- Email: [email protected]
Normocalcemic hyperparathyroidism: A Collaborative Endocrine Surgery Quality Improvement Program analysis
Surgery xxx (2019) 1e5
Contents lists available at ScienceDirect
Surgery journal homepage: www.elsevier.com/locate/surg
Normocalcemic hyperparathyroidism: A Collaborative Endocrine Surgery Quality Improvement Program analysis T.K. Pandian, MD, MPHa, Carrie C. Lubitz, MD, MPHa, Sarah H. Bird, BAb, Lindsay E. Kuo, MD, MBAc, Antonia E. Stephen, MDa,* a
Department of Surgery, Massachusetts General Hospital, Boston, MA Codman Center for Clinical Effectiveness in Surgery, Massachusetts General Hospital, Boston, MA c Department of Surgery, Temple University Lewis Katz School of Medicine, Philadelphia, PA b
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 2 June 2019 Available online xxx
Background: Normocalcemic primary hyperparathyroidism may be more challenging to cure compared with classical primary hyperparathyroidism. The aim of this study was to utilize a multi-institutional database to better characterize this condition. Methods: The Collaborative Endocrine Surgery Quality Improvement Program database was queried for all patients who underwent parathyroidectomy for sporadic primary hyperparathyroidism. Patient characteristics, operative details, pathology, and outcomes data were compared between patients with normocalcemic primary hyperparathyroidism and those with hypercalcemia. Results: Among 7,569 patients, 9.7% (733) were normocalcemic primary hyperparathyroidism. Mean age at surgery and sex were similar for normocalcemic primary hyperparathyroidism and primary hyperparathyroidism with hypercalcemia. The primary hyperparathyroidism with hypercalcemia cohort had a single parathyroid resected more frequently than the normocalcemic primary hyperparathyroidism group (73.3%% vs 47.5%, P < .05). Patients with normocalcemic primary hyperparathyroidism had a higher rate of subtotal (3.5 gland) resection (10.0% vs 4.7%, P < .05). Pathology reported a higher frequency of multigland hyperplasia in the normocalcemic primary hyperparathyroidism cohort (43.1% vs 21.9%, P <.05). In the normocalcemic primary hyperparathyroidism cohort, 47 patients (6.4%) underwent remedial surgery compared with 307 patients (4.5%) with primary hyperparathyroidism with hypercalcemia (P < .05). The rate of clinical concern for persistent hyperparathyroidism was similar between the 2 groups (P ¼ .09) but not reported in 25% overall. Conclusion: Patients with normocalcemic primary hyperparathyroidism have higher rates of multigland disease and remedial surgery compared with primary hyperparathyroidism with hypercalcemia. © 2019 Published by Elsevier Inc.
Introduction The diagnosis of primary hyperparathyroidism (PHPT) has classically been defined by hypercalcemia with inappropriately elevated parathyroid hormone (PTH) levels. However, variations in the biochemical profile of patients with primary hyperparathyroidism have been described as early as the 1950s.1e3 It has been postulated that PHPT may go through various phases with an early stage in which the PTH level is elevated but hypercalcemia has not
Presented at the 2019 Annual Meeting of the American Association of Endocrine Surgeons. * Reprint requests: Antonia E. Stephen, MD, Department of Surgery, Massachusetts General Hospital, Yawkey Center for Outpatient Care, 32 Fruit Street, Boston, MA 02114. E-mail address: [email protected] (A.E. Stephen). https://doi.org/10.1016/j.surg.2019.06.043 0039-6060/© 2019 Published by Elsevier Inc.
yet developed.4 More recently, however, it is believed that this state of elevated PTH and normal calcium levels may actually be a distinct entity and has been termed normocalcemic primary hyperparathyroidism (nPHPT).3e7 Normocalcemic primary hyperparathyroidism is defined by persistently normal total and ionized serum calcium levels in the presence of high serum PTH, without the presence of etiologies responsible for secondary elevation of PTH itself.3,8 It was formally recognized in the Third International Workshop of the Management of Primary Hyperparathyroidism in 2008.9 The American Association of Endocrine Surgeons (AAES) guidelines for primary hyperparathyroidism recommend measurement of ionized calcium levels in any patient suspected to have nPHPT5 to exclude patients who have hypercalcemia with normal serum total calcium levels. Current literature and knowledge on this particular condition is limited to single-institution experiences. As a result, data are
2
T.K. Pandian et al. / Surgery xxx (2019) 1e5
somewhat variable. In general, it appears that patients with nPHPT can still develop bone and renal sequelae similar to those with PHPT,4,7,10e12 thus necessitating discussion regarding surgery. In addition, nPHPT patients may be more likely to exhibit multigland disease,13e15 which may alter proposed operative approaches for this condition. However, not all data are fully consistent in these findings.3,4 These discrepancies may be due to the small cohorts of patients included in the available single-institution studies. Based on this, the aim of our study was to utilize a multi-institutional database to better characterize this condition and define the rate of multigland disease in this cohort.
