56 / 320
620
with efficacy is the ƒAUC/MIC. In a murine
pneumonia
model, ƒAUC/MIC ratios of 2.17 and 8.78 were required
to produce 1 and 2 log kill, respectively, against
Acinetobacter spp
(29). Using the data from the Phase
3 clinical trial in treatment of hospital acquired
pneumonia
, a ƒAUC/MIC
≥
0.9 was associated with an 8
fold higher probability of clinical success (30). After a
loading dose of 100mg followed by 50mg every 12 hours,
the steady state tigecycline AUC0-24 is
~
4.7mg*h/L.
Considering tigecycline protein binding is 80%, the
fAUC0-24 would be
~
0.94mg*h/L, which is similar to the
median fAUC0-24 observed during the hospital acquired
pneumonia
study, 1.08mg*h/L (range: 0.35-4.02). As
a result, standard doses of tigecycline achieve optimal
exposure using the clinical pharmacodynamic threshold
when the MIC is
~
1mg/L, or
~
0.5mg/L if the 1-log
CFU reduction target is applied. The FDA susceptibility
breakpoint is
≤
2mg/L, whereas the EUCAST breakpoint is
≤
1mg/L. Unfortunately, limited clinical data are available
to validate these observations, and variable outcomes
with standard dosing tigecycline have been reported.
A recent clinical trial of 55 patients with extensively
drug-resistant
A. baumannii
bacteremia compared
14 day mortality between a colistin/carbapenem and
colistin/tigecycline combination (31). Patients received
a standard tigecycline dosage. The colistin/tigecycline
combination was independently associated with excess
14 day mortality, but only in the subgroup of patients
with a tigecycline MICs greater than 2mg/L. Because of
poor clinical outcomes during the
pneumonia
registration
studies, doubling the dose of tigecycline to a 200mg
loading dose followed by 100mg every 12 hours has
become clinically fashionable to treat MDR gram-
negative bacteria. This aggressive dose improved clinical
cure (57.5% vs 30.4%, p=0.05) but not ICU mortality
(48.4% vs 66.6%, p=0.14) in critically ill patients with
CRAB and CRE infections (32). The majority of patients,
however, still received tigecycline in combination with a
second antibiotic such as colistin.
POLYMYXINS
Polymyxin B and colistin (polymyxin E) have re-emerged
into clinical practice because of their gram-negative activity
against MDR organisms. Both antibiotics were developed
during a time when pharmacodynamic studies were not
required nor widely understood for new compounds;
therefore, until a short time ago, package insert dosing
recommendations were largely incorrect. Contemporary
dosing regimens based on pharmacodynamic concepts have
only recently begun to be understood, and the majority of
available data has been contributed with colistin. Colistin
displays concentration-dependent killing, and most
studies suggest that the ƒAUC/MIC is best associated with
bactericidal activity (33). Using the murine, thigh infection
model, a ƒAUC/MIC of 12 was required to achieve a 2 log
reduction against
P. aeruginosa
and A.
baumannii
strains.
However, in the murine lung infection model, this ƒAUC/MIC
exposure increased to 48 for a 1 log reduction; furthermore,
2 of 3 A.
baumannii
strains tested never achieved this level
of killing with any exposure tested. Considering colistin
protein binding is approximately 50% in humans and
estimating exposure over 24 hours, average steady state
concentrations of 1 and 4mg/L correspond with ƒAUC/
MIC ratios of 12 and 48, respectively, when the colistin
MIC is 1mg/L. Notably, colistin induced nephrotoxicity is
concentration dependent and disproportionally increases
with concentrations greater than 2.5mg/L. It should
therefore become quickly apparent to the reader that the
exposures required for efficacy significantly overlap with
those that produce toxicity. Moreover, these required
exposures are at an MIC of only 1mg/L; greater exposures
are proportionally required for higher MICs. At the time of
writing, the Clinical Laboratory Standards Institute (CLSI)
and European Committee on Antimicrobial Susceptibility
Testing (EUCAST) were in discussions to harmonize
colistin breakpoints. EUCAST defines susceptibility against
P. aeruginosa
at
≤
4mg/L, and against
A. baumannii
and
Enterobacteriaceae
at
≤
2mg/L. CLSI uses
≤
2mg/L for the
non-fermenting gram-negatives, but has no breakpoint
defined for enterobacteriaceae. Based on contemporary
pharmacokinetic data from Garonzik and colleagues (34),
the European Medicines Agency (EMA) approved updated
dosing suggestions for patients with varying degrees of
renal function. This was followed by recommendations from
the US Food and Drug Administration (FDA). A summary of
these new dosing recommendations is provided in Table 2.
An ensuing simulation study compared the EMA and FDA
dosing recommendations with standard physician dosing
(35). Both EMA and FDA doses resulted in greater average
steady-state concentrations compared with physician
selected doses, and EMA dosing provided the highest
average concentrations across the creatinine
clearance
(CrCL) ranges. However, recommended dosing regimens
from both agencies were able to provide a high probability
of steady-state concentrations above 2mg/L when CrCL was
≥
80 ml/min. Therefore, caution is advised in using colistin
as monotherapy when patients have good kidney function,
MICs above 1mg/L, or both.
Although studies are still pending, polymyxin B is assumed
to have a similar pharmacodynamic profile to colistin in that
a ƒAUC/MIC of
~
12 is required for 2 log CFU reductions (33).
However, unlike colistin, polymyxin B is not a prodrug, thus
[REV. MED. CLIN. CONDES - 2016; 27(5) 615-624]