The present study was aimed to determine the clinicopathologic factors that predict false negative (FN) PET results in these patients.

A total of 116 breast lesions in 111 patients (pre-menopausal 45; perimenopausal 15; post-menopausal 51) with known or suspicious of breast cancer who underwent FDG-PET scans for staging, were included in this study. The median age was 52+/-11 years (range 32-79 years). All PET studies results were correlated with follow-up surgical pathology results. A cut off value of 2.5 was considered for positive or negative PET results. Univariate and multivariate analyses were performed to identify factors associated with FN results.

Of 116 breast lesions, 85 were malignant and 31 were benign on histopathology. Of the 85 malignant lesions, 41 were true positive (TP) and 44 were FN. Among the 31 benign lesions, 30 were true negative and one was false positive. There was significant difference in the tumor size (p=0.003) and tumor grade (p=0.001) in patients with TP and FN PET results. Multivariate logistic regression demonstrated that tumor size (< or =10 mm) and low tumor grade were independently associated with FN results. No significant relationship of FN PET results was found with age, menopausal status, tumor type, c-erbB-2, estrogen and progesterone receptors, sentinel lymph node or distant metastasis, parenchymal density and multifocality of primary breast tumor.

In present study, tumor size and tumor grade are independent factors that predict FDG-PET results. Smaller tumors (< or =10 mm) and low-grade tumors are strong predictor of FN FDG-PET results.