ABSTRACT: To estimate the global attributable fraction of human papillomavirus (HPV) in lung cancer, we provided updated information through a system review and meta-analysis. We did a literature search on PubMed, Ovid and Web of Science to identify case-control studies and cohort studies that detected HPV in lung carcinomas. We included studies that tested 30 or more cases and were published before Feb 28, 2017. We collected information about gender, smoking status, HPV detection methods, HPV types, materials and clinical features. If it was not possible to abstract the required information directly from the papers, we contacted the authors. A meta-analysis was performed to calculate the pooled effect sizes (OR/RR) with 95% confidence intervals (CI) including subgroup analysis and meta-regression to explore sources of heterogeneity, by Stata 13.0 software. 36 case-control studies, contributing data for 6,980 cases of lung cancer and 7,474 controls from 17 countries and one cohort study with 24,162 exposed and 1,026,986 unexposed from China were included. HPV infection was associated with cancer of lung, pooled OR was 3.64 (95% CI: 2.60-5.08), calculated with the random-effects model. Pooled OR for allogeneic case-control studies, self-matched case-control studies and nested case-control studies were 6.71 (95% CI: 4.07-11.07), 2.59 (95% CI: 1.43-4.69) and 0.92 (95% CI: 0.63-1.36), respectively. Pooled OR for HPV 16 and HPV 18 infection, were 3.14 (95% CI: 2.07-4.76) and 2.25 (95% CI: 1.49-3.40), respectively. We also found that HPV infection may be associated with squamous cell carcinoma, adenocarcinoma and small cell carcinoma. There is evidence that HPV infection, especially HPV 16 and HPV 18 infection, significantly increase the risk of lung cancer. Future research needs to focus attention toward whether an HPV vaccine can effectively reduce the incidence of lung cancer.
