- Email: [email protected]
Real-Time Super Selective Venous Sampling in Remedial Parathyroid Surgery
Real-Time Super Selective Venous Sampling in Remedial Parathyroid Surgery Amir H Lebastchi, MD, John E Aruny, MD, Patricia I Donovan, RN, MBA, Courtney E Quinn, Glenda G Callender, MD, Tobias Carling, MD, PhD, Robert Udelsman, MD, MBA
MD, MS,
Remedial cervical exploration for persistent or recurrent primary hyperparathyroidism can be technically difficult, but is expedited by accurate preoperative localization. We investigated the use of real-time super selective venous sampling (sSVS) in the setting of negative noninvasive imaging modalities. STUDY DESIGN: We performed a retrospective analysis of a prospective database incorporating real-time sSVS in a tertiary academic medical center. Between September 2001 and April 2014, 3,643 patients were referred for surgical treatment of primary hyperparathyroidism. Of these, 31 represented remedial patients who had undergone one (n ¼ 28) or more (n ¼ 3) earlier cervical explorations and had noninformative, noninvasive preoperative localization studies. RESULTS: We extended the use of the rapid parathyroid hormone assay in the interventional radiology suite, generating near real-time data facilitating onsite venous localization by a dedicated interventional radiologist. The predictive value of real-time sSVS localization was investigated. Overall, sSVS correctly predicted the localization of the affected gland in 89% of cases. Of 31 patients who underwent sSVS, a significant rapid parathyroid hormone gradient was identified in 28 (90%), localizing specific venous drainage of a culprit gland. All patients underwent subsequent surgery and were biochemically cured, with the exception of one who had metastatic parathyroid carcinoma. Three patients with negative sSVS were also explored and cured. CONCLUSIONS: Preoperative parathyroid localization is of paramount importance in remedial cervical explorations. Real-time sSVS is a sensitive localization technique for patients with persistent or recurrent primary hyperparathyroidism, when traditional noninvasive imaging studies fail. These results validate the utility and benefit of real-time sSVS in guiding remedial parathyroid surgery. (J Am Coll Surg 2015;220:994e1000. 2015 by the American College of Surgeons)
BACKGROUND:
Primary hyperparathyroidism (PHPT) is a common endocrine disease affecting 1 in 500 to 1,000 patients.1-3 Surgical management remains the sole curative treatment option. Index parathyroid operations for PHPT are associated with cure rates exceeding 95% and complication rates ranging between 1% and 2% when performed by experienced surgeons.4-6 However, a substantial number
of patients require remedial cervical exploration for persistent or recurrent PHPT after initial exploration. Such failure rates exist largely because the majority of parathyroid operations in the United States are performed by lowvolume surgeons who attain lower operative success rates.7-9 Remedial cervical exploration can be technically challenging due to scarring and obliteration of tissue planes, is associated with an increased risk of collateral injury, particularly to the recurrent laryngeal nerves, and the operative failure rate is increased. Given the increased complexity, the decision to reoperate is influenced by the ability to localize the culprit gland(s) preoperatively.10-12 First-line parathyroid localization studies consist of noninvasive imaging studies, including ultrasound, sestamibi, CT, and MRI scans. Patients with informative, noninvasive localization studies proceed to surgical exploration. However, there remains a subset of patients with nonlocalizing or discordant studies, who can benefit from
Disclosure Information: Nothing to disclose. Presented at the 95th Annual Meeting of the New England Surgical Society, Stowe, VT, September 2014. Received October 9, 2014; Revised December 22, 2014; Accepted January 5, 2015. From the Department of Surgery, Section of Endocrine Surgery (Lebastchi, Donovan, Quinn, Callender, Carling, Udelsman) and Department of Radiology, Section of Vascular and Interventional Radiology (Aruny), Yale University School of Medicine, New Haven, CT. Correspondence address: Robert Udelsman, MD, MBA, Department of Surgery, Yale University School of Medicine, 330 Cedar St, FMB 102, PO Box 208062, New Haven, CT 06510. email: [email protected]
ª 2015 by the American College of Surgeons Published by Elsevier Inc.
994
http://dx.doi.org/10.1016/j.jamcollsurg.2015.01.004 ISSN 1072-7515/15
Vol. 220, No. 6, June 2015
Lebastchi et al
Abbreviations and Acronyms
PHPT PTH sSVS SVS
¼ ¼ ¼ ¼
primary hyperparathyroidism parathyroid hormone super selective venous sample selective venous sampling
invasive preoperative localization techniques, such as selective venous sampling (SVS) and arteriography. We previously reported our experience with SVS and demonstrated the value of this localization modality in 6 patients in the reoperative setting.4 Our approach allowed for the extrapolation of the rapid parathyroid hormone (PTH) assay to the interventional radiology suite, which facilitated near real-time data during onsite venous localization. Use of onsite PTH analysis results in immediate feedback, guiding the interventionist and improving the success rate of this localization study. Additionally, the interventional radiologist can obtain samples from areas with a subtle gradient in an effort to increase the resolution by sampling smaller venous branches. Similarly, collecting additional blood samples from the area with the highest PTH gradient also increases the resolution of the study. Some authors refer to this sampling method as “super selective” venous sampling (sSVS), emphasizing the refinement of this modified technique compared with conventional SVS.13 The current cumulative series represents the experience from 31 patients who underwent real-time sSVS at our institution.
