Low-Dose Positron Emission Mammography: A Novel, Promising Technique for Breast Cancer Detection

See also article by Freitas et al in this issue.

David S. Barreto, MD, is an assistant professor in the breast imaging section in the department of radiology at The George Washington University School of Medicine and Health Science, Washington, DC. His research interests include, among others, high-risk screening breast imaging, minimally invasive breast biopsies, breast cancer cryoablation, and transgender breast imaging.

Jocelyn A. Rapelyea, MD, is a professor of radiology at The George Washington University School of Medicine and Health Sciences. Dr Rapelyea also serves as the associate director of breast imaging. She has received numerous honors and appointments throughout her career, including fellowship in the Society of Breast Imaging and the American College of Radiology. She has published abstracts, articles, and chapters and has made various national and international presentations.

Although mammography is widely used for breast cancer screening, it may not be as effective in detecting underlying disease in female individuals with dense breast tissue. Recent developments in advanced mammographic techniques, such as digital breast tomosynthesis and contrast-enhanced mammography, have addressed this issue and overcome some of the limitations of traditional mammography. Both modalities use low-dose x-rays, while contrast-enhanced mammography requires an additional injection of an intravenous contrast agent to identify underlying mammographic findings. These advanced techniques have shown promise in detecting additional cancers that mammography may have missed.

Similarly, the use of radionuclide injection needed in breast-specific gamma imaging (BSGI), molecular breast imaging (MBI), and positron emission mammography (PEM) imaging has demonstrated improved sensitivity and higher specificity compared with more traditional mammographic methods (1,2). Like contrast-enhanced MRI, radionuclide studies are focused on physiologic imaging and can be performed irrespective of breast density. However, the broad adoption of alternative radionuclide screening methods is slow due to the potential risk of whole-body radiation at high doses, as opposed to the localized risk with traditional screening methods like full-field digital mammography and digital breast tomosynthesis. The American College of Radiology recommends use of nuclear medicine breast-dedicated studies only as an alternative if a patient is unable to undergo breast MRI (3).

To alleviate concerns about radiation risk associated with radionuclide studies using conventional doses, studies within the last 10 years have focused on incremental dose reduction imaging protocols. In the study by Kuhn et al (4), BSGI doses were reduced from 32 mCi to 7 mCi in a series of 223 patients at increased risk for breast cancer, and no evidence of a difference was found in cancer detection between patients injected with the higher versus lower dose. Similarly, Rhodes and colleagues (5) conducted a study demonstrating that the combined use of mammography and MBI successfully reduced the administered dose of technetium 99m (^99mTc) sestamibi to 300 MBq (8 mCi) for 1585 patients. In this series, the authors found the addition of MBI as a supplemental screening method helped diagnose 75% of all cancers in women with extremely dense breast tissue and 92.9% in women with heterogeneously dense breast tissue (5). Previous phantom or simulation studies of PEM demonstrate that imaging can be performed at a lower dose of fluorine 18 fluorodeoxyglucose (¹⁸F-FDG), yielding similar diagnostic results without a significant difference in overall image quality. However, a more recent study that evaluated the clinical significance and effects of two administered doses of ¹⁸F-FDG (185 MBq vs 92.5 MBq) in two groups of patients diagnosed with breast cancer while keeping the same imaging duration found that the lesion-to-background ratio of radiotracer uptake was higher for invasive ductal carcinoma compared with ductal carcinoma in situ, and the image noise versus image counts increased for those images reconstructed with fewer counts and later time point imaging (6).

In this issue of Radiology: Imaging Cancer, Freitas et al performed a prospective comparative analysis of the efficacy of lower dose PEM imaging with concurrent contrast-enhanced breast MRI in the detection of breast cancer (7). Participants were assigned independently of breast density, tumor size, and histopathology cancer subtype. Twenty-five participants were included, of which 10 received a ¹⁸F-FDG dose of 5 mCi, 10 received a dose of 2 mCi, and five received a dose of 1 mCi. Participants underwent subsequent imaging using PEM after 1 hour, 2 hours, and 4 hours of the initial injection. PEM images were acquired similarly to mammographic views (craniocaudal and mediolateral oblique), with an average of 5 minutes for each projection (approximately 20 minutes for image acquisition). In this small series, low-dose PEM detected 96% of known index cancers and failed to detect one lobular index cancer (versus MRI, which detected 100%). PEM also showed a lower rate of false-positive findings than MRI (although statistically insignificant in this small series).

While there is published literature demonstrating the value of supplemental breast radionuclide imaging as an additional tool for the detection of early breast cancer in patients with dense fibroglandular tissue, additional use in evaluating the extent of disease, as well as monitoring response to neoadjuvant therapy, has also been shown to improve diagnostic accuracy when coupled with mammography (6). However, whole-body radiation associated with MBI and BSGI has prevented widespread use and broad acceptance. The low-dose PEM protocol evaluated by Freitas et al offers improved diagnostic capabilities for breast imaging while substantially reducing radiation exposure. As a result, it may become a more widely accepted adjuvant method for breast cancer screening and an alternative to patients unable to undergo other types of supplemental physiologic imaging due to renal disease, claustrophobia, or other contraindications to breast MRI.

The MBI activity dose currently used in clinical practice, using ^99mTc sestamibi, is approximately 6.5–8 mCi (8). In the current PEM system study, the authors used reduced ¹⁸F-FDG doses, ranging from 1 mCi to 5 mCi, which is comparable to breast mean glandular radiation doses of full-field digital mammography and contrast-enhanced mammography. Reduced doses did not affect the diagnostic accuracy of invasive breast cancer detection in this study. In addition, the reduced dose protocol demonstrated a lower rate of additional false-positive findings than MRI, corroborating results of previous studies, which demonstrated superior specificity of PEM compared with MRI (91% vs 86%) (9). Decreasing false-positive findings can also lead to other potential benefits, such as reduced patient anxiety and overall cost burden, as additional workups and biopsies may be rendered unnecessary. Another advantage of low-dose PEM is its shorter acquisition time (approximately 20 minutes) to complete the four standard views as compared with a full breast MRI protocol (approximately 30 minutes) and some protocols using MBI (between 28 and 40 minutes) (10).

The current study's small sample size and narrow selection of participants, including those with at least one biopsy-proven cancer, limits our understanding of the effectiveness of a lower dose PEM protocol for use with other clinical indications. Nevertheless, the results from this pilot study suggest the feasibility of effectively detecting invasive breast cancers using lower radiation doses and decreasing the rate of false-positive findings compared with other physiologic imaging methods.

Large-scale prospective clinical trials at this lower dose would be beneficial to aid specific PEM clinical use and demonstrate accuracy in diagnosing different sizes and histologic types of breast cancers. This is particularly important as one invasive lobular cancer was not detected using this novel technique, which could be related to the reduced lesion-to-background count in the lower administered dose. Further comparison with other breast radionuclide studies may help add information for clinical management. Overall, the new low-dose PEM protocol demonstrates promise as a future alternative adjunctive breast imaging tool in select patients with contraindications to other physiologic imaging modalities such as breast MRI.