Methods The Collaborative Endocrine Surgery Quality Improvement Program (CESQIP) database was created in 2012 by members of the AAES. The database is a Qualified Clinical Data Registry approved by the Centers for Medicare and Medicaid Services. It focuses on more than 300 perioperative variables specific to endocrine surgery for the purpose of quality improvement. Permission to use the CESQIP participant use file for parathyroid was granted through acceptance of a project proposal by the CESQIP Committee of the AAES and submission of a signed data use agreement. All data were deidentified. The CESQIP database from all 48 participating institutions from January 2014 to April 2018 was queried for patients who underwent parathyroidectomy for sporadic primary hyperparathyroidism. Patients with familial hyperparathyroidism and secondary or tertiary hyperparathyroidism are categorized separately in CESQIP and were excluded. Preoperative calcium levels were used to create two groupsdthose with normal calcium levels were classified as the normocalcemic primary hyperparathyroidism (nPHPT) group and those with elevated calcium levels as the classical primary hyperparathyroidism (PHPT) group. All patients without a recorded preoperative calcium level were excluded. Of note, calcium, PTH and 25-OH-vitamin D levels in CESQIP are reported as categorical values (normal, high, very high) based on each institutions’ laboratory reference range. The rationale for categorization of these variables into ranges of severity rather than on utilization of continuous values, is to allow for comparisons across institutions with diversity in their laboratory reference ranges. In CESQIP, preoperative values are based on the most recent laboratory value within 30 days of the operative date. Patient characteristics,
operative details, pathology, and outcomes data were compared between patients with nPHPT and those with PHPT. Persistent disease was the main focus of our outcome analysis and is recorded in CESQIP as “clinical concern for persistent hyperparathyroidism” and is determined based on physician assessment and interpretation of postoperative laboratory studies within the medical record of each patient. As is the case with preoperative values, numeric values for postoperative calcium and PTH are not available in the database. Statistical analysis All data were analyzed in terms of descriptive statistics and reported as proportions. We used Pearson’s c2 test for categorical variables with a significance level set at 0.05. Statistical analyses were performed using STATA (StataCorp. 2013. Stata Statistical Software: Release 13. StataCorp, College Station, TX, USA). Results In total, 7,634 patients who underwent parathyroidectomy for sporadic primary hyperparathyroidism were identified in CESQIP. Among these, 65 were excluded because of a lack of recorded preoperative calcium values. Of the remaining 7,569 patients, 9.7% (733) were nPHPT. As presented in Table I, patient characteristics were similar for nPHPT and PHPT groups in terms of age and body mass index. Preoperative PTH levels were not elevated in 13.2% vs 11.3% and high in 85.0% vs 87.7% for nPHPT and PHPT, respectively. Vitamin D levels (25-OH vitamin D) were low in 25.6% vs 39.1% and normal in 58.7% vs 48.4% for nPHPT and PHPT, respectively. Ionized calcium levels were not recorded in 72% of patients in the overall cohort, and thus this variable was not included in analysis. In the nPHPT cohort, 47 patients (6.4%) underwent remedial surgery compared with 307 patients (4.5%) with PHPT (P < .05). The PHPT cohort had a single parathyroid resected more frequently than the nPHPT group (73.3% vs 47.5%, P < .05). Similarly, patients with nPHPT had a higher rate of subtotal (3.5 gland) resection (10.0% vs 4.7%, P < .05). Pathology reported a higher frequency of multigland hyperplasia in the nPHPT cohort (43.1% vs 21.9%, P < .05). Operative characteristics are summarized in Table II. The rate of clinical concern for persistent hyperparathyroidism within 30 days postoperatively was similar between the 2 groups (4.0% vs 2.8%, P ¼ .09) but not
Table I Patient and preoperative characteristics
Age (y), mean ± SD Female sex Race Asian Black Hispanic Nonblack Unknown BMI > 40 Preoperative PTH Not elevated Elevated Not recorded Preoperative 25-OH-vitamin D Low Normal Not recorded
nPHPT (n ¼ 733)
PHPT (n ¼ 6,836)
P value
61.3 ± 11.9 81.4% (597)
60.4 ± 13.4 77.3% (5,286)
.08 < .05 < .05
1.9% (14) 4.1% (30) 3.3% (24) 84.2% (617) 6.5% (48) 8.0% (59)
1.2% (85) 8.3% (570) 6.8% (465) 76.1% (5,204) 7.5% (512) 9.5% (649)
13.2% (97) 85.0% (623) 1.8% (13)
11.3% (774) 87.7% (5,997) 1.0% (65)
25.6% (188) 58.7% (430) 15.7% (115)
39.1% (2,675) 48.4% (3,312) 12.4% (849)
.26 < .05
< .05
nPHPT, normocalcemic primary hyperparathyroidism; PHPT, hypercalcemic primary hyperparathyroidism.
T.K. Pandian et al. / Surgery xxx (2019) 1e5
3
Table II Operative characteristics nPHPT (n ¼ 733) Initial versus remedial surgery Initial Remedial Number of parathyroids removed 0 or Unknown 1e3 3.5 4 Pathology Unknown Multigland disease Solitary adenoma Cancer or atypical adenoma Normal
Calcium level Low Normal High Unknown PTH level Low Normal High Unknown
PHPT (n ¼ 6,836)
2.0% (15) 48.2% (353) 1.6% (12) 48.2% (353)
1.2% (83) 48.4% (3,306) 2.8% (188) 47.7% (3,259)
2.3% (17) 34.5% (253) 8.7% (64) 54.4% (399)
1.7% (119) 34.3% (2,343) 7.3% (499) 56.7% (3,875)
P value < .05
93.6% (686) 6.4% (47)
95.5% (6,529) 4.5% (307)
1.7% (12) 85.1% (623) 9.4% (69) 3.9% (29)
0.8% (52) 93.3% (6,378) 4.7% (318) 1.4% (88)
2.2% (16) 37.9% (278) 50.1% (367) 7.8% (57) 2.0% (15)
1.0% (71) 21.1% (1,439) 75.1% (5,134) 2.3% (155) 0.5% (37)
< .05
< .05
Table III Postoperative longitudinal (>30 days) laboratory data nPHPT (n ¼ 733)
PHPT (n ¼ 6,836)
P value < .05
.33
reported in 25% (n ¼ 1,893) overall. We were unable to correlate clinical concern for persistent hyperparathyroidism with laboratory data in the immediate postoperative period because postoperative calcium and PTH levels within 30 days postoperatively were not recorded in the database. Postoperative calcium and PTH levels beyond 30 days were unknown for 47.7% and 56.5% of the entire cohort, respectively. The available data, however, are summarized in Table III. As presented, there was a small difference in calcium, and no difference in PTH between nPHPT and PHPT groups. A sensitivity analysis excluding patients with normal and missing preoperative PTH values as well as low and missing preoperative vitamin D levels was also performed. This confirmed higher rates of subtotal gland resection, higher frequency of multigland hyperplasia, and a similar rate of persistent hyperparathyroidism within 30 days postoperatively for the nPHPT versus PHPT patients. The proportion of patients who underwent remedial surgery was higher in the nPHPT group compared with PHPT, however this was no longer statistically significant (6.2% vs 5.0%, P ¼ .32). Discussion In this large, multi-institutional, retrospective study, we found that patients with nPHPT are more often characterized by multigland disease on pathology and have a single parathyroid gland resected less often than patients with PHPT. In addition, more of the nPHPT patients in this data set underwent remedial surgery for hyperparathyroidism. To date, this is the largest reported experience with nPHPT in the literature. Earlier single-institution studies have also suggested that nPHPT is more often associated with multigland disease11,13e15 and that these patients had smaller parathyroid gland size13 or
