Substantial impact of altered pharmacokinetics in critically ill patients on the antibacterial effects of

32 Pages·2017·2.37 MB·English

AAC Accepted Manuscript Posted Online 6 March 2017
Antimicrob. Agents Chemother. doi:10.1128/AAC.02642-16
Copyright © 2017 American Society for Microbiology. All Rights Reserved.
1 Substantial impact of altered pharmacokinetics in critically ill patients on the
2 antibacterial effects of meropenem evaluated via the dynamic hollow-fiber infection
3 model
4 Phillip J. Bergen1, Jürgen B. Bulitta2, Carl M. J. Kirkpatrick1, Kate E. Rogers1,3, Megan J.
5 McGregor1, Steven C. Wallis4, David L. Paterson5, Roger L. Nation3, Jeffrey Lipman4,6,
D
6 Jason A. Roberts4,6,7, Cornelia B. Landersdorfer1,3,8,# o
w
n
7 1Centre for Medicine Use and Safety, Faculty of Pharmacy and Pharmaceutical Sciences, loa
d
8 Monash University, Melbourne, Victoria, Australia; 2Center for Pharmacometrics and ed

f
9 Systems Pharmacology, College of Pharmacy, University of Florida, Florida, USA; 3Drug ro
m

10 Delivery, Disposition and Dynamics, Monash Institute of Pharmaceutical Sciences, Monash h
t
t
p
11 University, Melbourne, Victoria, Australia; 4Burns, Trauma and Critical Care Research :/
/
a
12 Centre, The University of Queensland, Brisbane, Queensland, Australia; 5The University of a
c
.
a
13 Queensland Center for Clinical Research, Royal Brisbane and Women’s Hospital, Brisbane, s
m
14 Queensland, Australia; 6Royal Brisbane and Women’s Hospital, Brisbane, Queensland, .o
r
g
15 Australia; 7Centre for Translational Anti-infective Pharmacodynamics, The University of o/
n
16 Queensland, Brisbane, Queensland, Australia; 8Department of Pharmaceutical Sciences, J
a
n
17 School of Pharmacy and Pharmaceutical Sciences, State University of New York at Buffalo, u
a
r
y
18 Buffalo, New York, USA
1
0
,

19 Running title: Effect of meropenem clearance on antibacterial effects 2
0
1
20 Key words: augmented renal clearance, pharmacodynamic modeling, critically ill, 9
21 pharmacokinetics b
22 y

g
u
23 Parts of this study were presented at the American Society for Microbiology (ASM) Microbe
e
24 and Interscience Conference on Antimicrobial Agents and Chemotherapy (ICAAC) Congress, s
t
25 Boston, MA, 16 to 20 June 2016, and Australasian Society of Clinical and Experimental
26 Pharmacologists and Toxicologists (ASCEPT) and the Molecular Pharmacology of GPCRs
27 (MPGPCR) meeting, Melbourne, Australia, 27 to 30 November 2016.
28
29 #Corresponding Author. Cornelia B. Landersdorfer, PhD. 1Centre for Medicine Use and
30 Safety, Faculty of Pharmacy and Pharmaceutical Sciences, Monash University, Melbourne,
31 Australia. Phone: +61 -3-9903 9011; Fax: +61 -3-9903 9629; Email:
32 [email protected]
1
33 Abstract
34 Critically ill patients frequently have substantially altered pharmacokinetics compared to non-
35 critically ill patients. We investigated the impact of pharmacokinetic alterations on bacterial
36 killing and resistance for commonly used meropenem dosing regimens. A Pseudomonas
37 aeruginosa isolate (MIC 0.25 mg/L) was studied in the hollow-fiber infection model
meropenem
38 (inoculum ~107.5 CFU/mL; 10 days). Pharmacokinetic profiles representing critically ill D
o
w
39 patients with augmented renal clearance (ARC), normal, or impaired renal function n
lo
40 (creatinine clearances of 285, 120 or ~10 mL/min, respectively) were generated for three a
d
e
41 meropenem regimens (2g, 1g and 0.5g 8-hourly, 30 min infusion), plus 1g 12-hourly with d

f
r
o
42 impaired renal function. The time-course of total and less-susceptible populations and MICs m

h
43 were determined. Mechanism-based modeling (MBM) was performed using S-ADAPT. All t
t
p
:
44 dosing regimens across all renal functions produced similar initial bacterial killing (≤~2.5 //
a
a
45 log ). For all regimens subjected to ARC regrowth occurred after 7h. For normal and c
10 .
a
s
46 impaired renal function bacterial killing continued until 23 to 47h whereafter regrowth m
.
o
47 occurred with 0.5g and 1g regimens with normal renal function (fT>5×MIC 56% and 69%, rg
/

