Carcinogen bioassay occupies a central role in carcinogenic risk assessment. In order to maximize the value of bioassay data, established principles of bioassay design need to be carefully considered in developing the experimental protocol. Different designs are required for screening, dose response, and mechanistic studies. In all cases, however, basic principles of experimental design such as randomization, replication, and stratification need to be considered. In this article, we examine these and other aspects of carcinogen bioassay design, taking into account both theoretical and empirical results reported in the recent literature.
MontesanoRBartschHVainioHWilbournJYamasakiH. Long-term and short-term assays for carcinogens: A critical appraisal. IARC Scientific Publications, No. 83. Lyon: IARC.
2.
International Life Sciences Institute (ILSI).The selection of doses in chronic toxicity/carcinogenicity studies. In: GriceHC, ed. Current issues in toxicology. New York: Springer-Verlag; 1984: 6–49.
3.
ClaysonDBKrewskiDRMunroLC. The power and interpretation of the carcinogenicity bioassay. Regul Toxicol Pharmacol.1983;3: 329–348.
4.
BickisMG. Experimental design. In: ArnoldDLGriceHCKrewskiD, eds. In vivo toxicity testing: Principles, procedures and practices. New York: Academic Press; 1990: 113–147.
5.
GriceHCMunroICKrewskiDRBlumenthalH. In utero exposure in chronic toxicity/carcinogenicity studies. Food Cosmet Toxicol.1981;19: 373–379.
6.
HasemanJKHuffJERaoGNArnoldJEBoormanGAMcConnellEE. Neoplasms observed in untreated and corn oil gavage control groups of F344/N rats and (C57BL/6N × C3H/MeN)F1 (B6C3F1 mice. J Nat Cancer Inst.1985;75: 975–984.
7.
LagakosSMostellerF. A case study of statistics in the regulatory process: The FD&C Red No. 40 experiments. J Natl Cancer Inst.1981;66: 197–212.
8.
HasemanJK. Lack of cage effects on liver tumor incidence in B6C3F1 mice. Fundam Appl Toxicol.1988;10: 179–187.
9.
YoungSS. Are there location/cage/systematic nontreatment effects in long-term rodent studies? A question revisited. Fundam Appl Toxicol.1989;13: 183–188.
10.
HasemanJK. Commentary on “Are there location/cage/systematic nontreatment effects in long-term rodent studies? A question revisited.”Fundam Appl Toxicol.1989;13: 189–192.
11.
BickisMKrewskiD. The statistical design and analysis of the long-term carcinogenicity bioassay. In: ClaysonDKrewskiDMunroI., eds. Toxicological risk assessment, Vol. I. Boca Raton, FLA: CRC Press; 1985: 125–147.
12.
National Toxicology Program (NTP).Report of the ad hoc panel on chemical carcinogenesis testing and evaluation. Washington, DC: US Department of Health and Human Services; 1984.
13.
HasemanJHajianGCrumpKSelwynMPeaceK. Dual control groups in rodent carcinogenicity studies. In: PeaceKE, ed. Statistical issues in pharmaceutical drug development. New York: Marcel Dekker; 1989: 355–365.
14.
HasemanJK. False positive issues in carcinogenicity testing: an examination of 16 studies with dual control groups. In: Symposium on Long-Term Animal Carcinogenicity Studies: A Statistical Perspective. 4–6 March, 1985, Bethesda, Maryland. Washington, DC: American Statistical Association; 1985: 73–80.
15.
HasemanJKWinbushJSO'DonnellMWJr. Use of dual control groups to estimate false-positive rates in laboratory animal carcinogenicity studies. Fundam Appl Toxicol.1986;7: 573–584.
16.
DePassLRWeilCSBallantyneBLewisSCLoscoPEReidJBSimonGS. Influence of housing conditions for mice on results of a dermal oncogenicity bioassay. Fundam Appl Toxicol.1986;7: 601–688.
17.
GoldLSWrightCBernsteinLde VecianaM. Reproducibility of results in ‘near replicate’ carcinogenesis bioassays. J Nat Cancer Inst.1987;78: 1149–1158.
18.
AbdelbasitKMPlackettRL. Experimental design for binary data. J Am Statist Assoc.1983;78: 90–98.
AtkinsonAC. Recent developments in the methods of optimum and related experimental designs. International Statistical Review.1988;56: 99–116.
21.
LawlessJF. Some problems concerning experimental designs for extrapolation. In: ChaubeyYBDwivediTD, eds. Topics in applied statistics. Montreal: Concordia University; 1984: 357–366.
22.
BickisMKrewskiD. Statistical issues in the analysis of the long-term carcinogenicity bioassay in small rodents: An empirical evaluation of statistical decision rules. Fundam Appl Toxicol.1989;12: 202–221.
23.
