Abstract
1820
Introduction: Background and Objectives: 18F-fluorodeoxyglucose (FDG) PET/MRI uses the glycolysis system to functionally evaluate tumors by measuring tumor uptake of radioactive FDG. In contrast, the apparent diffusion coefficient (ADC) measures cellular densities. ADC measures feature a large overlap between benign and malignant tumors, which changes greatly depending on the type of equipment and region of interest (ROI). This study aimed to estimate the cut-off values of standardized uptake value (SUV) and ADC of normal breast and primary tumor breast tissues using PET/MRI images.
Methods: Twenty consecutive female patients with breast cancer were enrolled in this retrospective study between January and November 2017. Whole-body PET/MRI studies of 18F-FDG were performed with a scan protocol of 2 min/bed acquisition at 1 h post injection. Breast PET/MRI studies were performed with a scan protocol of 10 min/bed acquisition at 90 min post injection. ROIs for tumorous and normal breast tissues were the axis planes on which SUVmax existed. SUVearly was each SUVmax of the primary tumor and normal breast during whole-body acquisition. SUVdelay was each SUVmax in breast acquisition. ΔSUV was calculated by subtracting SUVearly from SUVdelay. We measured SUVearly, SUVdelay, ΔSUV, ADC, and the time intensity curve(TIC). Cut-off values were calculated from the ROC curve generated from the primary tumor and normal breast tissues.
Results: SUVearly, SUVdelay, ΔSUV, and ADC were significantly different between primary tumor and normal breast tissues, and the cut-off values were 2.3, 4.1, 1.38, and 0.000972, respectively. The sensitivity of SUVdelay was 0.95 and the specificity was 1.0. The sensitivity and specificity of ADC were 0.85 and 0.9, respectively. SUVdelay showed a tendency of small overlap when comparing SUVdelay and ADC. Comparing SUV of the primary tumor with ADC generated a correlation coefficient of 0.11, indicating that no correlation was observed.
Conclusions: The cut-off values for primary breast cancer tumorous tissue were 2.3 (SUVearly), 4.1 (SUVdelay), and 0.000972 (ADC). There was no correlation between SUVdelay and ADC. It is difficult to diagnose primary tumor with ADC alone. Calculating the cut-off value from each measured value is essential for diagnosis. Each cut-off value serves as a reference for diagnosis.