METHODS Institutional Review Board approval was obtained before the retrospective analyses of our prospective database. The medical records of all patients who underwent sSVS and subsequent surgical exploration for recurrent or persistent PHPT at a single institution between September 2001 and April 2014 were reviewed. A dedicated interventional radiologist (JEA) performed all selective venous localization studies, and there was no change in technique during the study period. After referral, all patients underwent a detailed history and physical examination, followed by confirmatory laboratory testing. Eligible patients had a biochemically confirmed diagnosis of PHPT (serum calcium >10.5 mg/dL [reference range 8.5 to 10.5 mg/dL] and intact PTH >65 pg/mL [reference range 10 to 65 pg/mL]), which persisted or recurred after previous parathyroid surgery. A meticulous review of all operative, imaging, pathologic, and biochemical studies was performed and additional noninvasive preoperative imaging was obtained, if necessary. If an imaging modality demonstrated unequivocal localization of a culprit parathyroid gland, the
Venous Sampling in Remedial Parathyroid Surgery
995
patients were explored without additional preoperative localization studies. When noninvasive imaging studies were negative or nonconvincing, the patients underwent sSVS in the interventional radiology suite using the rapid PTH assay as an adjunct, before remedial surgery. Real-time super selective venous sampling and arteriography The technical details of the procedure have been described previously.4 All procedures were performed using a standardized protocol. Briefly, after catheterization of the common femoral vein, blood from the iliac vein was obtained (baseline value) before extensive selective venous sampling of small venous branches from the neck and mediastinum. Venography is performed at every site of venous sampling. In addition, in super selective sampling, venograms were obtained in 2 planes to delineate the precise anatomic location of the vessel being sampled. In the setting of equivocal results, the interventional radiologist acquired additional samples in any area that demonstrated a subtle gradient. The investigation continued until a positive assay was obtained or no additional veins could be catheterized to reveal a significant gradient. A 2-fold elevated PTH value, as compared with the baseline PTH level obtained from the iliac vein, defined a positive localizing study.14 A negative localization study is one in which no useful localization information was obtained before the exploration. The results of sSVS were compared with the results of noninvasive localization methods, including ultrasound, sestamibi, CT, and MRI scans. The decision for initial side sampling takes into account operative records and pathologic findings from initial neck explorations, results from noninvasive imaging tests, and results from the prevenous arterial phase that might show a hypervascular focus. The venous anatomy of these patients is subject to vascular remodeling, as all patients underwent at least one earlier neck exploration. Accordingly, an arterial study precedes venous sampling. In brief, after the arterial injection, which might suggest a culprit gland, imaging is continued until the venous drainage is delineated. Occasionally, the major route of venous drainage from an arterial injection will be to the contralateral side. This information prompts venous sampling from this area. Remedial parathyroidectomy All remedial parathyroid operations were performed by the current endocrine surgeons at our institution after a thorough review of the angiography findings and PTH gradients with the interventional radiologist. An intraoperative PTH assay was used adjunctively in all cases, and an adequate reduction in PTH levels (50% drop from baseline), as well as drop into a normal PTH range after parathyroid gland resection allowed for termination
996
Lebastchi et al
Venous Sampling in Remedial Parathyroid Surgery
of the operation. Detailed contemporaneous illustrations were completed in the operating room in every case by the surgeon during remedial exploration. These drawings documented the exact site of the parathyroid gland(s) and other relevant structures. The accuracy of preoperative imaging was compared with intraoperative findings. Data analysis Standardized and prospective data acquisitions were obtained pre-, peri-, and postoperatively. The database comprised demographic information, signs and symptoms, biochemical parameters, preoperative imaging, venous localization data, operative details, pathologic findings, and postoperative follow-up information. GraphPad Prism software, version 6.0c (GraphPad Software, Inc.) was used for statistical analysis and graphical representation of data.
RESULTS Between 2001 and 2014, thirty-one patients with recurrent (3 patients [10%]) or persistent (28 patients [90%]) PHPT were referred for operative management in the setting of equivocal or negative preoperative imaging studies. These patients underwent sSVS using the rapid PTH assay in the interventional radiology suite. Demographic data are shown in Table 1. The majority were female (23 patients [74%]), with a mean age of 56 years (range 25 to 74 years). At the time of presentation, 28 patients (90%) had undergone 1 initial operation, and 3 patients had undergone 1 or more operations (10%). The patients were all symptomatic or had signs of disease (Table 1) and presented after their initial exploration with a mean SEM intact serum PTH level of 129 10.63 pg/ mL and serum calcium level of 11.2 0.11 mg/dL. Discordant or nonlocalizing noninvasive studies before sSVS included sestamibi (n ¼ 24), ultrasound (n ¼ 21), CT (n ¼ 22) and MRI (n ¼ 6). Intraoperative findings were correlated with preoperative localization studies. To investigate the value of the localization study for a correct prediction of a diseased parathyroid gland, we defined an accurate venous localization study if a gradient was appreciated preoperatively, the culprit parathyroid gland detected intraoperatively corresponded to the localization study, and resection of the above gland resulted in cure. If the study met all of these criteria, the localization study was considered to be a true-positive result. A diagnostic test was deemed false positive if the study demonstrated an abnormal gland at a site different than intraoperatively found, and false negative if the preoperative localization study failed to detect the abnormal gland. The sensitivity of a localization study was defined as the proportion of studies that correctly localized all abnormal
J Am Coll Surg
Table 1. Demographics, Signs, and Symptoms Characteristic
Age, y, mean (range) Sex, n (%) Male Female Preoperative calcium,* mg/dL, mean SEM Preoperative PTH,y pg/mL, mean SEM Earlier operations, n (%) 1 2 3 Persistent disease, n (%) Recurrent disease, n (%) Symptoms, n (%)z Fatigue Neurocognitive Gastrointestinal Asymptomatic Signs, n (%) Bone disease (osteopenia or osteoporosis)x Nephrolithiasis Cardiovascular diseasejj
56.1 (25e74) 8 (25.8) 23 (74.2) 11.2 0.11 129.5 10.63 28 2 1 28 3
(90) (7) (3) (90) (10)
21 17 12 1
(67.7) (54.8) (38.7) (3.2)
24 (77.4) 8 (25.8) 7 (22.6)
*Serum calcium reference range, 8.5 to 10.5 mg/dL. y Serum parathyroid hormone reference range, 10 to 65 pg/mL. z Many patients had more than one sign or symptom. x Bone disease is defined as radiographically confirmed osteopenia or osteoporosis. jj Cardiovascular disease includes a documented history of heart and/or vascular disease.
glands. Super-selective sampling based on subtle, but nonconfirmatory PTH gradients were performed in 10 of 31 patients, or 32% of cases. Surgical results All 31 patients undergoing sSVS underwent remedial exploration. Successful surgical treatment, defined by achieving normocalcemia (8.5 to 10.5 mg/dL) at the postoperative clinic visit (5 to 7 days after surgery), was achieved in 100% of these patients, except one who proved to have unresectable metastatic parathyroid carcinoma. There were no perioperative complications. Details about baseline PTH levels, the location of the maximum PTH level during sSVS, and the exact location during exploration are shown in Table 2 and Figure 1. Surgical intervention was performed using the sSVS gradient mapping, as well as previous operative records. However, if meticulous dissection did not reveal abnormal parathyroid tissue at the predicted site (false-positive gradient), exploration continued until the culprit gland was identified, resected, and an appropriate intraoperative PTH decrement was demonstrated (Fig. 2). In the 4 patients
Patient no.
