Objective  To compare the ¹⁸F-fluorodeoxyglucose (FDG) PET/CT imaging characteristics of primary gastric diffuse large B-cell lymphoma (GDLBCL) and gastric cancer, evaluate the value of ¹⁸F-FDG PET/CT in differential diagnosis of GDLBCL and gastric carcinoma.

Methods  The clinical data, imaging data and digestive endoscopy data of 71 primary GDLBCL and 50 gastric cancer patients with histopathologically confirmed in Nanjing Drum Tower Hospital, the Affiliated Hospital of Nanjing University Medical School from January 2018 to December 2022 were analyzed retrospectively. Patients with primary GDLBCL included 33 males and 38 females, aged 25–90 years, with a median age of 62 years. Patients with gastric cancer included 36 males and 14 females, aged 26–93 years, with a median age of 67 years. The features evaluated included lesion thickness of CT and PET/CT imaging type, maximum standardized uptake value (SUV_max), maximum thickness of lesion (THK_max), and ratio of SUV_max to THK_max (SUV_max/THK_max) between patients with primary GDLBCL and gastric cancer, and lesions involvement scope shown by digestive endoscopy and ¹⁸F-FDG PET/CT were compared. Differences in SUV_max, THK_max, and SUV_max/THK_max between groups were calculated by Mann-Whitney U test. Receiver operating characteristic (ROC) curve was used to evaluate the efficacy of quantitative indicators in differential diagnosis of primary GDLBCL and gastric cancer.

Results  Different ¹⁸F-FDG PET/CT imaging types were found in primary GDLBCL (type Ⅰ: 28 cases (28/71, 39.4%), type Ⅱ: 16 cases (16/71, 22.5%), and type Ⅲ: 27 cases (27/71, 38.0%)) and gastric cancer (type Ⅰ: 13 cases (13/50, 26.0%), type Ⅱ: 31 cases (31/50, 62.0%), and type Ⅲ: 6 cases (6/50, 12.0%)). The SUV_max, THK_max, and SUV_max/THK_max of primary GDLBCL lesions were higher than those of gastric cancer patients (16.8 (11.2, 23.8) vs.7.2 (4.8, 10.9), 2.2 (1.6, 3.6) cm vs.1.9 (1.5, 2.3) cm, 6.2 (5.0, 9.7) vs. 3.7 (2.4, 6.7)), and differences were statistically significant (Z=−6.342, −2.093, −4.541; all P<0.05). The ROC curve analysis results showed that the cut-off values of SUV_max, THK_max, and SUV_max/THK_max for differential diagnosis of primary GDLBCL and gastric cancer were 12.7 (area under curve(AUC): 0.839, sensitivity: 0.687, specificity: 0.863; P<0.001), 3.6 cm (AUC: 0.612, sensitivity: 0.284, specificity: 0.798; P<0.05), and 4.14 (AUC: 0.743, sensitivity: 0.861, specificity: 0.582; P<0.001). The lesion areas in the digestive endoscopy of four (4/35, 11.4%) patients with primary GDLBCL were smaller than that of ¹⁸F-FDG PET/CT and negative digestive endoscopy results in three (3/42, 7.1%) patients with gastric cancer, but the tumor lesions were visible on ¹⁸F-FDG PET/CT.

Conclusions  ¹⁸F-FDG PET/CT imaging types could help in the differential diagnosis of GDLBCL and gastric carcinoma. SUV_max=12.7 as cut-off value has high value in differential diagnosis of GDLBCL and gastric cancer. ¹⁸F-FDG PET/CT could identify more gastric lesions with GDLBCL and gastric cancer patients than digestive endoscopy.