Introduction Barrett’s Oesophagus (BO) is a precursor for Oesophageal Adenocarcinoma (OAC). Currently, there are no biomarkers in clinical practice that predict the malignant progression in BO. We sought to clarify the effectiveness of common genetic aberrations as potential biomarkers in this context through a systematic review and meta-analyses.

Methods MEDLINE, EMBASE and the Cochrane Library were searched by two independent reviewers to identify all clinical studies that assessed efficacy of p53, p16, Ki-67 and DNA content abnormalities as prognostic markers in BO. The main outcome measure was development of OAC. review and meta-analyses.

Results 104 clinical studies, with 12,353 samples were identified. Mutation [DOR 10.91 (2.46–48.42), sensitivity 47% (41–53%), specificity 92% (90–94%), PLR 4.71 (1.73–12.78), NLR 0.65 (0.50–0.85), AUC 0.7921] and loss of p53 [DOR 16.16 (6.09–42.88), sensitivity 31% (25–38%), specificity 98% (97–98%), PLR 6.66 (4.51–9.84), NLR 0.41 (0.11–1.47), AUC 0.9228] were found to be superior to the other p53 abnormalities (Loss of heterozygosity and Over-expression). Ki-67 was also noted to have high sensitivity in identifying high risk patients [DOR 5.54 (3.40–9.05), sensitivity 82% (78–85%), specificity 48% (44–52%), PLR 1.59 (1.20–2.10), NLR 0.42 (0.34–0.51), AUC 0.7607]. Aneuploidy [DOR 12.08 (8.09–18.03), sensitivity 53% (50–57%), specificity 87% (85–88%), PLR 4.26 (2.92–6.20), NLR 0.42 (0.32–0.55), AUC 0.8461], tetraploidy [DOR 5.87 (2.56–13.4), sensitivity 46% (39–53%), specificity 85% (82–87%), PLR 3.47 (1.98–6.05), NLR 0.65 (0.45–0.94), AUC 0.7926] and loss of Y chromosome [DOR 9.23 (4.34–19.63), sensitivity 68% (56–79%), specificity 80% (73–86%), PLR 2.67 (1.29–5.50), NLR 0.49 (0.35–0.68), AUC 0.8073] also predicted the malignant development with respectable accuracy but p16 aberrations [{Hypermethylation - DOR 3.00 (2.02–4.45), sensitivity 56% (51–61%), specificity 63% (59–66%), PLR 1.68 (1.35–2.08) (1.32–1.72), NLR 0.66 (0.54–0.80), AUC 0.6773}, {LOH - DOR 4.19 (1.85–9.53), sensitivity 55% (47–62%), specificity 54% (50–58%), PLR 1.84 (1.13–2.99), NLR 0.73 (0.60–0.88), AUC 0.6915}, {Mutation - DOR 1.61 (0.86–3.01), sensitivity 18% (13–24%), specificity 86% (83–89%), PLR 1.53 (0.97–3.42), NLR 0.96 (0.87–1.04), AUC 0.6890}, {Loss - DOR 2.18 (1.18–4.01), sensitivity 58% (53–63%), specificity 58% (53–63%), PLR 1.51 (1.08–1.95), NLR 0.74 (0.56–0.97), AUC 0.6314}] failed to demonstrate any advantage over the other biomarkers studied.

Conclusion Loss and mutation of p53 and Ki-67 effectively predict malignant progression in BO. A panel of biomarkers would be more suited to be included in surveillance programme. This will need to be confirmed in large, prospective clinical trials with cost- efficiency analyses.

Disclosure of Interest None Declared