weight.3 Multigland disease may decrease the success of preoperative localization15e17 and has implications for the technical difficulty of the operation and need for bilateral cervical exploration. It is unclear why exactly patients with nPHPT exhibit this pathologic pattern, but there is some speculation that PTH resistance in bone and kidney may lead to autonomous parathyroid function.12 Several authors have recommended using intraoperative PTH (IOPTH) in nPHPT patients to help guide appropriate resection.13,18,19 It should be noted that at least one study has shown that PTH value decay intraoperatively can be slowed in patients with a mild biochemical profile or multigland disease.19 Accordingly, interpretation of IOPTH should be considered in this context. Regardless of the etiology, our data corroborate reported findings of multigland disease, and we therefore agree that bilateral exploration be considered for all patients with nPHPT.13,14 If more than one preoperative imaging modality provides concordant evidence for single-gland disease, then a focused exploration with IOPTH can be utilized.20 However, even if bilateral cervical exploration is not initially planned, surgeons should have a low threshold to convert from a focused approach to four-gland exploration, especially if intraoperative PTH values suggest that single gland resection is not consistent with biochemical cure. Normocalcemic patients seemingly underwent remedial parathyroid surgery more often than the PHPT cohort. This was not statistically significant on sensitivity analysis; however, the absolute difference (1.9% on initial analysis vs 1.2% on sensitivity analysis) was similar. Although the rate of clinical concern for postoperative hyperparathyroidism within 30 days was similar for our 2 groups suggesting similar cure rates, 25% of patients had no information on this particular variable. Additionally, practice patterns for postoperative calcium and vitamin D supplementation are variable, so it is possible that patients who are routinely placed on these agents after surgery may have a transient biochemical picture of persistent hyperparathyroidism until supplementation is discontinued. Literature on surgical outcomes has suggested that there is no major difference in the cure rate of hyperparathyroidism between nPHPT and PHPT.3,13e15,21 These data, however, were based on single-institution studies, many of which utilized intraoperative PTH measurements. The CESQIP database includes data from 48 institutions with a more variable endocrine surgical practice and therefore may be more representative of national trends. A higher rate of remedial surgery suggests that a higher proportion of nPHPT patients have persistent or recurrent disease. This may be explained by more patients with multigland disease that are missed during an initial operation and further underscore the importance of considering 4-gland exploration and utilizing
4
T.K. Pandian et al. / Surgery xxx (2019) 1e5
intraoperative PTH monitoring in these patients. We did attempt to correlate clinical concern for postoperative hyperparathyroidism with laboratory data; however, a significant amount of information was not reported in CESQIP. Although Table III highlights a statistical difference between nPHPT and PHPT calcium levels beyond 30 days postoperatively, the absolute differences are small and must be considered in the context of nearly 50% of data being unknown. Our retrospective study is not without limitations. First, although large national databases such as this rely on standardized definitions for data extraction, these definitions in the early stages of an initiative such as CESQIP may require some refinement. In addition, unlike the National Surgery Quality Improvement Program that relies on a dedicated abstractor to record data, operative information can be entered into the CESQIP registry by surgeons, introducing some potential bias. Any data misclassification is expected to be nondifferential. Second, there was incomplete capture of variables of interest such as postoperative clinical concern for hyperparathyroidism, preoperative ionized calcium, bone disease, nephrolithiasis, renal function and others such as those identified in Tables I and II as “Unknown” or “Not Recorded.” Preoperative ionized calcium levels would have been especially helpful in confirming nPHPT patients. We suspect the lack of this laboratory value in the database likely reflects a combination of incomplete data entry and healthcare practitioners not obtaining it preoperatively. Long-term (>30 days postoperatively) data regarding persistent or recurrent hyperparathyroidism were not recorded in >40% of the cohort. Third, although we included only patients with sporadic primary hyperparathyroidism in our cohort, a significant portion of both the nPHPT and PHPT patients were noted to have low levels of 25OH-vitamin D levels and some had normal PTH. As noted in the Methods section of this report, preoperative lab values utilized in CESQIP are reported based on the most recent value within 30 days of surgery, and thus the low vitamin D and normal PTH values may be due to abstraction inconsistencies rather than patients with secondary hyperparathyroidism or those misclassified as having hyperparathyroidism. We also performed a sensitivity analysis excluding these patients, supporting abstraction consistencies. Unfortunately, glomerular filtration rate information is not available in CESQIP, which would be an additional way to identify secondary hyperparathyroidism patients miscategorized into the sporadic hyperparathyroidism group. Despite these limitations, we present the largest experience with nPHPT to date. It is clear that patients with nPHPT have higher rates of multigland disease compared with PHPT. In this data set, postoperative outcome data were limited by missing data and lack of long-term results. These data support consideration for multigland exploration, utilization of intraoperative PTH to guide surgery, and the need for longer follow-up in patients with nPHPT. Funding/Support No grant funding was utilized with respect to this research. Conflict of interest/Disclosure The CESQIP and the hospitals participating in CESQIP are the source of the data used herein. They have not verified and are not
