Simulation for Designing Clinical Trials: A Pharmacokinetic-Pharmacodynamic Modeling Perspective (Drugs and the Pharmaceutical Sciences, Vol 127)

Simulation for Designing Clinical Trials: A Pharmacokinetic-Pharmacodynamic Modeling Perspective......Page 2
Foreword......Page 5
Preface......Page 7
Contributors......Page 13
Contents......Page 10
1.1 BACKGROUND......Page 17
1.1.1 History of Clinical Trial Simulation......Page 18
1.1.2 FDA Opinion on Clinical Trial Simulation......Page 19
1.2 WHAT IS CLINICAL TRIAL SIMULATION?......Page 20
1.3 WHY SIMULATE CLINICAL TRIALS?......Page 21
1.4 CLINICAL TRIAL SIMULATION PLAN......Page 23
1.4.2 Simulation Team......Page 24
1.4.3.2 Covariate Distribution Models......Page 25
1.4.4 Model Evaluation Methods......Page 26
1.4.5.3 Differential Equation Solver......Page 27
1.4.7 Model-Building Assumptions......Page 28
1.4.8 Clinical Trial Design Scenarios......Page 29
1.5 CONCLUSION......Page 30
REFERENCES......Page 31
2.1 CLINICAL TRIAL SIMULATION MODELS......Page 33
2.3.1 IO Model Anatomy......Page 34
2.3.3.2 Population IO Model Simulation......Page 35
2.3.4.1 Group Parameter Model......Page 36
2.3.5.1 Individual Parameter Model......Page 38
2.3.5.2 Individual IO Model Simulation......Page 39
2.3.6.1 Observation Parameter Model......Page 41
2.5.1.2 Estimates from Data......Page 42
2.5.3 Posterior Distribution of Parameters......Page 43
REFERENCES......Page 44
3.1.1 Clinical Trial Simulation......Page 46
3.1.2 Definition of Terminology......Page 50
3.2 OVERVIEW OF AVAILABLE METHODS......Page 51
3.2.2 Resampling Methods......Page 53
3.2.3 Creating Multivariate Distributions of Covariates......Page 55
3.2.3.1 Univariate Distributions......Page 56
3.2.3.2 Joint Multivariate Distributions......Page 58
3.2.4 Discrete Functions......Page 61
3.2.4.3 Poisson Distribution......Page 62
3.4 QUALIFICATION OF COVARIATE DISTRIBUTION MODELS......Page 63
3.5 SUMMARY......Page 66
REFERENCES......Page 67
4.1 GOALS OF CLINICAL TRIAL SIMULATION......Page 69
4.2 DEFINITION OF EXECUTION MODELS......Page 70
4.3.1.1 Binomial and Multinomial Distributions......Page 71
4.3.2.1 Conditional Models as Discrete Markov Chains......Page 74
4.4.1 Models of Dose Taking and Dose Timing......Page 75
4.4.2 Markov Model of Medication Compliance......Page 76
4.5 CONSEQUENCES OF PROTOCOL DEVIATIONS......Page 79
4.7 AN ILLUSTRATIVE EXAMPLE......Page 80
4.8 SOFTWARE TOOLS......Page 81
4.9.1 Quetiapine......Page 83
4.9.2 Docetaxel......Page 84
REFERENCES......Page 85
5.1 INTRODUCTION......Page 87
5.2.1 Definition......Page 88
5.2.3 Predictive Models......Page 89
5.3.1 Model Evaluation......Page 90
5.3.2 Model Validation......Page 91
5.3.2.2 Internal Validation......Page 92
5.4.1 Graphical Displays......Page 100
5.4.3 Standardized Prediction Error......Page 102
5.5.1 A Descriptive Model Example......Page 103
5.5.2 A Predictive Model......Page 105
5.5.2.2 Model Evaluation Using the Bootstrap......Page 106
5.5.2.3 Predictive Performance: The Bootstrap Approach......Page 108
5.5.2.5 Implications......Page 112
