Abstract
Arsenic in drinking water is causally linked with cancer of the skin, lung, and urinary bladder, but there is very little data on a possible role for arsenic in the etiology of cervical cancer, a disease in which human papilloma virus is held to be a necessary but not sufficient cause. All histopathology results from cervical specimens from the National Institute of Cancer Research and Hospital (NICRH), Dhaka (1997-2015), and the Anowara Medical Services (2003-2015), both serving the whole of Bangladesh, were classified by cell type. Arsenic concentrations in well water in the thana of residence were estimated from systematic sampling carried out by the British Geological Survey. In a case-referent analysis arsenic estimates for cases of cervical cancer were compared with those found to have benign lesions. In this study, 3464 NICRH (CH) cervical specimens and 30,050 community medical service (CMS) specimens were available: 3329 (CH) and 899 (CMS) were recorded as malignant. Most were squamous cell carcinoma, of which 4.9% were poorly differentiated. Overall, there was no increase in cervical cancer with increasing arsenic concentration. Among those with squamous cell histology, a strong dose response was seen for poorly differentiated cancer with increasing arsenic exposure. The odds ratio increased monotonically, compared with exposure <10 μg/L, from 1.58 at 10 < 50 μg/L to 8.11 at >200 μg/L (p < 0.001). Given the high proportion of Bangladeshis using drinking water containing >50 μg/L of arsenic, the evidence that arsenic is implicated in cancer grade suggests a need for further investigation and the introduction of cervical screening in high arsenic areas.