Biomedical subjects

C J Portier

Publications and source records attributed to C J Portier.

At least 73 records · Page 4Linked to original sources

PubMed1993

An evaluation of some methods for fitting dose-response models to quantal-response developmental toxicology data.

The analysis of quantal-response developmental toxicology data by dose-response modeling is discussed, with emphasis on methods that avoid exact distributional assumptions. These methods (quasi-likelihood, bootstrapping, and jackknifing) are contrasted with analyses based on the beta-binomial distribution. For the resampling procedures, dose-response models are fit under a binomial likelihood. A justification for this choice of estimator in resampling plans is given, based on an extension of the standard results for asymptotic normality and consistency of maximum likelihood estimators. This justification depends only on the true distribution of the data having the usual binomial expectation. A quasi-likelihood approach is also considered, in which simple assumptions about the intralitter correlation structure are made. Quasi-likelihood methods are in theory asymptotically robust to misspecification of the intralitter correlation structure. The practical implications of these asymptotic results are evaluated in a simulation study.

Abnormalities, Drug-Induced↗

PubMed1993

A measure of tumorigenic potency incorporating dose-response shape.

Many researchers have considered the problem of ranking chemical agents based on their carcinogenic potency. Sawyer et al. (1984, Biometrics 40, 27-40) proposed a carcinogenic potency estimate that incorporates both intercurrent mortality and background tumor rates. Since then, many authors have either generalized the method outlined by Sawyer et al. or developed their own method based on a slightly different adjustment for treatment-related changes in survival. None of these methods, however, has estimated the shape of the dose-response curve and incorporated such an estimate in potency estimation. In this manuscript, a measure of tumorigenic potency is proposed that utilizes the estimated shape of the dose-response relationship, in addition to estimated dose effects, in order to rank chemicals on the basis of carcinogenic risk. Comparison of this new measure to that of Sawyer et al. is done using a large database of animal carcinogenicity experiments from the National Cancer Institute and the National Toxicology Program.

Animals↗

PubMed1993

An index of tumorigenic potency.

A new index of tumorigenic potency is derived using a survival-adjusted quantal response estimator given by Bailer and Portier (1988, Biometrics 44, 417-431). The small-sample bias and mean squared error properties of this estimator are explored using computer simulations of animal carcinogenicity experiments and are compared to the characteristics of the TD50, an estimator proposed by Sawyer et al. (1984, Biometrics 40, 27-40). It is observed that this new index tends to have smaller bias and mean squared error relative to the TD50 over varying shapes of the tumor onset distribution, levels of tumor lethality, and levels of treatment lethality.

Animals↗

PubMed1992

Dose-response relationships for chronic exposure to 2,3,7,8-tetrachlorodibenzo-p-dioxin in a rat tumor promotion model: quantification and immunolocalization of CYP1A1 and CYP1A2 in the liver.

The mechanisms responsible for the braod spectrum of effects of 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) are not entirely clear but seem to involve an initial interaction with the Ah receptor. A major uncertainty in risk assessment for TCDD is the lack of adequate dose-response relationships following chronic exposure to TCDD. Induction of cytochrome P-450 enzymes (CYP1A1 and CYP1A2) is one of the most sensitive responses to TCDD and its structural analogues. We have used a two-stage model for hepatocarcinogenesis in female Sprague-Dawley rats to evaluate dose-response relationships for induction of CYP1A1 and CYP1A2 in diethylnitrosamine-initiated as well as in noninitiated rats. After initiation with a single dose of diethylnitrosamine, TCDD was administered biweekly by p.o. gavage at doses equivalent to 3.5, 10.7, 35.7, and 125 ng/kg/day for 30 weeks. CYP1A1 and CYP1A2 concentrations were quantified in hepatic microsomes by radioimmunoassay and localized in hepatic tissue slices by immunohistochemical techniques. Radioimmunoassay data revealed a maximum induction of 200-fold for CYP1A1 and 10-fold for CYP1A2 and there were no statistically significant differences between initiated and noninitiated rats. Induction at the lowest dose (3.5 ng/kg/day) was 20-fold for CYP1A1 and 3-fold for CYP1A2. Mathematical analysis indicates that the best fit of the induction data are inconsistent with a threshold for this response. There was a linear relationship between administered dose and TCDD liver concentration over the entire dose range of the study. This indicates that induction of CYP1A2 does not significantly alter the distribution of TCDD in our chronic dosing regimen. Immunolocalization of CYP1A1 and CYP1A2 revealed the same localization and induction pattern for both isozymes in the cytoplasm of hepatocytes. However, the hepatic distribution pattern was not uniform with the most intense staining observed around central veins. These studies help to clarify dose-response relationships for dioxin-mediated effects and demonstrate different sensitivity of hepatocytes to the effects of TCDD.

