Chapter
Introduction: Don’t Just Dose—Choose a Specific Target Goal, Suited to the Patient’s Need, and Dose to Hit It Most Precisely
1 Ways of Thinking—Qualitative and Quantitative
2 Graphical Plots and Optical Illusions
3 Other Illusions Sharing This Feature of Perception
3.1 The Concept of “Half-Time”
3.2 The Saying That “One Has to Lose Two-Thirds to Three-Quarters of One’s Renal Function Before the Serum Creatinine Begin...
3.3 The Saying “Get a ‘Peak’ Aminoglycoside Serum Sample Half an Hour After the End of the Infusion”
3.4 “Therapeutic Ranges” of Serum Drug Concentrations
4 General Remarks About Dosing
I. Basic Techniques for Individualized Therapy
1 Basic Pharmacokinetics and Dynamics for Clinicians
1.1 Excretion Is Usually Proportional to Amount or Concentration
1.2 Accumulation Takes Place by the Mirror Image of Elimination
1.3 Suiting Loading and Maintenance Doses to Each Other
1.4 The Basic Idea—Dose and Half-Time—They Let You Control the Total Amount of Drug You Permit the Patient to Have in the B...
1.5 Events Following a Change in Daily Maintenance Dose
1.6 Events Following a Change in Excretion Rate
1.7 Separating Elimination Into Renal and Nonrenal Components
1.8 Adding More Compartments for a More Realistic Pharmacokinetic Model
1.9 Output Equations: Describing the Observations
1.10 Parameterizing The Model: Volume and Clearance or Volume and Rate Constant?
1.11 The Clearance Community in PK
1.12 A Current Clinical Issue: “Augmented Renal Clearance” in the ICU
1.13 Properties of Systems: Observability, Identifiability, and Controllability
1.14 Nonlinear Drug Systems
1.14.1 Michaelis–Menten Systems
2 Describing Drug Behavior in Groups of Patients
2.1 Early Approaches to Modeling
2.1.1 The Naïve Pooling Approach
2.1.2 The Standard Two-Stage (S2S) Approach
2.1.3 The Iterative Two-Stage Bayesian (IT2B) Approach
2.2 True Population Modeling Approaches
2.2.1 Parametric Models With Assumed and Constrained Parameter Distributions
2.2.2 Nonparametric (NP) Population Models With Unconstrained Parameter Distributions
3 Developing Maximally Precise Dosage Regimens for Patients—Multiple Model (MM) Dosage Design
3.1 Again, Select a Specific Target, Not a Range
3.2 The Separation Principle
3.3 The Way Around the Separation Principle: Multiple Model Dosage Design
4 Optimizing Laboratory Assay Methods for Individualized Therapy
4.1 Introduction: Wrong Weighting of Data, Wrong PK Models, Wrong Doses
4.2 Percent Coefficient of Variation is Not the Correct Measure
4.3.1 Calculating the Assay CV%
4.3.2 Calculating the Reciprocal of the Variance
4.4 Results: Application to Real Assay Data
4.4.1 Examining an Assay for Vancomycin
4.4.3 Another Example: Voriconazole
4.4.4 An Example of Gentamicin
4.5 Discussion: LLOQ is an Illusion
4.5.1 Within-Day and Between-Day Variability
4.5.2 Defining the Lower Limit of Assay Detection (LLOD) is Improved
4.6.1 Reciprocal of Measurement Variance Is the Correct Measure of Assay Precision
4.6.2 The Assay Error Polynomial Provides a Practical Way to Store the Error Data
4.6.3 Relationship to Other Clinical Sources of Error in the Therapeutic Environment
4.6.4 One Can Stop Censoring Low Measurements
4.6.5 Other Sources of Uncertainty in the Clinical Environment
5 Evaluation of Renal Function
5.1 Classical Estimation of Creatinine Clearance (CCr), Based on Urinary Excretion
5.1.1 Estimation of CCr Without a Urine Specimen, Based on a Single SCr
5.2 Problems With Estimates of CCr Using Only a Single Serum Creatinine (SCr) Sample
5.3 Estimating CCr From a Pair of SCr Samples at Known Times
5.3.1 A Dynamic Mass-Balance Model of Total Body Creatinine Over Time
5.3.2 Calculation of Daily Creatinine Production (P)
5.3.2.1 Age and Creatinine Production
5.3.2.2 Chronic Uremia and Creatinine Production
5.3.3 Calculation of Daily Creatinine Excretion
5.4 The Final Overall Formula
5.5 When Did the Patient’s Renal Function Change?
5.6 Uncertainties in the Gold Standard Measurement of Creatinine Clearance
5.7 Comparison of Estimated Versus Measured Creatinine Clearance
5.8 Comparison With Cockcroft–Gault Estimation When SCr is Stable
5.9 Should Ideal Body Weight Be Used Instead of Total Body Weight?
5.10 Changing SCr—The Direct Clinical Link Between the Patient’s Changing Renal Function and Drug Behavior
II. The Clinical Software
6 Using the BestDose Clinical Software—Examples With Aminoglycosides
6.1 Introduction—The BestDose Clinical Software
6.2 Two Representative Drugs—Amikacin and Gentamicin
6.3 Planning the Initial Regimen
6.4 Analyzing a Gentamicin Patient’s Existing Data, and Developing the Adjusted Regimen
6.6 Planning the New Adjusted Dosage Regimen
7 Monitoring the Patient: Four Different Bayesian Methods to Make Individual Patient Drug Models
7.2 But First, Weighted Nonlinear Least Squares Regression
7.3 Using Bayes’ Theorem in Analyzing Data, Using Parametric PK Models
7.3.1 MAP Bayesian Analysis
7.4 Bayesian Analysis for Nonparametric (NP) Models
7.5 Hybrid Bayesian Analysis
7.6 The Interacting Multiple Model (IMM) Bayesian Approach to Unstable ICU Patients
7.7 Using the Augmented Population Model From the Hybrid as the Bayesian Prior for Subsequent IMM Analysis
Appendix: More Detail on Nonparametric Bayesian Analysis
8 Monitoring Each Patient Optimally: When to Obtain the Best Samples for Therapeutic Drug Monitoring
8.2 Optimizing Therapeutic Drug Monitoring (TDM) Protocols and Policies
8.3 D-Optimal Design and Its Variants
8.4 D-Optimal Times Also Depend Upon the Dosage Format
8.4.1 Intermittent Intravenous (IV) Infusion
8.4.2 Continuous IV Infusion
8.4.3 A Loading Followed by a Maintenance Infusion
8.5 Multiple Model Optimal (MMopt) design
8.6 New Specific Clinical Tasks That Can Also Be Optimized with WEIGHTED MMopt (wMMopt)
9 Optimizing Individualized Drug Therapy in the ICU
9.3 Apparent Volume of Distribution, Drug Elimination, and Clearance
9.4 Increased and Changing V and “Augmented Renal Clearance” in ICU Patients
9.5 Tracking Drug Behavior Optimally in Unstable Patients
9.6 An Illustrative Chronic Dialysis Patient With Sepsis
9.7 IMM Analysis of the Patient’s Data
9.8 Another Patient, Highly Unstable, With High Intraindividual Variability
9.9 Two New Moves to Further Improve the IMM Approach
10 Quantitative Modeling of Diffusion Into Endocardial Vegetations, the Postantibiotic Effect, and Bacterial Growth and Kill
