Reconciling animal and human data in a cancer risk assessment of acrylonitrile

Cancer dose–response assessment for acrylonitrile based upon rodent brain tumor incidence: Use of epidemiologic, mechanistic, and pharmacokinetic support for nonlinearity

“…The interspecies comparison is considered an improvement over previous comparisons between the epidemiology and animal data for AN (Collins and Strother, 1999;Schulz et al, 2001;Ward and Starr, 1993) for several reasons. First, the comparison across species is made in terms of internal dose, and therefore accounts for potential species diVerences in the kinetics of AN.…”

“…On the other hand, cancer bioassays in rats and mice consistently identify AN as a multi-site carcinogen following oral and inhalation exposures to high concentrations (>20 ppm in air; >30 ppm in water) Johannsen and Levinskas, 2002a,b;Maltoni et al, 1977Maltoni et al, , 1988Quast et al, 1980a,b). Several studies have been conducted in an attempt to evaluate this apparent discrepancy in a quantitative manner (Collins and Strother, 1999;Schulz et al, 2001;Ward and Starr, 1993). In comparing the results from eleven epidemiologic studies of AN-exposed workers to predicted brain tumor response from a rodent drinking water bioassay data, Ward and Starr (1993) concluded that, subject to certain caveats, the upper bound estimate of the inhalation cancer potency for AN is too high to be consistent with the human experience.…”

“…Human internal dose measures were predicted using information on exposure intensity for each decile or quartile of the corresponding cohorts. Because interspecies comparisons are sensitive to adjustments made to exposure for duration (Schulz et al, 2001), comparisons were also made using these internal doses after adjusting for less-than-lifetime exposures by multiplying each dose by the mean exposure duration for each cohort expressed as a fraction of lifetime (i.e., mean duration/70 years). The resulting internal dose measures for workers in each exposure decile or quartile span more than two orders of magnitude (Table 1).…”

“…The authors concluded that the epidemiology evidence indicates that AN is not a potent CNS tumorigen. Schulz et al (2001) compared the CNS cancers predicted from the most appropriate inhalation bioassay in rats to the CNS cancers observed in three recent, large epidemiologic studies. After adjusting the exposures for less than lifetime duration in the epidemiologic studies, the authors reported that the rodent-based predictions were well within their 95% conWdence intervals for CNS cancer mortality rates from the epidemiology studies.…”

Cytochrome P450 2E1 (CYP2E1) is Essential for Acrylonitrile Metabolism to Cyanide: Comparative Studies Using CYP2E1-Null and Wild-Type Mice

“…Early epidemiological studies have suggested that AN may increase the incidence of lung, colon, and stomach cancers among exposed workers (Thiess and Fleig, 1978;Blair et al, 1998). However, recent studies suggested that there is no significant association between human exposure to AN and carcinogenicity (Benn and Osborne 1998;Swaen et al, 1998;Wood et al, 1998;IARC, 1999;Marsh et al, 2001;Schulz et al, 2001). Animal studies demonstrated that AN is mutagenic, teratogenic, and a multisite carcinogen in rats and mice (Strother et al, 1988;IARC, 1999;Saillenfait and Sabate, 2000;NTP, 2001;Ghanayem et al, 2002).…”