Risk assessment for PFOA and kidney cancer based on a pooled analysis of two studies.

Abstract

Perfluorooctanoic acid (PFOA) has been associated with kidney cancer in human studies. We conducted a pooled analysis of two large studies of PFOA and renal cell carcinoma (RCC, the most common type of kidney cancer); one from the National Cancer Institute (NCI) (324 cases and controls), and a second from the C8 Science Panel (103 cases and 511 controls). Serum PFOA levels were estimated a median of 8years before diagnosis. Analyses were conducted via conditional logistic regression. Lifetime risk of kidney cancer per unit serum PFOA concentration and per unit dose were calculated. The 25th, 50th and 75th percentiles of serum PFOA levels were 4.8, 7.3, and 23.9ng/ml for the pooled analysis. The preferred model for the pooled datawas a two-piece linear spline model (knot at 12.5ng/ml serum PFOA); the log odds of RCC increased 0.1349 per 1ng/ml increase in serum PFOA up to the knot (eg, an OR of 2.02 (1.45-2.80) from the median to the knot), and was flat thereafter. The estimated lifetime excess risk (cancer slope factor) with an exposure of 1ng/ml was 0.0018, similar to the excess risk of 0.0026 recently reported by CalEPA based on different methods. Assuming a serum half-life of 2.3years and a distribution volume of 170ml/kg for PFOA, our results are equivalent to 0.0128 per ng/kg/d of PFOA intake. To limit excess lifetime kidney cancer risk to 1/1,000,000, our data suggest a limit of 0.0015ng/L (0.0015 ppt) for PFOA in drinking water, similar to CalEPA's proposed Public Health Goal and the new US EPA Drinking Water Health Advisory. Our results correspond reasonably well with cancer slope factors developed by other investigators using published summary data, and suggest drinking water limits similar to new recommendations by the US EPA.