Re: risk assessment of internal cancers from arsenic in drinking water. (Correspondence).

Arsenic in drinking water was the first environmental health problem to be recognized. It has been regulated for 100 years, first by the British Royal Commission on Arsenic Poisoning of 1903 (1), then by individual states in the United States, by the U.S. Department of Health in 1942, and since 1970, by the U.S. Environmental Protection Agency (U.S. EPA) (2). The scientific basis of all arsenic standards, including the present 50 [micro]g/L drinking water standard, is the toxic threshold concentration of around 250 [micro]g/L (1). Arsenic-related skin cancer has that same threshold (2). Morales et al. (3) should have acknowledged the contradiction of their bladder cancer interpretation with the much stronger skin cancer evidence.

The U.S. EPA decided to regulate chemicals with a procedure called "Risk Assessment" that extrapolates risk below the threshold to a point of 1/[10.sup.-6] risk as the regulatory target, or roughly 10,000 times below the threshold concentration. The 50 [micro]g/L standard already is as strict as is feasible. The U.S. EPA did not apply risk assessment to arsenic for 30 years and continued to use the 50 [micro]g/L standard. Now, in a controversial proposal, the U.S. EPA proposes to lower the standard 5-fold, using as official justification (4) the risk assessment by Morales et al. (3). The authors (3) state that

   ... our analysis suggests that the current standard of 50 [micro]g/L is
   associated with a substantial increased risk of cancer and is not
   sufficiently protective of human health.

The evidence presented by Morales et al. (3) in Table 5 of their paper can be interpreted as showing no increased risk in the 0-400 [micro]g/L concentration range. Numbers of cases, with the standardized mortality ratio (SMR; shown in parentheses) for the ranges 0-50, 50-100, 100-200, 200-300, and 300-400 are 26 (10.0), 12 (4.2), 12 (10.5), 8 (7.7), and 6 (7.5), respectively, for bladder cancer, and 30 (1.6), 31 (1.4), 21 (2.4), 24 (3.1), and 12 (2.0), respectively, for lung cancer. Bladder cancer is uniformly elevated relative to the control population, but neither the raw case numbers nor the associated mortality ratios show a positive trend in the 0-400 [micro]g/L range. The authors (3) admit that the "computed SMRs display a large amount of noise." A better interpretation for the data would be that there is an unexplained increase in bladder cancer. Above 600 [micro]g/L, both bladder and lung cancer are positively correlated with arsenic, but may be confounded by smoking.

Regarding the threshold, Morales et al. (3) used two models to estimate the effective level of 1/100 risk (E[D.sub.01]]); values for male bladder cancer were 395 and 351 [micro]g/L. The E[D.sub.01] is roughly identical with the limit and the threshold of real risk. Both estimates are supported by the data and are in line with the Royal Commission safe level near 250 [micro]g/L (1) and my interpretation that the threshold for bladder and lung cancer is > 400 [micro]g/L.

Scientists have a special obligation if their interpretations are used by governments for regulation. That obligation requires clarification of the statement by Morales et al. (3) that "... the current standard of 50 [micro]g/L is associated with a substantial increased risk of cancer" (p. 655). That statement is not supported by the data. It should also be kept in mind that two previous advisory panels on arsenic recommended that risk should not be extrapolated for arsenic (5). The most recent of these reports states that "[the U.S.] EPA has not requested, nor has the subcommittee endeavored to provide, a formal risk assessment for arsenic in drinking water" (6; p. 253).

The U.S. EPA has formally adopted the 10 [micro]g/L arsenic standard since this letter was submitted (7).

Gerhard Stohrer
Risk Policy Center
Larchmont, New York
E-mail: [e-mail]

REFERENCES AND NOTES

(1.) Royal Commission on Arsenic Poisoning. London:Wyman and Sons, 1903.

(2.) Stohrer G. Arsenic: opportunity for risk assessment. Arch Toxicol 65:525-531 (1991).

(3.) Morales KH, Ryan L, Kuo TL, Wu MM, Chen CJ. Risk Assessment of Internal Cancers from Arsenic in Drinking Water. Environ Health Perspect 108:655-661 (2000).

(4.) U.S. Environmental Protection Agency. Water supply: National primary drinking water regulations--arsenic; maximum contaminant level. Fed Reg 65:63027-63035 (2000).

(5.) U.S. EPA. EPA Science Advisory Board's Review of the Arsenic Issues Relating to the Phase II Proposed Regulation from the Office of Drinking Water. EPASABERC 89038. Washington, DC:U.S. Environmental Protection Agency, 1989.

(6.) National Research Council. Arsenic in Drinking Water. Washington, DC:National Academy Press, 1999.

(7.) Seelye KQ. EPA to adopt Clinton arsenic standard. New York Times 1 November 2001;A18.