Pharmacodynamics: the methods In vitro models Animal models Clinical studies Population studies With the support of Wallonie-Bruxelles-International 3B-1
Pharmacodynamics: the methods "un peu de tout " 3B-2
In vitro dynamic models Dilution models Diffusion models Hybrid models Physiologic models Intracellular models Adapted from J. Mouton, 4th ISAP Educational Workshop, 2001 3B-3
Dilution models: a simple, useful system... Fresh Medium Pump Inflow = Clearance Waste Drug Conc. T 1/2 = 0.693 * V/Cl Peak:MIC AUC:MIC Ratio Time Above MIC MIC Time V Sampling of Drug, Bacteria Adapted from M.N. Dudley, ISAP / FDA Workshop, March 1st, 1999 3B-4
Diffusion models Membranes (hollow fibers) dialyzers (artificial kidneys) Adapted from M.N. Dudley, ISAP / FDA Workshop, 1999 3B-5
Some models can be very complex 3B-6
The goal is to mimic potentially useful and achievable serum concentration variations 120 Ceftazidime (mg/l) 90 60 30 0 0 6 12 18 24 30 36 Time (h) intermittent dosing continuous infusion Adapted from J. Mouton, 4th ISAP Educational Workshop, 2001 3B-7
Why in vitro dynamic models... The goal is to establish basic relationships between drug exposure (PK) and effect (PD) PK/PD parameters for efficacy to apply across species, models, for combinations, etc... Basis of dosage in phase II trials Limitations: Experimental conditions (laboursome; contamination; ) Usually only 1 or 2 days (effects fade after 12-24 h) absence of host factors (includ. protein binding and metabolism)... Adapted from J. Mouton, 4th ISAP Educational Workshop, 2001 3B-8
Animal models neutropenic mouse rabbit (endocarditis) rat, guinea pig,... The main advantage is the possibility to explore a VERY large array of dosing regimens so as dissociate PK covariables (C max vs AUC ) explore the PK conditions of failure Adapted from W.A. Craig, 2d ISAP Educational Workshop, 2000 3B-9
Dissociating PK covariables: see what are C max, time above MIC and AUC with a once-a-day (qd) schedule of a given dose Concentration, ng/ml 800 700 600 500 400 300 200 100 C max / MIC T > MIC qd dosing dose = 1 MIC 90 0 0 4 8 12 16 20 24 Time, hours Adapted from F. O. Ajayi, ISAP-FDA Workshop, 1999 3B-10
Now see what are C max, time > MIC and AUC/MIC if increase the dose without changing the schedule C max / MIC Concentration, ng/ml 800 700 600 500 400 300 200 100 C max / MIC T > MIC qd dosing dose = 1.4 Peak/MIC T > MIC AUC / MIC MIC 90 0 0 4 8 12 16 20 24 Time, hours Adapted from F. O. Ajayi, ISAP-FDA Workshop, 1999 3B-11
But see how C max, time > MIC and AUC/MIC become dissociated if the SAME DAILY dose is given with a different schedule (here: divided in 3 administrations) Concentration, ng/ml 800 700 600 500 400 300 200 100 Peak/MIC T > MIC AUC / MIC = tid dosing C max / MIC MIC 90 T > MIC 0 0 4 8 12 16 20 24 Time, hours Adapted from F. O. Ajayi, ISAP-FDA Workshop, 1999 3B-12
A typical animal model to establish which PK parameter is associated with efficacy Use neutropenic murine thigh-and lung-infection models Evaluate 20-30 different dosing regimens (5 different total doses given at 4-6 different dosing intervals) Measure efficacy from change in Log 10 CFU per thigh or lung at the end of 24 hours of therapy Correlate efficacy with various pharmacodynamic parameters (Time above MIC, peak/mic, 24-Hr AUC/MIC) Adapted from W.A. Craig, 2d ISAP Educational Workshop, 2000 3B-13
