Abstract
447
Objectives: Identification of pulmonary metastases is crucial for appropriate therapy in patients with a history of colorectal cancer. The aim of this study is to describe and find diagnostically important FDG PET/CT findings of pulmonary metastases.
Methods: 55 nodules from 34 colorectal cancer patients, who underwent pulmonary metastasectomy from June 2003 to April 2005, were studied retrospectively (M:F=24:10, Age=62±10 years). The findings of PET were categorized as SUV, the nature of FDG accumulation and CT findings were categorized in terms of size, marginal characteristics, the shape, calcification, cavitation, and Hounsfield unit. The malignancy or benignity of the nodules was determined according to pathologic results (n=41) and clinical follow-up (n=14).
Results: Univariate analysis showed that SUV, focal FDG accumulation, margin, nodule shape, and diffuse calcification were different between malignant and benign nodules(p<0.05) (SUV 3.73 vs 1.00; focal FDG accumulation 61.5% vs 6.3%; well margin, 76.9% vs 50.0%; round shape, 97.4% vs 62.5%; diffuse calcification, 0 % vs 18.8%), but SUV was the only predictor by multiple logistic analyses. Diagnostic accuracy and the best cut-off value of SUV varied with the size of nodules substantially. In the case of less than 8 mm (n=20), AUC was only 0.74 (cut-off >1.2 g/ml, sensitivity 70, specificity 90) whereas greater than 8 mm, AUC was 0.9052. In the cases of small nodule, the low sensitivity was improved to 90% by considering margin and shape of the nodule but specificity decreased to 30%. The best cut-off values were linear with the lesion size (6-8 mm: 1.4 g/ml, 12-14 mm: 2.3 g/ml, 18-20 mm: 3.6 g/ml).
Conclusions: FDG SUV was the best predictor in the setting of pulmonary metastases from colorectal cancer, but the cut-off value should be adjusted to the size of the lesion.
- Society of Nuclear Medicine, Inc.