Importance of the loading dose: CMS vs Polymyxin B
Lena E. Friberg
Pharmacometrics research groupDepartment of Pharmaceutical BiosciencesUppsala UniversitySweden Not
to be
copie
d and
distr
ibuted
to ot
hers
withou
t the p
ermiss
ion of
the a
uthor
How should CMS (colistin) and Polymyxin B be administered?
Dose
Bacterialkill
Toxicity
Maintenance dosing?Need for loading dose?
Conc
PK
PD
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
How should CMS (colistin) be administered?
Dose(CMS)
Bacterialkill
Toxicity
Maintenance dosing?Need for loading dose?CMS
Colistin
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
• Low degradation of CMS to colistin during work-up (< 1%)• Important for characterization of initial concentrations,
i.e. when CMS is high and colistin is low• Measure colistin A and colistin B (only)• ”CMS” = Colistin(A+B) after hydrolysis – Colistin(A+B) before hydrolysis
Analysis of CMS and colistin in biological samples
Colistin
CMS
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
1st Dose 4th - 8th Dose
0.05
0.10
0.20
0.50
1.00
2.00
5.00
10.00
20.00
0 2 4 6 8 0 2 4 6 8Time after dose (h)
Con
cent
ratio
n (µ
M)
CompoundCMSColistin
Pharmacokinetics of CMS and colistin 3MU q8h
Colistin
CMSp
”CMS” plasmaCMS DoseF =1 CMSc
All modeling performedusing molar units
t1/2 of 14 h
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
A loading dose of CMS suggested
0
1
2
3
0 12 24 36 48 60 72 84 96
Col
istin
Con
cent
ratio
n (m
g/L)
Time after first dose (hours)
3 MU q8h9 MU + 4.5 MU q12h12 MU + 4.5 MU q12h9 MU (2h infusion) + 4.5 MU q12h12 MU (2h infusion) + 4.5 MU q12h3 MU q3h + 3 MU q8h
0
1
2
3
0 12 24 36 48 60 72 84 96
Col
istin
Con
cent
ratio
n (m
g/L)
Time after first dose (hours)
3 MU q8h9 MU + 4.5 MU q12h12 MU + 4.5 MU q12h9 MU (2h infusion) + 4.5 MU q12h12 MU (2h infusion) + 4.5 MU q12h3 MU q3h + 3 MU q8h
0
10
20
30
40
0 6 12 18 24
CM
S C
once
ntra
tion
(mg/
L)
Time after first dose (hours)
3 MU q8h9 MU + 4.5 MU q12h12 MU + 4.5 MU q12h9 MU (2h infusion) + 4.5 MU q12h12 MU (2h infusion) + 4.5 MU q12h3 MU q3h + 3 MU q8h
0
10
20
30
40
0 6 12 18 24
CM
S C
once
ntra
tion
(mg/
L)
Time after first dose (hours)
3 MU q8h9 MU + 4.5 MU q12h12 MU + 4.5 MU q12h9 MU (2h infusion) + 4.5 MU q12h12 MU (2h infusion) + 4.5 MU q12h3 MU q3h + 3 MU q8h
Plachouras et al., AAC, 2009
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
1st Dose 4th - 8th Dose
0.05
0.10
0.20
0.50
1.00
2.00
5.00
10.00
20.00
0 2 4 6 8 0 2 4 6 8Time after dose (h)
Con
cent
ratio
n (µ
M)
CompoundCMSColistin Colistin
CMSp
”CMS” plasmaCMS DoseF =1
CMSc
A loading dose of 6MU (+ 3MU q8h)
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
AROFF ARON
Bgrow Bresting
+
+
+
-
Is application of a loading dose likely toaffect bacterial killing?
Mohamed et al., AAC 2012, WCoP 2012
B = BacteriaAR = Adaptive resistance
Time after first dose (h)
PKPD-model based on in vitro experiments
(P. Aeroguinosa)CMS Dose
PK-model based on studies in critically ill
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
When to start maintenance dosing? 3MU q8h or 4.5MU q12h?