Relationship Between Peak/MIC Ratio and Efficacy for Cefotaxime against Klebsiella pneumoniae in a Murine Pneumonia Model (after W.A. Craig * ) Bacterial growth Bacterial killing Log 10 CFU per Lung at 24 Hours 10 9 8 7 6 5 0.1 1 10 100 1000 10000 Peak/MIC Ratio No correlation with peak / MIC ratio!! * 2d ISAP Educational Workshop, Stockholm, Sweden, 2000 3B-14
Relationship Between 24-Hr AUC/MIC and Efficacy for Cefotaxime against Klebsiella pneumoniae in a Murine Pneumonia Model (after W.A. Craig * ) Bacterial growth Bacterial killing Log 10 CFU per Lung at 24 Hours 10 9 8 7 6 5 3 10 30 100 300 1000 3000 24-Hour AUC/MIC Ratio No correlation with AUC / MIC ratio!! * 2d ISAP Educational Workshop, Stockholm, Sweden, 2000 3B-15
Relationship Between Time Above MIC and Efficacy for Cefotaxime against Klebsiella pneumoniae in a Murine Pneumonia Model (after W.A. Craig * ) Bacterial growth Bacterial killing Log10 CFU per Lung at 24 Hours 10 9 8 7 6 5 R 2 = 94% 0 20 40 60 80 100 Time Above MIC (Percent) Excellent correlation with time above MIC!! * 2d ISAP Educational Workshop, Stockholm, Sweden, 2000 3B-16
End-points of animal models Bacterial counts static dose 50 % effect E max Mortality Log10 CFU per Thigh at 24 Hrs 9 8 7 6 5 P50 1 Log Kill Static Dose Emax 10 30 100 300 Dose (mg/kg/6 hrs) Recovery of resistant bacteria * 2d ISAP Educational Workshop, Stockholm, Sweden, 2000 3B-17
Demonstrated advantages of animal models Is the magnitude of the parameter required for efficacy the same in different animal species? YES Does the magnitude of the parameter vary with: 1. the dosing regimen? NO 2. different drugs within the same class? NO 3. different organisms? Minimal 4. different sites of infection (e.g. blood, lung, peritoneum, soft tissue)? NO, but... Adapted from W.A. Craig, 2d ISAP Educational Workshop, 2000 3B-18
PK/PD of fluoroquinolones in clinics Demonstration of the role of the 24h-AUC / MIC ratio In nosocomial pneumonia Forrest et al., AAC, 1993 3B-19
Link between 24h-AUC /MIC and clinical success clinical outcome AUC / MIC AUC F. O. Ajayi, ISAP-FDA Workshop, 1999 3B-20
24h AUC / MIC : what were the data of the Forrest et al's study? Parameter No.Pat. % CureMicrob. % CureClinical MIC (mg/l) <0,125 28 82 79 0,125-0,25 13 75 69 succes 0.5 14 54 79 1 9 33 44 failures 2 2 0 0 24h AUC / MIC 0-125 19 32 failures 42 125-250 16 81 88 success 250-1000 14 79 71 1000-5541 15 87 80 Forrest et al., AAC, 1993 3B-21
AUC/CMI =125 : a magical number?? 125 was the limit below which failure rates became unacceptable based either on a large MIC or on a low dosage (AUC is proportional to the dosage) 3B-22
Is 125 good for all?? For S. pneumoniae, it all depends on your immune status 100 100 Mortality (%) 80 60 40 20 Emax at 30... Mortality (%) 80 60 40 20 Emax at 125... 0 0 1 2.5 5 10 25 50 100 3 10 30 100 300 1000 24 Hr AUC/MIC 24 hr AUC/MIC non-neutropenic mice Neutropenic mice 3B-23
Why are the conclusions of the clinical trials apparently (sometimes and apparently) contradictory? insufficient separation of covariables only one or a few dosage regimens not enough true failures Pathologies pas assez sévères study design intercurrent variables influencing outcome and not recognized as such unsufficient or inappropriate collection of PK data only peaks or troughs... Correct but uncomplete conclusion No conclusion possible Conclusions of poor value (shed confusion ) 3B-24