responsible for the statistical validity of the data analysis or the conclusions derived by the authors. The conclusions, findings, and opinions expressed by the authors do not necessarily reflect the official position of the AAES or CESQIP. Use of CESQIP data does not imply endorsement by AAES or CESQIP. The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
References 1. Forster J, Monchik JM, Martin HF. A comparative study of serum ultrafiltrable, ionized, and total calcium in the diagnosis of primary hyperparathyroidism in patients with intermittent or no elevation in total calcium. Surgery. 1988;104: 1137e1142. 2. Wills MR, Pak CY, Hammond WG, et al. Normocalcemic primary hyperparathyroidism. Am J Med. 1969;47:384e391. 3. Kiriakopoulos A, Petralias A, Linos D. Classic primary hyperparathyroidism versus normocalcemic and normohormonal variants: Do they really differ? World J Surg. 2018;42:992e997. 4. Lowe H, McMahon DJ, Rubin MR, et al. Normocalcemic primary hyperparathyroidism: Further characterization of a new clinical phenotype. J Clin Endocrinol Metab. 2007;92:3001e3005. 5. Campbell MJ. The definitive management of primary hyperparathyroidism who needs an operation? JAMA. 2017;317:959e968. 6. Siperstein AE, Shen W, Chan AK, et al. Normocalcemic hyperparathyroidism. Biochemical and symptom profiles before and after surgery. Arch Surg. 1992;127: 1157e1156; discussion 1161e1163. 7. Stuart HC, Harvey A, Pasieka JL. Normocalcemic hyperparathyroidism: Preoperatively a disease, postoperatively cured? Am J Surg. 2014;207: 673e680. 8. Cusano NE, Maalouf NM, Wang PY, et al. Normocalcemic hyperparathyroidism and hypoparathyroidism in two community-based nonreferral populations. J Clin Endocrinol Metab. 2013;98:2734e2741. 9. Bilezikian JP, Brandi ML, Eastell R, et al. Guidelines for the management of asymptomatic primary hyperparathyroidism: Summary statement from the fourth international workshop. J Clin Endocrinol Metab. 2014;99: 3561e3569. 10. Pawlowska M, Cusano NE. An overview of normocalcemic primary hyperparathyroidism. Curr Opin Endocrinol Diabetes Obes. 2015;22:413e421. 11. Koumakis E, Souberbielle JC, Sarfati E, et al. Bone mineral density evolution after successful parathyroidectomy in patients with normocalcemic primary hyperparathyroidism. J Clin Endocrinol Metab. 2013;98: 3213e3220. 12. Pierreux J, Bravenboer B. Normocalcemic primary hyperparathyroidism: A comparison with the hypercalcemic form in a tertiary referral population. Horm Metab Res. 2018;50:797e802. 13. Trinh G, Rettig E, Noureldine SI, et al. Surgical management of normocalcemic primary hyperparathyroidism and the impact of intraoperative parathyroid hormone testing on outcome. Otolaryngol Head Neck Surg. 2018;159: 630e637. 14. Lim JY, Herman MC, Bubis L, et al. Differences in single gland and multigland disease are seen in low biochemical profile primary hyperparathyroidism. Surgery. 2017;161:70e77. 15. Traini E, Bellantone R, Tempera SE, et al. Is parathyroidectomy safe and effective in patients with normocalcemic primary hyperparathyroidism? Langenbeck’s Arch Surg. 2017;402:344e345. 16. Siperstein A, Berber E, Barbosa GF, et al. Predicting the success of limited exploration for primary hyperparathyroidism using ultrasound, sestamibi, and intraoperative parathyroid hormone: Analysis of 1158 cases. Ann Surg. 2008;248:420e426. 17. Siprov a H, Frys ak Z, Sou cek M. Primary hyperparathyroidism, with a focus on management of the normocalcemic form: To treat or not to treat? Endocr Pract. 2016;22:294e301. 18. Wade TJ, Yen TWF, Amin AL, et al. Surgical management of normocalcemic primary hyperparathyroidism. World J Surg. 2012;36:761e766. 19. Schneider DF, Burke JF, Ojomo KA, et al. Multigland disease and slower decline in intraoperative PTH characterize mild primary hyperparathyroidism. Ann Surg Oncol. 2013;20:4205e4211. 20. Kebebew E, Hwang J, Reiff E, et al. Predictors of single-gland vs multigland parathyroid disease in primary hyperparathyroidism: A simple and accurate scoring model. Arch Surg. 2006;141:777e782. 21. Sho S, Kuo EJ, Chen AC, et al. Biochemical and skeletal outcomes of parathyroidectomy for normocalcemic (incipient) primary hyperparathyroidism. Ann Surg Oncol. 2019;26:539e546.
T.K. Pandian et al. / Surgery xxx (2019) 1e5
5
Discussion Dr Christopher R. McHenry (Cleveland, OH): How did you define multiglandular disease? Was it based on persistent PTH elevation intraoperatively? Or was this based on visible appearance of abnormal glands? Dr T.K. Pandian: That is a great question. Unfortunately, with this large database, it is not granular enough to determine exactly how each institution defined multigland disease. The abstractors from our institution used pathology results. Basically, if a pathologist noted hyperplasia, they would categorize that as multigland disease. But if a surgeon entered that information, it could potentially be different. Dr Janice Pasieka (Calgary, AB, Canada): Do you have the preoperative localization results for your 2 groups? If they had more multigland disease, did they have a higher rate of nonlocalization preoperatively, and therefore were at risk of needing a more extensive operation? Dr T.K. Pandian: That’s a great point. Unfortunately, we don’t have those data. Some of that information is available
in CESQIP, so it’s certainly something we can go back and look at. Dr Herb Chen (Birmingham, AL): I may have misread this, but when you showed your two groups, was there a higher incidence of suspected cancer in the lower calcium group? We normally think of parathyroid cancer as being associated with a very high calcium and high PTH, so that doesn’t go together very well. Dr T.K. Pandian: Yes, absolutely. It was a curious finding. I don’t know that we have a great explanation for that. Some of this may be misclassification errors. Dr David Schneider (Madison, WI): Great use of CESQIP data. One question I had is whether you have the capability to examine this by institution in a deidentified manner. I suspect in this particular disease process there’s wide variability in who even gets surgery, and some centers may choose not to operate on these patients at all. So it might be interesting to look at that variability, and it might add something to your paper as well. Dr T.K. Pandian: Thank you. That’s a good suggestion.
