Diagnostic efficacy of contrast-enhanced gastric ultrasonography in staging gastric cancer: a meta-analysis

Abstract

Background: As comprehensive surgical management for gastric cancer becomes increasingly specialized and standardized, the precise differentiation between ≤T1 and ≥T2 gastric cancer before endoscopic intervention holds paramount clinical significance.

Objective: To evaluate the diagnostic efficacy of contrast-enhanced gastric ultrasonography in differentiating ≤T1 and ≥T2 gastric cancer.

Methods: PubMed, Web of Science, and Medline were searched to collect studies published from January 1, 2000 to March 16, 2023 on the efficacy of either double contrast-enhanced gastric ultrasonography (D-CEGUS) or oral contrast-enhanced gastric ultrasonography (O-CEGUS) in determining T-stage in gastric cancer. The articles were selected according to specified inclusion and exclusion criteria, and the quality of the included literature was assessed using the Quality Assessment of Diagnostic Accuracy Studies-2 scale. Meta-analysis was performed using Stata 12 software with data from the 2 × 2 crosslinked tables in the included literature.

Results: In total, 11 papers with 1124 patients were included in the O-CEGUS analysis, which revealed a combined sensitivity of 0.822 (95% confidence interval [CI] = 0.753-0.875), combined specificity of 0.964 (95% CI = 0.925-0.983), and area under the summary receiver operating characteristic (sROC) curve (AUC) of 0.92 (95% CI = 0.89-0.94). In addition, five studies involving 536 patients were included in the D-CEGUS analysis, which gave a combined sensitivity of 0.733 (95% CI = 0.550-0.860), combined specificity of 0.982 (95% CI = 0.936-0.995), and AUC of 0.93 (95% CI = 0.91-0.95). According to the I² and P values ​​of the forest plot, there was obvious heterogeneity in the combined specificities of the included papers. Therefore, the two studies with the lowest specificities were excluded from the O-CEGUS and D-CEGUS analyses, which eliminated the heterogeneity among the remaining literature. Consequently, the combined sensitivity and specificity of the remaining studies were 0.794 (95% CI = 0.710-0.859) and 0.976 (95% CI = 0.962-0.985), respectively, for the O-CEDUS studies and 0.765 (95% CI = 0.543-0.899) and 0.986 (95% CI = 0.967-0.994), respectively, for the D-CEGUS studies. The AUCs were 0.98 and 0.99 for O-CEGUS and D-CEGUS studies, respectively.

Conclusion: Both O-CEGUS and D-CEGUS can differentiate ≤T1 gastric cancer from ≥T2 gastric cancer, thus assisting the formulation of clinical treatment strategies for patients with very early gastric cancer. Given its simplicity and cost-effectiveness, O-CEGUS is often favored as a staging method for gastric cancer prior to endoscopic intervention.

Keywords: Contrast agent; Gastric cancer; Gastric ultrasonography; Meta-analysis; T-stage.

Fig. 1

Flow chart of literature screening

Fig. 2

Risk of bias and applicability concerns. A Quality evaluation of the included studies (n = 14). Red indicates high risk, green indicates low risk, and yellow indicates incomplete information that cannot be assessed. B Authors’ judgments about each domain presented as percentages across the included studies

Fig 3

The combined sensitivity of O-CEGUS/D-CEGUS for discriminating ≤T1 and ≥T2 gastric cancer. a The sensitivity forest plot of the 11 included O-CEGUS–related papers. b The sensitivity forest plot of the five included D-CEGUS–related papers. c Sensitivity forest plots of nine O-CEGUS–related papers after excluding the sources of heterogeneity. d Sensitivity forest plots of four D-CEGUS–related papers after exclusion the sources of heterogeneity

Fig. 4

The combined specificity of O-CEGUS/D-CEGUS for discriminating ≤T1 and ≥T2 gastric cancer. a The specificity forest plot of the 11 included O-CEGUS-related papers. b The specificity forest plot of the five included D-CEGUS–related papers. c Specificity forest plots of nine O-CEGUS–related papers after excluding the sources of heterogeneity; Figure 4d: Specificity forest plots of four D-CEGUS–related papers after excluding the sources of heterogeneity

Fig. 5

The sROC curves of O-CEGUS/D-CEGUS for discriminating ≤T1 and ≥T2 gastric cancer. a The sROC curves of the 11 included O-CEGUS–related papers. b The sROC curves of the five included D-CEGUS–related papers. c The sROC curves of nine O-CEGUS–related papers after excluding the sources of heterogeneity. d The sROC curves of four D-CEGUS–related papers after excluding the sources of heterogeneity. An empty circle represents each study’s sensitivity/specificity. A filled black circle represents the summary point for sensitivity/specificity. Dotted closed line, 95% confidence interval of the summary point; dashed closed line, 95% prediction region

Fig. 6

Funnel plot for publication bias among the included studies. a Funnel plot of 11 papers related to O-CEGUS. b Funnel plot of five papers related to D-CEGUS. c Funnel plot of nine papers related to O-CEGUS after excluding the sources of heterogeneity. d Funnel plot of four papers related to D-CEGUS after excluding the sources of heterogeneity. Each red circle represents a study