Evaluation of ^sup 18^F-FDG PET in patients with advanced, metastatic, or recurrent gastric cancer

PET with 18F-FDG has been widely used in oncology, but its application for stomach neoplasms has been limited. The aim of this study was to evaluate the visual diagnostic accuracy of (18)F-FDG PET for advanced, metastatic, or recurrent gastric cancer and to generate semiquantitative values for lesions. 18F-FDG PET scans were obtained on 42 patients (29 men, 13 women; age, 27-78 y; median age, 63 y): 20 patients with a PT931/04 scanner and 22 patients with a SET2400W scanner. The PT931/04 has a spatial resolution of 6.0 mm at full width at half maximum (FWHM) and covers 15 cm above and below the targeted lesion, and the SET2400W has a spatial resolution of 3.9 mm at FWHM and images the entire body. All PET images were interpreted visually, and tracer uptakes were quantitated as standardized uptake values (SUVs) on SET2400W images. The sensitivity, specificity, and accuracy as a whole were as follows: 71%, 74%, and 73%, respectively, with the SET2400W scanner and 47%, 79%, and 62%, respectively, with the PT931/04 scanner. Values were high for primary lesions, liver, lymph node, and lung metastases, but were low for bone metastases, ascites, peritonitis, and pleuritis carcinomatoses. SUVs were 8.9 +/- 4.2 (primary lesions, 19 patients/19 lesions), 6.5 +/- 2.2 (liver, 9/55), 6.1 +/- 2.5 (lymph nodes, 14/38), 6.5 +/- 1.8 (abdominal wall, 4/7), 3.9 +/- 2.0 (bone, 3/27), and 4.7 +/- 2.6 (lung, 2/3). Comparing SUVs and histologic findings for 17 untreated patients, values for well-differentiated and moderately differentiated adenocarcinomas versus poorly differentiated adenocarcinomas and signet ring cell carcinomas were 13.2 +/- 6.3 (4/4) versus 7.7 +/- 2.6 (13/13) (P < 0.05) for the primary lesions, 7.0 +/- 2.4 (5/39) versus 5.6 +/- 2.8 (2/2) for the liver, and 5.5 +/- 1.9 (9/28) versus 8.8 +/- 3.3 (3/8) (P < 0.05) for the lymph nodes. Our results indicate that 18F-FDG PET is a useful diagnostic modality for advanced, metastatic, or recurrent gastric cancer but not for detecting bone metastases, peritonitis, or pleuritis carcinomatoses. 18F-FDG uptake by gastric cancers is relatively high but does not parallel histopathologic features of malignancy.