The accuracy of cell-free DNA blood-based tests in the detection of gastric cancer: An exploratory meta-analysis.

Abstract

e16158 Background: Gastric Cancer (GC) is the 5th most common type of cancer worldwide, accounting for around one million new cases per year. Although early detection of GC is critical in improving survival outcomes for patients, current screening methods, including upper endoscopies and tissue biopsies, are often invasive and cost prohibitive. Recently, however, several studies have emerged studying the use of cell-free DNA (cfDNA) blood-based tests as a non-invasive tool to screen for GC, which could allow for higher compliance and earlier detection in patients. The aim of this meta-analysis is to assess the diagnostic accuracy of cfDNA blood-based tests as a tool to detect GC. Methods: An extensive literature search was conducted across several medical databases, including EMBASE, Cochrane, and MEDLINE. Only studies testing the diagnostic accuracy of cfDNA blood-based tests specifically in the context of GC were included in the study. Data on validity metrics were extracted and pooled using the random effects model. I-squared statistics were used to estimate heterogeneity between study-specific estimates. Results: In total, this exploratory meta-analysis included 12 studies and 1,752 patients. The pooled sensitivity and specificity of the cfDNA blood-based tests for GC detection was found to be 0.823 (95% CI: 0.758-0.887) and 0.87 (95% CI: 0.802-0.917) respectively (Figure 1). The area under the curve (AUC) for the summary receiver operating characteristic (SROC) indicates high overall accuracy of the cfDNA test as a diagnostic tool (Figure 2). Furthermore, the pooled negative likelihood ratio (NLR) and pooled positive likelihood ratio (PLR) were found to be 0.196 (95% CI: 0.137-0.280) and 8.66 (95% CI: 5.85-12.84) respectively. The pooled Diagnostic Odds Ratio (DOR) was found to be 33.126 (95% CI: 15.021-73.054), again indicating the high accuracy of the cfDNA blood-based test. Conclusions: Overall, this meta-analysis demonstrates that cfDNA blood-based tools can serve as a highly accurate diagnostic tool for the detection of GC. Future studies are warranted to compare the performance of cfDNA to other diagnostic methods in reducing GC-related mortality. Summary of studies included in meta-analysis. Study Study Type Patients Controls Age (In healthy group) Sensitivity Specificity TP TN FP FN AUC 1 23456789101112 CCCCCCCCCCCCCCCCCCCCCCCC 106 10625549030110120351245096 118 118255985341398039925088 51.35 51.355961.6±12.6 59.063.79 ± 6.76 57.3 ± 10.3 50.04344959.2 ± 7.6 56.1 ± 11.3 86.79 83.9656756796.6788.293.977.178.9688.682 90.68 93.0788639194.1192.195.1592.491.8194.389 92 110144128212811327983179 107 8922373832977636843378 11 1732242913438210 14 81113141117826417 0.92 (0.88-0.96) 0.93 (0.89-0.960.73 (0.58–0.87) 0.7840.86 (0.86–0.94) 0.990.96 (0.94–0.98) 0.987 (0.985–0.99)0.86 (0.74-0.95)0.94 (0.91-0.97) 0.950.88 (0.79-0.96)