48 fC /MIC <2); the emergence of less-susceptible populations (≥32-fold increases in MIC) o
min n

J
49 accompanied all regrowth. Bacterial counts remained suppressed across 10 days with a
n
u
50 normal (2g 8-hourly regimen) and impaired (all regimens) renal function (fT>5×MIC ≥82%, a
r
y
51 fCmin/MIC ≥2). The MBM successfully described bacterial killing and regrowth for all renal 1
0
,
52 functions and regimens simultaneously. Optimized dosing regimens including extended 2
0
1
53 infusions and/or combinations, supported by MBM and Monte Carlo simulations, should be 9

b
54 evaluated in the context of ARC to maximize bacterial killing and suppress resistance y

g
u
55 emergence. e
s
t
2
56 INTRODUCTION
57 Effective antibiotic therapy is a crucial determinant of survival for patients with
58 serious infections in an intensive care unit (ICU) (1, 2). Serious infections caused by
59 Pseudomonas aeruginosa are increasing in prevalence in ICUs and present a substantial
60 problem (3-5). Mortality rates exceed 50% for specific patient groups such as those with
61 septic shock (6), with early administration of effective antibiotic therapy substantially D
o
w
62 improving survival (1, 7). Meropenem is a common treatment for P. aeruginosa infections (5, n
lo
63 8). It is well-recognized that for β-lactam antibiotics such as meropenem the fraction of a a
d
e
64 dosing interval where unbound concentrations exceed 1× MIC of the infecting pathogen d

f
r
o
65 (fT>MIC) should be optimized to achieve near-maximal bacterial killing (9, 10). Furthermore, it m

h
66 has become evident that higher exposures, e.g. unbound concentrations remaining above 4 t
t
p
:
67 or 5× MIC throughout the dosing interval, or a minimum unbound concentration (fCmin) of ≥4× //a
a
68 MIC are required for suppression of emergence of resistance (11-14). The emergence of c
.
a
s
69 antibiotic resistance in patients occurs in ~10% of antibiotic courses (15), with sub-optimal m
.
o
70 dosing an important contributor (16). r
g
/

o
71 Attainment of PK/PD targets is challenging in critically ill patients who can have n

J
a
72 significantly altered antibiotic pharmacokinetics compared to non-critically ill patients (17, 18). n
u
a
73 In particular, patients with augmented renal clearance (ARC) have been demonstrated to r
y

1
74 exhibit substantially increased meropenem clearances and thereby decreased meropenem 0
,

2
75 exposures (17, 19-21). This results in a significant risk of sub-therapeutic antibiotic 0
1
9
76 concentrations. However, the impact of a wide range of renal functions including ARC on
b
y
77 bacterial killing and resistance emergence has not been quantified for meropenem. At g
u
e
78 present, clinicians still rely predominantly on the product information when selecting dosing s
t
79 regimens instead of adjusting the doses for patients with ARC (22, 23).
80 The hollow-fiber infection model (HFIM) can simulate the time-course of antibiotic
81 concentrations with a specific elimination half-life at high inocula over long durations, both of
82 which are typically not possible with in vivo models. Results from the HFIM are also well
3
83 correlated with clinical endpoints for bacterial killing and time course of emergence of
84 resistance (24). This study used the HFIM to characterize the effect of different meropenem
85 exposures, administered as standard short-term infusions, likely to occur in critically ill
86 patients on the bacterial killing and emergence of resistance for P. aeruginosa. Specifically,
87 we generated unbound concentration-time profiles consistent with those observed in patients
D
88 with augmented renal clearance (ARC), and those with normal and impaired renal function
o
w
89 following intravenous (i.v.) administration of four commonly used dosing regimens. We n
lo
a
90 additionally developed a mechanism-based mathematical model (MBM) to quantitatively d
e
d
91 characterize the relationships between meropenem concentrations, bacterial killing and
f
r
o
92 regrowth over time for a wide range of studied renal functions and dosing regimens. m

h
t
t
93 p
:
/
/
a
a
94 MATERIALS AND METHODS
c
.
a
s
95 Antibiotic, media, bacterial isolate and susceptibility testing m
.
o
r
g
96 Meropenem (lot number MAUS1025; Fresenius Kabi, NSW, Australia) stock
/

o
97 solutions were prepared as described previously (25). Viable counting was performed on n