No. of samples
Highest PTH, pg/mL
sSVS exact vein
Operative finding, exact location
Laterality correct
Location correct
Cure
68 72 57 59 34 68 40 49 73 37 53 25 55 55 58 69 72 66 58 57 26 64 55 63 52 66 61 56 74 38 56
M F F F M M F F F M F F F M F F F F M F F F F F F F M F M F F
22 29 21 25 27 21 47 27 40 42 24 43 41 36 28 24 35 33 22 32 26 28 32 25 16 19 36 33 36 25 19
4,150 7,035 18,670 449 1,950 11,420 614 3,440 2,358 231 340 4,713 123 198 249 7,160 324 1,884 2,583 420 576 8,460 897 1,629 2,026 152 2,639 2,855 5,615 1,505 1,114
R superior thyroid L inferior thyroid L middle thyroid L brachiocephalic R thyroid ima R middle thyroid R superior thyroid R vertebral L internal mammary R internal jugular SVC R thyroid ima R subclavian SVC SVC R superior thyroid L vertebral trunk (low) R anterior jugular base R and L superior thyroid branch R brachiocephalic R internal jugular (C6) R inferior thyroid lateral branch L mediastinal R thymic R vertebral (mid trunk) SVC Azygos mid cardiac L branch R anterior jugular R lateral thymic R thyroid ima Mid L brachiocephalic
R superior, retroesophageal Not identified/ligation L superior Thymus R superior and inferior R inferior R superior R inferior/retroesophageal Thymus Left superior R intrathymic R intrathyroidal Right superior R intrathyroidal Mediastinum R superior L upper ectopic L intrathymic L superior R inferior R superior R superior and inferior Mediastinum R thymus R intrathyroidal Thymus Metastatic spread of L thorax R inferior R inferior R inferior L carotid sheath
Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No No Yes Yes No Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes
Yes No Yes Yes Yes Yes Yes Yes Yes No Yes Yes No No Yes Yes Yes No Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes No
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes
Venous Sampling in Remedial Parathyroid Surgery
Sex
Lebastchi et al
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Age, y
Vol. 220, No. 6, June 2015
Table 2. Venous Sampling Results
F, female; L, left; M, male; PTH, parathyroid hormone; R, right; sSVS, super selective venous sampling; SVC, superior vena cava. 997
998
Lebastchi et al
Venous Sampling in Remedial Parathyroid Surgery
J Am Coll Surg
Figure 1. Correlation of venous sampling results with intraoperative findings. Numbers with circles refer to individualized patients, corresponding to Table 2. White and gray circles demonstrate the location of intravenous sampling sites with the maximum parathyroid hormone concentration during the diagnostic selective venous sampling. Green and red circles represent gland location during operative exploration.
with false-positive venous localization studies, operative intervention was ultimately successful. Additionally, 3 patients underwent cervical exploration despite insignificant venous gradients on preoperative sSVS (false-negative study; venous gradient <2). These 3 patients were also cured after remedial exploration.
Selective venous sampling Venous sampling was positive in 28 of 31 patients (90%). No adverse events occurred from the catheterization method. The mean sampling site number per procedure was 30 (range 16 to 47 samples). Results of venous localization studies are demonstrated in Table 3. The sensitivity (86%) and positive predictive value (93%) for correct precise preoperative localization of this test far exceeded the results of all other noninvasive diagnostic tests (Table 3). Noninvasive preoperative imaging modalities Sensitivity and positive predictive value data for each diagnostic method were analyzed and are depicted in Table 3.
Figure 2. Preoperative, baseline, and intraoperative intact parathyroid hormone (PTH) levels in patients who underwent remedial cervical exploration after venous localization. Preoperative levels were obtained 1 week before surgery (Preoperative). Intraoperative levels were obtained on the day of surgery before incision (Baseline), and then at the time of gland excision (T0), and 5 minutes (T5), 10 minutes (T10), and 15 minutes (T15) afterwards. The normal range for the intact PTH assay is 10 to 65 pg/mL.
DISCUSSION Remedial cervical exploration for persistent or recurrent PHPT presents both diagnostic and technical challenges.9-11,15,16 This study demonstrates the ability of real-time sSVS to localize residual abnormal parathyroid glands in patients with persistent or recurrent PHPT, after other noninvasive localization studies have failed. It also illustrates the advantages of this modified technique using a rapid PTH assay as an adjunct in the interventional radiology suite. The extension of the use of the rapid PTH assay from the operating room to the angiography suite provides the interventionist with critical information
Lebastchi et al
Vol. 220, No. 6, June 2015
Table 3.
Venous Sampling in Remedial Parathyroid Surgery
999
Comparison of Effectiveness of Localization by Preoperative Radiographic Study
Underwent study and operation, n Negative study, n (%) Positive study, n (%) Incorrect localization, n (%) Correct localization, n (%) Sensitivity, % Positive predictive value, %
MRI
Ultrasound
Sestamibi
CT
Real-time sSVS
6 4 (66.7) 2 (33.3) 1 (50) 1 (50) 20.0 50.0
21 15 (71.4) 6 (28.6) 2 (33.3) 4 (66.6) 21.5 66.7
24 16 (72.7) 8 (27.3) 5 (62.5) 3 (37.5) 15.8 37.5
22 9 (40.9) 13 (59.1) 4 (30.8) 9 (69.2) 50.0 69.2
31 3 (9.7) 28 (90.3) 3 (10.7) 25 (89.3) 86.2 92.6
Correct localization determined by percentage of positive studies with correct localization. sSVS, super selective venous sampling.
within 18 minutes of blood sampling. These “real-time” PTH values allow termination of the study when adequate information is obtained, or continuation if the study remains noninformative. Real-time data influenced venous sampling in essentially every patient because the PTH assay results directed the angiographer to sitespecific locations. Additionally, the interventional radiologist can obtain samples from areas with a subtle gradient in an effort to increase the fidelity by sampling smaller venous branches. Similarly, collecting additional blood samples from the area with the highest PTH gradient also increases the resolution of the study. Some authors refer to this sampling method as “super selective,” emphasizing the refinement of this modified technique compared with conventional SVS techniques.13 It is important to note that even when a specific gradient is identified, the contralateral side is always sampled because multisite disease can be encountered. Preoperative noninvasive imaging studies are always used before remedial parathyroid surgery. Despite technical advancements, the results of such studies can be noninformative in the reoperative setting, as they often fail to demonstrate adequate localization in a subset of patients. Real-time conventional SVS has been shown to be the most sensitive parathyroid localization procedure.16-18 Using a PTH gradient of 2-fold or greater to define a positive result, we achieved sensitivity and positive predictive values of 86% and 93%, respectively. Because all noninvasive imaging modalities focus on the cervical and mediastinal areas, they fail to identify PTH-secreting tissue outside of these regions, as exemplified in patient number 27. This patient presented with severe hypercalcemia 18 years after initial parathyroid surgery for parathyroid carcinoma and negative imaging studies including ultrasound, 4D CT, MRI, and sestamibi scans. Real-time sSVS demonstrated an elevated PTH gradient in the azygos vein, and a subsequent CT of the lower thorax demonstrated systemic presentation of parathyroid carcinoma on the left hemi-diaphragm. This diagnostic sSVS study, not only localized ectopic parathyroid
tissue where initial noninvasive localization procedures failed, it also obviated inappropriate cervical exploration. Despite these encouraging results, several limitations are noted. First, there are inherent limitations to studies evaluating a single institution’s experience. Because a single, dedicated interventional radiologist performed all sSVS, the application of the same principles universally by other, less experienced interventionists affects its reproducibility. Another limitation is that although we have complete short-term follow-up for all patients, long-term data are lacking. However, we contacted all of our patients and have received long-term follow-up in 18. Of these, 17 patients with mean follow-up of 96 months remain eucalcemic. The one patient with metastatic parathyroid carcinoma remains hypercalcemic. Lastly, sSVS is an expensive, invasive study subject to potential complications, including bleeding, infection, and patient discomfort. Complications were not encountered in the current series.