Animals↗

PubMed1992

Birth and death/differentiation rates of papillomas in mouse skin.

A simple one-stage model for the formation of papillomas is used to evaluate data from a series of initiation/promotion/stop-promotion experiments performed on NMRI mice. These experiments used a single application of 100 nmol 7,12-dimethylbenz[a]anthracene followed by twice weekly applications of 5 nmol 12-O-tetradecanoylphorbol-13-acetate. It is shown that, on the basis of this model, the data considered here provide no evidence for a growth advantage of initiated cells versus normal cells. Although model estimates always have to be treated with utmost caution, they can at least be interpreted in a qualitative way. In consequence, it has to be concluded either that the mechanism upon which the model is based fails to describe properly the growth behaviour of initiated cells, or that initiated cells do not possess a growth advantage over normal cells, even in the presence of a promoter.

9,10-Dimethyl-1,2-benzanthracene↗

PubMed1992

Concordance of carcinogenic response between rodent species: potency dependence and potential underestimation.

The use of average qualitative concordance between two bioassay endpoints is considered, with emphasis directed at agreement between rats and mice from results of long-term carcinogenicity studies. It is noted that concordance varies as a function of the underlying potency or toxicity of the chemicals over which the averaging is performed. Thus, the averaging process dilutes large observed concordances from potent chemicals, and possibly inflates lower observed concordances from weakly active chemicals. Stratification over some measure of potency is suggested as a method for taking these effects into account. Statistical simulations of concordance analyses limited to low-potency ranges are employed to examine the concordance measure in greater detail. It is seen that at low potencies, observed concordance is consistently underestimated, reaching maximum levels of only about 80%.

Animals↗

PubMed1991

The application of a multistage model that incorporates DNA damage and repair to the analysis of initiation/promotion experiments.

In a previous article, a multistage model of carcinogenesis was introduced that takes into account the role of DNA damage, DNA repair, and cell replication on the incidence of malignancies. For this model the number of detectable clones of initiated cells is derived and model parameters are estimated using data arising from a two-stage skin-painting experiment in mice. The data from this experiment are interpretable in terms of the cellular events involved in initiation and promotion.

Animals↗

PubMed1991

Should the presence of carcinogens in breast milk discourage breast feeding?

Pollutant chemicals are commonly found in human milk at levels that would prevent its sale as a commercial food for infants. The chemicals found most commonly are dichlorodiphenyl-dichloroethene, polychlorinated biphenyls, dieldrin, chlordane, heptachlor, and polychlorinated dibenzodioxins. In general, the regulatory levels for these chemicals have been set to prevent cancer in adult humans from lifetime exposure. We compared lives saved in the postneonatal period by breast feeding to the estimated excess cancer deaths attributable to the contaminants in breast milk. The results of this analysis suggest that only extreme levels of contaminants in breast milk represent more of a hazard than failure to breast feed, but clinical considerations in individual cases might override this conclusion. Our analysis depends on assumptions about how the chemicals might cause cancer in humans and on whether breast feeding prevents some postneonatal mortality. Noncarcinogenic hazards from chemical exposure, other hazards from breast feeding such as transmission of viruses, and benefits of breast feeding other than reduction in mortality were not considered.