J
a
98 cation-adjusted Mueller-Hinton agar (CAMHA; 25 mg/L Ca2+ and 12.5 mg/L Mg2+; University n
u
a
99 of Melbourne, Australia). Drug-containing agar plates were prepared by adding meropenem r
y

1
100 stock solution to CAMHA (lot number 5030699; BD, Sparks, MD, USA). HFIM studies were 0
,

2
101 performed using cation-adjusted Mueller-Hinton broth (CAMHB; lot number 3322206; BD, 0
1
9
102 Sparks, MD, USA) containing 20 - 25 mg/L Ca2+ and 10 - 12.5 mg/L Mg2+. Clinical isolate
b
y

103 P. aeruginosa 1280 (meropenem MIC 0.25 mg/L, Royal Brisbane and Women’s Hospital, g
u
e
104 Australia) was obtained from a critically ill patient with a soft tissue infection. MICs prior to s
t
105 and following drug exposure were determined using the CLSI agar dilution method (26).
106 Susceptibility and resistance was defined as an MIC ≤2 mg/L and >8 mg/L, respectively (27).
107
108
4
109 Hollow-fiber in vitro infection model (HFIM)
110 A HFIM (36°C) was used to evaluate changes in bacterial burden and suppression of
111 resistance with four commonly used meropenem dosing regimens. Cellulosic cartridges
112 (Batch No. B720170715b; FiberCell Systems Inc., Frederick, MD, USA) were used. An initial
113 inoculum of ~107.5 CFU/mL was prepared as described previously (25). Experiments were
D
114 conducted over 10 days with periodic sampling (1.0 to 1.5 mL) for viable counting (Table 1). o
w
n
115 Samples were twice centrifuged (4,000 g, 5 min) and re-suspended in saline to reduce lo
a
d
116 antibiotic carryover, and manually plated on CAMHA containing no drug (100 µL plated; limit e
d

117 of counting, 1.0 log10 CFU/mL) and meropenem at 5× and 10× MIC (see emergence of fro
m
118 resistance below).
h
t
t
p
119 :/
/
a
a
120 Experimental design and simulated meropenem regimens c.
a
s
m
121 The free (non-protein-bound) plasma concentration-time profiles based on a .
o
r
g
122 population pharmacokinetic model for meropenem in critically ill patients (protein binding 2%)
/

o
123 (20) and a fraction excreted unchanged in urine of 79% of a meropenem dose in subjects n

J
a
124 with normal renal function (28) were first simulated using Berkeley Madonna (version 8.3.18). n
u
a
125 Profiles were simulated for patients with different drug clearances due to altered renal r
y

1
126 function, namely patients with ARC (creatinine clearance [CL ] ~285 mL/min, meropenem 0
CR
,

2
127 clearance 34.0 L/h, half-life (t ) 0.6h), normal renal clearance (CL 120 mL/min, 0
1/2 CR
1
9
128 meropenem clearance 16.3 L/h, t 1.1h) and renal impairment (CL ~10 mL/min,
1/2 CR b
y

129 meropenem clearance 4.1 L/h, t 4h) (Table 1). For each level of renal function the in-silico g
1/2 u
e
130 simulations evaluated dosing regimens of 2g, 1g and 0.5g meropenem administered via a 30 s
t
131 min i.v. infusion 8-hourly (Table 1) (29). A further 1g regimen administered via a 30 min i.v.
132 infusion 12-hourly, as used in patients with renal impairment, was also evaluated via
133 simulation. The final simulated meropenem concentration-time profiles were then generated
134 in the HFIM. No loading dose was administered. Untreated growth controls were included.
5
135
136
137 Mutation frequency and emergence of resistance studies
138 Mutation frequencies were determined at baseline (0h) and 23, 47, 71, 95, 119, 143,
D
139 167, 191, 215 and 239h to quantify less-susceptible subpopulations. A 200 µL sample of
o
w
140 appropriately diluted log-phase growth suspension was manually plated onto CAMHA n
lo
a
141 containing no antibiotic and meropenem at 5× and 10×MIC (i.e. 1.25 and 2.5 mg/L d
e
d
142 meropenem). Given the potential for a low number of less-susceptible bacteria in samples
f
r
o
143 from untreated controls and some treatment regimens, 200 µL was used to increase m