5.6 SUMMARY......Page 114
REFERENCES......Page 115
6.1 INTRODUCTION......Page 119
6.2 CHOICE OF SOFTWARE PLATFORM......Page 120
6.3 BASIC CONCEPTS OF SIMULATION......Page 121
6.3.1 Give the Virtual Subjects Genetic, Demographic, Physical, and Pathophysiological Characteristics......Page 122
6.3.2 Determine How the Virtual Subjects Will Behave During the Trial......Page 123
6.3.5 Repeat the Simulation Many Times......Page 124
6.3.6 Calculate Precision, Accuracy, and Power......Page 125
6.4 COMPUTER-GENERATED RANDOM NUMBERS......Page 126
6.4.1 Adequacy of Random Number Generators......Page 127
6.4.2 Generating Uniform Random Numbers......Page 128
6.4.3.1 Simulating a Coin Flip (Or Any Situation with Two Outcomes)......Page 129
6.4.3.2 Simulating n-way Branches......Page 130
6.4.5 Generating Normally Distributed Random Numbers......Page 132
6.4.5.2 Generating a Pair of Correlated Normally Distributed Random Variables......Page 133
6.4.6 Generating Log Normally Distributed Random Numbers......Page 134
6.4.6.2 Generating Log Normally Distributed Random Numbers Whose Logarithms Have a Specific Mean and Standard Deviation......Page 135
6.5.1 Solving Pharmacokinetic Models Analytically......Page 136
6.5.2.1 Euler’s Method......Page 137
6.5.2.3 Runge-Kutta Methods......Page 138
6.5.2.4 Other Methods......Page 139
6.6 SIMULATING PHARMACODYNAMIC EFFECTS......Page 140
6.6.2 Simulating Discrete Outcomes......Page 141
6.6.3 Simulating Survival Outcomes......Page 142
6.7 CONCLUSIONS......Page 143
REFERENCES......Page 144
7.1 INTRODUCTION......Page 145
7.2.2 Simulated Experiment Analysis......Page 147
7.2.2.1 A Population Pharmacokinetic Study Example: Determination of Power......Page 148
7.2.2.2 An Efficacy Study Example: Dosage Optimization......Page 150
7.2.3.1 Sources of Multiplicity......Page 152
7.2.3.2 Error Rate and Control......Page 153
7.2.3.3 Multiplicity Adjustment Approaches......Page 154
7.3 COMMUNICATION OF ANALYSIS RESULTS......Page 157
REFERENCES......Page 162
8.1 INTRODUCTION......Page 166
8.2 DEFINITIONS AND SENSITIVITY MEASURES......Page 167
8.3 SENSITIVITY ANALYSIS OF PHARMACOKINETIC AND PHARMACODYNAMIC MODELS......Page 170
8.4 SENSITIVITY ANALYSIS AND CLINICAL TRIAL SIMULATION AND DESIGN......Page 175
8.6 APPENDIX......Page 180
REFERENCES......Page 182
9.1 INTRODUCTION......Page 185
9.1.1 Motivation......Page 186
9.1.2 Optimal Design of Experiments in PK/PD......Page 187
9.1.3 Simulation in PK/PD and Drug Development......Page 188
9.2.1 Information and Goals......Page 190
9.2.2 Data Dependence Versus Data Independence......Page 191
9.2.4 Analytical Optimization......Page 192
9.2.7 Sufficient, Best, and Optimal Designs......Page 195
9.3.1 Parameter Estimation Precision......Page 196
9.3.2 Fisher Information Matrix......Page 197
9.3.3 Parameter Estimation Accuracy......Page 198
9.3.4 Parameter Estimation Efficiency......Page 199
9.4.1.1 Grid Methods......Page 200
9.4.2.2 Simplex Method......Page 201
9.4.2.4 Simulated Annealing......Page 202
9.5 THE SIMULATION EXPERIMENT......Page 203
