More accurate than MRI measurement of tumor size in breast cancer by using the peri-tumoral halo uptake layer method of the 18F-FDG PET/CT scan.

Abstract

To evaluate the reliability of a method using the peri-tumoral halo layer (PHL) for assessing tumor size in breast cancer patients on the fluorine-18-fluorodeoxy glucose positron emission tomography/computed tomography (18F-FDG PET/CT) scan compared to MRI and pathology. Among 121 patients with breast cancer who underwent both 18F-FDG PET/CT and MRI between March 2013 and June 2016, 59 patients were included in this study. Exclusion criteria were as follows: history of neoadjuvant therapy, history of pre-operative mammotome, insufficient pathologic/radiologic size report, clustered tumor, positive tumor resection margin, 18F-FDG non-avid tumor. The PHL was examined by two nuclear medicine physicians. Tumor sizes (longest diameters) on 18F-FDG PET/CT were estimated using margins defined as the inner line of the PHL. Pathologic tumor sizes were utilized as reference standards. The PHL of each tumor was most commonly designated as the 20%-30% band of the maximum standardized uptake value (SUVmax) it exhibited an inverse correlation with tumor SUVmax. Tumor size on 18F-FDG PET/CT showed a more linear correlation with pathology than that on MRI (r2=0.91 vs 0.65). In Bland-Altman analysis, 18F-FDG PET/CT showed significantly lower bias in size difference relative to pathology, compared with MRI (0.6±9.6cm vs. -1.9±17.3cm). Fluorine-18-FDG PET/CT showed more accurate T staging with pathology, especially in T3 cases, than MRI. A method of tumor size determination, using PHL on 18F-FDG PET/CT, showed more linear relationship and smaller size differences with pathology than MRI (average 0.6 vs. 1.9cm). It provides sufficient reliability and reproducibility for measuring tumor size in breast cancer.