Recent Advances in Quantitative Methods in Cancer and Human Health Risk Assessment (Wiley Series in Probability and Statistics)

Human health risk assessment involves the measuring of risk of exposure to disease, with a view to improving disease prevention. Mathematical, biological, statistical, and computational methods play a key role in exposure assessment, hazard assessment and identification, and dose-response modelling. Recent Advances in Quantitative Methods in Cancer and Human Health Risk Assessment is a comprehensive text that accounts for the wealth of new biological data as well as new biological, toxicological, and medical approaches adopted in risk assessment. It provides an authoritative compendium of state-of-the-art methods proposed and used, featuring contributions from eminent authors with varied experience from academia, government, and industry. Provides a comprehensive summary of currently available quantitative methods for risk assessment of both cancer and non-cancer problems. Describes the applications and the limitations of current mathematical modelling and statistical analysis methods (classical and Bayesian). Includes an extensive introduction and discussion to each chapter. Features detailed studies of risk assessments using biologically-based modelling approaches. Discusses the varying computational aspects of the methods proposed. Provides a global perspective on human health risk assessment by featuring case studies from a wide range of countries. Features an extensive bibliography with links to relevant background information within each chapter. Recent Advances in Quantitative Methods in Cancer and Human Health Risk Assessment will appeal to researchers and practitioners in public health & epidemiology, and postgraduate students alike. It will also be of interest to professionals working in risk assessment agencies. Human health risk assessment involves the measuring of risk of exposure to disease, with a view to improving disease prevention. Mathematical, biological, statistical, and computational methods play a key role in exposure assessment, hazard assessment and identification, and dose-response modelling. Recent Advances in Quantitative Methods in Cancer and Human Health Risk Assessment is a comprehensive text that accounts for the wealth of new biological data as well as new biological, toxicological, and medical approaches adopted in risk assessment. It provides an authoritative compendium of state-of-the-art methods proposed and used, featuring contributions from eminent authors with varied experience from academia, government, and industry. Provides a comprehensive summary of currently available quantitative methods for risk assessment of both cancer and non-cancer problems. Describes the applications and the limitations of current mathematical modelling and statistical analysis methods (classical and Bayesian). Includes an extensive introduction and discussion to each chapter. Features detailed studies of risk assessments using biologically-based modelling approaches. Discusses the varying computational aspects of the methods proposed. Provides a global perspective on human health risk assessment by featuring case studies from a wide range of countries. Features an extensive bibliography with links to relevant background information within each chapter. Recent Advances in Quantitative Methods in Cancer and Human Health Risk Assessment will appeal to researchers and practitioners in public health & epidemiology, and postgraduate students alike. It will also be of interest to professionals working in risk assessment agencies. [Ed.] Recent Advances in Quantitative Methods in Cancer and Human Health Risk Assessment......Page 4 Contents......Page 10 Contributors......Page 22 Preface......Page 28 Introduction......Page 32 I CANCER AND HUMAN HEALTH RISK ASSESSMENT Introductory remarks......Page 34 1.1 The risk assessment paradigm......Page 38 1.2 Hazard identification......Page 40 1.3.1 Different objectives, different data sets, different approaches......Page 41 1.3.2 Extrapolations in dose-response assessment......Page 42 1.3.3 Safety assessment......Page 