9.5.2 Design and Simulation......Page 204
9.5.4 Databases and Optimization......Page 205
9.6.3 Time Requirements......Page 206
9.9 FUTURE CHALLENGES......Page 207
9.10 SUMMARY......Page 208
REFERENCES......Page 209
10.1 INTRODUCTION......Page 213
10.2 BACKGROUND......Page 215
10.3.1.2 Database Management......Page 216
10.3.1.7 Technical Issues......Page 217
10.3.1.8 Audit Trail of M&S Procedure......Page 218
10.4 GENERAL POINTS TO CONSIDER......Page 219
REFERENCES......Page 221
11.2.1 Stochastic Simulations......Page 223
11.2.2 Mechanistic Simulations......Page 224
11.3.2 Design......Page 225
11.3.4 Meta-Analysis......Page 226
11.4.1.1 The Straight Line......Page 227
11.4.2 Objective Functions......Page 228
11.4.2.2 Weighted Least Squares......Page 229
11.5.2 ModelMaker......Page 230
11.6.1 Model......Page 231
11.6.3 Analysis......Page 232
11.7.1 Workshop Outline......Page 233
11.7.2 Workshop Variations Depending on Students’ Previous Knowledge......Page 235
11.7.3 Workshop Objectives and Experience......Page 236
REFERENCES......Page 237
12.1 INTRODUCTION......Page 239
12.2 CLINICAL TRIAL SIMULATION—A SWISS ANALOGY......Page 240
12.3.1 Pre-/Nonclinical Development......Page 241
12.3.2 Early Clinical Development......Page 243
12.3.3 Late Clinical Development......Page 244
12.3.4 The Learn-Confirm Paradigm......Page 246
12.3.5 Perceptions of M&S by the Nonmodelers in the Industry......Page 247
12.4.1 A Team Game......Page 249
12.5 MODELING AND SIMULATION: PROSPECTS......Page 251
REFERENCES......Page 252
13.1 INTRODUCTION......Page 257
13.2 ORIGINS OF MODELING AND SIMULATION IN PHYSIOLOGY......Page 258
13.2.1 Pioneering Work on the Control of Blood Pressure......Page 259
13.3 DEVELOPMENT OF CONTROL THEORY......Page 260
13.3.2 Analysis of Feedback Control in Other Areas of Physiology......Page 261
13.3.3 Another Illustration of the Time-Scales Concept in Physiological Regulation......Page 262
13.4 THE UDRS NONLINEARITY......Page 264
13.4.2 UDRS in the Pupillary Re•ex to Light......Page 265
13.4.4 UDRS at the Level of the Integrated Baroreceptor Re•ex......Page 266
13.4.5 UDRS in Other Receptors......Page 267
13.4.6 UDRS in the Pharmacodynamics of Nifedipine......Page 268
13.4.8 Summary of the Key Dynamic Features of the UDRS Nonlinearity......Page 269
13.6.1 First Objective: Identifying Particular Temporal Patterns of Drug Input That Can Trigger Hazardous Effects......Page 270
13.6.2 Second Objective: Providing Crucial Constraints on PD Modeling......Page 272
13.6.3 Third Objective: Understanding How PD May Change as the Duration of Exposure to the Drug Lengthens......Page 273
13.8 HARD VS. SOFT NONLINEARITIES......Page 274
13.9 HOW MANY DIFFERENT TEMPORAL PATTERNS ARE ENOUGH?......Page 275
13.10 MINIMAL CASETTE OF TEMPORAL PATTERNS OF DRUG INPUT FOR ROBUST PD MODELING......Page 276
13.11 STARTLING PD OF GONADOTROPIN RELEASING HORMONE AND ITS ROLE IN THE CONTROL OF OVULATION......Page 277
13.13 CONCLUSION......Page 278
REFERENCES......Page 279
14.1 INTRODUCTION......Page 282