CONCLUSIONS Super selective venous sampling with real-time rapid PTH assay is a safe and effective method of preoperative localization before remedial parathyroid surgery. However, noninvasive studies should always be performed initially and, when deemed to be sufficient, sSVS is not required. In cases where noninvasive techniques fail to localize an abnormal parathyroid gland, sSVS can be used to increase operative success. We achieved cure in all patients (except one with metastatic parathyroid carcinoma) who were explored after sSVS, even if the study was negative, emphasizing the importance of surgeon’s experience for successful parathyroid surgery. Author Contributions Study conception and design: Lebastchi, Aruny, Udelsman Acquisition of data: Lebastchi, Donovan Analysis and interpretation of data: Aruny, Quinn, Callender, Carling, Udelsman
1000
Mazzaglia
J Am Coll Surg
Invited Commentary
Drafting of manuscript: Lebastchi Critical revision: Udelsman
18. Miller DL, Doppman JL, Krudy AG, et al. Localization of parathyroid adenomas in patients who have undergone surgery. Part II. Invasive procedures. Radiology 1987;162[1 Pt 1]:138e141.
REFERENCES 1. Silverberg SJ, Bilezikian JP. Evaluation and management of primary hyperparathyroidism. J Clin Endocrinol Metab 1996;81:2036e2040. 2. Adami S, Marcocci C, Gatti D. Epidemiology of primary hyperparathyroidism in Europe. J Bone Miner Res 2002;17: N18eN23. 3. Boonstra CE, Jackson CE. Serum calcium survey for hyperparathyroidism: results in 50,000 clinic patients. Am J Clin Pathol 1971;55:523e526. 4. Udelsman R, Aruny JE, Donovan PI, et al. Rapid parathyroid hormone analysis during venous localization. Ann Surg 2003; 237:714e719; discussion 719e721. 5. Udelsman R. Surgery in primary hyperparathyroidism: the patient without previous neck surgery. J Bone Miner Res 2002;17[Suppl 2]:N126eN132. 6. Udelsman R, Lin Z, Donovan P. The superiority of minimally invasive parathyroidectomy based on 1650 consecutive patients with primary hyperparathyroidism. Ann Surg 2011; 253:585e591. 7. Thompson GB, Grant CS, Perrier ND, et al. Reoperative parathyroid surgery in the era of sestamibi scanning and intraoperative parathyroid hormone monitoring. Arch Surg 1999; 134:699e704; discussion 705. 8. Saunders BD, Wainess RM, Dimick JB, et al. Who performs endocrine operations in the United States? Surgery 2003;134: 924e931; discussion 931. 9. Shen W, Duren M, Morita E, et al. Reoperation for persistent or recurrent primary hyperparathyroidism. Arch Surg 1996; 131:861e867; discussion 867e869. 10. Prescott JD, Udelsman R. Remedial operation for primary hyperparathyroidism. World J Surg 2009;33:2324e2334. 11. Udelsman R, Donovan PI. Remedial parathyroid surgery: changing trends in 130 consecutive cases. Ann Surg 2006;244:471e479. 12. Wang TS, Udelsman R. Remedial surgery for primary hyperparathyroidism. Adv Surg 2007;41:1e15. 13. Gimm O, Arnesson LG, Olofsson P, et al. Super-selective venous sampling in conjunction with quickPTH for patients with persistent primary hyperparathyroidism: report of five cases. Surg Today 2012;42:570e576. 14. Jones JJ, Brunaud L, Dowd CF, et al. Accuracy of selective venous sampling for intact parathyroid hormone in difficult patients with recurrent or persistent hyperparathyroidism. Surgery 2002;132:944e950; discussion 950e951. 15. Yen TW, Wang TS, Doffek KM, et al. Reoperative parathyroidectomy: an algorithm for imaging and monitoring of intraoperative parathyroid hormone levels that results in a successful focused approach. Surgery 2008;144:611e619; discussion 619e621. 16. Reidel MA, Schilling T, Graf S, et al. Localization of hyperfunctioning parathyroid glands by selective venous sampling in reoperation for primary or secondary hyperparathyroidism. Surgery 2006;140:907e913; discussion 913. 17. Sugg SL, Fraker DL, Alexander R, et al. Prospective evaluation of selective venous sampling for parathyroid hormone concentration in patients undergoing reoperations for primary hyperparathyroidism. Surgery 1993;114:1004e1009; discussion 1009e1010.
Invited Commentary Peter J Mazzaglia, Providence, RI
MD, FACS
The Endocrine Surgery and Interventional Radiology divisions at Yale University have summarized more than a decade’s experience with surgical management of persistent and recurrent hyperparathyroidism, specifically highlighting the role that selective venous sampling plays in preoperative localization of the enlarged parathyroid gland. The study expands on previously published work and examines patients for whom traditional modalities of parathyroid localization, such as sestamibi, ultrasound, and 4-dimensional CT, are negative or equivocal. They described the performance of super-selective venous sampling in 31 such patients while using real time rapid parathyroid hormone testing, which enabled the interventional radiologist to zero in on the anatomic location of the enlarged gland and provide detailed mapping for the operating surgeon. Using a cutoff gradient of twice baseline, 28 patients had positive studies. At surgery, the enlarged gland was correctly identified in 24 of 28 of these patients. Despite the 4 false positive and 3 false negative selective venous sampling studies, there was 100% operative success with use of the intraoperative parathyroid hormone assay. This illustrates the well-known adage that the most important localizing study is the localization of an experienced endocrine surgeon.1 The expertise, labor intensiveness, and cost required to perform the selective venous sampling procedure with simultaneous rapid parathyroid hormone testing would necessarily limit this type of evaluation to only the most specialized interventional radiology departments. Also, because of its invasive nature, this type of parathyroid localization should be limited to those patients with persistent or recurrent hyperparathyroidism in whom repeat noninvasive imaging, performed at a center of expertise in endocrine surgery, has been unrevealing. REFERENCE 1. Doppman JL. Reoperative parathyroid surgery; localization procedures. Prog Surg 1986;18:117e132.
