A novel technique to improve the diagnostic yield of negative sestamibi scans

Minimally invasive parathyroidectomy is successful in achieving cure for most patients with primary hyperparathyroidism. Most surgeons rely on preoperative imaging as part of the workup for localization. Ultrasonography and sestamibi are the 2 most commonly used preoperative imaging studies. When these 2 studies are positive and concordant the preoperative localization is straightforward. However, when ≥1 of these studies is negative, the preoperative localization is suspect. We hypothesize that the yield of useful localizing information from "negative" sestamibi scans can be increased in certain situations. Specifically, in cases where the thyroid lobe length seen on sestamibi is discordant from the lobe length of the ultrasonography, this often represents a "hidden" parathyroid adenoma. If our hypothesis is correct, this could lead to decreased resource utilization in cases of nonlocalized parathyroid adenomas.

We retrospectively analyzed our database of patients with primary hyperparathyroidism who underwent parathyroidectomy from 2005 to 2011. The anteroposterior views of early phase sestamibi were analyzed for thyroid lobe lengths. A ratio of the length of the right lobe to left lobe was calculated. The thyroid lobe lengths on ultrasonography were measured and similar ratios were calculated. The difference in ratios between sestamibi and ultrasonography was calculated for each patient. A difference in ratios from sestamibi and ultrasonography that corresponded with a "hidden" parathyroid on the side of the additional length on sestamibi at the time of surgery was considered a positive finding. When the difference in ratios from the 2 images did not correspond with a "hidden" parathyroid at the time of operation, it was considered a negative finding.

There were 59 patients with single-gland disease, negative sestamibi, and images available for review. There were 32 patients (54%) with the positive finding of a "hidden" parathyroid corresponding with a difference in thyroid lobe length ratios from sestamibi and ultrasonography. The overall mean difference in ratios between sestamibi and ultrasonography was 0.37 ± 0.32. The mean ratio difference in the group of patients with a negative "hidden" parathyroid was 0.11 ± 0.02, and the mean ratio difference in the group of patients with a positive "hidden" parathyroid was 0.58 ± 0.05 (P < .001). When a difference in ratios of ≥0.23 was obtained, this predicted a "hidden" parathyroid with a sensitivity of 93.8% and specificity of 85.2%. There were 39 patients with multigland disease, negative sestamibi, and images available for review. None of these patients had a ratio difference of ≥0.23. The mean ratio difference for patients with multigland disease was significantly lower than that of the single-gland disease (0.08 ± 0.06 vs 0.37 ± 0.32; P < .001).

Discordance between thyroid lobe lengths on the early phase sestamibi compared with ultrasonography has led to successful preoperative identification of parathyroid adenomas, even though the sestamibi was traditionally read as negative. This finding has not been previously described, seems to be reliable, and can lead to improved preoperative localization and decreased resource utilization in this subset patients.