h
144 sensitivity (limit of counting, 0.7 log CFU/mL). For determination of total and less- t
10 t
p
:
/
145 susceptible counts at baseline, log-phase growth suspension was used and enumeration /
a
a
146 performed manually after 24h incubation at 36°C. The log10 (mutation frequency) (log10MF) c.a
s
147 was determined as: m
.
o
r
g
148 log MF = log (CFU/mL on antibiotic-containing agar) - log (CFU/mL on antibiotic-free agar)
10 10 10 /

o
n
149 Although some viable counts were too low to quantify colonies on antibiotic-containing agar, J
a
n
150 these samples provided information on the upper limit of the log MF. To confirm changes in u
10
a
r
151 MIC from baseline, MICs were determined at 47 and 191h using a subset of at least three y

1
0
152 colonies from antibiotic-free and antibiotic-containing plates. ,

2
0
1
153 9

b
y
154 Measurement of meropenem concentrations g
u
e
s
155 Duplicate (1 mL) samples were collected periodically (Table 1) in cryovials and t
156 stored at -80°C. Concentrations were measured at ambient temperature using a validated
157 high-performance liquid chromatography-tandem mass spectrometry (HPLC-MS/MS)
158 method. The HPLC-MS/MS system comprised two Shimadzu Nexera2 LC-30AD liquid
159 chromatographs, a SIL-30ACMP autosampler and an 8030+ triple quadrupole detector
6
160 (Shimadzu, Tokyo, Japan). To a 2.5 µL sample, 10 µL of 0.5 μg/mL [d6]-meropenem
161 (internal standard) and 30 µL of acetonitrile were added before centrifuging (5 min, 13,000 g).
162 A 30 µL volume of supernatant was removed to an autosampler vial and 0.5 μL injected onto
163 a SeQuant ZIC-HILIC 5.0 μm column (2.1 x 50 mm) (Merck, Darmstadt, Germany). The
164 mobile phase consisted of 80% acetonitrile, 20% water and 0.1% formic acid delivered
D
165 isocratically at 0.3 mL/min. Meropenem was monitored by positive mode electrospray at o
w
166 MRMs of 383.50→68.15; 383.50→114.15 and 383.50→141.10. [d6]-meropenem was n
lo
a
167 monitored in positive mode at 390.10→147.10 and 390.10→73.90. The run time was 3.2 d
e
d
168 minutes and assay range 0.2 to 100 mg/L; samples were diluted when expected
f
r
o
169 concentrations exceeded the upper limit of quantification. Unknown samples were assayed m

h
170 in batches alongside calibration and quality control samples and results subjected to batch tt
p
:
/
171 acceptance criteria. The assay method was validated for linearity, lower limit of quantification /a
a
c
172 (LLOQ), precision and accuracy using both the US Food and Drug Administration (FDA) and .
a
s
m
173 European Medicines Evaluation Agency (EMEA) criteria (30, 31). The precision was within
.
o
174 4.9% and accuracy 8.1% at the concentrations tested (0.60, 4.0 and 80 mg/L). rg
/

o
n
175
J
a
176 Pharmacodynamic modeling n
u
a
r
177 Modeling was conducted using importance sampling (pmethod=4) in parallelized S- y

1
0
178 ADAPT with SADAPT-TRAN (32-34). All pharmacokinetic and pharmacodynamic ,

2
0
179 observations were fitted simultaneously. Models were evaluated based on the S-ADAPT 1
9

180 objective function (-1×log-likelihood), standard diagnostic plots, visual predictive checks and b
y