MD, MS,
Remedial cervical exploration for persistent or recurrent primary hyperparathyroidism can be technically difficult, but is expedited by accurate preoperative localization. We investigated the use of real-time super selective venous sampling (sSVS) in the setting of negative noninvasive imaging modalities. STUDY DESIGN: We performed a retrospective analysis of a prospective database incorporating real-time sSVS in a tertiary academic medical center. Between September 2001 and April 2014, 3,643 patients were referred for surgical treatment of primary hyperparathyroidism. Of these, 31 represented remedial patients who had undergone one (n ¼ 28) or more (n ¼ 3) earlier cervical explorations and had noninformative, noninvasive preoperative localization studies. RESULTS: We extended the use of the rapid parathyroid hormone assay in the interventional radiology suite, generating near real-time data facilitating onsite venous localization by a dedicated interventional radiologist. The predictive value of real-time sSVS localization was investigated. Overall, sSVS correctly predicted the localization of the affected gland in 89% of cases. Of 31 patients who underwent sSVS, a significant rapid parathyroid hormone gradient was identified in 28 (90%), localizing specific venous drainage of a culprit gland. All patients underwent subsequent surgery and were biochemically cured, with the exception of one who had metastatic parathyroid carcinoma. Three patients with negative sSVS were also explored and cured. CONCLUSIONS: Preoperative parathyroid localization is of paramount importance in remedial cervical explorations. Real-time sSVS is a sensitive localization technique for patients with persistent or recurrent primary hyperparathyroidism, when traditional noninvasive imaging studies fail. These results validate the utility and benefit of real-time sSVS in guiding remedial parathyroid surgery. (J Am Coll Surg 2015;220:994e1000. 2015 by the American College of Surgeons)
BACKGROUND:
Primary hyperparathyroidism (PHPT) is a common endocrine disease affecting 1 in 500 to 1,000 patients.1-3 Surgical management remains the sole curative treatment option. Index parathyroid operations for PHPT are associated with cure rates exceeding 95% and complication rates ranging between 1% and 2% when performed by experienced surgeons.4-6 However, a substantial number
of patients require remedial cervical exploration for persistent or recurrent PHPT after initial exploration. Such failure rates exist largely because the majority of parathyroid operations in the United States are performed by lowvolume surgeons who attain lower operative success rates.7-9 Remedial cervical exploration can be technically challenging due to scarring and obliteration of tissue planes, is associated with an increased risk of collateral injury, particularly to the recurrent laryngeal nerves, and the operative failure rate is increased. Given the increased complexity, the decision to reoperate is influenced by the ability to localize the culprit gland(s) preoperatively.10-12 First-line parathyroid localization studies consist of noninvasive imaging studies, including ultrasound, sestamibi, CT, and MRI scans. Patients with informative, noninvasive localization studies proceed to surgical exploration. However, there remains a subset of patients with nonlocalizing or discordant studies, who can benefit from
Disclosure Information: Nothing to disclose. Presented at the 95th Annual Meeting of the New England Surgical Society, Stowe, VT, September 2014. Received October 9, 2014; Revised December 22, 2014; Accepted January 5, 2015. From the Department of Surgery, Section of Endocrine Surgery (Lebastchi, Donovan, Quinn, Callender, Carling, Udelsman) and Department of Radiology, Section of Vascular and Interventional Radiology (Aruny), Yale University School of Medicine, New Haven, CT. Correspondence address: Robert Udelsman, MD, MBA, Department of Surgery, Yale University School of Medicine, 330 Cedar St, FMB 102, PO Box 208062, New Haven, CT 06510. email: [email protected]
ª 2015 by the American College of Surgeons Published by Elsevier Inc.
994
http://dx.doi.org/10.1016/j.jamcollsurg.2015.01.004 ISSN 1072-7515/15
Vol. 220, No. 6, June 2015
Lebastchi et al
Abbreviations and Acronyms
PHPT PTH sSVS SVS
¼ ¼ ¼ ¼
primary hyperparathyroidism parathyroid hormone super selective venous sample selective venous sampling
invasive preoperative localization techniques, such as selective venous sampling (SVS) and arteriography. We previously reported our experience with SVS and demonstrated the value of this localization modality in 6 patients in the reoperative setting.4 Our approach allowed for the extrapolation of the rapid parathyroid hormone (PTH) assay to the interventional radiology suite, which facilitated near real-time data during onsite venous localization. Use of onsite PTH analysis results in immediate feedback, guiding the interventionist and improving the success rate of this localization study. Additionally, the interventional radiologist can obtain samples from areas with a subtle gradient in an effort to increase the resolution by sampling smaller venous branches. Similarly, collecting additional blood samples from the area with the highest PTH gradient also increases the resolution of the study. Some authors refer to this sampling method as “super selective” venous sampling (sSVS), emphasizing the refinement of this modified technique compared with conventional SVS.13 The current cumulative series represents the experience from 31 patients who underwent real-time sSVS at our institution.
