Back to the Top
Date: Mon, 09 Jun 2003 15:13:15 -0500
From: "Rob Ariano"
Subject: Question for PharmPK - Antimicrobial Dosage
Hi all, I would like to enquire with the group about the decision
making process involved in selecting a dosing regimen for a new antimicrobial in
humans. Obviously toxicological studies in animals assist with selecting
potentially tolerable regimens in humans, however, do investigators about to
embark on a large scale clinical trial with the new antibiotic use PK/PD
surrogate targets to design the single dosing regimen for all patients? If you
are working with a new fluoroquinolone for pseudomonal infections, do you first
study dosing regimens which result in P/MIC targets of >8:1, or AUC/MIC > 125:1
or both, based on MIC90's? (ie. after toxicology studies have shown these
concentrations to be easily tolerable by the majority of humans).
Thanks, Very best regards,
Robert Ariano, Pharm.D.,BCPS
Clinical Pharmacist Critical Care
St.Boniface General Hospital;
& Associate Professor of Pharmacy,
& Medicine, University of Manitoba,
204-237-2050 Phone 204-237-2165 FAX
rariano.aaa.sbgh.mb.ca
www.sbgh.mb.ca
Back to the Top
Date: Mon, 09 Jun 2003 15:13:15 -0500
From: "Rob Ariano"
Subject: Question for PharmPK - Antimicrobial Dosage
Hi all, I would like to enquire with the group about the decision
making process involved in selecting a dosing regimen for a new antimicrobial in
humans. Obviously toxicological studies in animals assist with selecting
potentially tolerable regimens in humans, however, do investigators about to
embark on a large scale clinical trial with the new antibiotic use PK/PD
surrogate targets to design the single dosing regimen for all patients? If you
are working with a new fluoroquinolone for pseudomonal infections, do you first
study dosing regimens which result in P/MIC targets of >8:1, or AUC/MIC > 125:1
or both, based on MIC90's? (ie. after toxicology studies have shown these
concentrations to be easily tolerable by the majority of humans).
Thanks, Very best regards,
Robert Ariano, Pharm.D.,BCPS
Clinical Pharmacist Critical Care
St.Boniface General Hospital;
& Associate Professor of Pharmacy,
& Medicine, University of Manitoba,
204-237-2050 Phone 204-237-2165 FAX
rariano.aaa.sbgh.mb.ca
www.sbgh.mb.ca
Back to the Top
From: GLDrusano.-at-.aol.com
Date: Sun, 15 Jun 2003 08:43:23 -0400
To: david.-a-.boomer.org
Subject: Re: PharmPK Antimicrobial Dosage
The following message was posted to: PharmPK
HI!
This is relatively straightforward. There are several tasks that need to be accomplished to make this decision:
1) What is the goal of therapy? This needs to be answered in the context of the patient group being treated. If the group is not overly ill (e.g. Fine class I-III CAP), then using an exposure target that achieves stasis is fine (no pun intended). If the patient group being treated has vetilator-associated pneumonia, then an exposure that drives a 2-3 log10 (CFU/ml) kill is more appropriate.
As to whether one chooses Peak/MIC or AUC/MIC for a fluoroquinolone probably is determined by the bacterial burden at the infection site. It is likely that, for fluoroquinolones, when one finds peak/MIC as the pharmacodynamically-linked variable, it is because the number of organisms at the primary infection site substantially exceeds (by an order of magnitude, or more) the inverse of the mutational frequency to resistance. When this occurs, there will be a high probability that there will be clones in the population that are resistant. This is the circumstance where animal studies often show the Peak/MIC linkage. We have recently examined this problem for a fluoroquinolone in a mouse model and defined and prospectively validated an exposure target to shut off amplification of the resistant subpopulation. This is In Press at the Journal of Clinical Investigation.
Please also remember that the human studies do not have the breakpoints etched in stone. An AUC/MIC exposure target of 125 as set forth by Alan Forrest works well, but is fully stochastic. By this I mean that if the patient achieves 120 instead of 125, the probability of failure changes very little. Our group has studied a similar set of patients prospectively and shown for a different fluoroquinolone that the AUC/MIC breakpoint is between 87 and 110. This is a function of the way classification and regression tree analysis works. Further, NOT all organisms were created equal. For Gram-negative bacilli, somewhere around 90-125 for an AUC/MIC target works just fine. For pneumococcus, a free drug target of around 30 is a quite reasonable value.
2) What is the variability in PK for the new drug? This aspect is handled by performing a Monte Carlo simulation. This allows the variability in exposure to be discerned for the choice of a specific dose.
3) What is the protein binding? Our group measures free drug, but also performs MBC testing in 90% human serum to ascertain the impact of protein binding on antimicrobial activity. In this evaluation, we use arithmetic, not geometric diluting series to get a better handle on the impact of binding. Once we know this, then the exposures are corrected for the impact of binding (in the Monte Carlo simulation above).
4) What is the distribution of MIC values for the pathogens of interest. We do NOT use MIC90's, as this is but a single point on the distribution. We use the protein-binding-corrected exposure from the Monte Carlo simulation to evaluate how well the dose simulated will work. For the case of fluoroquinolones, we take the AUC values and divide by the lowest MIC seen in the distribution. We then check how frequently the resulting AUC/MIC ratio attains the target (AUC/MIC = 30, 90, 110, 125, whatever) if the organism has that specific MIC. We then repeat this process for the next highest MIC (0.25 mg/L, 0.5 mg/L, 1.0 mg/L, etc) and calculate target attainment rates for each. This gives a curve of target attainment by MIC (for an example, see Drusano et al in AAC for the drug evernimicin). This allows the rational decision for a breakpoint value (95% target attainment - good for dear old Mom; 87% target attainment - close enough for government work for your mother-in-law).
Further, since we know the full distribution of pathogen MIC's, we can take a weighted average (an expectation in my jargon) over the distribution. That is, take the target attainment rate at a specific MIC, then multiply it times the fraction of the entire population of organisms at that MIC. Do this for all relevant MIC's, add them up and this will give the target attainment rate FOR THE POPULATION for that specific dose (e.g. at an MIC = 0.5 mg/L, the target attainment rate is 0.95 and 24% of the organisms have an MIC for the drug being studied of 0.5 mg/L. Multiply 0.95 x 0.24, giving 0.228. This is repeated for all other MIC's, their organism distribution values and target attainment rates and summed up.)
Now, we have the performance for a specific goal of therapy for a specific dose. Want to try another dose or goal of therapy? Go ahead. The technique does NOT make a judgement. It just helps to put the judgemnet on a rational basis.
I hope this helps.
All the best,
George Drusano
PharmPK Discussion List Archive Index page
Copyright 1995-2010 David W. A. Bourne (david@boomer.org)