to identify an abnormal parathyroid in an ectopic position.

IOPTH

Overall, IOPTH was accurate in 47 of 60 (78.4%) patients with solitary tumors. It was helpful in all 14 patients with primary sporadic HPT and multiple abnormal parathyroid glands. It was only valuable in 45% of patients with secondary, tertiary, or familial disease. Although our numbers are relatively small, we do not believe that IOPTH is sufficiently reliable to recommend it for patients with secondary, tertiary, or familial disease at this time. We thought that it was worthwhile to assess IOPTH use in these diseases to evaluate possible ways to prevent persistent disease in patients with more than four glands. We believe that in these patients the parathyroid glands might secrete intermittently or have different set-points for PTH secretion. Further studies to explain these differences, such as using different PTH fragments for assays, or to establish different criteria for PTH levels for patients with multiple abnormal parathyroid glands are indicated. Reliance on IOPTH results without using good surgical judgment can unnecessarily prolong an operation. For example, it may encourage the surgeon to continue to explore when all of the abnormal parathyroid glands have already been identified. We and others recommend repeating the PTH assay in patients whose PTH level has not fallen by more than 50% at 10 min, because it may fall more slowly in some patients.

Limited parathyroid surgery

With good quality preoperative localization and IOPTH monitoring, limited exploration parathyroid surgery becomes practical if based on proper patient selection. The best candidates are patients with primary sporadic HPT (without familial HPT) who have a preoperative localization study that shows a characteristic solitary adenoma and no thyroid nodules on physical examination or, if necessary, preoperative ultrasound. Patients with secondary or tertiary HPT, receiving lithium therapy, or with a family history of HPT would not be candidates. Multiple abnormal parathyroid glands are the rule rather than the exception, and a solitary hot spot on sestamibi scanning may be misleading. Based on the above criteria, 64% of patients who present with HPT at our medical center would have been candidates for focal or limited operation. In contrast, 36% of the patients referred to us for parathyroidectomy would have been considered ineligible for localized parathyroid surgery. Secondary, tertiary, or familial disease accounted for 33%, negative sestamibi scans accounted for 25%, and coexisting thyroid pathology and mediastinal disease accounted for the remaining 17% of patients who were ineligible for a focused operation. This high percentage of ineligible patients is probably due to the tertiary referral pattern that is seen at this institution. Localizing studies should be performed by surgeons and radiologists with experience, because better results are obtained (4).

From the data collected in this study, using selective criteria 93% of patients undergoing limited parathyroid operation by sestamibi criteria and who had an appropriate drop in IOPTH would have had a successful outcome. The general rule for limited parathyroid surgery is that sestamibi scanning, in essence, suggests to the surgeon where to start the exploration, whereas IOPTH suggests when the operation has been successfully accomplished. This is true in most, but certainly not all, patients.

Is a limited approach as successful or cost-effective as the traditional open bilateral exploration? Given the 95–98% success rate reported for the traditional approach, it would be difficult for a new approach to be better. The expected success for a limited approach, being defined as those candidates who have a 50% drop in PTH levels 10 min after tumor removal, may have other benefits compared with open bilateral exploration. This includes minimized dissection, time saving from elimination of exploration of normal parathyroid glands and therefore decreased operating charges, possibly less pain, and shorter length of hospital stay (6, 7). Data on limited operations with IOPTH in noncontrolled studies note a decrease in operating room time by 30 min/patient and a potential saving of $1000.00/patient (7). Whether recurrent laryngeal nerve injury would be lower or higher is questionable, as only one nerve would be at risk in patients having a focal approach, but visualization of the nerve is more difficult with this limited approach. The recurrent laryngeal is at greater risk when the parathyroid tumor is an upper gland, because the nerve is often very close to the parathyroid capsule.

Balancing these benefits is the increased cost of IOPTH due to the need for trained personnel, the probe, and the preoperative scan. The charge of each scintigraphy study is approximately $500.00. If the injection is synchronized with surgery, the compliment of γ-probe adds nothing to the cost if the hospital has purchased the probe for other surgical uses. Because the peak time for visualization varies among individuals, however, a single sestamibi scan performed just before operation may, on occasion, be suboptimal. The timing of the injection is sometimes problematic in a busy operating room. In the past a flat charge for IOPTH ranged from $500–1000/kit; however, newer competitive companies have been able to reduce this cost to $100–300/kit (Diagnostic Products, Los Angeles, CA). High volume centers can reduce this cost even further, using one kit for three or four patients if the patients are scheduled on the same day. Therefore, the cost of this combined imaging and biochemical technology would be between $600–800.

Theoretical calculation of the success of limited incision parathyroid surgery based on this series leads us to make the following recommendations. Preoperative sestamibi is useful to determine whether a patient with sporadic HPT is a candidate for a limited surgery and indicates to the surgeon where to start the exploration. It also occasionally reveals an abnormal parathyroid gland in an unusual site. The sestamibi probe adds little to the preoperative sestamibi localization. It is effective in confirming that the removed tissue is usually a parathyroid adenoma, although some other tumors also have increased counts. It may also help us in directing the dissection through a small incision. IOPTH is useful in determining when to stop the operation and lowers the risk of leaving other adenomatous or hyperplastic parathyroid glands in place. When PTH remains elevated, the assay should be repeated as the PTH level sometimes does not fall within 10 min after a successful operation, as reported by others (8–10). By applying these guidelines we believe that limited parathyroid surgery may be a viable and safe approach for patients with sporadic primary HPT who have a positive preoperative localizing study when only one abnormal parathyroid gland is available and there is no significant coexisting thyroid abnormality. More controlled studies are needed, however, to demonstrate the validity of each of these new techniques.

Conclusion

The use of preoperative sestamibi scanning and IOPTH enables the focused approach in 64% of patients, with an acceptable 93% success rate in contrast to a 98% success rate with bilateral exploration, including all patients with sporadic primary hyperparathyroidism. For patients with familial primary HPT, secondary and tertiary HPT and coexisting thyroid pathology, equivocal or negative sestamibi scans, or scans suggesting multiple glands, a bilateral exploration remains the treatment of choice, because the success rate with IOPTH would only be approximately 63%.

Acknowledgements

We acknowledge the tremendous assistance of Ray Darbyshire, technician at Department of Nuclear Medicine, and our colleagues in Endocrinology who helped us care for these patients. We are most indebted to Ms. Phyllis Easter (secretary at Wake Forest University Baptist Medical Center) and Ms. Kate Poole (administrative assistant at University of California-San Francisco) for the excellent typing and manuscript preparation.

This work was supported in part by Mount Zion Health Systems and the Friends of Endocrine Surgery.

Abbreviations:

 
• HPT,

  Sporadic hyperparathyroidism;

 
• IOPTH,

  intraoperative PTH.