Breast Feeding↗

PubMed1991

An evaluation of the Rai and Van Ryzin dose-response model in teratology.

The underlying assumptions of the Rai and Van Ryzin dose-response model for reproductive toxicological data are evaluated on the basis of existing experimental data. The model under consideration is unusual in its use of litter size to completely account for extra-binomial variation in the data by associating litter size with reproductive outcome. The experimental data show that controlling litter size is not sufficient to account for the litter-to-litter variability in responses. It is also shown that the two linear components of the Rai and Van Ryzin model are inappropriate. For the component which applies to the dam, the data suggest a strong nonlinearity, supported by rejection of the linear model via statistical hypothesis tests. In the component involving litter size, a relationship with dose is not apparent. The litter size parameters offer considerable potential for bias in estimation; bias which is at least partly masked by the model having good prediction characteristics due to the increased number of parameters. A simulation study is presented to illustrate how the Rai and Van Ryzin model can exaggerate litter size effects on the probability of response when the simulated data arise from a model involving a nonlinear dam component, common to this type of data, and no effect of litter size.

Abnormalities, Drug-Induced↗

PubMed1991

A multistage model of carcinogenesis incorporating DNA damage and repair.

Mathematical models of carcinogenesis are one tool used for research into the mechanisms of carcinogenesis and for assessing risks from exposure to carcinogenic agents. Recent research into carcinogenic mechanisms has focused on the role of cell replication in fixing damage to cellular DNA and increasing the incidence of malignancies. The most commonly used multistage model of carcinogenesis does not explicitly account for DNA damage and the contribution of cell replication rates on the transformation of cells through the various stages. In this manuscript, a generalized multistage model of carcinogenesis is developed in which DNA damage, cell replication, and DNA repair are explicitly included.

Animals↗

PubMed1991

Distinguishing between models of carcinogenesis: the role of clonal expansion.

Several multistage models have been proposed to describe the process of carcinogenesis. The model attributed to Armitage and Doll (1954, Brit. J. Cancer 8, 1-12) assumes that a normal cell is transformed into a malignant cell by k greater than or equal to 1 fundamental biological events. The model usually attributed to Moolgavkar and colleagues (1979, Math. Biosci. 47, 54-77; 1981, J. Natl. Cancer Inst. 66, 1037-1052) assumes that carcinogenesis is a two-stage process, taking into account the ability of premalignant cells to proliferate. In this paper, we investigate our ability to distinguish between these models using animal carcinogenicity data. Three different approaches are considered: one based on actual tumor incidence data from National Toxicology Program experiments and two based on simulated data. The results show that both models will adequately fit most tumor incidence data. This implies that we must be cautious in accepting the biological basis of either of these models simply because it "fits the data." Suggestions for experimental designs and toxicological endpoints (other than tumor incidence) which might provide for better discrimination between models are given.

Animals↗

PubMed1990

A note on fitting one-compartment models: non-linear least squares versus linear least squares using transformed data.

Drug concentrations in one-compartment systems are frequently modeled using a single exponential function. Two methods of estimation are commonly used for determining the parameters of such a model. In the first method, non-linear least-squares regression is used to calculate the parameters. In the second method, the data are first transformed by a logarithmic function, and then the log-concentration data are fit using linear least-squares regression. The assumptions for fitting these models are discussed with special emphasis on which data points are most influential in determining parameter values. The similarities between fitting a linear regression model to the log-concentration data and fitting a weighted regression model to the original data are noted. An example is presented that illustrates the differences in fitting a model to the log-transformed data versus fitting unweighted and weighted models to the original-scale data.

Animals↗

PubMed1990

Two-stage models of carcinogenesis, classification of agents, and design of experiments.

The implications of a clonal two-stage model of carcinogenesis on the design and analysis of 2-year in vivo tumorigenesis experiments are addressed. Using a simple classification scheme for labelling test agents as initiators, promoters, and completers, it is shown that the standard experimental design has very little ability to differentiate between these different modes of action. Even when chemicals in one class (e.g., promoters) follow a highly nonlinear dose-response relationship and chemicals in another class (e.g., initiators) follow a linear dose-response relationship, it is difficult to reject one mode of action versus the other. A simple modification of the design using age-dependent dosing schemes produces patterns of tumor incidence which are unique to the particular class, making it slightly easier to differentiate between the assumed mechanisms of action. The implications of this finding on study design are discussed.

