Evaluation criteria promise better PET performance

The key to reliable detection of the spread of breast cancer to lymph nodes may lie in the administration of a radiotracer, according to research conducted at Cedars-Sinai Medical Center in Los Angeles. Variability from one PET scan to another may be related more to the way exams are done than the technology.

Axillary nodes with standardized uptake values of 2.3 or greater were 15 times more likely to contain metastases than nodes with SUVs of less than 2.3, according to a study published in the August issue of Archives of Surgery. In fact, with the SUV threshold set at 2.3, 60% of patients with axillary metastases were identified. All of those who tested positive were accurately diagnosed. And there were no false positives.

"This suggests that if physicians are going to perform a PET scan before surgery for systemic staging, an SUV should be calculated," said Dr. Edward H. Phillips, executive vice chairman of Cedars-Sinai's surgery department, chief of the general surgery division, and director of the Saul and Joyce Brandman Breast Center.

The spread of breast cancer cells to lymph nodes under the arm is a significant indicator of the stage of the disease and a critical determinant of patient prognosis. Lymph node biopsies or even surgical removal of the node often are needed to determine whether metastasis has occurred.

PET scans with fluorine-18 fluorodeoxyglucose hold the potential to noninvasively identify invasive breast cancers and their metastases.

"FDG-PET is not a perfect test, but by quantifying the cancer-cell activity in the axillary lymph nodes, it can become a more accurate test," Phillips said. "As we continue to see advances in technologies, I believe we will be able to identify smaller and smaller tumors with greater accuracy."

Although Cedars-Sinai's threshold for the uptake of FDG was 2.3, that number is expected to vary from one institution to another because of technical and calibration factors, Phillips said. Therefore, each PET center is advised to develop its own reference values.

In the study, researchers reviewed the medical records of 462 women with invasive breast cancer who underwent FDG-PET for staging between Nov. 1, 2001 and Aug. 31, 2005. The researchers focused on 51 women (with 54 invasive cancers) on whom PET scanning was performed before they underwent axillary lymph node surgery or chemotherapy. The PET scans indicated activity in the axillary area for 32 (59%) of the breast cancer cases. SUVs ranged from 0.7 to 11; 20 tumors had a value of 2.3 or greater, while 34 had a value of less than 2.3.

Further refining their study, the researchers set the SUV threshold at 2.3. Using this threshold, the scans had a sensitivity of 60% and a specificity of 100%. Positive predictive value with an SUV of 2.3, therefore, was 100%.

It was determined that an SUV greater than 2.3 appears sufficient to diagnose lymph node involvement, suggesting that if a PET SUV is validated with 100% specificity, chemotherapy can be initiated or the surgeon can proceed directly to axillary lymph node dissection to control the spread of the disease.