14.2.2 Modeling Methods......Page 283
14.2.3.1 Model Fit......Page 284
14.2.2 Simulation Model Validation......Page 285
14.3.1 Data......Page 287
14.3.2 Modeling Methods......Page 288
14.3.3 Results......Page 289
14.4 POPULATION PK SUBSTUDY DESIGN......Page 290
14.5.1 Methods......Page 291
14.5.2 Results......Page 293
14.6 DISCUSSION......Page 298
REFERENCES......Page 299
15.1 INTRODUCTION......Page 300
15.2 DEVELOPMENT OF THE DRUG MODEL......Page 302
15.3 EXPLORATION OF THE LIKELY DOSE-RESPONSE RELATIONSHIP FOR THE NCE......Page 307
15.4 STUDY POWER CONSIDERATIONS......Page 308
15.5 POWER TO DETECT A DOSE-RESPONSE RELATIONSHIP......Page 312
15.6 ABILITY TO SELECT DOSES FOR FUTURE DEVELOPMENT......Page 314
15.7 DISCUSSION......Page 321
REFERENCES......Page 322
16.1 INTRODUCTION......Page 324
16.2 MODELS......Page 325
16.3 PK MODEL PARAMETER SPECIFICATION......Page 326
16.3.1 Absorption Lag Time......Page 327
16.4.1 Rate Constant of the Effect Compartment......Page 328
16.4.2 Logistic Model Parameters......Page 329
16.5 SIMULATIONS......Page 330
16.6.1 Optimal Design......Page 331
16.6.2 Sample Size Determination......Page 333
16.7 DISCUSSION......Page 334
REFERENCES......Page 335
17.1 INTRODUCTION......Page 337
17.2 BACKGROUND—THE MATHEMATICAL MODEL......Page 338
17.3 LITERATURE STUDIES AND DATA FOR MODEL CHECKING......Page 340
17.3.2 Fitting of Experimental Data......Page 341
17.4 MODEL-BASED PREDICTIONS—ASSUMPTIONS......Page 344
17.5.1.2 Pharmacodynamics......Page 346
17.5.2 Model-Based Predictions: Extrapolating to Patients from the Healthy Volunteer Data......Page 352
17.6.1 A Patho-Physiologically Based, Mechanistic PK/PD Model......Page 353
17.6.2 Experimental Fitting of Patient Data......Page 355
17.6.3 Extrapolations to Patients from the Healthy Volunteer Data......Page 356
17.7 CONCLUSION......Page 358
REFERENCES......Page 359
18.1 INTRODUCTION......Page 362
18.2.1 Study Design......Page 363
18.2.4 Population Analysis of Pharmacokinetic Data......Page 364
18.2.6 Population Analysis of Combined Pharmacokinetic and Pharmacodynamic Data......Page 365
18.3.1 Population Analysis of Pharmacokinetic Data......Page 366
18.3.2 Population Analysis of Pharmacodynamic Data During Placebo Administration......Page 367
18.3.3 Population Analysis of Combined Pharmacokinetic and Pharmacodynamic Data......Page 368
18.3.4 Monte Carlo Simulation and Analysis......Page 369
18.4 DISCUSSION......Page 370
REFERENCES......Page 375
19.1.1 Clinical Problem and Goals......Page 377
19.1.2 Theoretical Background and Rationale for the Use of Simulation......Page 381
19.3.1 Simulation Plan......Page 385
19.3.2 Simulation Model......Page 386
19.3.5 Model Evaluation......Page 387
19.3.6.2 Updating Scheme for Population Hyper-Priors......Page 388
19.3.7 Assessment of Simulation Results......Page 389
19.4.1 Updating Scheme for Population Priors......Page 390
19.4.2 Effect of PK Sampling Times on Individual Parameter Estimates......Page 393
19.5 CONCLUSIONS AND DISCUSSION......Page 395
REFERENCES......Page 396