Tobacco Documents Online

Endeavouring new shores in the estimation and assessment of the cancer risk by environmental materials (abstract) Pr Erich Hecker (As Pr. Hecker's paper arrived too late to be included in the program of the morning session, it will be presented in the afternoon).

Endeavouring new shores in the estimation and assesment of the cancer risk by environmental materials E. Hecker German Cancer Research Center, Research Program 3: Risk Factors of Cancer and Cancer Prevention, Division: Mechanisms of Tumori- genesis, Im Neuenheimer Feld 280, W-6900 Heidelberg, Germany Processes of environmental cancerogenesis1 may be surveyed in general mathematical terms [1], relating the dependent variable "response of the host by tumors" RT to all independent variables involved (see expr. (1) below), namely "observation time" r, "protocol" Peprf [i.e. the pattern of exposure to gnvironmental RT - f (r, Peprf , H) (1) principal Z:isk factor(s) (eprf)] and "host" H. on the background of exp. (1), using as experimental model mouse back skin (host/target tissue), quantitative dose/effect relations were established comparatively for the two principal processes of environmental cancerogenesis - i.e. for solitary cancerogenesis (SC)1 and conditiona7l cancerogenesis (CC) (type: initiation/pro- motion)l. As risk factors the PAH-type2 solitary cancerogen (and initiator) 7.12-dimethylbenz[a]anthracene (DMBA, doses ds) and three DTE-type3 conditional cancerogens (type: tumor promoter, doses dp), such as TPA4 and the (indole) alkaloid-type teleocidin were used in the highly standardized "protocol 16" [1,3]. It includs colony outbred initiated female NI+RI mice (di = 100 nmole of DMBA), and computer assisted evaluation of response data [3]. r was chosen to be identical with time t of chronic exposure to the eprf (in weeks). For five dose groups ds or dp (48 mice each), respectively - in addition to tumor rates and tumor yield - the I The terminology used in this paper follows the definitions of terms used frequently in chemical cancerogenesis as compiled and revisited under the auspices of the Deutsche Forschungsgemein- schaft (DFG), Bonn-Bad Godesberg, by an ad hoc group of toxi- cologists, pathologists, biochemists and representatives of public health authorities engaged in cancer research and/or preventive governmental legislation. In its original language it was published within the series of DFG documentations, Verlag Chemie, Weinheim/Bergstra8e. For the English version of it, including an English/German overview of the terms compiled and revised, see loc.cit. [2]. N Ln 0 E, ~ w CC C714

-2- cumulative tumor incidence F was read over at least 24 weeks as quantitative measures of RT. The time/effect functions F(t) including confidence intervals were established by a general Weibull function; median latency times t5p were read and calculated for all dose groups. It may be noted that from the statistical point of view it is most preferable to use the median latency time t50 as an experimental measure for RT. For the case of human exposure realistic incidence rates lay of course much below 50 % (e.g. between 1/105 - 100/105 or 1/103). Therefore, the doses ds or dp used in experimental models generally are much higher than it is usually the case in human exposure, equal sensitivity provided. For SC by the eprf DMBA [ds = 100(26,5); 5D(12,8); 50(12,8); 25(6,4); 10(2,56) and 5(1,28) nmoles(Ag)/mouse) the log/log plot of experimental t50 versus ds was linear within narrow confidence intervals following eq. (2) log t50 =- 1/n log ds + const', i.e. ds • t~0 = const5 (2) Thus the classical expectation for solitary cancerogens was reproduced, for DMBA for the first time (4]. In particular this result may be taken as a "positive control" to apply protocol 16 also to establish the quantitative dose/effect relations for the non-classical conditional cancerogens (type: tumor promoters). For CC each of the four promoters were used as eprf in dose ranges dp comparable to that used for DMBA. The log/log plot of ex- perimental t5p versus dp was non-linear in each of the four cases and with narrow confidence intervals. They were shown to follow a novel type of dose/effect function [4], eq. (3) log t5p = 109 (aondst') + log tmin , i.e. p' t (t50'tmin)(dp'dt) = const (3) In addition to t5p and dp, the novel function accounts for a minimal latency time tmin (of appearance of the first tumor in anyone dose group dp) and a threshold dose dt. The results presented for SC and CC in the experimenta3l model of mouse skin may be summarized in a descriptive general theory of 2 4 polyc~clic -4romatic hydrocarbon, 3 _diterpene _ester, i.e. 12-0-tetradecanoylphorbol-13-acetate (TPA), 3-0-tetra decanoylingenol (3-TI) and simplexin (SIM) n is a system(model) immanent constant called "amplification exponent". For the present purposes it may be neglected setting n = 1. Otherwise it my be determined experimentally for any one special model [see slope eq. (2)]. 5 Z5ai171387

