Systematic literature searches were conducted until July 2010. Meta-analysis was performed to estimate prognostic effects of p53 alterations on patient outcomes in HCC. Sensitivity and subgroup analyses were also conducted in the meta-analysis.
Thirty-seven studies (7 tumour p53 mutation, 23 tumour p53 expression and 7 serum anti-p53 antibodies) were included in the systematic review and meta-analysis. The average percentages of p53 mutation, p53 expression upregulation and anti-p53 antibody level elevation in HCC patients were 31.5 % , 35.0 % and 23.8 % , respectively. Tumour p53 alterations were associated significantly with poor patient outcomes in HCC: the summary hazard ratio (HR) of mutant p53 versus wild type p53 phenotype was 2.58 [95 % confidence interval (CI): 1.96-3.41] for OS and 3.19 [95 % CI: 2.21-4.60] for RFS, respectively; and the summary HR of upregulated p53 expression versus low/undetectable p53 expression was 1.68 [95 % CI: 1.49-1.90] for OS and 1.89 [95 % CI: 1.34-2.66] for RFS, respectively. However, elevated serum anti-p53 antibody was only associated with poor OS in HCC group with a high proportion of (?50 % ) of hepatitis C virus (HCV) infection [HR: 1.92; 95 % CI: 1.30-2.85]. Moreover, sensitivity analyses showed that the results of meta-analyses were not altered.
HCC patients with p53 mutation and upregulated expression in tumour tissue have a shorter OS and RFS than patients with wild type p53 and low/undetectable p53 expression. However, the prognostic value of serum anti-p53 antibody is required to be further examined.