Value and impact factors of multidetector computed tomography in diagnosis of preoperative lymph node metastasis in gastric cancer: A PRISMA-compliant systematic review and meta-analysis

BACKGROUND: Multidetector computed tomography (MDCT) exhibited wide ranges of sensitivities and specificities for lymph node assessment of gastric cancer (GC) in several individual studies. This present meta-analysis was carried out to evaluate the value of MDCT in diagnosis of preoperative lymph node metastasis (LNM) and to explore the impact factors that might explain the heterogeneity of its diagnostic accuracy in GC. METHODS: A comprehensive search was conducted to collect all the relevant studies about the value of MDCT in assessing LNM of GC within the PubMed, Cochrane library and Embase databases up to Feb 2, 2016. Two investigators independently screened the studies, extracted data, and evaluated the quality of included studies. The sensitivity, specificity, and area under ROC curve (AUC) were pooled to estimate the overall accuracy of MDCT. Meta-regression and subgroup analysis were carried out to identify the possible factors influencing the heterogeneity of the accuracy. RESULTS: A total of 27 studies with 6519 subjects were finally included. Overall, the pooled sensitivity, specificity, and AUC were 0.67 (95% CI: 0.56–0.77), 0.86 (95% CI: 0.81–0.90), and 0.86 (95% CI: 0.83–0.89), respectively. Meta-regression revealed that MDCT section thickness, proportion of serosal invasion, and publication year were the main significant impact factors in sensitivity, and MDCT section thickness, multiplanar reformation (MPR), and reference standard were the main significant impact factors in specificity. After the included studies were divided into 2 groups (Group A: studies with proportion of serosa-invasive GC subjects ≥50%; Group B: studies with proportion of serosa-invasive GC subjects <50%), the pooled sensitivity in Group A was significantly higher than in Group B (0.84 [95% CI: 0.75–0.90] vs 0.55 [95% CI: 0.41–0.68], P < .01). For early gastric cancer (EGC), the pooled sensitivity, specificity, and AUC were 0.34 (95% CI: 0.15–0.61), 0.91 (95% CI: 0.84–0.95), and 0.83 (95% CI: 0.80–0.86), respectively. CONCLUSION: To summarize, MDCT tends to be adequate to assess preoperative LNM in serosa-invasive GC, but insufficient for non-serosa-invasive GC (particularly for EGC) owing to its low sensitivity. Proportion of serosa-invasive GC subjects, MDCT section thickness, MPR, and reference standard are the main factors influencing its diagnostic accuracy.