Geographic origin is a significant determinant of human papillomavirus prevalence in oesophageal squamous cell carcinoma: systematic review and meta-analysis.

BACKGROUND: Since the first reports in 1982 suggesting an aetiological role for human papillomavirus (HPV) in a subset of oesophageal squamous cell carcinomas (ESCC), the literature reporting HPV detection in ESCC has expanded rapidly. However no formal meta-analysis of this literature has been published yet. The objective of this study was to perform a systematic review and formal meta-analysis of the literature reporting HPV detection in ESCC.
METHODS: MEDLINE and Current Contents were searched through March 2012. The effect size was calculated as event rates and their 95% confidence interval (95% CI), with homogeneity testing using Cochran's Q and I² statistics. Meta-regression was used to test the impact of study-level covariates (HPV detection method, geographic origin of study) on effect size, and potential publication bias was estimated using funnel plot symmetry (Begg and Mazumdar rank correlation, Egger's regression, and Duval and Tweedie's trim and fill method).
RESULTS: Of the 1177 abstracts found, 152 studies were determined to be eligible for this meta-analysis. These 152 studies covered a total of 10,234 ESCC cases, analysed by different HPV detection methods in different geographic regions. Of these 10,234 cases, 3135 (30.6%) tested HPV-positive, translating to an effect size of 0.372 (95% CI 0.360-0.384; fixed effects model) and 0.290 (95% CI 0.251-0.31; random effects model). When stratified by HPV detection technique, there was a significant heterogeneity between the studies, but importantly, the between-strata summary comparison was not significant (random effects model; p = 0.440). In contrast, there was significant heterogeneity between the studies from the different geographic regions. In the maximum likelihood meta-regression, HPV detection method was not a significant study-level covariate, in contrast to the geographic origin of the study, which had a significant impact (p = 0.00005) on the summary effect size estimates. No evidence for significant publication bias was found in funnel plot symmetry testing. In the sensitivity analysis, all meta-analytic results appeared robust to all (n = 151) one-by-one study removals.
CONCLUSIONS: These meta-analysis results indicate that the reported wide variability in HPV detection rates in ESCC is not due to the HPV detection techniques, but is explained by the geographic origin of the study. These data substantiate the recently elaborated concept that ESCC might have a different aetiology in low-incidence and high-incidence geographic regions, HPV playing an important role only in the latter.