OBJECTIVE: Currently available imaging modalities, including CT, ultrasound, and MRI are unsatisfactory in the detection of metastatic gastric cancer, especially lymph node metastases and peritoneal spread. The aim of this study is to evaluate FDG-PET in the diagnosis of primary as well as metastatic gastric cancer. METHOD: All patients (18M/5F; mean age 62, range 33-81) with gastric cancer referred for FDG-PET scan from 1/1/97 to 3/20/98 were studied. PET images of the neck, chest, and abdomen were acquired using a dedicated whole body scanner. A final diagnosis was reached in 21 patients by histology, surgical findings, or clinical follow-up. RESULTS: All but 1 primary tumor (12 out of 13) and both cases with local recurrence showed a high degree of FDG uptake, with mean SUV of 8.9 (range 4.8-17.6). The one false negative (FN) occurred in a poorly controlled diabetic with blood sugar of >400 at the time of study. Six cases with prior gastrectomy were true negative (TN) in the region of the stomach. This translates to a sensitivity of 93%, specificity of 100%, and accuracy of 95% for the primary site. In contrast, for intra-abdominal lymph node (LN) stations PET was true positive (TP) in 2, FN in 7, false positive (FP) in 1, and TN in 32, yielding a sensitivity of 22% and accuracy of 81%. PET detected only 2 out of 9 perigastric lymph node metastases. For the 8 patients in this series who had documented M1 disease, PET detected two with liver metastasis, one with colon metastasis, and one with submandibular LN. However, four cases of peritoneal spread were missed. CONCLUSION: FDG-PET is highly sensitive in detecting the primary lesion in gastric cancer and shows promise in the detection of liver and extra-abdominal metastasis. However, PET appears limited in the detection of perigastric lymphadenopathy as distinct from the primary tumor and in assessing peritoneal spread. Our preliminary assessment suggests that staging of gastric cancer with FDG-PET scanning is relatively effective at detecting distant metastatic disease and will complement standard staging methods such as laparoscopy, which are more effective at staging local nodal spread and peritoneal disease.
OBJECTIVE: Currently available imaging modalities, including CT, ultrasound, and MRI are unsatisfactory in the detection of metastatic gastric cancer, especially lymph node metastases and peritoneal spread. The aim of this study is to evaluate FDG-PET in the diagnosis of primary as well as metastatic gastric cancer. METHOD: All patients (18M/5F; mean age 62, range 33-81) with gastric cancer referred for FDG-PET scan from 1/1/97 to 3/20/98 were studied. PET images of the neck, chest, and abdomen were acquired using a dedicated whole body scanner. A final diagnosis was reached in 21 patients by histology, surgical findings, or clinical follow-up. RESULTS: All but 1 primary tumor (12 out of 13) and both cases with local recurrence showed a high degree of FDG uptake, with mean SUV of 8.9 (range 4.8-17.6). The one false negative (FN) occurred in a poorly controlled diabetic with blood sugar of >400 at the time of study. Six cases with prior gastrectomy were true negative (TN) in the region of the stomach. This translates to a sensitivity of 93%, specificity of 100%, and accuracy of 95% for the primary site. In contrast, for intra-abdominal lymph node (LN) stations PET was true positive (TP) in 2, FN in 7, false positive (FP) in 1, and TN in 32, yielding a sensitivity of 22% and accuracy of 81%. PET detected only 2 out of 9 perigastric lymph node metastases. For the 8 patients in this series who had documented M1 disease, PET detected two with liver metastasis, one with colon metastasis, and one with submandibular LN. However, four cases of peritoneal spread were missed. CONCLUSION:FDG-PET is highly sensitive in detecting the primary lesion in gastric cancer and shows promise in the detection of liver and extra-abdominal metastasis. However, PET appears limited in the detection of perigastric lymphadenopathy as distinct from the primary tumor and in assessing peritoneal spread. Our preliminary assessment suggests that staging of gastric cancer with FDG-PET scanning is relatively effective at detecting distant metastatic disease and will complement standard staging methods such as laparoscopy, which are more effective at staging local nodal spread and peritoneal disease.