Contents lists available at ScienceDirect
Surgery journal homepage: www.elsevier.com/locate/surg
Normocalcemic hyperparathyroidism: A Collaborative Endocrine Surgery Quality Improvement Program analysis T.K. Pandian, MD, MPHa, Carrie C. Lubitz, MD, MPHa, Sarah H. Bird, BAb, Lindsay E. Kuo, MD, MBAc, Antonia E. Stephen, MDa,* a
Department of Surgery, Massachusetts General Hospital, Boston, MA Codman Center for Clinical Effectiveness in Surgery, Massachusetts General Hospital, Boston, MA c Department of Surgery, Temple University Lewis Katz School of Medicine, Philadelphia, PA b
a r t i c l e i n f o
a b s t r a c t
Article history: Accepted 2 June 2019 Available online xxx
Background: Normocalcemic primary hyperparathyroidism may be more challenging to cure compared with classical primary hyperparathyroidism. The aim of this study was to utilize a multi-institutional database to better characterize this condition. Methods: The Collaborative Endocrine Surgery Quality Improvement Program database was queried for all patients who underwent parathyroidectomy for sporadic primary hyperparathyroidism. Patient characteristics, operative details, pathology, and outcomes data were compared between patients with normocalcemic primary hyperparathyroidism and those with hypercalcemia. Results: Among 7,569 patients, 9.7% (733) were normocalcemic primary hyperparathyroidism. Mean age at surgery and sex were similar for normocalcemic primary hyperparathyroidism and primary hyperparathyroidism with hypercalcemia. The primary hyperparathyroidism with hypercalcemia cohort had a single parathyroid resected more frequently than the normocalcemic primary hyperparathyroidism group (73.3%% vs 47.5%, P < .05). Patients with normocalcemic primary hyperparathyroidism had a higher rate of subtotal (3.5 gland) resection (10.0% vs 4.7%, P < .05). Pathology reported a higher frequency of multigland hyperplasia in the normocalcemic primary hyperparathyroidism cohort (43.1% vs 21.9%, P <.05). In the normocalcemic primary hyperparathyroidism cohort, 47 patients (6.4%) underwent remedial surgery compared with 307 patients (4.5%) with primary hyperparathyroidism with hypercalcemia (P < .05). The rate of clinical concern for persistent hyperparathyroidism was similar between the 2 groups (P ¼ .09) but not reported in 25% overall. Conclusion: Patients with normocalcemic primary hyperparathyroidism have higher rates of multigland disease and remedial surgery compared with primary hyperparathyroidism with hypercalcemia. © 2019 Published by Elsevier Inc.
Introduction The diagnosis of primary hyperparathyroidism (PHPT) has classically been defined by hypercalcemia with inappropriately elevated parathyroid hormone (PTH) levels. However, variations in the biochemical profile of patients with primary hyperparathyroidism have been described as early as the 1950s.1e3 It has been postulated that PHPT may go through various phases with an early stage in which the PTH level is elevated but hypercalcemia has not
Presented at the 2019 Annual Meeting of the American Association of Endocrine Surgeons. * Reprint requests: Antonia E. Stephen, MD, Department of Surgery, Massachusetts General Hospital, Yawkey Center for Outpatient Care, 32 Fruit Street, Boston, MA 02114. E-mail address: [email protected] (A.E. Stephen). https://doi.org/10.1016/j.surg.2019.06.043 0039-6060/© 2019 Published by Elsevier Inc.
yet developed.4 More recently, however, it is believed that this state of elevated PTH and normal calcium levels may actually be a distinct entity and has been termed normocalcemic primary hyperparathyroidism (nPHPT).3e7 Normocalcemic primary hyperparathyroidism is defined by persistently normal total and ionized serum calcium levels in the presence of high serum PTH, without the presence of etiologies responsible for secondary elevation of PTH itself.3,8 It was formally recognized in the Third International Workshop of the Management of Primary Hyperparathyroidism in 2008.9 The American Association of Endocrine Surgeons (AAES) guidelines for primary hyperparathyroidism recommend measurement of ionized calcium levels in any patient suspected to have nPHPT5 to exclude patients who have hypercalcemia with normal serum total calcium levels. Current literature and knowledge on this particular condition is limited to single-institution experiences. As a result, data are
2
T.K. Pandian et al. / Surgery xxx (2019) 1e5
somewhat variable. In general, it appears that patients with nPHPT can still develop bone and renal sequelae similar to those with PHPT,4,7,10e12 thus necessitating discussion regarding surgery. In addition, nPHPT patients may be more likely to exhibit multigland disease,13e15 which may alter proposed operative approaches for this condition. However, not all data are fully consistent in these findings.3,4 These discrepancies may be due to the small cohorts of patients included in the available single-institution studies. Based on this, the aim of our study was to utilize a multi-institutional database to better characterize this condition and define the rate of multigland disease in this cohort.