g
181 plausibility of the parameter estimates (14, 35, 36). Models with one, two and three bacterial u
e
s
182 subpopulations, and which described the effect of meropenem via direct bacterial killing, t
183 inhibition of successful bacterial replication, or both, were evaluated (36-38).
184 The final model incorporated three subpopulations that were susceptible (CFU ),
S
185 intermediate (CFU) or resistant (CFU ) to meropenem. The size of the total inoculum
I R
186 (log ) and the log mutation frequencies for the intermediate (logMF) and resistant
CFU0 10 I
7
187 (logMF ) subpopulations were estimated parameters. Thus the estimated subpopulations did
R
188 not directly reflect bacterial counts on meropenem-containing agar plates at 5× and 10×MIC.
189 Each subpopulation contained bacteria in two states: those preparing for replication (CFU )
1
190 and those immediately before replication (CFU ) (38, 39). The total concentration of viable
2
191 bacteria was CFU :
ALL
D
o
(cid:1829)(cid:1832)(cid:1847) =(cid:1829)(cid:1832)(cid:1847) +(cid:1829)(cid:1832)(cid:1847) +(cid:1829)(cid:1832)(cid:1847) +(cid:1829)(cid:1832)(cid:1847) +(cid:1829)(cid:1832)(cid:1847) +(cid:1829)(cid:1832)(cid:1847)
(cid:3002)(cid:3013)(cid:3013) (cid:3020)(cid:2869) (cid:3020)(cid:2870) (cid:3010)(cid:2869) (cid:3010)(cid:2870) (cid:3019)(cid:2869) (cid:3019)(cid:2870) w
n
lo
192 The initial condition of the CFUS population was calculated as the difference between a
d
e
193 the total inoculum (CFU0, i.e. the value of CFUALL at time = 0h) and the initial conditions of d

f
r
194 the CFU and CFU populations. All bacteria were initialized in state 1, and the initial o
I R m

195 conditions for CFU , CFU and CFU were 0. h
S2 I2 R2 t
t
p
:
/
196 The concentration of meropenem-susceptible bacteria in state 1 (CFUS1) was: /a
a
c
(cid:1856)(cid:1829)(cid:1832)(cid:1847) .a
(cid:3020)(cid:2869)=(cid:1844)(cid:1831)(cid:1842) ∙ (cid:3435)1−(cid:1835)(cid:1866)ℎ (cid:3439) ∙ (cid:1863) ∙ (cid:1829)(cid:1832)(cid:1847) − (cid:1863) ∙ (cid:1829)(cid:1832)(cid:1847) s
(cid:1856)(cid:1872) (cid:3019)(cid:3032)(cid:3043),(cid:3020) (cid:2870)(cid:2869) (cid:3020)(cid:2870) (cid:2869)(cid:2870),(cid:3020) (cid:3020)(cid:2869) m
.
o
r
197 where k was the replication rate constant, k the growth rate constant and REP the g
21 12,S
/

o
198 replication factor: n

J
a
(cid:1829)(cid:1832)(cid:1847) n
(cid:1844)(cid:1831)(cid:1842)=2 ∙ (cid:3436)1− (cid:3002)(cid:3013)(cid:3013) (cid:3440) u
(cid:1829)(cid:1832)(cid:1847) + (cid:1829)(cid:1832)(cid:1847) a
(cid:3040)(cid:3028)(cid:3051) (cid:3002)(cid:3013)(cid:3013) r
y

1
199 At low CFU , REP approached 2, representing a 100% probability of successful 0
ALL ,

2
200 replication. As CFU approached the maximum population size (CFU ), REP approached 0
ALL max 1
9
201 1, i.e. a 50% probability of successful replication with the total viable count remaining b
y

202 constant. g
u
e
s
203 The effect of meropenem was best described as inhibition of successful replication t
204 (Inh ):
Rep,S
(cid:1835)(cid:1865)(cid:1853)(cid:1876) ∙ (cid:1829) (cid:3035)(cid:3036)(cid:3039)(cid:3039),(cid:3020)
(cid:3019)(cid:3032)(cid:3043),(cid:3020)
(cid:1835)(cid:1866)ℎ =
(cid:3019)(cid:3032)(cid:3043),(cid:3020) (cid:1835)(cid:1829)50 (cid:3035)(cid:3036)(cid:3039)(cid:3039),(cid:3020) + (cid:1829) (cid:3035)(cid:3036)(cid:3039)(cid:3039),(cid:3020)
(cid:3019)(cid:3032)(cid:3043),(cid:3020)
8
205 where C was the meropenem concentration in the growth medium, Imax the maximum
Rep,S
206 inhibition of successful replication, IC50 the concentration causing 50% of the maximum
Rep,s
207 inhibition effect, and hill the hill coefficient. An Inh of 0.50 resulted in net stasis whereas
,s Rep,S
208 an Inh of >0.50 led to bacterial killing due to the elimination of bacteria that replicate
Rep,S
209 unsuccessfully.
D
210 The concentration of meropenem-susceptible bacteria in state 2 (CFU ) was: o
S2 w
n
lo
211 (cid:3031)(cid:3004)(cid:3007)(cid:3022)(cid:3268)(cid:3118) =− (cid:1863) ∙ (cid:1829)(cid:1832)(cid:1847) + (cid:1863) ∙ (cid:1829)(cid:1832)(cid:1847) a
(cid:3031)(cid:3047) (cid:2870)(cid:2869) (cid:3020)(cid:2870) (cid:2869)(cid:2870),(cid:3020) (cid:3020)(cid:2869) d
e
d
212 The bacterial concentrations of the intermediate and the resistant subpopulations f
r
o
m
213 were described in a similar manner. The biological variability between the viable count

