Author(s): Zia S, Sippel RS, Chen H
Abstract Share this page
Abstract BACKGROUND: Preoperative localization is the first step towards minimally invasive targeted parathyroidectomy. While there are data emphasizing that surgeon experience optimizes operative outcomes, the role of the radiologist's experience in successful preoperative imaging is unclear. We hypothesized that the accuracy of sestamibi scanning for primary hyperparathyroidism is dependent upon surgeon interpretation and radiologist volume. METHODS: Between January 2000 to August 2009, 1,255 patients underwent parathyroidectomy for hyperparathyroidism at our institution. Of these, 763 had sestamibi scans for primary hyperparathyroidism. All scans were reviewed by surgeons and radiologists blinded, and were correlated with the operative findings and pathological reports. Radiologists were grouped into high volume (>50 cases/year, HV-RAD) or low volume (<50 cases/year, LV-RAD) based upon a database of >6,000 parathyroid cases reported by 89 regional hospitals. RESULTS: Of the 763 patients, 77 \% were female and the mean age was 60 years. Mean baseline calcium and parathyroid hormone levels were 11.2 ± 0.03 mg/dl and 133 ± 3.27 pg/ml, respectively. The sensitivity of the surgeon (93 \%) was higher than both HV (83 \%) and LV (72 \%) radiologists. Importantly, the positive predictive values were similar: 96 \% for surgeon, 93 \% for HV-RAD, and 98 \% for LV-RAD. As a result, out of 99 scans which were correctly read by the surgeon but not by radiologist, 84 were read as negative by radiologist, 11 on the wrong side of the neck, and 4 on the same side but the wrong gland. CONCLUSIONS: Surgeon interpretation and radiologist volume increase the likelihood of successful preoperative sestamibi parathyroid localization for primary hyperparathyroidism. We recommend that imaging be reviewed by experienced parathyroid surgeons rather than relying on radiological interpretation alone.
This article was published in Ann Surg Oncol
and referenced in Journal of Surgery