44 1.3.4 Modelling to estimate risk at low doses......Page 47 1.3.5 Uncertainty and human variation......Page 54 II BIOLOGICAL ASPECTS OF CARCINOGENESIS Introductory remarks......Page 58 2.1 Introduction......Page 62 2.3 Biomarkers......Page 63 2.3.1 Biomarkers of exposure......Page 64 2.3.2 Biomarkers of susceptibility......Page 65 2.4 Validation of biomarkers......Page 66 2.4.2 Genetic and statistical analysis......Page 67 2.5 Factors influencing cancer risk......Page 68 2.5.3 Carcinogen metabolism......Page 69 2.5.6 Immune status......Page 70 2.6.1 Microarrays and toxicogenomics......Page 71 2.6.3 Promising directions for cancer diagnosis and cancer biomarker discovery......Page 72 2.7 Conclusions......Page 73 3.1 Introduction......Page 76 3.1.1 Studies investigating genetic polymorphisms as lung cancer risk factors......Page 77 3.2.1 Planning of the study......Page 79 3.2.2 Laboratory analyses......Page 83 3.2.3 Statistical analyses......Page 85 3.3.1 N-Acetyltransferases (NAT1 and NAT2) and lung cancer risk......Page 89 3.3.3 Myeloperoxidase and lung cancer risk......Page 90 3.3.4 CYP3A4 and CYP3A5 and lung cancer risk......Page 91 3.4 Discussion......Page 92 Acknowledgements......Page 95 4.1 Introduction......Page 96 4.2 Models of human carcinogenesis......Page 97 4.2.1 Prostate cancer......Page 98 4.2.2 Colorectal cancer......Page 101 4.2.3 Endometrial cancer......Page 104 4.3 The multistage mouse skin carcinogenesis model......Page 106 4.4 Epilogue......Page 110 5.1 Introduction......Page 112 5.2.1 Colorectal cancer......Page 113 5.2.2 The role of genomic instability in colon cancer......Page 115 5.2.3 Barrett’s esophagus......Page 116 5.2.4 Intermediate lesions......Page 117 5.3.1 Model building......Page 118 5.3.2 Mathematical development and the hazard function......Page 120 5.4.2 Age-specific incidence......Page 123 5.4.3 Colorectal cancer in the SEER registry......Page 124 5.5 Summary......Page 127 6.1 Introduction......Page 130 6.2 The concept of hormesis......Page 132 6.3 Chemical hormesis......Page 133 6.3.1 The -shaped and -shaped dose-response curve......Page 134 6.3.2 Critical issues in low-dose extrapolation......Page 136 6.3.3 The evaluation of dose-response relationship......Page 138 6.4 Radiation hormesis......Page 139 6.5 Concluding remarks......Page 141 III MODELING FOR CANCER RISK ASSESSMENT Introductory remarks......Page 144 7.1.1 Physiologically based pharmacokinetic models......Page 148 7.1.2 Model formulation......Page 150 7.1.3 Data sources......Page 151 7.2.1 Metabonomics......Page 152 7.3 The next generation of physiological models......Page 154 7.4 Discussion and conclusions......Page 156 Acknowledgements......Page 157 8.1 Introduction......Page 158 8.1.3 Stereological aspects in the evaluation of liver focal lesions......Page 159 8.2.1 The multistage model with clonal expansion......Page 162 8.3.1 Model-based approach to study the mode of action of chemicals......Page 164 8.3.2 Model-based approach to study the process of formation and growth of liver foci......Page 165 8.4 Conclusions......Page 168 9.1 Background......Page 170 9.2 Conventional approach to modeling carcinogenesis......Page 172 9.3 State space modeling using sampling techniques......Page 173 9.4 Self-organizing algorithm for state space modeling......Page 175 9.5.1 State space model for cell labeling......Page 176 9.5.2 State space model of carcinogenesis......Page 180 9.6 A computing procedure for the three-stage model......Page 185 9.7 Discussion......Page 187 Appendix: Simulation programs......Page 189 10.1 Introduction......Page 194 10.2.3 Mayo Lung Project (MLP)......Page 195 10.2.5 Early Lung Cancer Action Project......Page 196 10.2.8 National Lung Screening Trial (NLST)......Page 197 10.3.1 Natural history of disease......Page 198 10.3.2 Critical parameters......Page 199 10.3.3 Mortality versus case fatality versus survival......Page 200 10.4.2 Simulation modeling of the Mayo Lung Project......Page 201 10.4.4 Modeling the outcome of the Czechoslovak screening study......Page 202 10.4.6 Markov model of helical CT-screened and non-screened cohort......Page 203 10.5.1 