18F]Fluorodeoxyglucose Positron Emission Tomography–Computed Tomography in Breast Cancer: When… and When Not?

Abstract

Diagnostic imaging plays an important role in the care of patients with breast cancer and is used for breast cancer detection, diagnosis, staging, and therapeutic response evaluation. Advances in imaging technology, especially relatively newer technologies such as magnetic resonance imaging (MRI) and positron emission tomography without or with integrated computed tomography (PET or PET/CT) provide powerful diagnostic tools, but also generate questions and controversy regarding where and when to use these new modalities. Early studies of new imaging approaches most often come from centers that are scientifically invested in the development of the technology. Typically, small and highly selected patient populations are evaluated, and these preliminary studies often indicate excellent diagnostic performance. Later on, when large multicenter trials are performed with more clinically representative patient populations, the new test invariably performs less well. Nevertheless, the excitement surrounding early results from a new imaging technology and the increasing public access to these early results fuel the desire by patients for access to new forms of diagnostic imaging. Physicians and patients share the wish for diagnostic certainty in excluding cancer spread at the time of diagnosis and may reason that a purely diagnostic procedure, even if untested, can do no harm. These factors, combined with the challenge and expense of carrying out larger scale diagnostic imaging trials, can lead to the early adoption of new imaging studies into clinical practice, often before their performance and optimal use have been fully determined. However, the use of a new imaging technology in populations of patients unlikely to benefit from its use can add anxiety, expense, unnecessary procedures, and delay in care. Equally important, inappropriate use in the wrong patient population can obfuscate benefit in other clinical settings where the imaging might be helpful. These dilemmas have been aptly demonstrated in the application of [F]fluorodeoxyglucose (FDG) PET and PET/CT to early breast cancer staging. Early studies of FDG PET for breast cancer axillary staging (reviewed in Eubank et al) demonstrated high sensitivity and specificity (85%–90% or better for both) and created considerable enthusiasm for FDG PET as a noninvasive method for detecting axillary nodal metastases. However, these studies were heavily biased towards patients with advanced disease, often with larger and more aggressive cancers in which the disease burden in the lymph nodes was larger and better suited for detection by FDG PET compared with early-stage disease. A subsequent multicenter trial cast doubt on these early studies, and more recent single-center trials performed in the era of sentinel lymph node mapping showed that, compared with sentinel lymph node biopsy, the sensitivity of FDG PET and PET/CT for axillary nodal metastases was as low as 20%–40%. In the article that accompanies this editorial, Pritchard et al add to this body of evidence. In a prospective, four-center study of 325 patients in Ontario with stages I and II breast cancer and clinically negative axilla who underwent FDG PET and PET/CT for axillary nodal staging, the authors found nearly 100% specificity, but sensitivity of only 24%, for FDG PET. In addition, in 13 patients who were suspected of having distant metastases by FDG PET, 10 had false-positive findings, and only three were confirmed to have stage IV disease by biopsy or clinical follow-up. Although there were some minor limitations to the study, including a lack of central PET scan review and the inclusion of some patients scanned with older technology, the Pritchard study confidently shows both FDG PET/CT’s limited sensitivity for axillary nodal metastases and limited yield for distant disease in early-stage breast cancer. The results for distant metastasis detection echo earlier studies of systemic staging using CT and bone scan that showed a low prevalence of distant metastases (5% or less in stage I breast cancer) and a notable number of false-positive findings. Although there have been some exceptions, the majority of recent studies and systematic reviews are in agreement with the results of the Pritchard study and suggest a low diagnostic yield for FDG PET/CT in patients with stage I and early stage II breast cancer. This evidence underlies the strong recommendation in the current National Comprehensive Cancer Network (NCCN) consensus guidelines that systemic staging, including FDG PET/CT, is not indicated for early-stage breast cancer in the absence of signs or symptoms suggesting metastasis. Despite existing evidence and guidelines, there is anecdotal evidence that the use of imaging for systemic staging in early-stage breast cancer in the medical community is quite variable. Recent evidence from the ADVICE consortium in the Puget Sound region shows two trends relevant to the work of Pritchard et al. First, the trend in use of JOURNAL OF CLINICAL ONCOLOGY E D I T O R I A L S VOLUME 30 NUMBER 12 APRIL 2