Carrie B. Hruska PhD, Abstract Co-Author: Institutional license agreement, Gamma Medica Ideas, Inc

Christina M. Tello-Skjerseth MD, Presenter: Nothing to Disclose

Cindy L. Tortorelli MD, Abstract Co-Author: Nothing to Disclose

Robert Watson Maxwell MD, Abstract Co-Author: Member, Maxwell Publishing Company LLC Stockholder, Maxwell Publishing Company LLC

Amy Lynn Conners MD, Abstract Co-Author: Nothing to Disclose

Eric M Wagenaar, Abstract Co-Author: Nothing to Disclose

Michael K. O'Connor PhD, Abstract Co-Author: Research grant, General Electric Company Patent royalties, Gamma Medica Ideas, Inc

To evaluate the diagnostic performance of low-dose Molecular Breast Imaging (MBI) performed with 2-8 mCi Tc-99m sestamibi.

MBI studies from 32 women who had participated in IRB-approved, HIPAA-compliant MBI research protocols were selected for review. Studies comprised 17 patients with 20 lesions visible on MBI (8 cancers and 12 benign) and 15 patients with negative MBI studies. Median tumor size was 1.0 cm (range 0.6-4.3 cm). Histopathological diagnosis of all lesions was determined from more severe of core-biopsy or surgical excision. MBI studies were performed on a dedicated dual-head CZT gamma camera with low-dose modifications to collimation and energy window (LumaGem, Gamma Medica). Patients received 8-mCi Tc-99m sestamibi; imaging began < 5 minutes post-injection. CC and MLO views of each breast were acquired for 10 min/view, in 4 frames of 2.5 min each. Frames were summed to generate MBI studies with count densities equivalent to injected doses of 2, 4, 6, and 8 mCi. Four radiologists independently performed blinded, random-order reads of all studies in 4 sessions (in order of 2 mCi, 4 mCi, 6 mCi, 8 mCi), separated by > 1 week, assigning a 1-5 assessment which parallel BI-RADs for each breast. Scores of 3 or higher were considered positive. Sensitivity, specificity, and ROC analysis was performed for each dose level. Intra-reader agreement kappa between assessments at 8 mCi and each lowered dose level was calculated. Count density (cts/pixel) was measured in normal background breast tissue.  

Count density was perceived as adequate diagnostic quality for all dose levels. The sensitivity, specificity and area under curve (AUC) of ROC was similar at all dose levels for all readers. Kappa statistics indicated that intra-reader agreement between assessments of 8 mCi MBI and each lowered dose was substantial (>0.6) to nearly perfect (>0.8).

With low-dose modifications, MBI performed on a dual-head CZT system can achieve excellent diagnostic performance at doses of 2-8 mCi. Count density obtained at doses of 4-6 mCi (700-1060 cts/cm2) was similar to that obtained on MBI with conventional doses of 20 mCi (~870 cts/cm2), prior to dose-reduction modifications.

Given that MBI at 4 mCi presents radiation risks comparable to screening mammography, low-dose MBI may be a viable adjunct to mammography for screening in dense breasts.

Hruska, C, Tello-Skjerseth, C, Tortorelli, C, Maxwell, R, Conners, A, Wagenaar, E, O'Connor, M, Low-Dose Molecular Breast Imaging (MBI): Initial Patient Results.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11001562.html