Methods The Collaborative Endocrine Surgery Quality Improvement Program (CESQIP) database was created in 2012 by members of the AAES. The database is a Qualified Clinical Data Registry approved by the Centers for Medicare and Medicaid Services. It focuses on more than 300 perioperative variables specific to endocrine surgery for the purpose of quality improvement. Permission to use the CESQIP participant use file for parathyroid was granted through acceptance of a project proposal by the CESQIP Committee of the AAES and submission of a signed data use agreement. All data were deidentified. The CESQIP database from all 48 participating institutions from January 2014 to April 2018 was queried for patients who underwent parathyroidectomy for sporadic primary hyperparathyroidism. Patients with familial hyperparathyroidism and secondary or tertiary hyperparathyroidism are categorized separately in CESQIP and were excluded. Preoperative calcium levels were used to create two groupsdthose with normal calcium levels were classified as the normocalcemic primary hyperparathyroidism (nPHPT) group and those with elevated calcium levels as the classical primary hyperparathyroidism (PHPT) group. All patients without a recorded preoperative calcium level were excluded. Of note, calcium, PTH and 25-OH-vitamin D levels in CESQIP are reported as categorical values (normal, high, very high) based on each institutions’ laboratory reference range. The rationale for categorization of these variables into ranges of severity rather than on utilization of continuous values, is to allow for comparisons across institutions with diversity in their laboratory reference ranges. In CESQIP, preoperative values are based on the most recent laboratory value within 30 days of the operative date. Patient characteristics,
operative details, pathology, and outcomes data were compared between patients with nPHPT and those with PHPT. Persistent disease was the main focus of our outcome analysis and is recorded in CESQIP as “clinical concern for persistent hyperparathyroidism” and is determined based on physician assessment and interpretation of postoperative laboratory studies within the medical record of each patient. As is the case with preoperative values, numeric values for postoperative calcium and PTH are not available in the database. Statistical analysis All data were analyzed in terms of descriptive statistics and reported as proportions. We used Pearson’s c2 test for categorical variables with a significance level set at 0.05. Statistical analyses were performed using STATA (StataCorp. 2013. Stata Statistical Software: Release 13. StataCorp, College Station, TX, USA). Results In total, 7,634 patients who underwent parathyroidectomy for sporadic primary hyperparathyroidism were identified in CESQIP. Among these, 65 were excluded because of a lack of recorded preoperative calcium values. Of the remaining 7,569 patients, 9.7% (733) were nPHPT. As presented in Table I, patient characteristics were similar for nPHPT and PHPT groups in terms of age and body mass index. Preoperative PTH levels were not elevated in 13.2% vs 11.3% and high in 85.0% vs 87.7% for nPHPT and PHPT, respectively. Vitamin D levels (25-OH vitamin D) were low in 25.6% vs 39.1% and normal in 58.7% vs 48.4% for nPHPT and PHPT, respectively. Ionized calcium levels were not recorded in 72% of patients in the overall cohort, and thus this variable was not included in analysis. In the nPHPT cohort, 47 patients (6.4%) underwent remedial surgery compared with 307 patients (4.5%) with PHPT (P < .05). The PHPT cohort had a single parathyroid resected more frequently than the nPHPT group (73.3% vs 47.5%, P < .05). Similarly, patients with nPHPT had a higher rate of subtotal (3.5 gland) resection (10.0% vs 4.7%, P < .05). Pathology reported a higher frequency of multigland hyperplasia in the nPHPT cohort (43.1% vs 21.9%, P < .05). Operative characteristics are summarized in Table II. The rate of clinical concern for persistent hyperparathyroidism within 30 days postoperatively was similar between the 2 groups (4.0% vs 2.8%, P ¼ .09) but not
Table I Patient and preoperative characteristics
Age (y), mean ± SD Female sex Race Asian Black Hispanic Nonblack Unknown BMI > 40 Preoperative PTH Not elevated Elevated Not recorded Preoperative 25-OH-vitamin D Low Normal Not recorded
nPHPT (n ¼ 733)
PHPT (n ¼ 6,836)
P value
61.3 ± 11.9 81.4% (597)
60.4 ± 13.4 77.3% (5,286)
.08 < .05 < .05
1.9% (14) 4.1% (30) 3.3% (24) 84.2% (617) 6.5% (48) 8.0% (59)
1.2% (85) 8.3% (570) 6.8% (465) 76.1% (5,204) 7.5% (512) 9.5% (649)
13.2% (97) 85.0% (623) 1.8% (13)
11.3% (774) 87.7% (5,997) 1.0% (65)
25.6% (188) 58.7% (430) 15.7% (115)
39.1% (2,675) 48.4% (3,312) 12.4% (849)
.26 < .05
< .05
nPHPT, normocalcemic primary hyperparathyroidism; PHPT, hypercalcemic primary hyperparathyroidism.
T.K. Pandian et al. / Surgery xxx (2019) 1e5
3
Table II Operative characteristics nPHPT (n ¼ 733) Initial versus remedial surgery Initial Remedial Number of parathyroids removed 0 or Unknown 1e3 3.5 4 Pathology Unknown Multigland disease Solitary adenoma Cancer or atypical adenoma Normal
Calcium level Low Normal High Unknown PTH level Low Normal High Unknown
PHPT (n ¼ 6,836)
2.0% (15) 48.2% (353) 1.6% (12) 48.2% (353)
1.2% (83) 48.4% (3,306) 2.8% (188) 47.7% (3,259)
2.3% (17) 34.5% (253) 8.7% (64) 54.4% (399)
1.7% (119) 34.3% (2,343) 7.3% (499) 56.7% (3,875)
P value < .05
93.6% (686) 6.4% (47)
95.5% (6,529) 4.5% (307)
1.7% (12) 85.1% (623) 9.4% (69) 3.9% (29)
0.8% (52) 93.3% (6,378) 4.7% (318) 1.4% (88)
2.2% (16) 37.9% (278) 50.1% (367) 7.8% (57) 2.0% (15)
1.0% (71) 21.1% (1,439) 75.1% (5,134) 2.3% (155) 0.5% (37)
< .05
< .05
Table III Postoperative longitudinal (>30 days) laboratory data nPHPT (n ¼ 733)
PHPT (n ¼ 6,836)
P value < .05
.33
reported in 25% (n ¼ 1,893) overall. We were unable to correlate clinical concern for persistent hyperparathyroidism with laboratory data in the immediate postoperative period because postoperative calcium and PTH levels within 30 days postoperatively were not recorded in the database. Postoperative calcium and PTH levels beyond 30 days were unknown for 47.7% and 56.5% of the entire cohort, respectively. The available data, however, are summarized in Table III. As presented, there was a small difference in calcium, and no difference in PTH between nPHPT and PHPT groups. A sensitivity analysis excluding patients with normal and missing preoperative PTH values as well as low and missing preoperative vitamin D levels was also performed. This confirmed higher rates of subtotal gland resection, higher frequency of multigland hyperplasia, and a similar rate of persistent hyperparathyroidism within 30 days postoperatively for the nPHPT versus PHPT patients. The proportion of patients who underwent remedial surgery was higher in the nPHPT group compared with PHPT, however this was no longer statistically significant (6.2% vs 5.0%, P ¼ .32). Discussion In this large, multi-institutional, retrospective study, we found that patients with nPHPT are more often characterized by multigland disease on pathology and have a single parathyroid gland resected less often than patients with PHPT. In addition, more of the nPHPT patients in this data set underwent remedial surgery for hyperparathyroidism. To date, this is the largest reported experience with nPHPT in the literature. Earlier single-institution studies have also suggested that nPHPT is more often associated with multigland disease11,13e15 and that these patients had smaller parathyroid gland size13 or