Modeling the NLST trial......Page 204 10.5.2 Modeling the effects of long-term mass screening by CT scan......Page 205 10.6 Comparison of models and concluding remarks......Page 206 11.1 Optimal screening......Page 210 11.2 A comprehensive model of cancer natural history......Page 213 11.2.2 Tumor growth......Page 214 11.2.4 Tumor detection......Page 215 11.3 Formula for the screening efficiency functional......Page 217 11.4 The data and parameter estimation......Page 219 11.5 Numerical experiments......Page 220 11.6 Discussion......Page 223 Acknowledgement......Page 224 IV STATISTICAL APPROACHES FOR CARCINOGENESIS STUDIES Introductory remarks......Page 226 12.1 Introduction......Page 230 12.2 Models......Page 231 12.3.1 Parametric estimation......Page 232 12.3.2 Semiparametric estimation......Page 233 12.3.3 Modified partial likelihood estimator......Page 235 12.4 Goodness-of-fit for the Cox model against the cross-effects models......Page 236 12.5 Examples......Page 238 12.5.3 Lung cancer radiotherapy......Page 239 Appendix: Proof of Theorem 12.1......Page 240 13.1 Introduction......Page 244 13.2 Elements of the dose-response assessment......Page 246 13.2.1 Exposure and dose......Page 247 13.2.2 Response......Page 248 13.2.3 Biomarkers......Page 249 13.3 Dose-response models......Page 250 13.3.1 Qualitative dose-response analysis......Page 251 13.3.2 Quantitative dose-response analysis......Page 252 13.4 Dose-response models in risk assessment......Page 260 13.4.1 Model search......Page 261 13.4.3 Linear versus nonlinear low-dose extrapolation......Page 262 13.4.4 Point of departure......Page 263 13.5 Dose-Response modeling of 2,3,7,8-tetrachlorodibenzo-p-dioxin......Page 264 13.5.2 Toxicokinetic dose-response models......Page 265 13.5.3 Laboratory animal responses......Page 266 13.5.4 Human response......Page 267 13.6 Concluding remarks......Page 269 14.1 Introduction......Page 272 14.2 Use by regulatory agencies......Page 274 14.3.1 Types of models......Page 275 14.3.3 Goodness of fit......Page 277 14.3.4 Lower confidence limit......Page 278 14.3.5 Experimental design, dose selection and response metrics......Page 280 14.4 Literature survey......Page 282 14.5 Software and calculation example......Page 284 15.1 Introduction......Page 288 15.2 Representations of uncertainty......Page 289 15.3 Causes of uncertainty......Page 290 15.4.1 Model uncertainty......Page 291 15.5 Quantifying uncertainty......Page 293 15.5.2 Model uncertainty......Page 294 15.6.2 Accounting for variability to reduce uncertainty: multilevel modelling......Page 296 15.7 Conclusion......Page 298 Acknowledgements......Page 299 16.2 On the optimal design theory......Page 300 16.2.1 Binary outcome designs......Page 302 16.2.3 Definition of the optimal design......Page 304 16.3 The Michaelis–Menten model......Page 305 16.3.1 D-optimal design of the Michaelis–Menten model......Page 307 16.3.2 Applications of the D-optimal design......Page 309 16.4 Dose extrapolation designs......Page 310 16.4.1 The Weibull model......Page 311 Acknowledgements......Page 312 V SPECIFIC MODELING APPROACHES FOR HEALTH RISK ASSESSMENT Introductory remarks......Page 314 17.1 Introduction......Page 316 17.2 Basic principles......Page 318 17.2.1 Synergism......Page 319 17.2.2 Additivity......Page 320 17.2.3 Modeling additivity......Page 322 17.2.5 The additive model reconsidered......Page 323 17.3 Design of mixture experiments......Page 325 17.3.1 The 2(k) factorial design......Page 326 17.3.4 q-component mixture model: response surface analysis......Page 327 17.3.6 Response trace plots......Page 330 17.4 Discussion......Page 331 18.1 Introduction......Page 332 18.2 Experimental designs and models for whole mixture......Page 334 18.3.1 Relative potency factor method......Page 338 18.3.2 A dose addition model for mixtures......Page 341 18.4 Component-based risk assessment for quantitative response data......Page 345 18.5 Discussion......Page 348 19.1 Introduction......Page 350 19.2 The model......Page 352 19.3 Estimation of model parameters......Page 355 19.4 Data analysis......Page 356 Acknowledgements......Page 360 VI CASE STUDIES......Page 