-3- environmental chemical cancerogenesis (4] developed out of expr. (1) . If the results above are generalized for environmental materials in the widest possible sense, certain theoretical conclusions may be drawn, concerning e.g. mechanistic interpretations. They may assist, for example, to inaugurate short term assays as surrogates of long term testing for cancerogenic activity and will be dis- cussed in terms of "facts and fictions". From a generalization of the result obtained by the mouse skin model important practical consequences follow: (1) the Delany Amendment in its original text, applied in legislative practice to all kinds of environmental materials, more or less worldwide, does not reflect anymore the state of the art in biomedical science, especially of the theory of modern preventive toxicology. In fact it is outdated as illustrated by the cyclamat/saccharin case in the USA. (2) as an updated alternative for estimation and assessment of the cancer risk a novel and more sophisticated approach is proposed. (i) environmental materials may represent either one of two categories of risk factors of cancer with an experimentally defined, but graded potential of cancer risk: - first order risk factors of cancer: solitary cancerogens genotoxic cancerogens - second order risk factors of cancer: conditional cancerogens non-genotoxic cancerogens of type - tumor promoters (DTE, teleocidin) - growth stimulating hormones (estrogens, androgens) - others to be specified. (ii) estimation of the risk of cancer may take place individually for each singular case of environmental materials based upon classification into the categories of risk factor(s) above. (iii) environmental materials may be evaluated as to risk and benefit; no such material should be prohibited solely because it was shown to be solitary or conditional cancero- genic in animal experiments, as the Delany Amendment at least suggests and as infact it is partly practised. The proposed novel approach for estimation and assessment of the cancer Visk by environmental materials provides enough of flexi- bility to be extended or corrected by results of future research and may be practiced immediately. For example, in our laboratory the implications of the novel approach are presently investigated in trials to estimate and assess the risk of cancer by second order risk factors (promoters of the DTE-type) in certain environmental model situations.

-4- Ref erences : [1] R. Schmidt, E. Hecker, Biological assys for irritant, tumor-initiating and tumor-promoting activities, J Cancer Res Clin Oncol ;_I.4, 516-524 (1989). [2] K.E. Appel, G. FUrstenberger, H.J. Hapke, E. Hecker, A.G. Hildebrand, W. Koransky, F. Marks, H.G. Neumann, F.K. Ohnesorge and R. Schulte-Hermann, Guest Editorial, Chemical cancerogenesis: definitions of frequently used terms, J Cancer Res Clin Oncol jj&, 232-236 (1990). [3] L. Edler, R. Schmidt, E. Weber, F. Rippmann and E. Hecker, Biological assays for irritant, tumor-initiating and promoting activities III. Computer-assisted management and validation of biodata generated by standardized intiation/promotion protocols in skin of mice, J Cancer Res Clin Oncol ~, 205- 216 (1991). [4] E. Hecker and F. Rippmann, Outline of a descriptive general theory of environmental chemical cancerogenesis - experimental threshold doses for tumor promoters, Mechanism of environ- mental threshold doses for tumor promoters, Edited by A. Kappas, Plenum Press, New York (1990). N CJ'I ~ ~ . ~ -.3 ~ W CQ ~