A meta-analysis on diagnostic accuracy of serum HLA-G level for breast cancer

Introduction: According to the role of human leukocyte antigen (HLA)-G in tumor progression and tumor escape from immune system, this meta-analysis on diagnostic accuracy studies is designed to reach a pooled accuracy for this biomarker. Methods: The present work is meta-analysis on diagnostic accuracy studies using preferred reporting items for systematic reviews and meta-analyses (PRISMA) guideline. Evidence level of this study is 2a according to Oxford Centre for Evidence-Based Medicine. All documents comparing serum level of HLA-G both in breast cancer and in healthy controls using receiver operating characteristics (ROC) curve with reporting area under ROC curve (AUC) were eligible for inclusion. Results: Five articles including 754 participants were eligible for quantitative synthesis. The range of AUC of the selected studies was 0.735-0.953. The pooled AUC was 0.923 (95% confidence interval [CI] 0.888-0.959) based fixed effect model and 0.880 (95% CI 0.785-0.975) based on random effect model (P <0.001). I2 was 71.6% (P =0.007) and showed heterogeneity. Conclusion: This meta-analysis updated the level of evidence for using serum HLA-G in diagnosis of breast cancer. However this piece of evidence cannot be used as a diagnostic tool. This biomarker can be used for investigation of recurrence and response to treatment. Further studies are suggested to complete this evidence gap.

Keywords:

HLA-G , Meta-analysis , Breast Cancer , Diagnostic Accuracy Studies