362 20.1 Introduction......Page 366 20.2 Experiments, data and statistical issues......Page 368 20.2.1 Preprocessing the data......Page 369 20.2.2 Data structure and notation......Page 370 20.3 Results......Page 371 20.3.2 Classifying the test set: validating the chosen set of differentially expressed genes......Page 372 20.3.3 Hotelling’s T(2) statistic......Page 373 20.4 Discussion......Page 374 Acknowledgements......Page 376 21.1 Medical problem......Page 378 21.2 Prediction model and advantage of designing the experiment......Page 380 21.3 Optimal experimental design approach......Page 381 21.4 Solution......Page 384 22.1 Introduction......Page 388 22.2.1 System description......Page 390 22.2.2 System modeling......Page 391 22.3.1 Definitions......Page 393 22.3.3 Fallacies of asymptotic maximum likelihood estimation......Page 394 22.3.4 Exact logistic regression......Page 396 22.3.6 Numerical aspects......Page 398 22.4 Neurophysiological example revisited......Page 399 22.5 Discussion......Page 402 23.1 Introduction......Page 404 23.2 Patients......Page 406 23.3.1 Selection of variables......Page 408 23.3.4 Approaches to missing values......Page 409 23.4.1 Classification C......Page 410 23.5 Discussion......Page 414 23.6 Concluding remarks......Page 415 24.1 Introduction......Page 416 24.2.2 Statistics......Page 417 24.3 Results......Page 419 24.4 Discussion......Page 424 Acknowledgements......Page 427 25.1 Introduction......Page 428 25.2 Materials and methods......Page 429 25.3 Results......Page 430 25.4 Discussion......Page 431 References......Page 434 Index......Page 484 WILEY SERIES IN PROBABILITY AND STATISTICS......Page 498

Human health risk assessment involves the measuring of risk of exposure to disease, with a view to improving disease prevention. Mathematical, biological, statistical, and computational methods play a key role in exposure assessment, hazard assessment and identification, and dose-response modelling.

Recent Advances in Quantitative Methods in Cancer and Human Health Risk Assessment is a comprehensive text that accounts for the wealth of new biological data as well as new biological, toxicological, and medical approaches adopted in risk assessment. It provides an authoritative compendium of state-of-the-art methods proposed and used, featuring contributions from eminent authors with varied experience from academia, government, and industry.

• Provides a comprehensive summary of currently available quantitative methods for risk assessment of both cancer and non-cancer problems.
• Describes the applications and the limitations of current mathematical modelling and statistical analysis methods (classical and Bayesian).
• Includes an extensive introduction and discussion to each chapter.
• Features detailed studies of risk assessments using biologically-based modelling approaches.
• Discusses the varying computational aspects of the methods proposed.
• Provides a global perspective on human health risk assessment by featuring case studies from a wide range of countries.
• Features an extensive bibliography with links to relevant background information within each chapter.

Recent Advances in Quantitative Methods in Cancer and Human Health Risk Assessment will appeal to researchers and practitioners in public health & epidemiology, and postgraduate students alike. It will also be of interest to professionals working in risk assessment agencies.

In his comprehensive volume on environmental risk assessment, Paustenbach (2002) characterizes risk assessment as the description of the likelihood of adverse or unwanted responses to exposure and states as its goal the estimation of that likelihood after assembling and assessing all scientific information regarding toxicology, human experience, and environmental fate and exposure. Research into quantitative methods in cancer and human health risk assessment continues to grow, and the methodology is now used widely in practice. This book brings together the current state-of-the-art in mathematical, statistical and computational methods in cancer and human health risk assessment, with a focus on practical implementation.