Using 18F-FDG PET/CT to Estimate the Length of Gross Tumor and Involvement of Lymph Nodes in Esophagogastric Junction Carcinoma

Abstract

To determine the optimal method to estimate length of gross tumor and involvement of lymph nodes of using 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT) in esophagogastric junction carcinoma (EGJC) verified by pathologic examination. Twenty patients with newly diagnosed and untreated EGJC were enrolled, with all of them then being treated with radical surgery. Every patient underwent a whole-body 18F-FDG PET/CT imaging within a week before the operation. The length of the gross tumor on PET/CT was measured using different methods: standardized uptake value (SUV) 1.5 – 5.5 in intervals of 1.0, and 10% – 50% of maximum standardized uptake value (SUVmax) on 18F-FDG PET/CT in intervals of 10%, and recorded as L1.0 to L5.0, and L10% to L50%, respectively. Measurements of gross tumor length were made at operation before the esophagogastric junction specimen was removed (in-vivo, Lvivo), and also with the specimen lying free (contracted, Lvitro). The shrinkage ratio for the primary tumor was calculated as Lvitro/Lvivo, and was then used to calculate the pathological length of tumor (Lpath). We observed the measurable lymph nodes on PET/CT preoperative, labeled them during operation, and examined for pathology. The t test and Wilcoxon Signed Ranks Test were used to assess the differences between groups, Spearman correlation and Pearson correlation was performed to assess the relationship between groups. A receiver operating characteristic curve was to determine the cut-off SUVmax for detecting lymph node metastases. Lpath was 6.87 ± 2.25 cm, L30% and L2.5 were 6.61 ± 1.76 cm and 7.56 ± 1.89 cm, respectively. L30% was more approximate to Lpath than the other % SUVmax, L2.5 was more approximate to Lpath than the other absolute SUV. The difference between L30% and L2.5 was not significantly different (p = 0.107). Two hundred eight lymph nodes were included in our study, 158 lymph nodes were negative, and 50 lymph nodes were positive. The 18F-FDG PET/CT had the best diagnostic performance for lymph nodes at the cut-off SUV of 2.7, the sensitivity, specificity, positive predictive value, negative predictive value, false-negative rate (FNR), false-positive rate (FPR), and accuracy of 18F-FDG PET/CT in detecting lymph node metastases were 72%, 83.54%, 58.06%, 90.41%, 28%, 16.46%, and 80.77%. To identify an SUV threshold for the distinction of benign lymph nodes from malignant lymph nodes, we plotted the error rates of positive and negative interpretations against a hypothetical SUV threshold. The minimal FNR+FPR was achieved at an SUV of 2.7. A 30% SUVmax and SUV 2.5 threshold of 18F-FDG PET/CT provided closest estimation of the length of gross tumor in EGJC. The 18F-FDG PET/CT had the best diagnostic efficiency at the cut-off SUVmax of 2.7 for detecting lymph node metastases in EGJC.