Aging↗

PubMed1990

Biologically based models for risk assessment.

The modelling problems associated with the estimation of risks from long-term chemical exposures at low dose levels represent a statistical and mathematical challenge with special relevance to environmental research. Determining an adequate model for estimating the relationship between dose and response is critical to reducing potential bias in the risk estimation process. This paper discusses the various assumptions and models used in carcinogenic risk assessment. The emphasis is on our ability to accurately determine the magnitude of the carcinogenic risk, the shape of the dose-response relationship and the overall variability of the risk estimates.

Animals↗

PubMed1989

Testing for increased carcinogenicity using a survival-adjusted quantal response test.

The linear trend test in proportions is frequently used to analyze the results of animal carcinogenicity experiments. This test has two major advantages over other frequently used tests; it is easily understood and it is simple to calculate. This test, however, fails to correct for treatment-related differences in survival across the experimental groups. A test which is a simple modification of the linear trend test in proportions and which has the same advantages is proposed to correct for differences in survival. The results of this modified test are compared to those of the linear trend test in proportions, the incidental tumor test, the logistic regression score test, the life table test, and the truncated trend test using information on the incidence of combined alveolar/bronchiolar adenomas or carcinomas in female B6C3F1 mice exposed to vinylcyclohexene diepoxide.

Adenoma↗

PubMed1989

Variability of safe dose estimates when using complicated models of the carcinogenic process. A case study: methylene chloride.

Advances in understanding carcinogenesis have led to the development of mathematical models that have biologically interpretable parameters. These models utilize more of the available scientific data than the empirical models routinely employed for quantifying carcinogenic risk. They also require consideration of sources of uncertainty in risk estimates that were previously ignored, such as animal-to-animal variability of physiological and pharmacological constants. A numerical technique is proposed for studying the consequences of incorporating the intrapopulation variability of biologically interpretable parameters into the risk assessment process. To demonstrate the technique, the variability of safe dose estimates for exposure to methylene chloride is considered. The results suggest that intrapopulation variability of the model parameters can increase the variability of safe dose estimates an appreciable amount.

Animals↗

PubMed1989

Two-stage models of tumor incidence for historical control animals in the National Toxicology Program's carcinogenicity experiments.

The tumor incidence rate is modeled for various tumors in a database of control animals [Fischer 344 rats and (C57BL/6 x C3H)F1 mice] originally developed by the national Toxicology Program and recently augmented to include additional sacrifice data by Portier et al. (1986). These rates are assumed to follow a clonal two-stage model of carcinogenesis where cells in the various tissues are allowed to be in one of three states arbitrarily defined as a "normal" state, an "initiated" state, and a "tumor present" state. The parameters of this clonal two-stage model have a direct interpretation with regard to the mechanism of action and in some cases may suggest a common mechanism for various tumors in the different sex/species groups. Also, the risk of dying (from all causes) in tumor-bearing animals is compared to the risk of dying in non-tumor-bearing animals via estimates of relative risk. In general, it was found that the risk of dying was elevated for most tumors. The purpose of this analysis is to provide estimates of baseline tumor rates and relative risks, which are useful in the design and analysis of future carcinogenicity experiments.

Animals↗

PubMed1989

A note on approximating the cumulative distribution function of the time to tumor onset in multistage models.

The general multistage theory of carcinogenesis is a very special type of interconnected birth and death process in which the final state is absorbing and all states except the first one are empty at time zero. An approximation proposed by Whittemore and Keller (1978, SIAM Review 20, 1-30) is assessed. It is shown that the adequacy of this approximation depends on the number of malignant cells resulting from a single normal cell. If more than one malignant cell is likely to occur, the approximation will fail.

Animals↗