Abstract:
Objective To evaluate the predictive value of intra-tumor metabolic heterogeneity index (HI) of primary tumor on preoperative 18F-fluorodeoxyglucose (FDG) PET/CT for occult lymph node metastasis (OLM) in gastric adenocarcinoma.
Methods The clinical data of 79 patients with gastric adenocarcinoma who underwent preoperative 18F-FDG PET/CT examination in the First Affiliated Hospital of Zhengzhou University from January 2016 to December 2022 were retrospectively analyzed. The participants included 62 males and 17 females, aged (63.8±9.0) years. All patients underwent radical resection of gastric adenocarcinoma within 1 month after 18F-FDG PET/CT imaging. According to the results of postoperative histopathological examination, they were divided into OLM positive group (n=39) and OLM negative group (n=40). The age, gender, primary tumor site, differentiation degree, Lauren classification, pathological T stage, carbohydrate antigen 199 (CA199), and carcinoembryonic antigen (CEA) were compared between the two groups. Chi-square test, two independent sample t test, and Mann-Whitney U test were used to compare the clinical characteristics and 18F-FDG PET/CT metabolic parameters of patients with gastric adenocarcinoma between groups. Univariate and multivariate Logistic regression models were used to analyze independent risk factors for predicting OLM. Receiver operating characteristic (ROC) curve was used to analyze the diagnostic efficacy of HI for OLM.
Results Gender, differentiation degree, and pathological T stage significantly differed between the OLM positive group and the OLM negative group (χ2=3.903, 6.061, 6.361, all P<0.05). The HI-2 of primary lesions in the OLM positive group was significantly higher than that in the OLM negative group (4.98 (2.68, 8.44) vs. 2.61 (1.84, 4.23)), and the difference was statistically significant (Z=−3.178, P=0.001). However, the maximum standardized uptake value (SUVmax), mean standardized uptake value (SUVmean), and HI-1 of primary lesions in the OLM negative group were significantly higher than those in the OLM positive group (6.91 (5.11, 10.64) vs. 5.59 (4.46, 7.51), 3.65 (3.25, 4.64) vs. 3.33 (3.06, 3.85), (0.29±0.14) vs. (0.23±0.12)), and the differences were statistically significant (Z=−2.000, −2.001, t=2.096; all P<0.05). Univariate Logistic regression model analysis showed that differentiation degree (poorly differentiated) (OR=4.037, 95%CI: 1.295–12.585, P=0.016), pathological T stage (T3–T4) (OR=4.080, 95%CI: 1.310–12.709, P=0.015), HI-1 (OR=0.025, 95%CI: 0.001–0.992, P=0.045) and HI-2 (>4.962) (OR=7.368, 95%CI: 2.385–22.764, P<0.001) were the risk factors for OLM. Multivariate Logistic regression model analysis showed that pathological T stage (T3–T4) (OR=4.780, 95%CI: 1.238–18.458, P=0.023) and HI-2 (>4.962) (OR=6.893, 95%CI: 1.922–24.718, P=0.003) were independent risk factors for OLM in patients with gastric adenocarcinoma. ROC curve analysis showed that the area under the curve of HI-2 for predicting OLM was 0.708 (95%CI: 0.237–0.483, P=0.001). When the optimal cut-off value of HI-2 was 4.962, the sensitivity and specificity for the diagnosis of OLM were 51.3% (20/39) and 87.5% (35/40), respectively.
Conclusion Preoperative 18F-FDG PET/CT intra-tumor metabolic HI of primary tumor has predictive value for OLM in gastric adenocarcinoma, and HI-2 is an independent risk factor for OLM.