weight.3 Multigland disease may decrease the success of preoperative localization15e17 and has implications for the technical difficulty of the operation and need for bilateral cervical exploration. It is unclear why exactly patients with nPHPT exhibit this pathologic pattern, but there is some speculation that PTH resistance in bone and kidney may lead to autonomous parathyroid function.12 Several authors have recommended using intraoperative PTH (IOPTH) in nPHPT patients to help guide appropriate resection.13,18,19 It should be noted that at least one study has shown that PTH value decay intraoperatively can be slowed in patients with a mild biochemical profile or multigland disease.19 Accordingly, interpretation of IOPTH should be considered in this context. Regardless of the etiology, our data corroborate reported findings of multigland disease, and we therefore agree that bilateral exploration be considered for all patients with nPHPT.13,14 If more than one preoperative imaging modality provides concordant evidence for single-gland disease, then a focused exploration with IOPTH can be utilized.20 However, even if bilateral cervical exploration is not initially planned, surgeons should have a low threshold to convert from a focused approach to four-gland exploration, especially if intraoperative PTH values suggest that single gland resection is not consistent with biochemical cure. Normocalcemic patients seemingly underwent remedial parathyroid surgery more often than the PHPT cohort. This was not statistically significant on sensitivity analysis; however, the absolute difference (1.9% on initial analysis vs 1.2% on sensitivity analysis) was similar. Although the rate of clinical concern for postoperative hyperparathyroidism within 30 days was similar for our 2 groups suggesting similar cure rates, 25% of patients had no information on this particular variable. Additionally, practice patterns for postoperative calcium and vitamin D supplementation are variable, so it is possible that patients who are routinely placed on these agents after surgery may have a transient biochemical picture of persistent hyperparathyroidism until supplementation is discontinued. Literature on surgical outcomes has suggested that there is no major difference in the cure rate of hyperparathyroidism between nPHPT and PHPT.3,13e15,21 These data, however, were based on single-institution studies, many of which utilized intraoperative PTH measurements. The CESQIP database includes data from 48 institutions with a more variable endocrine surgical practice and therefore may be more representative of national trends. A higher rate of remedial surgery suggests that a higher proportion of nPHPT patients have persistent or recurrent disease. This may be explained by more patients with multigland disease that are missed during an initial operation and further underscore the importance of considering 4-gland exploration and utilizing
4
T.K. Pandian et al. / Surgery xxx (2019) 1e5
intraoperative PTH monitoring in these patients. We did attempt to correlate clinical concern for postoperative hyperparathyroidism with laboratory data; however, a significant amount of information was not reported in CESQIP. Although Table III highlights a statistical difference between nPHPT and PHPT calcium levels beyond 30 days postoperatively, the absolute differences are small and must be considered in the context of nearly 50% of data being unknown. Our retrospective study is not without limitations. First, although large national databases such as this rely on standardized definitions for data extraction, these definitions in the early stages of an initiative such as CESQIP may require some refinement. In addition, unlike the National Surgery Quality Improvement Program that relies on a dedicated abstractor to record data, operative information can be entered into the CESQIP registry by surgeons, introducing some potential bias. Any data misclassification is expected to be nondifferential. Second, there was incomplete capture of variables of interest such as postoperative clinical concern for hyperparathyroidism, preoperative ionized calcium, bone disease, nephrolithiasis, renal function and others such as those identified in Tables I and II as “Unknown” or “Not Recorded.” Preoperative ionized calcium levels would have been especially helpful in confirming nPHPT patients. We suspect the lack of this laboratory value in the database likely reflects a combination of incomplete data entry and healthcare practitioners not obtaining it preoperatively. Long-term (>30 days postoperatively) data regarding persistent or recurrent hyperparathyroidism were not recorded in >40% of the cohort. Third, although we included only patients with sporadic primary hyperparathyroidism in our cohort, a significant portion of both the nPHPT and PHPT patients were noted to have low levels of 25OH-vitamin D levels and some had normal PTH. As noted in the Methods section of this report, preoperative lab values utilized in CESQIP are reported based on the most recent value within 30 days of surgery, and thus the low vitamin D and normal PTH values may be due to abstraction inconsistencies rather than patients with secondary hyperparathyroidism or those misclassified as having hyperparathyroidism. We also performed a sensitivity analysis excluding these patients, supporting abstraction consistencies. Unfortunately, glomerular filtration rate information is not available in CESQIP, which would be an additional way to identify secondary hyperparathyroidism patients miscategorized into the sporadic hyperparathyroidism group. Despite these limitations, we present the largest experience with nPHPT to date. It is clear that patients with nPHPT have higher rates of multigland disease compared with PHPT. In this data set, postoperative outcome data were limited by missing data and lack of long-term results. These data support consideration for multigland exploration, utilization of intraoperative PTH to guide surgery, and the need for longer follow-up in patients with nPHPT. Funding/Support No grant funding was utilized with respect to this research. Conflict of interest/Disclosure The CESQIP and the hospitals participating in CESQIP are the source of the data used herein. They have not verified and are not