Mohamed et al., AAC 2012
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
A loading dose of 9MU (+4.5MU q12h)
Nephrotoxicity – P-12 (Recommend to commence maintenance dosing at 24h after 9MU load)
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
New data:PK of CMS and Colistin after 9MU
Karaiskos et al., unpublished data
3MUq8h0.25h inf
6MU Load, 3MUq8h0.25h inf
9MU Load, 4.5MUq12h0.5h inf
9MU Load, 4.5MUq12h1h inf
0.050.100.200.501.002.005.00
10.0020.00
0.050.100.200.501.002.005.00
10.0020.00
1st Dose
4th - 8th Dose
0 2 4 6 8 10 12 0 2 4 6 8 10 12 0 2 4 6 8 10 12 0 2 4 6 8 10 12Time after dose (h)
Con
cent
ratio
n (µ
M)
CompoundCMSColistin
Mohamed et al., 2012
Plachouras et al., 2009
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
Model fit without re-estimatingparameters
Red lines are medians/trends of observationsPink fields are 95%CI of median from datasets simulated from the model
Colistin concunderpredicted
CMS concunderpredicted initially
Previous published model
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
Derivativesc
Colistin
CMSp
Derivativesp
CMS DoseF (Study1&2) = 78%
”CMS” plasma
CMSc
Current best model
”Derivatives” DoseF (All studies) = 22%
All 3 studies:• Same CMS brand (Norma, Greece)• Similar patient population• Same assay• Calculations made in the same way
Analyze Colistin A & B (only)• Proportions of A & B to total CMS differ?
Differences in CMS brands:P-6 and R. Nation’s talk
Other studies of PK of CMS & colistin in critically ill:O-5 & P-9
F (Study3) = 146%
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
Current best model explainsobserved data much better
Red lines are medians/trends of observationsPink fields are 95%CI of median from datasets simulated from the model
Current best modelEarlier published model
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
CMS: V1 – dependent on Body wtCL – dependent on CrCL
Individualization of CMS dosing by covariates
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
0
5
10
15
20
0 4 8 12 16 20 24Time after dose (h)
CM
S C
once
ntra
tion
(mg/
L)
Body Weight507090
How does a WT-V1 relationship for CMS affect concentrations and bacterial killing?
Mohamed et al., WCoP 2012
AROFF ARON
Bgrow Bresting
+
++
-
B = BacteriaAR = Adaptive resistance
Loading dose proportional to weightV1 for CMS proportional to weight
but V1 has limited influence on conc-time profiles
Weight-independent maintenance dose
Weight-basedloading dose
0.5
1.0
1.5
2.0
2.5
3.0
0 4 8 12 16 20 24Time after dose (h)
Col
istin
con
cent
ratio
n (m
g/L)
Body Weight507090
10 1̂
10 2̂
10 3̂
10 4̂
10 5̂
10 6̂
0 4 8 12 16 20 24Time after dose (h)
Bact
eria
l cou
nt (C
FU/m
l)
Body Weight507090
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
How should Polymyxin B be administered?
Dose
Bacterialkill
Toxicity
Maintenance dosing?Need for a loading dose?
Polymyxin Bconc
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
Pharmacokinetics of Polymyxin B
PolymyxinBpPolymyxinBc
Polymyxin B: CL & V terms – dependent on Body wtNo dependence on CrCL (Supported by Zavascki et al., CID, 2008, Abdelraouf, AAC, 2012)
Studied at steady-state(doses and infusion duration varied)
t1/2β – 12h
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
Simulations of dosing regimens
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
How should Polymyxin B be administered?
Dose
Bacterialkill
Toxicity
PK and Effects after a loading dose?Polymyxin B
conc
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
CMS / Colistin• Slow formation + Long half-life→ A clear need for a loadingdose
• Flat fixed: 9MU CMS (a weight-based loading dose may result in underexposure in low weight patients)
• Some PKPD-information available
• Variability in composition
Polymyxin B• Long half-life
→ A loading dose is recommended
• Weight-based: 2-2.5 mg/kg
• In need of more PKPD information
• Less variability in composition(?)
ConclusionsLoading dose in Critically ill patients
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
• Otto Cars• Matti Karvanen• Ami Fazlin Mohamed• Salim Bouchene• David Khan• Elisabet Nielsen• Britt Jansson
AIDA project (Health-F3-2011-278348)Swedish Foundation of StrategicResearch
• Diamantis Plachouras• Ilias Karaiskos• Helen Giamarellou• K Pontikis• E Papadomichelakis • A Antoniadou • A Armaganidis • E Apadomichelakis• I Tsangaris • G Poulakou• F Kontopidou
Acknowlegdements
Not to
be co
pied a
nd di
stribu
ted to
othe
rs
withou
t the p
ermiss
ion of
the a
uthor
Top Related