Population approaches : Doctor or Regulator? In clinical therapy, we would like to give optimal dose to each individual patient for the particular disease Individualized therapy In new drug assessment / development, we would like to know the overall probability for a population of an appropriate response to a given drug and proposed regimen Population-based recommendations H. Sun, ISAP-FDA Workshop, 1999 3B-25
Obtaining population cumulative frequencies 3B-26 100 90 80 70 60 50 40 Frequency 30 20 10 0 2.231 2.604 2.977 3.350 3.724 4.097 4.470 4.843 5.216 5.590 5.963 6.336 6.709 7.083 7.456 7.829 Drug Concentration (ug/ml) 100 90 80 70 60 cumulative effect (%) Quantal drug concentration effects 140 120 100 80 60 40 20 0 Frequency Cumulative %, MIC=2 Cumulative %, MIC=4 Cumulative %, MIC=5 50 40 30 20 10 0 Frequency 0.00 2.00 3.39 4.52 5.65 6.77 7.90 9.03 10.15 11.28 12.41 13.54 Tmic 100% 90% 80% 70% % Subject Quantal T>MIC plots H. Sun, ISAP-FDA Workshop, 1999 60% 50% 40% 30% 20% 10% 0%
Monte Carlo simulations 3B-27
Monte Carlo Simulation : the basics randomly" generating at least 10,000 scenarios of PK and PD parameters that could be seen in patients Determining what the PK/PD values would be under each of the 10,000 scenarios Forming a histogram of those results. This represents a discrete approximation for the probability distribution of the data. Monte Carlo simulation allows us to make use of prior knowledge of how a target population handles a specific drug to predict how well that drug will perform clinically at the dose chosen for clinical trials 3B-28
Monte Carlo Simulation How is this done? Through use of data from a population PK study, a sampling distribution is set up think of every body in the world in a bucket from which you randomly select a large number of subjects, each of whom knows their PK parameter values. This allows the pertinent PK parameters to be calculated for all the subjects you then only need to apply your pertinent PD parameter!! Modified from: G. Drusano, Joint ISAP/ECCMID Symposium, Glasgow, UK, May 11th, 2003 3B-29
Monte Carlo simulation for pneumococci (based on AUC/MIC) f f AUC 1. Patients' PK distribution MIC 2. Bacteria MIC distribution 3B-30
Monte Carlo simulation for pneumococci (based on AUC/MIC) f f patients AUC MIC broth f AUC / MIC 3. Simulated AUC/ MIC distribution 3B-31
Monte Carlo simulation for pneumococci (based on AUC/MIC) f f 1. patients AUC MIC 2. broth 3. Simulation f AUC / MIC 4. Solve the equations for the AUC values of 3 quinolones f 0 Cipro 50 100 200 300 Levo Moxi AUC / MIC 3B-32
Monte Carlo simulation for pneumococci (based on AUC/MIC) The results are obvious f 0 Cipro 50 100 200 300 Levo Moxi AUC / MIC 3B-33
Another look at Monte-Carlo simulations : Levofloxacin Vs S. pneumoniae Probability 0.05 0.045 0.04 0.035 0.03 0.025 0.02 0.015 0.01 0.005 0 Levofloxacin Certainty is only 80% to get values of AUC:MIC ratio higher than 30 0 50 100 150 200 250 300 350 400 AUC:MIC ratio Preston SL, Drusano GL et at. AAC 1998;42:1098-1104; Ambrose PG, Grasela D. ICAAC 1999 Ambrose PG et al Chapter 17 in Antimicrobial Pharmacodynamics in Theory and Clinical Prectice, eds Nightingale CH, Murakawa T, Ambrose PG. 2002. Marcel Decker, NY 3B-34
Those methods allow to know that for each antibiotic PK PD Therapeutic effects Dosing C max AUC half-life dose response Time E max Toxic effects 3B-35
Those methods allow to know that for each antibiotic PK PD Therapeutic effects Dosing C max AUC half-life dose response Time E max Toxic effects We now will tell you what these methods show. Section 3 c 3B-36