responsible for the statistical validity of the data analysis or the conclusions derived by the authors. The conclusions, findings, and opinions expressed by the authors do not necessarily reflect the official position of the AAES or CESQIP. Use of CESQIP data does not imply endorsement by AAES or CESQIP. The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
References 1. Forster J, Monchik JM, Martin HF. A comparative study of serum ultrafiltrable, ionized, and total calcium in the diagnosis of primary hyperparathyroidism in patients with intermittent or no elevation in total calcium. Surgery. 1988;104: 1137e1142. 2. Wills MR, Pak CY, Hammond WG, et al. Normocalcemic primary hyperparathyroidism. Am J Med. 1969;47:384e391. 3. Kiriakopoulos A, Petralias A, Linos D. Classic primary hyperparathyroidism versus normocalcemic and normohormonal variants: Do they really differ? World J Surg. 2018;42:992e997. 4. Lowe H, McMahon DJ, Rubin MR, et al. Normocalcemic primary hyperparathyroidism: Further characterization of a new clinical phenotype. J Clin Endocrinol Metab. 2007;92:3001e3005. 5. Campbell MJ. The definitive management of primary hyperparathyroidism who needs an operation? JAMA. 2017;317:959e968. 6. Siperstein AE, Shen W, Chan AK, et al. Normocalcemic hyperparathyroidism. Biochemical and symptom profiles before and after surgery. Arch Surg. 1992;127: 1157e1156; discussion 1161e1163. 7. Stuart HC, Harvey A, Pasieka JL. Normocalcemic hyperparathyroidism: Preoperatively a disease, postoperatively cured? Am J Surg. 2014;207: 673e680. 8. Cusano NE, Maalouf NM, Wang PY, et al. Normocalcemic hyperparathyroidism and hypoparathyroidism in two community-based nonreferral populations. J Clin Endocrinol Metab. 2013;98:2734e2741. 9. Bilezikian JP, Brandi ML, Eastell R, et al. Guidelines for the management of asymptomatic primary hyperparathyroidism: Summary statement from the fourth international workshop. J Clin Endocrinol Metab. 2014;99: 3561e3569. 10. Pawlowska M, Cusano NE. An overview of normocalcemic primary hyperparathyroidism. Curr Opin Endocrinol Diabetes Obes. 2015;22:413e421. 11. Koumakis E, Souberbielle JC, Sarfati E, et al. Bone mineral density evolution after successful parathyroidectomy in patients with normocalcemic primary hyperparathyroidism. J Clin Endocrinol Metab. 2013;98: 3213e3220. 12. Pierreux J, Bravenboer B. Normocalcemic primary hyperparathyroidism: A comparison with the hypercalcemic form in a tertiary referral population. Horm Metab Res. 2018;50:797e802. 13. Trinh G, Rettig E, Noureldine SI, et al. Surgical management of normocalcemic primary hyperparathyroidism and the impact of intraoperative parathyroid hormone testing on outcome. Otolaryngol Head Neck Surg. 2018;159: 630e637. 14. Lim JY, Herman MC, Bubis L, et al. Differences in single gland and multigland disease are seen in low biochemical profile primary hyperparathyroidism. Surgery. 2017;161:70e77. 15. Traini E, Bellantone R, Tempera SE, et al. Is parathyroidectomy safe and effective in patients with normocalcemic primary hyperparathyroidism? Langenbeck’s Arch Surg. 2017;402:344e345. 16. Siperstein A, Berber E, Barbosa GF, et al. Predicting the success of limited exploration for primary hyperparathyroidism using ultrasound, sestamibi, and intraoperative parathyroid hormone: Analysis of 1158 cases. Ann Surg. 2008;248:420e426. 17. Siprov a H, Frys ak Z, Sou cek M. Primary hyperparathyroidism, with a focus on management of the normocalcemic form: To treat or not to treat? Endocr Pract. 2016;22:294e301. 18. Wade TJ, Yen TWF, Amin AL, et al. Surgical management of normocalcemic primary hyperparathyroidism. World J Surg. 2012;36:761e766. 19. Schneider DF, Burke JF, Ojomo KA, et al. Multigland disease and slower decline in intraoperative PTH characterize mild primary hyperparathyroidism. Ann Surg Oncol. 2013;20:4205e4211. 20. Kebebew E, Hwang J, Reiff E, et al. Predictors of single-gland vs multigland parathyroid disease in primary hyperparathyroidism: A simple and accurate scoring model. Arch Surg. 2006;141:777e782. 21. Sho S, Kuo EJ, Chen AC, et al. Biochemical and skeletal outcomes of parathyroidectomy for normocalcemic (incipient) primary hyperparathyroidism. Ann Surg Oncol. 2019;26:539e546.
T.K. Pandian et al. / Surgery xxx (2019) 1e5
5
Discussion Dr Christopher R. McHenry (Cleveland, OH): How did you define multiglandular disease? Was it based on persistent PTH elevation intraoperatively? Or was this based on visible appearance of abnormal glands? Dr T.K. Pandian: That is a great question. Unfortunately, with this large database, it is not granular enough to determine exactly how each institution defined multigland disease. The abstractors from our institution used pathology results. Basically, if a pathologist noted hyperplasia, they would categorize that as multigland disease. But if a surgeon entered that information, it could potentially be different. Dr Janice Pasieka (Calgary, AB, Canada): Do you have the preoperative localization results for your 2 groups? If they had more multigland disease, did they have a higher rate of nonlocalization preoperatively, and therefore were at risk of needing a more extensive operation? Dr T.K. Pandian: That’s a great point. Unfortunately, we don’t have those data. Some of that information is available
in CESQIP, so it’s certainly something we can go back and look at. Dr Herb Chen (Birmingham, AL): I may have misread this, but when you showed your two groups, was there a higher incidence of suspected cancer in the lower calcium group? We normally think of parathyroid cancer as being associated with a very high calcium and high PTH, so that doesn’t go together very well. Dr T.K. Pandian: Yes, absolutely. It was a curious finding. I don’t know that we have a great explanation for that. Some of this may be misclassification errors. Dr David Schneider (Madison, WI): Great use of CESQIP data. One question I had is whether you have the capability to examine this by institution in a deidentified manner. I suspect in this particular disease process there’s wide variability in who even gets surgery, and some centers may choose not to operate on these patients at all. So it might be interesting to look at that variability, and it might add something to your paper as well. Dr T.K. Pandian: Thank you. That’s a good suggestion.