Clinicopathological and prognostic significance of cancer stem cell markers CD44 and CD133 in patients with gastric cancer: A comprehensive meta-analysis with 4729 patients involved

BACKGROUND: In recent years, CD44 and CD133 have been identified as 2 common used cancer stem cell (CSC) markers in gastric cancer. However, the clinicopathological and prognostic value of these markers in gastric cancer remains controversial; moreover, there is lack of comparison of these 2 markers’ roles in clinical applications. A systematic review and meta-analysis was conducted to elucidate these markers’ clinicopathological features and association with prognosis in patients with gastric cancer. METHODS: Eligible studies were identified and odds ratios (ORs), hazard ratios (HRs) with 95% confidence intervals (CIs) were estimated. Heterogeneity and sensitivity were analyzed as well. Publication bias was assessed using funnel plots and Egger tests. RESULTS: The meta-analysis included 26 studies involving 4729 patients. High expression of CD44 was associated with Lauren type (intestinal type) (OR, 1.53 [95% CI, 1.02–2.30]; P = 0.038) and lymphatic vessel invasion (OR, 1.36 [95% CI, 1.06–1.76]; P = 0.021). CD133 overexpression was related to high TNM stage (III/IV) (OR, 3.18 [95% CI, 2.48–4.07]; P = 0.000), high depth of invasion (T3/T4) (OR, 2.97 [95% CI, 2.20–4.03]; P = 0.000), lymph node metastasis (OR, 2.82 [95% CI, 2.16–3.69]; P = 0.000), vascular invasion (OR, 6.71 [95% CI, 1.63–27.63]; P = 0.008), and distant metastasis (OR, 2.32 [95% CI, 1.64–3.29]; P = 0.000). In addition, survival analysis demonstrated a significant association between CD44, as well as CD133 and poor 5-year overall survival (HR, 1.87 [95% CI, 1.55–2.26]; P = 0.000; HR, 2.07 [95% CI, 1.76–2.44]; P = 0.000, respectively). CONCLUSION: These data suggest that upregulated expression of CD44 and CD133 correlates with several clinicopathological features and poor prognosis. Since the related features do not overlap, combined detection of CD44 and CD133 expression can be an especially effective tool for pathological diagnosis and prognostic prediction of gastric cancer patients in clinical applications.