h
t
214 curves was assumed to be log-normally distributed, except for parameters constrained tp
:
/
/
215 between 0 and 1 that were logistically transformed. Between-curve variability was set to 15% a
a
c
216 coefficient of variation during the end of the estimation (35). An additive residual error model .a
s
m
217 was used to fit the viable counts. For observations <100 CFU/mL the number of
.
o
r
218 colonies/plate was directly fitted via a previously developed residual error model (39). g
/

o
n
219 J
a
n
u
220 RESULTS a
r
y

1
221 Typical examples of observed, targeted and fitted concentration-time profiles are 0
,

2
222 shown in the supplementary material (Figure S1). Measured meropenem concentrations 0
1
9
223 adequately matched the target profiles produced in the in-silico design simulations (Figures
b
y
224 S1 and 1). The fitted meropenem concentration-time profiles were used in the MBM. g
u
e
225 Changes in total and less-susceptible bacterial populations are shown in Figures 2 and 3. s
t
226 Changes in mutation frequencies and MICs are presented in Tables 2 and 3. The log MF
10
227 before treatment (0h) was -5.56 to <-6.88 on agar containing 5× and 10× the meropenem
228 MIC (Table 2). Very few colonies grew on antibiotic-containing agar prior to treatment
229 (Figure 3).
9
230 At all levels of renal function (CL ) the three growth controls were virtually
CR
231 superimposable, with less-susceptible populations (i.e. those growing in the presence of
232 meropenem at 5× and 10× MIC) plateauing at ~3 - 4 log CFU/mL by ~48 - 72h (Figure 3).
10
233 Log MFs and MICs were also relatively stable, with MICs generally within one (occasionally
10
234 two) 2-fold dilution of baseline when colonies were detected. At each CL , bacterial killing
CR
D
235 over the first 7h was similar (~1.5 - 2.5 log killing; Figure 2) with all meropenem-containing
10 o
w
236 regimens. However, for all regimens subject to ARC little or no further bacterial killing n
lo
a
237 occurred with regrowth to within ~1.5-log10 of control values by 48h (0.5 and 1g regimens) d
e
d
238 and 96h (2g regimen). In each case less-susceptible populations emerged rapidly growing to
f
r
o
239 within <~2-log of the total population, with MICs at each concentration (5× and 10× MIC) m

h
240 increasing to 16 mg/L (0.5g regimen) and 32 mg/L (1 and 2g regimens) by 191h. For normal tt
p
:
/
241 and impaired renal function, all regimens demonstrated essentially identical bacterial killing /a
a
c
242 up to 23h (<1-log difference between all regimens). At normal renal function rapid regrowth .
a
s
243 with near-complete replacement of susceptible by less-susceptible bacteria (Figure 3) and a m
.
o
244 concomitant increase in mutation frequency and MIC (Tables 2 and 3) occurred with the 0.5g rg
/

o
245 regimen after 23h. At normal renal function similar regrowth and resistance emergence n

J
246 occurred with the 1g regimen, although regrowth was delayed until ~48h, whereas killing a
n
u
247 continued and/or regrowth remained suppressed across 10 days with the 2g regimen. With a
r
y

248 impaired renal function all regimens (including 1g 12-hourly) suppressed regrowth across 10 1
0
,

249 days. In each case where regrowth was suppressed no viable bacteria were detected on at 2
0
1
250 least one occasion from 143h onwards; viable bacteria (maximum of 1.54 log CFU/mL; no 9
10
b
y
251 less-susceptible bacteria) were detected in only three of the five regimens at 239h.
g
u
e
252 The developed MBM was able to successfully describe simultaneously the s
t
253 meropenem concentrations and viable bacterial counts for all renal functions and dosing
254 regimens that were evaluated in the HFIM (Figures S1, 1 and 4). Three subpopulations were
255 required to fit the observed bacterial counts, with the intermediate and resistant
256 subpopulations growing more slowly than the susceptible subpopulation (k -1, Table 4). A
12
10