METHODS Institutional Review Board approval was obtained before the retrospective analyses of our prospective database. The medical records of all patients who underwent sSVS and subsequent surgical exploration for recurrent or persistent PHPT at a single institution between September 2001 and April 2014 were reviewed. A dedicated interventional radiologist (JEA) performed all selective venous localization studies, and there was no change in technique during the study period. After referral, all patients underwent a detailed history and physical examination, followed by confirmatory laboratory testing. Eligible patients had a biochemically confirmed diagnosis of PHPT (serum calcium >10.5 mg/dL [reference range 8.5 to 10.5 mg/dL] and intact PTH >65 pg/mL [reference range 10 to 65 pg/mL]), which persisted or recurred after previous parathyroid surgery. A meticulous review of all operative, imaging, pathologic, and biochemical studies was performed and additional noninvasive preoperative imaging was obtained, if necessary. If an imaging modality demonstrated unequivocal localization of a culprit parathyroid gland, the
Venous Sampling in Remedial Parathyroid Surgery
995
patients were explored without additional preoperative localization studies. When noninvasive imaging studies were negative or nonconvincing, the patients underwent sSVS in the interventional radiology suite using the rapid PTH assay as an adjunct, before remedial surgery. Real-time super selective venous sampling and arteriography The technical details of the procedure have been described previously.4 All procedures were performed using a standardized protocol. Briefly, after catheterization of the common femoral vein, blood from the iliac vein was obtained (baseline value) before extensive selective venous sampling of small venous branches from the neck and mediastinum. Venography is performed at every site of venous sampling. In addition, in super selective sampling, venograms were obtained in 2 planes to delineate the precise anatomic location of the vessel being sampled. In the setting of equivocal results, the interventional radiologist acquired additional samples in any area that demonstrated a subtle gradient. The investigation continued until a positive assay was obtained or no additional veins could be catheterized to reveal a significant gradient. A 2-fold elevated PTH value, as compared with the baseline PTH level obtained from the iliac vein, defined a positive localizing study.14 A negative localization study is one in which no useful localization information was obtained before the exploration. The results of sSVS were compared with the results of noninvasive localization methods, including ultrasound, sestamibi, CT, and MRI scans. The decision for initial side sampling takes into account operative records and pathologic findings from initial neck explorations, results from noninvasive imaging tests, and results from the prevenous arterial phase that might show a hypervascular focus. The venous anatomy of these patients is subject to vascular remodeling, as all patients underwent at least one earlier neck exploration. Accordingly, an arterial study precedes venous sampling. In brief, after the arterial injection, which might suggest a culprit gland, imaging is continued until the venous drainage is delineated. Occasionally, the major route of venous drainage from an arterial injection will be to the contralateral side. This information prompts venous sampling from this area. Remedial parathyroidectomy All remedial parathyroid operations were performed by the current endocrine surgeons at our institution after a thorough review of the angiography findings and PTH gradients with the interventional radiologist. An intraoperative PTH assay was used adjunctively in all cases, and an adequate reduction in PTH levels (50% drop from baseline), as well as drop into a normal PTH range after parathyroid gland resection allowed for termination
996
Lebastchi et al
Venous Sampling in Remedial Parathyroid Surgery
of the operation. Detailed contemporaneous illustrations were completed in the operating room in every case by the surgeon during remedial exploration. These drawings documented the exact site of the parathyroid gland(s) and other relevant structures. The accuracy of preoperative imaging was compared with intraoperative findings. Data analysis Standardized and prospective data acquisitions were obtained pre-, peri-, and postoperatively. The database comprised demographic information, signs and symptoms, biochemical parameters, preoperative imaging, venous localization data, operative details, pathologic findings, and postoperative follow-up information. GraphPad Prism software, version 6.0c (GraphPad Software, Inc.) was used for statistical analysis and graphical representation of data.
RESULTS Between 2001 and 2014, thirty-one patients with recurrent (3 patients [10%]) or persistent (28 patients [90%]) PHPT were referred for operative management in the setting of equivocal or negative preoperative imaging studies. These patients underwent sSVS using the rapid PTH assay in the interventional radiology suite. Demographic data are shown in Table 1. The majority were female (23 patients [74%]), with a mean age of 56 years (range 25 to 74 years). At the time of presentation, 28 patients (90%) had undergone 1 initial operation, and 3 patients had undergone 1 or more operations (10%). The patients were all symptomatic or had signs of disease (Table 1) and presented after their initial exploration with a mean SEM intact serum PTH level of 129 10.63 pg/ mL and serum calcium level of 11.2 0.11 mg/dL. Discordant or nonlocalizing noninvasive studies before sSVS included sestamibi (n ¼ 24), ultrasound (n ¼ 21), CT (n ¼ 22) and MRI (n ¼ 6). Intraoperative findings were correlated with preoperative localization studies. To investigate the value of the localization study for a correct prediction of a diseased parathyroid gland, we defined an accurate venous localization study if a gradient was appreciated preoperatively, the culprit parathyroid gland detected intraoperatively corresponded to the localization study, and resection of the above gland resulted in cure. If the study met all of these criteria, the localization study was considered to be a true-positive result. A diagnostic test was deemed false positive if the study demonstrated an abnormal gland at a site different than intraoperatively found, and false negative if the preoperative localization study failed to detect the abnormal gland. The sensitivity of a localization study was defined as the proportion of studies that correctly localized all abnormal
J Am Coll Surg
Table 1. Demographics, Signs, and Symptoms Characteristic
Age, y, mean (range) Sex, n (%) Male Female Preoperative calcium,* mg/dL, mean SEM Preoperative PTH,y pg/mL, mean SEM Earlier operations, n (%) 1 2 3 Persistent disease, n (%) Recurrent disease, n (%) Symptoms, n (%)z Fatigue Neurocognitive Gastrointestinal Asymptomatic Signs, n (%) Bone disease (osteopenia or osteoporosis)x Nephrolithiasis Cardiovascular diseasejj
56.1 (25e74) 8 (25.8) 23 (74.2) 11.2 0.11 129.5 10.63 28 2 1 28 3
(90) (7) (3) (90) (10)
21 17 12 1
(67.7) (54.8) (38.7) (3.2)
24 (77.4) 8 (25.8) 7 (22.6)
*Serum calcium reference range, 8.5 to 10.5 mg/dL. y Serum parathyroid hormone reference range, 10 to 65 pg/mL. z Many patients had more than one sign or symptom. x Bone disease is defined as radiographically confirmed osteopenia or osteoporosis. jj Cardiovascular disease includes a documented history of heart and/or vascular disease.
glands. Super-selective sampling based on subtle, but nonconfirmatory PTH gradients were performed in 10 of 31 patients, or 32% of cases. Surgical results All 31 patients undergoing sSVS underwent remedial exploration. Successful surgical treatment, defined by achieving normocalcemia (8.5 to 10.5 mg/dL) at the postoperative clinic visit (5 to 7 days after surgery), was achieved in 100% of these patients, except one who proved to have unresectable metastatic parathyroid carcinoma. There were no perioperative complications. Details about baseline PTH levels, the location of the maximum PTH level during sSVS, and the exact location during exploration are shown in Table 2 and Figure 1. Surgical intervention was performed using the sSVS gradient mapping, as well as previous operative records. However, if meticulous dissection did not reveal abnormal parathyroid tissue at the predicted site (false-positive gradient), exploration continued until the culprit gland was identified, resected, and an appropriate intraoperative PTH decrement was demonstrated (Fig. 2). In the 4 patients
Patient no.