FarrarDBCrumpKS. Exact statistical tests for any carcinogenic effect in animal bioassays. Fundam Appl Toxicol.1988;11: 652–663.
24.
LouisTA. Analysis of a random sample of 2-Year carcinogen bioassays from the NCI data base. Fund Appl Toxicol.1989;12: 222–231.
25.
WestfallPHYoungSS. P Value adjustments for multiple tests in multivariate binomial models. J Amer Statist Assoc.1989;84: 780–786.
26.
HasemanJKTharringtonECHuffJEMcConnellEE. Comparison of site-specific and overall incidence analysis of 81 recent National Toxicology Program carcinogenicity studies. Reg Tox Pharm.1986;6: 155–170.
27.
GartJJKrewskiDLeePNTaroneREWahrendorfJ. Statistical methods in cancer research. III. The design and analysis of long-term animal experiments. IARC Scientific Publications No. 79. Lyon: IARC.
28.
GartJJTaroneRE. On the efficiency of age-adjusted tests in animal carcinogenicity experiments. Biometrics.1987;43: 235–244.
29.
McKnightB. A guide to the statistical analysis of long-term carcinogenicity bioassays. Fundam Appl Toxicol.1988;10: 355–364.
30.
PortierCJ. Life table analysis of carcinogenicity experiments. J Am Coll Toxicol.1988;7: 575–582.
31.
McKnightBCrowleyJ. Tests for differences in tumor incidence based on animal carcinogenesis experiments. J Amer Statist Assoc.1984;79: 639–648.
32.
LagakosSWLouisTA. Use of tumor lethality to interpret experiments lacking cause-of-death data. Appl Statist.1988;37: 167–179.
33.
ArcherLERyanLM. Accounting for misclassification in the cause-of-death test for carcinogenicity. J Amer Statist Assoc.1989;84: 787–791.
HasemanJK. Exact sample sizes for use with the Fisher–Irwin test for 2 × 2 tables. Biometrics.1978;34: 106–109.
36.
WaltersDE. In defence of the arc sine approximation. Statistician.1979;28: 219–222.
37.
ChenHJ. Sample size determinations when two binomial proportions are very small. Commun Stat-Theor Meth.1984;A13: 2707–2712.
38.
DobsonAJGebskiV. Sample sizes for comparing two independent proportions using the continuity-corrected arc sine transformation. Statistician.1986;35: 51–53.
PocockSJ. Group sequential methods in the design and analysis of clinical trials. Biometrika.1977;64: 191–199.
48.
ElashoffRMBealS. Two-stage screening designs applied to chemical-screening problems with binary data. Ann Rev Biophys Bioeng.1976;5: 561–587.
49.
ElashoffRMPrestonDL. An investigation of the properties of some two sample two-stage designs. In: KrishnaiahPR, ed. Applications of statistics. North-Holland, Amsterdam: North-Holland; 1977;407–432.
50.
ElashoffRMPrestonDLFearsTR. Comparison and evaluation of some experimental designs for use in carcinogen screening. J Natl Cancer Inst.1979;62: 1209–1219.
51.
SomsAP. An algorithm for the discrete Fisher's permutation test. J Amer Statist Assoc.1977;72: 662–664.
52.
KrewskiDBrennanJBickisM. The power of the Fisher permutation test in 2 × k tables. Commun Stat-Simul Comp.1984;B13: 433–448.
53.
National Research Council (NRC).Complex mixtures: Methods for in vivo toxicity testing. Washington DC: National Academy Press; 1988.
54.
KodellRL. Modeling the joint action of toxicants: Basic concepts and approaches, In: LandauEWellingtonDG, eds. ASA/EPA Conferences on interpretation of environmental data: 1. Current assessment of combined toxicant effects. Washington, DC: United States Environmental Protection Agency; 1986: 1–8.
55.
FearsTRElashoffRMSchneidermanMA. The statistical analysis of a carcinogen mixture experiment. II. Carcinogens with different target organs, N-Methyl-N'-Nitro-N-Nitrosoguanidine, N-Bytyl-N-(4-Hydroxybutyl)Nitrosamine, Dipentyl-Nitrosamine, and Nilotieacetic acid. Toxic and Indus Health.1988;4: 221–255.
56.
FearsTRElashoffRMSchneidermanMA. The statistical analysis of a carcinogen mixture experiment. II. Carcinogens with different target systems, Aflatoxin B1, N-Bytyl-N-(4-Hydroxybutyl)Nitrosamine, Lead acetate and thiouracil. Toxic and Indust Health.1989;5: 1–23.
57.
ElashoffRMFearsTRSchneidermanMA. Statistical analysis of a carcinogen mixture experiment. I Liver carcinogens. J Natl Cancer Inst.1987;79: 509–526.
58.
