Abstract
Abstract
Background
LncRNA H19 is upregulated considerably in gastric cancer and participates in its apoptosis, proliferation, migration, invasion, and metastasis. The objective of this meta-analysis is to demonstrate its diagnostic and prognostic values for gastric cancer.
Methods
PubMed, Embase and Web of Science were searched for relevant records published from the inception to July 2023. Quality In Prognosis Studies (QUIPS) was used to assess the risk of bias of prognostic studies, while Quality Assessment of Diagnostic Accuracy Studies−2 (QUADAS-2) was used to assess that of diagnostic studies. A random effect model synthesised the overall sensitivity, specificity and diagnostic odds ratio (DOR) of included diagnostic studies. Deek’s funnel plot and symmetric test were used to measure publication bias. Narrative synthesis was used to summarize prognostic studies. In addition, a bivariate diagnostic random-effects meta-analysis using the Reitsma approach with restricted maximum likelihood (REML) estimation was performed as a hierarchical analysis.
Results
17 out of 264 searched records from 3 databases were included in the meta-analysis. The pooled effect estimates of 9 diagnostic studies were 0.7472 (0.6592–0.8188) for sensitivity, 0.8169 (0.6922–0.8984) for specificity, and 12.66 (5.42–29.58) for DOR without publication bias. Nine out of 11 included prognostic studies reported a statistically significant relationship between lncRNA H19 and poor prognosis of gastric cancer. The additional bivariate model yielded an area under the summary ROC curve (AUC) of 0.818, with a positive likelihood ratio (LR+) of 3.286, a negative likelihood ratio (LR−) of 0.348, and a diagnostic odds ratio (DOR) of 10.10. The between-study standard deviations were 0.584 for logit sensitivity and 0.614 for logit false-positive rate, with a correlation of −0.472.
Conclusions
LncRNA H19 had diagnostic and prognostic values for gastric cancer. Given the high risk of bias, substantial heterogeneity, and limited number of studies, these findings should be interpreted cautiously.