William C. Lavely MD, Presenter: Nothing to Disclose

Sibyll Goetze MD, Abstract Co-Author: Nothing to Disclose

Kent P. Friedman MD, Abstract Co-Author: Nothing to Disclose

Jeffrey P. Leal BS, Abstract Co-Author: Nothing to Disclose

Samuel P Pak MD, Abstract Co-Author: Nothing to Disclose

Zhe Zhang PhD, Abstract Co-Author: Nothing to Disclose

Elizabeth Garrett-Mayer PHD, Abstract Co-Author: Nothing to Disclose

Alan Dackiw MD, Abstract Co-Author: Nothing to Disclose

Martha A Zeiger MD, Abstract Co-Author: Nothing to Disclose

Harvey A. Ziessman MD, Abstract Co-Author: Nothing to Disclose

et al, Abstract Co-Author: Nothing to Disclose

This interim analysis evaluated early and late acquisition of Tc-99m sestamibi planar, SPECT and SPECT/CT images in patients with primary hyperparathyroidism, seeking to determine the optimal imaging protocol for pre-operative parathyroid adenoma localization.

Sixty-seven sequential patients were referred for Tc-99m sestamibi parathyroid scintigraphy; 40 underwent surgery and were analyzed. At 15 minutes and 2 hours post-injection, anterior, RAO and LAO planar and SPECT/CT images were acquired. Six image sets were anonymized. Two reading groups (two nuclear medicine physicians in each) read the six image sets as well as combinations of two image sets (19 different readings) to determine the presence or absence of an adenoma at 13 locations. Each image set or combination set was scored using a 5 point scale for certainty of focus and location. Surgical location was the gold standard. Data were analyzed using logistic regression with inferences based on accuracy (AUC of ROC curve), PPV and NPV, as well as correlation with patient characteristics. Scores of 1-3 were considered negative and 4-5 positive.

The kappa value comparing reading groups was 0.72; lowest kappa values were for planar images while highest kappa values were for combinations which included SPECT and SPECT/CT. Overall lateralization AUC, PPV and NPV were 84.8%, 99.3% and 74.1%, respectively. Overall precise localization AUC, PPV and NPV were 74.9%, 75.3% and 95.5%, respectively. Highest AUC was found in combination image sets which included SPECT and SPECT/CT at either early or late time points; the most accurate image sets were combinations of early SPECT/CT-late planar and early SPECT/CT-late SPECT (P<0.05 in 9 of 19 image sets). Statistically significant PPVs were found only in combination image sets that included SPECT or SPECT/CT; the image set with the highest PPV was early SPECT-late SPECT/CT (PPV = 85.3%). Significant correlations occurred between certainty of focus and mass of gland (P<0.0001), location (P<0.0001), gender (P=0.0005) and PTH level (P<0.0001).

Parathyroid localization using Tc-99m sestamibi scintigraphy should include both early and late SPECT and/or SPECT/CT acquisition.

Lavely, W, Goetze, S, Friedman, K, Leal, J, Pak, S, Zhang, Z, Garrett-Mayer, E, Dackiw, A, Zeiger, M, Ziessman, H, et al, , Early and Delayed Planar, SPECT and SPECT/CT Scintigraphy with Tc-99m Sestamibi for the Localization of Parathyroid Adenoma in Patients with Primary Hyperparathyroidism: An Interim Analysis.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4414045.html