No. of samples
Highest PTH, pg/mL
sSVS exact vein
Operative finding, exact location
Laterality correct
Location correct
Cure
68 72 57 59 34 68 40 49 73 37 53 25 55 55 58 69 72 66 58 57 26 64 55 63 52 66 61 56 74 38 56
M F F F M M F F F M F F F M F F F F M F F F F F F F M F M F F
22 29 21 25 27 21 47 27 40 42 24 43 41 36 28 24 35 33 22 32 26 28 32 25 16 19 36 33 36 25 19
4,150 7,035 18,670 449 1,950 11,420 614 3,440 2,358 231 340 4,713 123 198 249 7,160 324 1,884 2,583 420 576 8,460 897 1,629 2,026 152 2,639 2,855 5,615 1,505 1,114
R superior thyroid L inferior thyroid L middle thyroid L brachiocephalic R thyroid ima R middle thyroid R superior thyroid R vertebral L internal mammary R internal jugular SVC R thyroid ima R subclavian SVC SVC R superior thyroid L vertebral trunk (low) R anterior jugular base R and L superior thyroid branch R brachiocephalic R internal jugular (C6) R inferior thyroid lateral branch L mediastinal R thymic R vertebral (mid trunk) SVC Azygos mid cardiac L branch R anterior jugular R lateral thymic R thyroid ima Mid L brachiocephalic
R superior, retroesophageal Not identified/ligation L superior Thymus R superior and inferior R inferior R superior R inferior/retroesophageal Thymus Left superior R intrathymic R intrathyroidal Right superior R intrathyroidal Mediastinum R superior L upper ectopic L intrathymic L superior R inferior R superior R superior and inferior Mediastinum R thymus R intrathyroidal Thymus Metastatic spread of L thorax R inferior R inferior R inferior L carotid sheath
Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Yes Yes No No Yes Yes No Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes
Yes No Yes Yes Yes Yes Yes Yes Yes No Yes Yes No No Yes Yes Yes No Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes No
Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes
Venous Sampling in Remedial Parathyroid Surgery
Sex
Lebastchi et al
01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Age, y
Vol. 220, No. 6, June 2015
Table 2. Venous Sampling Results
F, female; L, left; M, male; PTH, parathyroid hormone; R, right; sSVS, super selective venous sampling; SVC, superior vena cava. 997
998
Lebastchi et al
Venous Sampling in Remedial Parathyroid Surgery
J Am Coll Surg
Figure 1. Correlation of venous sampling results with intraoperative findings. Numbers with circles refer to individualized patients, corresponding to Table 2. White and gray circles demonstrate the location of intravenous sampling sites with the maximum parathyroid hormone concentration during the diagnostic selective venous sampling. Green and red circles represent gland location during operative exploration.
with false-positive venous localization studies, operative intervention was ultimately successful. Additionally, 3 patients underwent cervical exploration despite insignificant venous gradients on preoperative sSVS (false-negative study; venous gradient <2). These 3 patients were also cured after remedial exploration.
Selective venous sampling Venous sampling was positive in 28 of 31 patients (90%). No adverse events occurred from the catheterization method. The mean sampling site number per procedure was 30 (range 16 to 47 samples). Results of venous localization studies are demonstrated in Table 3. The sensitivity (86%) and positive predictive value (93%) for correct precise preoperative localization of this test far exceeded the results of all other noninvasive diagnostic tests (Table 3). Noninvasive preoperative imaging modalities Sensitivity and positive predictive value data for each diagnostic method were analyzed and are depicted in Table 3.
Figure 2. Preoperative, baseline, and intraoperative intact parathyroid hormone (PTH) levels in patients who underwent remedial cervical exploration after venous localization. Preoperative levels were obtained 1 week before surgery (Preoperative). Intraoperative levels were obtained on the day of surgery before incision (Baseline), and then at the time of gland excision (T0), and 5 minutes (T5), 10 minutes (T10), and 15 minutes (T15) afterwards. The normal range for the intact PTH assay is 10 to 65 pg/mL.
DISCUSSION Remedial cervical exploration for persistent or recurrent PHPT presents both diagnostic and technical challenges.9-11,15,16 This study demonstrates the ability of real-time sSVS to localize residual abnormal parathyroid glands in patients with persistent or recurrent PHPT, after other noninvasive localization studies have failed. It also illustrates the advantages of this modified technique using a rapid PTH assay as an adjunct in the interventional radiology suite. The extension of the use of the rapid PTH assay from the operating room to the angiography suite provides the interventionist with critical information
Lebastchi et al
Vol. 220, No. 6, June 2015
Table 3.
Venous Sampling in Remedial Parathyroid Surgery
999
Comparison of Effectiveness of Localization by Preoperative Radiographic Study
Underwent study and operation, n Negative study, n (%) Positive study, n (%) Incorrect localization, n (%) Correct localization, n (%) Sensitivity, % Positive predictive value, %
MRI
Ultrasound
Sestamibi
CT
Real-time sSVS
6 4 (66.7) 2 (33.3) 1 (50) 1 (50) 20.0 50.0
21 15 (71.4) 6 (28.6) 2 (33.3) 4 (66.6) 21.5 66.7
24 16 (72.7) 8 (27.3) 5 (62.5) 3 (37.5) 15.8 37.5
22 9 (40.9) 13 (59.1) 4 (30.8) 9 (69.2) 50.0 69.2
31 3 (9.7) 28 (90.3) 3 (10.7) 25 (89.3) 86.2 92.6
Correct localization determined by percentage of positive studies with correct localization. sSVS, super selective venous sampling.
within 18 minutes of blood sampling. These “real-time” PTH values allow termination of the study when adequate information is obtained, or continuation if the study remains noninformative. Real-time data influenced venous sampling in essentially every patient because the PTH assay results directed the angiographer to sitespecific locations. Additionally, the interventional radiologist can obtain samples from areas with a subtle gradient in an effort to increase the fidelity by sampling smaller venous branches. Similarly, collecting additional blood samples from the area with the highest PTH gradient also increases the resolution of the study. Some authors refer to this sampling method as “super selective,” emphasizing the refinement of this modified technique compared with conventional SVS techniques.13 It is important to note that even when a specific gradient is identified, the contralateral side is always sampled because multisite disease can be encountered. Preoperative noninvasive imaging studies are always used before remedial parathyroid surgery. Despite technical advancements, the results of such studies can be noninformative in the reoperative setting, as they often fail to demonstrate adequate localization in a subset of patients. Real-time conventional SVS has been shown to be the most sensitive parathyroid localization procedure.16-18 Using a PTH gradient of 2-fold or greater to define a positive result, we achieved sensitivity and positive predictive values of 86% and 93%, respectively. Because all noninvasive imaging modalities focus on the cervical and mediastinal areas, they fail to identify PTH-secreting tissue outside of these regions, as exemplified in patient number 27. This patient presented with severe hypercalcemia 18 years after initial parathyroid surgery for parathyroid carcinoma and negative imaging studies including ultrasound, 4D CT, MRI, and sestamibi scans. Real-time sSVS demonstrated an elevated PTH gradient in the azygos vein, and a subsequent CT of the lower thorax demonstrated systemic presentation of parathyroid carcinoma on the left hemi-diaphragm. This diagnostic sSVS study, not only localized ectopic parathyroid
tissue where initial noninvasive localization procedures failed, it also obviated inappropriate cervical exploration. Despite these encouraging results, several limitations are noted. First, there are inherent limitations to studies evaluating a single institution’s experience. Because a single, dedicated interventional radiologist performed all sSVS, the application of the same principles universally by other, less experienced interventionists affects its reproducibility. Another limitation is that although we have complete short-term follow-up for all patients, long-term data are lacking. However, we contacted all of our patients and have received long-term follow-up in 18. Of these, 17 patients with mean follow-up of 96 months remain eucalcemic. The one patient with metastatic parathyroid carcinoma remains hypercalcemic. Lastly, sSVS is an expensive, invasive study subject to potential complications, including bleeding, infection, and patient discomfort. Complications were not encountered in the current series.
