| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From the Department of Surgery, Memorial Sloan-Kettering Cancer Center, New York, New York.
Correspondence: Address correspondence to: Elisa Rush Port, MD, Department of Surgery, Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, MRI-1026, New York, New York, 10021; Fax: 212-794-5812; E-mail: eport{at}mskcc.org
Positron emission tomography (PET) scanning relies on the differential uptake by tumor cells of a radiolabeled glucose, FDG (18F-2-fluoro-2-deoxy-D-glucose), and by generating a whole-body image, can image both the primary tumor and its regional and distant metastases. Although the management algorithms of lymphoma, esophageal, lung, and colorectal cancers increasingly incorporate PET as a tool for pre- and posttreatment evaluation, the role of PET in breast cancer remains undefined.
Whereas early experience found that PET could identify virtually all large breast cancers and known metastases, its main limitation has proved to be its lack of sensitivity in detecting smaller-volume disease. Currently, PET has the resolution to identify only those tumor foci 5 mm or larger. This is a significant limitation and effectively excludes PET from a role in breast cancer screening, where mammography has dramatically increased the proportion of breast cancers diagnosed as duct carcinoma in situ (DCIS) or as very small invasive lesions, and where both ultrasound (US) and magnetic resonance imaging (MRI) (both highly sensitive) play a growing role. Whereas PET seems comparably sensitive in imaging known cancers (93% in one study1), and will occasionally yield an incidental diagnosis of cancer when done to evaluate an unrelated malignancy, its role in population screening is unknown and its cost would seem prohibitive. Current efforts, therefore, focus on defining the role of PET for staging, both regional lymph nodes and distant sites.
For axillary lymph node staging, Greco et. al.2 found that PET was both sensitive (identifying 68 of 72 patients who were node positive, 94%, corresponding to a false–negative rate of 6%) and specific (identifying 82 of 95 node-negative patients, 86%, corresponding to a false–positive rate of 14%), and expressed the hope that PET might identify patients who could avoid axillary lymphadenectomy (ALND). Schirrmeister et al.1 reported much less encouraging results (sensitivity of 79% and specificity of 92%), as have Wahl et. al.3 (sensitivity of 61% and specificity of 80%), Guller et al.4 (sensitivity of 43% and specificity of 94%), and others5–7 (sensitivity 20% to 79% and specificity 57% to 100%).
This trend continues in a report from Lovrics et al.8, in this issue of the Annals of Surgical Oncology, of a study comparing preoperative PET with ALND in 90 patients, 72 of whom also had a SLN biopsy. PET lacked sensitivity, detecting only 36% of patients who were node positive on ALND (with pathologic analysis by hematoxylin and eosin [H&E] staining alone), and only 27% of those who were SLN positive (with pathologic analysis by H&E, serial sections, and anticytokeratin staining). Compared with ALND (with routine pathology) and SLN biopsy (with enhanced pathology), PET had false–negative rates of 64% and 72%, respectively. Compared with either ALND or with SLN biopsy, which many institutions, including ours, consider to be a new standard of care, PET is so insensitive that a negative result simply cannot be believed, and surgical staging is required regardless.
Can a positive result be believed? Lovrics et al. report a specificity of 96% to 97%, corresponding to a false-positive rate of only 3% to 4%. This suggests two possibilities. On the plus side, patients with a positive PET scan could proceed directly to ALND without first having SLN biopsy, with obvious logistic advantages for the surgeon and the operative schedule. On the minus side, a small proportion of patients would incur the morbidity of ALND (and the added disadvantage of routine pathologic analysis) when a SLN biopsy (with the advantage of enhanced pathology) might have sufficed. For a positive PET scan to replace SLN biopsy, what false–positive rate is acceptable: 3% to 4% as reported by Lovrics et al.8 or 20% as reported by Wahl et al.3? Patients will accept considerable treatment-related morbidity for small levels of benefit,9 but will they be equally willing to accept this morbidity if it results from a false-positive test result? Before a positive PET scan can bypass SLN biopsy, further study is needed, both to confirm diagnostic accuracy on a wider basis and to establish cost-effectiveness as well. A low cost and widely available alternative, axillary US with US-guided fine needle aspiration10 offers the dual advantages of (1) a preoperative cytologic diagnosis and (2) low cost.
Is there a role for PET in finding distant metastases? In asymptomatic patients with stage I to II breast cancer, the yield of unexpected distant metastases from bone scan and computed tomography (CT) is close to zero. One would expect the yield of PET in this setting to be small as well. Schirrmeister et al.1 reported (among 117 women with largely stage I to II disease) that PET identified distant metastases missed by conventional staging in 3 patients (2.5%). Van der Hoeven et al.11 reported on PET for staging locally advanced (stage III) breast cancer (LABC); among 48 patients with LABC, PET suggested metastases in 10; of these, 1 was a false-positive finding, 5 were indeterminate, and 4 (8%) proved to be metastases. These authors emphasize that the upside of PET for LABC (switching patients from stage III to stage IV treatment protocols) must be balanced against a downside (the delay and uncertainty involved in evaluating equivocal PET findings). To date no studies have systematically compared PET with conventional staging (bone scan and CT) in the same group of patients, or have addressed cost-effectiveness.
Positron emission tomography has promise in several other areas. First, using the new technology of fusion imaging, a combined PET–CT study can localize FDG-avid lesions more accurately than PET alone, with the possibility of CT-guided biopsy (of note, the CT images obtained in a PET–CT study are not as detailed as those obtained by a dedicated CT scanner). Second, PET allows quantitation of glucose uptake as a standardized uptake value (SUV) and high SUV can correlate with known features of poor prognosis,12 and provide insight into other areas of tumor biology. Third, as with other imaging modalities, PET can track response to therapy and as a whole-body, tumor-specific study can prove over time to have advantages over bone scan, CT, and MRI. Finally, PET may be the ideal vehicle for a group of novel target-specific tracers, such as labeled estrogen-receptor or HER2-neu ligands, giving insight into tumor heterogeneity and mechanisms of drug resistance.
In summary, the early promise that PET would find a central role in the diagnosis, staging, and follow-up for patients with breast cancer is currently unfulfilled. PET offers no clear advantage over standard methods of screening; it is insufficiently sensitive to play a role in lymph node staging, and its role in detecting unsuspected systemic disease requires further definitive studies. PET may prove most useful in the pre- or posttreatment evaluation of those few patients in whom systemic metastasis is suspected but for whom the results of conventional assessment (CT, bone scan, MRI, and radiography) remain ambiguous. PET scanning will almost certainly play a significant in breast cancer management, but most likely in a small subset of patients. Do not give up your gamma probe.
Received for publication July 19, 2004. Accepted for publication July 22, 2004.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
