Background: Patients with primary hyperparathyroidism and baseline intraoperative parathyroid hormone levels in the normal range are challenging. This study compares the predictive value of a commonly used intraoperative parathyroid hormone algorithm, a software model for cure prediction, and surgeon judgment in this population.
Methods: This was a retrospective review of consecutive patients who underwent parathyroidectomy for primary hyperparathyroidism at a single institution from March 2013 to October 2014.
Results: Of 541 operative patients, 114 (21.1%) had a mean normal baseline intraoperative parathyroid hormone of ≤69 pg/mL (median 59.0 ± 10.3; range 26-69). Of the 114 patients, 93 (81.6%) were women, median age was 61 years (range 18-88). Overall, 107/108 (99.1%) patients were cured; 47 (41.2%) patients had single adenomas, 16 (14%) had double adenomas, and 51 (44.7%) had multigland hyperplasia. Using the 50% decline algorithm, a correct prediction was made in 86 (75.4%) patients. Using the computer software, a correct prediction was made in 88 (77.2%) patients. Surgeon judgment, however, was 99.1% accurate.
Conclusion: Patients with normal baseline intraoperative parathyroid hormone have a high incidence of multigland disease (58.8%), greater than reported previously. Current software modeling and the 50% decline algorithm are insufficient to predict cure in this population; intraoperative parathyroid hormone interpretation combined with operative findings and surgical judgment yield optimal outcomes.
Copyright © 2016 Elsevier Inc. All rights reserved.