CONCLUSIONS Super selective venous sampling with real-time rapid PTH assay is a safe and effective method of preoperative localization before remedial parathyroid surgery. However, noninvasive studies should always be performed initially and, when deemed to be sufficient, sSVS is not required. In cases where noninvasive techniques fail to localize an abnormal parathyroid gland, sSVS can be used to increase operative success. We achieved cure in all patients (except one with metastatic parathyroid carcinoma) who were explored after sSVS, even if the study was negative, emphasizing the importance of surgeon’s experience for successful parathyroid surgery. Author Contributions Study conception and design: Lebastchi, Aruny, Udelsman Acquisition of data: Lebastchi, Donovan Analysis and interpretation of data: Aruny, Quinn, Callender, Carling, Udelsman
1000
Mazzaglia
J Am Coll Surg
Invited Commentary
Drafting of manuscript: Lebastchi Critical revision: Udelsman
18. Miller DL, Doppman JL, Krudy AG, et al. Localization of parathyroid adenomas in patients who have undergone surgery. Part II. Invasive procedures. Radiology 1987;162[1 Pt 1]:138e141.
REFERENCES 1. Silverberg SJ, Bilezikian JP. Evaluation and management of primary hyperparathyroidism. J Clin Endocrinol Metab 1996;81:2036e2040. 2. Adami S, Marcocci C, Gatti D. Epidemiology of primary hyperparathyroidism in Europe. J Bone Miner Res 2002;17: N18eN23. 3. Boonstra CE, Jackson CE. Serum calcium survey for hyperparathyroidism: results in 50,000 clinic patients. Am J Clin Pathol 1971;55:523e526. 4. Udelsman R, Aruny JE, Donovan PI, et al. Rapid parathyroid hormone analysis during venous localization. Ann Surg 2003; 237:714e719; discussion 719e721. 5. Udelsman R. Surgery in primary hyperparathyroidism: the patient without previous neck surgery. J Bone Miner Res 2002;17[Suppl 2]:N126eN132. 6. Udelsman R, Lin Z, Donovan P. The superiority of minimally invasive parathyroidectomy based on 1650 consecutive patients with primary hyperparathyroidism. Ann Surg 2011; 253:585e591. 7. Thompson GB, Grant CS, Perrier ND, et al. Reoperative parathyroid surgery in the era of sestamibi scanning and intraoperative parathyroid hormone monitoring. Arch Surg 1999; 134:699e704; discussion 705. 8. Saunders BD, Wainess RM, Dimick JB, et al. Who performs endocrine operations in the United States? Surgery 2003;134: 924e931; discussion 931. 9. Shen W, Duren M, Morita E, et al. Reoperation for persistent or recurrent primary hyperparathyroidism. Arch Surg 1996; 131:861e867; discussion 867e869. 10. Prescott JD, Udelsman R. Remedial operation for primary hyperparathyroidism. World J Surg 2009;33:2324e2334. 11. Udelsman R, Donovan PI. Remedial parathyroid surgery: changing trends in 130 consecutive cases. Ann Surg 2006;244:471e479. 12. Wang TS, Udelsman R. Remedial surgery for primary hyperparathyroidism. Adv Surg 2007;41:1e15. 13. Gimm O, Arnesson LG, Olofsson P, et al. Super-selective venous sampling in conjunction with quickPTH for patients with persistent primary hyperparathyroidism: report of five cases. Surg Today 2012;42:570e576. 14. Jones JJ, Brunaud L, Dowd CF, et al. Accuracy of selective venous sampling for intact parathyroid hormone in difficult patients with recurrent or persistent hyperparathyroidism. Surgery 2002;132:944e950; discussion 950e951. 15. Yen TW, Wang TS, Doffek KM, et al. Reoperative parathyroidectomy: an algorithm for imaging and monitoring of intraoperative parathyroid hormone levels that results in a successful focused approach. Surgery 2008;144:611e619; discussion 619e621. 16. Reidel MA, Schilling T, Graf S, et al. Localization of hyperfunctioning parathyroid glands by selective venous sampling in reoperation for primary or secondary hyperparathyroidism. Surgery 2006;140:907e913; discussion 913. 17. Sugg SL, Fraker DL, Alexander R, et al. Prospective evaluation of selective venous sampling for parathyroid hormone concentration in patients undergoing reoperations for primary hyperparathyroidism. Surgery 1993;114:1004e1009; discussion 1009e1010.
Invited Commentary Peter J Mazzaglia, Providence, RI
MD, FACS
The Endocrine Surgery and Interventional Radiology divisions at Yale University have summarized more than a decade’s experience with surgical management of persistent and recurrent hyperparathyroidism, specifically highlighting the role that selective venous sampling plays in preoperative localization of the enlarged parathyroid gland. The study expands on previously published work and examines patients for whom traditional modalities of parathyroid localization, such as sestamibi, ultrasound, and 4-dimensional CT, are negative or equivocal. They described the performance of super-selective venous sampling in 31 such patients while using real time rapid parathyroid hormone testing, which enabled the interventional radiologist to zero in on the anatomic location of the enlarged gland and provide detailed mapping for the operating surgeon. Using a cutoff gradient of twice baseline, 28 patients had positive studies. At surgery, the enlarged gland was correctly identified in 24 of 28 of these patients. Despite the 4 false positive and 3 false negative selective venous sampling studies, there was 100% operative success with use of the intraoperative parathyroid hormone assay. This illustrates the well-known adage that the most important localizing study is the localization of an experienced endocrine surgeon.1 The expertise, labor intensiveness, and cost required to perform the selective venous sampling procedure with simultaneous rapid parathyroid hormone testing would necessarily limit this type of evaluation to only the most specialized interventional radiology departments. Also, because of its invasive nature, this type of parathyroid localization should be limited to those patients with persistent or recurrent hyperparathyroidism in whom repeat noninvasive imaging, performed at a center of expertise in endocrine surgery, has been unrevealing. REFERENCE 1. Doppman JL. Reoperative parathyroid surgery; localization procedures. Prog Surg 1986;18:117e132.