Background: This study sought to systematically assess the diagnostic and prognostic value of serum amyloid A (SAA) in gastric cancer (GC).
Methods: PubMed, Embase, EBSCO, CNKI, and the Cochrane Library databases were searched for eligible studies. Extracted data were analyzed to determine diagnostic parameters and the summary receiver operating characteristic (SROC). Pooled hazard ratios (HR) and odds ratios (ORs) with their corresponding 95% confidence intervals (95% CIs) were calculated to summarize the effects.
Results: Six articles in English that met the inclusion criteria were identified. Meta-analysis of the included studies indicated high sensitivity of 0.84 (95% CI: 0.77 - 0.89) and moderate specificity of 0.61 (95% CI: 0.55 - 0.67) of SAA, with a diagnostic odds ratio (DOR) of 8.17 (95% CI: 4.82 - 13.86). The area under the receiver operating characteristic curve was 0.77. Survival analysis showed that SAA was associated with shorter survival time (HR = 4.42, p = 0.000; I2 = 0.0%). Stratified analyses showed that the assay of SAA from serum harbored higher efficacy than that from serum + plasma (AUC: 0.81 vs. 0.77), and SAA testing also achieved a better diagnostic performance in Asians than in Caucasians (AUC: 0.77 vs. 0.50).
Conclusions: Collectively, our analyses suggest that SAA can be used as a clinical auxiliary reference index for the diagnosis and prognosis prediction of GC; however, this diagnostic method is not independently sufficient. Our findings require confirmation in a larger prospective study.