Objectives Data on the role of arsenic in renal cancer are suggestive but inconclusive. The present analysis aimed to determine whether renal cancers were more likely in Bangladeshi villagers exposed to high arsenic concentration in well water and, if so, whether this excess was limited to transitional cell cancers (TCC) or occurred also for renal cell cancers (RCC).

Methods Histology/cytology results from renal biopsies carried out at a single clinic in Dhaka, Bangladesh, from January 2008 to October 2011 were classified into four groups: RCC, TCC, other malignancy and benign. Patients aged ≥18 years using hand-pumped well water were identified by questionnaire, blind to diagnosis. Arsenic concentration was estimated from British Geological Survey reports for administrative area (thana) of residence. In a case-referent design (with benign results as referents), ORs were calculated by multilevel logistic regression adjusted for confounding. Time since well installation and smoking were examined by stratification.

Results Among 1489 cases included, 896 were RCC, 90 TCC and 503 benign. Arsenic concentration was estimated for 301 thanas with 63% of cases and 40% referents with arsenic concentration ≥50 µg/L (p<0.001). Risk increased monotonically with arsenic concentration ≥50 µg/L for both cell types (RCC and TCC). Risk estimates were greater in thana with early well installation where risk was increased for RCC in exposure stratum 10<50 µg/L (OR=2.47 95% CI 1.52 to 4.01). Stratification by ‘ever smoked’ confirmed the presence of risk in non-smokers.

Conclusions The relationship between arsenic concentration and both RCC and TCC suggests that arsenic is a causal factor in renal cancer.