10.2 Diffusion Into Endocardial Vegetations
10.2.1 Simulated Vegetations of Various Diameters
10.3 Simulating a Small Microorganism
10.4 Modeling Bacterial Growth and Kill
10.4.1 General Considerations
10.5 An Illustrative Case
10.5.1 Further Analysis of the Model
11 Individualizing Digoxin Therapy
11.2 The Population Model of Digoxin
11.3 Implications for Dosage
11.4 Adjusting Initial Dosage to Body Weight and Renal Function
11.5 Variability in Response: The Need for Monitoring Serum Concentrations and Dosage Adjustment
11.5.1 Protocols for Monitoring Serum Concentrations
11.6 The Very Wide Spectrum of Serum Digoxin Concentrations and Patient Responses
11.7 Management of Patients with Atrial Fibrillation and Flutter
11.8 An Illustrative Patient
11.9 Another Patient Who Converted Three Times but Relapsed
11.10 Another Case—A Very Large, Heavy Patient Who Did Not Convert
11.11 Ratios Between Central and Peripheral Compartments
11.12 The Effect of Serum Potassium
11.13 A Very Relevant Patient
III. Clinical Applications of Individualized Therapy
12 Optimizing Single-Drug Antibacterial and Antifungal Therapy
12.2 Minimum Inhibitory Concentration
12.4.1 Pharmacokinetic and Pharmacodynamic Relationships
12.6 Use of Therapeutic Drug Management and Multiple Model Bayesian Adaptive Control of Dosage Regimens
12.7 Problems with Trough-Only Sampling
12.8 An Illustrative Patient
12.9 Issues in Fitting Data
12.10 The Approach to the Patient
12.12 An Illustrative Patient
12.13 Evaluation of Dosage Guidelines
12.14 Another Illustrative Patient
13 Combination Chemotherapy With Anti-Infective Agents
13.1 Why Employ Combination Therapy?
13.2 Increased Spectrum of Empirical Coverage
13.3 Increased Bacterial Kill With Additive or Synergistic Interaction
13.4 What Are Synergy, Additivity, and Antagonism?
13.5 Suppression of Amplification of Less-Susceptible Subpopulations
13.6 Suppression of Protein Expression (If One Agent Is a Protein Synthesis Inhibitor)
13.6.1 Why Not Use Combination Therapy?
14 Controlling Antiretroviral Therapy in Children and Adolescents with HIV Infection
14.2 Pharmacokinetics (PK)
14.3 Pharmacodynamics (PD)
14.3.1 Drug Exposure and Efficacy
14.3.1.1 Nucleoside Reverse Transcriptase Inhibitors (NRTIs)
14.3.1.2 Nonnucleoside Reverse Transcriptase Inhibitors (NNRTIs)
14.3.1.3 Protease Inhibitors (PIs)
14.3.1.4 Integrase Strand Inhibitors (INSTIs) and Entry Inhibitors (EI)
14.3.2 Drug Exposure and Toxicity
14.4 Pharmacogenomics (PG)
14.5 ARV Therapeutic Drug Monitoring/Management (TDM)
14.6 Multiple-Model Bayesian Adaptive Control: Case Examples
14.8 Patient 1. General Techniques and the Need for Nonstandard Dosage Schedules
14.9 Patient 2. Unsuspected Impaired Clearance: Patients Needing Smaller Doses Than Usual
14.10 Patient 3. Low Concentrations: Underdosing or Poor Adherence?
14.11 Patient 4. Adolescents: Should They Get Adult Doses?
14.12 Patient 5. Extrapolating From Adults to Children
15 Individualizing Tuberculosis Therapy
15.1 Introduction: The WHO and Public Health Approach to Anti-TB Drug Dosing: One-Size-Fits-All
15.2 The Rationale for Dose Individualization of Anti-TB Drugs
