Back to Annual Meeting Abstracts
Primary Hyperparathyroidism with Negative Imaging: A Significant Clinical Problem
Heather Wachtel*, Edmund K. Bartlett*, Rachel R Kelz*, Isadora Cerullo*, Giorgos C Karakousis*, Douglas L Fraker
Hospital of the University of Pennsylvania, Philadelphia, PA
Objectives: Preoperative imaging plays an increasingly important role in the evaluation of primary hyperparathyroidism (PHPT), and surgical referral may be predicated upon successful imaging. We sought to compare the outcomes for patients undergoing parathyroidectomy for PHPT by imaging results.
Methods: We performed a retrospective study of patients undergoing initial parathyroidectomy for PHPT using a prospectively maintained database (1997-2012). Patients were classified as non-localized (NL) when preoperative imaging failed to identify affected gland(s) and localized (L) if successful. Primary outcome was biochemical cure. Cohort comparison was performed. Propensity score was developed to match NL to L (1:1). Conditional logistic regression determined factors associated with cure in the matched cohort.
Results: Of 1999 patients analyzed, 40% (n=798) were NL. Compared to L, NL had smaller glands (269 versus 542mg, p<0.001), lower rates of single adenoma (77.3 versus 86.3%, p<0.001) and higher proportions of hyperplasia (12.0 versus 5.4%, p<0.001). The cure rates were clinically similar between the NL and L groups (97.6 versus 99.0%, p=0.014). Eighty-two percent of NL patients (n=657) were successfully matched to L controls. In the matched cohort, localization remained significantly associated with cure (OR=2.5, 95%CI=1.0-6.0), but the attributable risk to NL was just 1.5% (95% CI=0.7-2.8%). On multivariate subgroup analysis of NL, the presence of single adenoma was predictive of cure.
Conclusions: Smaller parathyroids and a higher incidence of multiglandular disease are associated with NL. However, NL minimally impacts the rate of surgical cure. Referral for surgical evaluation should be based on biochemical diagnosis rather than localization by imaging.
Back to Annual Meeting Abstracts