To the Editor:
Olsen and Ley have commented on our findings for prostate cancer incidence among workers exposed to perfluorooctanoic acid (PFOA) at a DuPont plant in West Virginia.1 As they note, we found a modest positive trend in incidence with increasing PFOA exposure (P = 0.11 on the basis of a linear test of categorical data), with rate ratios of 1.00, 1.81, 2.45, and 1.88 by quartile of estimated cumulative serum PFOA levels (129 cases). We stated in our report that these positive findings found “some support” in earlier reports by Vieira et al2 (incidence, case-control), Eriksen et al3 (incidence, cohort), and Lundin et al4 (mortality, cohort at the 3M Cottage Grove plant).2–4 We now also note one further report, a case-control study in which higher serum PFOA concentrations were associated with significantly higher prostate cancer risk, but only among those with a family history of prostate cancer.5
It is certainly true that not all studies found a positive association between PFOA and prostate cancer. As Olsen and Ley note, a later follow-up (Raleigh et al)6 of a larger cohort that included the cohort studied by Lundin et al., found little association between PFOA and prostate cancer incidence (rate ratios = 1.00, 0.80, 0.85, 0.89, and 1.11) (188 exposed cases, 253 nonexposed referent cases, no trend test presented), which argues against a true association.
We agree that prostate cancer incidence is a better measure than mortality, and we should have noted the negative findings of Raleigh et al6 in our report. Nevertheless, analyses of prostate cancer incidence can also have limitations; for example, if the referent group in Raleigh et al, who were workers at a different 3M plant, were different in important ways from the group of exposed workers (such as if they had higher rates of prostate cancer screening or a different prevalence of prostate cancer risk factors [eg, race or other chemical exposures], which were not controlled for in the models), their prostate cancer rates would not be directly comparable. If Raleigh et al6 had used their lowest quartile of cumulative PFOA air levels as their referent (analogous to what we did), thereby conducting a solely internal analysis, their rate ratios would have been 1.00, 1.06, 1.11, and 1.39, showing a modest positive trend.
We also note that the exposure in Raleigh et al6 was estimated air levels, not directly comparable to our modeled serum levels. The correlation between our modeled and observed serum levels for more than 2000 measurements was a respectable 0.67. No such correlation was presented in Raleigh et al, making it impossible to know how well the modeled air concentrations compared with actual exposure, although these authors did present some serum data, indirectly supporting their broad exposure rankings.
As pointed out by Olsen and Ley, in the International Agency for Research on Cancer's recent classification of PFOA as a group 2B carcinogen (possibly carcinogenic), the International Agency for Research on Cancer found the strongest human evidence to date from findings for testicular and kidney cancer, not prostate cancer. Nevertheless, it should be noted that, in general, the epidemiologic studies of PFOA remain rather sparse, and are inconsistent. Further studies of the highly exposed occupational cohorts at 3M and DuPont, as well as other populations exposed to PFOA above low background levels, will help clarify this situation.
Kyle Steenland, PhD
Andrea Winquist, PhD, MD
The Rollins School of Public Health,
Emory University, Atlanta, Ga.
1. Steenland K, Zhao L, Winquist A. A cohort incidence study of workers exposed to perfluorooctanoic acid (PFOA). Occup Environ Med. 2015. pii: oemed-2014-102364. doi:10.1136/oemed-2014-102364. [Epub ahead of print].
2. Vieira VM, Hoffman K, Shin HM, Weinberg JM, Webster TF, Fletcher T. Perfluorooctanoic acid exposure and cancer outcomes in a contaminated community: a geographic analysis. Environ Health Perspect. 2013;121:318–323.
3. Eriksen KT, Sorensen M, McLaughlin JK, et al. Perfluorooctanoate and perfluorooctanesulfonate plasma levels and risk of cancer in the general Danish population. J Natl Cancer Inst. 2009;101:605–609.
4. Lundin JI, Alexander BH, Olsen GW, Church TR. Ammonium perfluorooctanoate production and occupational mortality. Epidemiology. 2009;20:921–928.
5. Hardell E, Kärrman A, van Bavel B, Bao J, Carlberg M, Hardell L. Case-control study on perfluorinated alkyl acids (PFAAs) and the risk of prostate cancer. Environ Int. 2014;63:35–39.
6. Raleigh KK, Alexander BH, Olsen GW, et al. Mortality and cancer incidence in ammonium perfluorooctanoate production workers. Occup Environ Med. 2014;71:500–506.