15.2.1 Pharmacokinetic Variability
15.2.2 Pharmacodynamic Variability
15.2.3 Pharmacokinetic–Pharmacodynamic Relationships
15.2.4 Conclusion: One Size Cannot Fit All
15.3 How to Individualize Anti-TB Drug Regimens
15.3.2 Therapeutic Drug Monitoring (TDM) of Anti-TB Drugs
15.3.3 Bayesian Dose Individualization
16 Individualizing Transplant Therapy
16.2 Calcineurin Inhibitors (CNI)
16.2.1.1 Cyclosporine A pharmacokinetics
16.2.1.2 Therapeutic drug monitoring (TDM) of cyclosporine A
16.2.1.3 Pharmacokinetic modeling of CsA
16.2.1.4 Clinical validation
16.2.2.1 Tacrolimus pharmacokinetics
16.2.3 Sirolimus and Everolimus
16.2.3.1 mTORi pharmacokinetics
16.2.4.1 Pharmacokinetics of mycophenolic acid
16.2.4.2 Exposure-effect relationships: mycophenolate
16.2.4.3 Multiple linear regression (MLR) equations for AUC estimation
16.2.4.4 Pharmacokinetic modeling: mycophenolate
16.2.4.5 Bayesian estimation of mycophenolate AUC
16.2.4.6 Clinical utility of mycophenolate mofetil TDM
17 Individualizing Dosage Regimens of Antineoplastic Agents
17.1 History and Current Status
17.1.5 Targeted Therapies
18 Controlling Busulfan Therapy in Children
19 Individualizing Antiepileptic Therapy for Patients
19.1.1 Patient Data and Methods
19.2 Population Modeling: Results
19.2.1.1 Modeling of CBZ pharmacokinetics after completion of an autoinduction period
19.2.1.2 Results for CBZ-polytherapy
19.2.1.3 PK modeling of postinduction CBZ and its main metabolite
19.3.2 Statistical Analysis
19.3.3 Results and Discussion
19.4 More Complex Nonlinear PK Models
19.4.1 Time-Dependent CBZ Pharmacokinetics During Autoinduction
19.4.1.1 Patient data and PK analysis
19.4.2 Phenytoin Concentration-Dependent, or Saturable, Pharmacokinetics
19.4.2.1 Patient data and PK analysis
19.5 Indications for TDM and Individualizing AED Dosage
20 Individualizing Drug Therapy in the Elderly
20.2 Highlights of Some Biological Aspects of Aging
20.3 Pharmacodynamic Changes in the Elderly and Their Therapeutic Implications
20.4 The Renal Aging Process and Its Pharmacokinetic Consequences
20.5 A Special Case: Intraindividual Variability in the Elderly
20.6 Conclusions and Perspectives
21 The Present and Future State of Individualized Therapy
21.1 Models of Large, Nonlinear Systems of Multiple Interacting Drugs
21.1.1 Optimizing Combination Drug Therapy
21.2 Equations Without Constant Coefficients
21.2.1 Obstacles to Progress: Institutionalized Ritualistic Behavior, in Which the Patient Is Nothing and Ritual Is Everything
21.3 The Pharmaceutical Industry, Doses, Patients, and Missed Opportunities
21.4 The Pharmaceutical Industry and Clinical Trials
21.4.1 Teaching Pharmacokinetics
21.4.2 Teaching Decision Analysis
21.5 Bayes’ Theorem and Medical Decisions
21.6 The Two-Armed Bandit
21.7 Conclusion—Monitor Each Patient Optimally and Control the System Optimally
Individualized Drug Therapy for Patients
：Basic Foundations, Relevant Software and Clinical Applications
Disclaimer: Any content in publications that violate the sovereignty, the constitution or regulations of the PRC is not accepted or approved by CNPIEC.
