Occupational asbestos exposure and ovarian cancer: updated systematic review

Abstract Background The association between asbestos exposure and ovarian cancer has been questioned given the possible misdiagnosis of peritoneal mesothelioma as ovarian cancer. Aims To update a systematic review on ovarian cancer risk in women occupationally exposed to asbestos, exploring the association with the time since first exposure and the duration of exposure. Methods We searched PubMed from 2008 onwards, screened previous systematic reviews, combined standardized mortality ratios (SMR) using random effect models and quantified heterogeneity using the I2 statistic. To assess tumour misclassification, we compared the distribution of observed excess ovarian cancers (OEOC) to that expected (EEOC) from the distribution of peritoneal cancers in strata of latency and exposure duration. Results Eighteen publications (20 populations), including a pooled analysis of 21 cohorts, were included. The pooled SMR was 1.79 (95% confidence interval 1.38–2.31), with moderate heterogeneity between studies (I2 = 42%), based on 144 ovarian cancer deaths/cases. The risk was increased for women with indirect indicators of higher exposure, longer duration and latency, and lower for chrysotile than for crocidolite exposure. The effect of duration and latency could not be completely disentangled, since no multivariate analysis was available for time-related variables. The dissimilarity index between OEOC and EEOC for the time since first exposure was small suggesting a similar pattern of risk. Conclusions While some misclassification between ovarian and peritoneal cancers cannot be excluded, the observed excess risk of ovarian cancer should be added to the overall disease burden of asbestos.