WahrendorfJZentgrafRBrownCC. Optimal designs for the analysis of interactive effects of two carcinogens or other toxicants. Biometrics.1981;37: 45–54.
59.
KrewskiDColinDDewanjiA. Statistical considerations in the design of toxicological experiments with chemical mixtures. In: ClaysonDBMunroICShubikPSwenberJA eds. Progress in predictive Toxicology. Elsevier Science Publishers, B.V. Amsterdam: 1990;335–348.
60.
GoldLSSawyerCBMagawRBackmanGMde VeciannaMLevinsonRHooperNKHavenderWRBernsteinLPetoRPikeMCAmesBN. A carcinogenic potency database of the standardized results of animal bioassays. Environmental Health Perspectives.1984:58: 9–319.
61.
KrewskiDKovarJ. Low dose extrapolation under single parameter dose–response models Commun Statis-Simula Computa. 1982;11: 27–45.
62.
PortierCHoelD. Optimal design of the chronic animal bioassay. J Toxicol Environ Health.1983;12: 1–19.
63.
KrewskiDKovarJGBickisM. Optimal experimental designs for low dose extrapolation. II. The case of nonzero background. In: ChaubeyYPDwivediTD, eds. Topics in applied statistics. Concordia University; Montreal: 1984; 167–191.
64.
KrewskiDMurdochDDewanjiA. Statistical modeling and extrapolation of carcinogenesis data. In: MoolgavkarSHPrenticeRL, eds. Modern statistical methods in chronic disease epidemiology. New York: Wiley-Interscience; 1986: 259–282.
65.
ChernoffH. Sequential analysis and optimal design. Philadelphia: Society for Industrial and Applied Mathematics; 1972.
66.
GaylorDWChenJJKodellRL. Experimental designs of bioassays due for screening and low dose extrapolation. Risk Anal.1985;5: 9–16.
67.
PortierCHoelD. Design of animal carcinogenicity studies for goodness-of-fit of multistage models. Fundam Appl Toxicol.1984;4: 949–959.
68.
ChambersEACoxDR. Discrimination between alternative binary response models. Biometrika.1967;54: 573–578.
69.
CrumpKS. Designs for discriminating between binary dose response models with applications to animal carcinogenicity experiments. Commun Stat-Theor Meth.1982;A11: 379–394.
70.
KrewskiDGoddardMJMurdochDJ. Statistical considerations in the interpretation of negative carcinogenicity data. Reg Toxicol Pharmacol.1989;9: 5–22.88.
71.
HoelDGHasemanJKHoganMDHuffJMcConnellEE. The impact of toxicity on carcinogenicity studies: Implications for risk assessment. Carcinogenesis.1988;9: 2045–2052.
72.
HoelDGKaplanNLAndersonMW. Implication of nonlinear kinetics on risk estimation in carcinogenesis. Science.1983;219: 1032–1037.
73.
AndersenMEClewellHJIIIGargasMLSmithFAReitzRH. Physiologically based pharmacokinetics and the risk assessment for methylene chloride. Toxicol Appl Pharmacol.1987;87: 185–205.
74.
CohenSMEllweinLB. Cell growth dynamics and DNA alterations in carcinogenesis. In: MoolgavkarSH, ed. Scientific issues in quantitative cancer risk assessment. Birkhauser: Boston; 1990;116–135.
75.
Food and Chemical Toxicology. Saccharin: Current status. Food Chem Toxicol.1985;23: 417–546.
76.
International Agency for Research on Cancer (IARC) (1984). Models, mechanisms and etiology of tumor promotion. IARC Monographs, No 56, IARC, Lyon.
77.
SlagaTJ. SENCAR mouse skin tumorgenesis model versus other strains and stocks of mice. Environmental Health Perspectives.1986;68: 27–32.
78.
SchererEFeringaAWEmmelotPL. Initiation-promotion-initiation. Induction of neoplastic foci within islands of precancerous liver cells in the rat. In: BarzsonyiMLapisKDayNEYamasakiH, Eds. Models, mechanisms and etiology of tumor promotion. Lyon: International Agency for Research on Cancer; 1984;.
79.
MoolgavkarSH. Carcinogenesis modeling: From molecular biology to epidemiology. Ann Rev Public Health.1986;7: 151–169.
80.
CairnsT. The ED01 study: Introduction, objectives and experimental design. J Environ Pathol Toxicol.1980;3: 1–7.
81.
LittlefieldNAGreenmanDLFarmerJH. Effect of continuous and discontinued exposure to 2-AAF on urinary bladder hyperplasia and neoplasia. J Environ Pathol Toxicol.1980;3: 35–54.
82.
ArnoldDLKrewskiDRJunkinsDBMcGuirePFMoodieCAMunroIC. Reversibility of ethylenethiourea-induced thyroid lesions. Toxicol Appl Pharmacol.1983;67: 264–273.
