Purpose: To evaluate [fluorine-18]2-deoxy-2-fluoro-D-glucose (FDG) positron emission tomography (PET) of the axilla as a screening test for detecting regional spread of breast cancer.
Materials and methods: High-dose FDG PET of the axilla was successfully performed in 50 patients (age range, 36-79 years) with breast cancer before 52 axillary lymph node dissections. Two additional patients had scans that were uninterpretable because of intense myocardial activity that obscured the axilla.
Results: The sensitivity and negative predictive value were both 95%, the specificity was 66%, and the overall accuracy was 77%. The only false-negative PET scan was obtained in the largest patient, who had a low-quality scan.
Conclusion: Patients with negative PET scans had such a low risk for axillary lymph node metastases that axillary dissection was not warranted. Patients with positive PET scans required dissection to confirm the presence and determine the number of positive lymph nodes. Had this algorithm been used to select patients for dissection, approximately $120,000 in charges ($2,300 per patient) would have been saved and 22 patients would have been spared the morbidity of axillary lymph node dissection. Within this study population, PET scans of the axilla were interpreted with sufficient sensitivity for PET to serve as a cost-effective screening test for axillary lymph node metastases.