PAGE 2010 amox 2010 amox... · 2018. 6. 7. ·...
1
HighDose Amoxicillin Pharmacokine5cs (PK) and Pharmacodynamics (PD) in Children Michael Neely 1 , Michael Reed 2 , Roger Jelliffe 1 1 University of Southern California Laboratory of Applied PharmacokineEcs, 134B Alcazar Ave, Los Angeles, CA 90033 USA 2 The Rebecca D. Considine Clinical Research InsEtute, Akron Children's Hospital, One Perkins Square, Akron, OH 44308 USA Objec5ves • We are unaware of any published PK data addressing oral amoxicillin (amox) doses >25 mg/kg in children. • Since amox absorption is saturable, 1,2 we wished to estimate the percentage of the dosing interval that the plasma amox concentration would exceed S. pneumoniae (pneumococcal) breakpoint minimum inhibitory concentrations (T>MIC), and the percent of children with the target T>MIC of at least half the dosing interval after various amox doses, up to the commonly used daily dose of 90 mg/kg. #1802 Conclusions • We extended a published adult model of saturable oral amox absorption to children. • Simulation from the model suggests that amox plasma exposure in children is dose proportional at least to a dose of 45 mg/kg. At a median dose of 864 mg (Kma), bioavailability is expected to be 50% of maximal. • A resistance breakpoint of ≥2 mg/L for oral amox is appropriate at approved doses (4045 mg/kg/day). • With a higher daily dose of 90 mg/kg, by percentage target attainment of a T>MIC for at least half the dosing interval, more resistant organisms could likely be treated, especially with dosing three times daily. • ConTirmation of this simulation study is desirable. www.lapk.org Methods • We adapted a published adult fourcompartment model (Figure 1) with saturable, delayed, timelimited amox absorption through an absorptive compartment into a central compartment with linear elimination and transfer to/from a peripheral compartment. 1,2 The original model was derived from 6 adult patients Tirst given a single dose of amox 500 mg IV, with intensive 10hour postdose PK sampling. A twocompartment model was Titted to the data. Each subject was then given single oral amox doses of 500 mg and 3000 mg with the same sampling after each dose as for the IV dose. The order of oral doses was random and separated by a week minimum. For the oral model, volume and intercompartmental rate transfer constants were Tixed for each participant according to their estimates from the IV model. • We logtransformed and allometrically scaled the reported model parameters as the initial ranges prior to Titting the data. No parameter in our model was Tixed, including volume. • Individual data after oral dosing from the original study were used to Tit the model using nonparametric methods implemented in MMUSCPACK (available from www.lapk.org ), with standard visual/numerical checks. • Upon comparison of modelsimulated proTiles with published Cmax and AUC ranges for doses of 15 and 25 mg/kg in children, 3 volume and elimination terms were multiplied by populationconstant age functions to accurately simulate the adult and pediatric data. • Finally, singledose T>MIC were simulated in 1000 representative 15 month, 12 kg children for varying amox doses and dose intervals. Results References 1. Piotrovskij VK, Paintaud G, Alván G, Trnovec T. Modeling of the saturable timeconstrained amoxicillin absorption in humans. Pharm. Res. 1994;11(9):13461351. 2. Paintaud G, Alván G, Dahl ML, et al. Nonlinearity of amoxicillin absorption kinetics in human. Eur. J. Clin. Pharmacol. 1992;43(3):283288. 3. Fonseca W, Hoppu K, Rey LC, Amaral J, Qazi S. Comparing pharmacokinetics of amoxicillin given twice or three times per day to children older than 3 months with pneumonia. Antimicrob. Agents Chemother. 2003;47(3):9971001. Presented at the 19th PopulaEon Approach Group Europe MeeEng, Berlin, Germany, June 812, 2010 Acknowledgements/Declara5ons Support from NIHNIAID 1 K23 AI076106 (MN) and NIHNBIB, R01 EB005803 (MN and RJ) is gratefully acknowledged. None of the authors has a Tinancial or commercial conTlict of interest. Ra AgeMult1 K10 = K10s / (AgeMult1*kg 0.25 ) dX(1)/dt = Ksys*X(4) (K12 + K10)*X(1) + K21*X(2) dX(2)/dt = K12*X(1) K21*X(2) dX(3)/dt = Ra dX(4)/dt = Ra Ksys*X(4) OUTPUT AgeMult2 Y(1) = X(1) / (AgeMult2*V1s*kg) 3 4 V1 2 K12 K21 Ksys Ra Dose Tlag K10 Figure 1 Structural model and defining equaEons. ! if ≤2 years 3.5 1.25*Age if >2 to ≤20 years 1 if >20 years X(3)*Vma/(Kma + X(3)) if Time < Tabs if Time ≥ Tabs ! ! 4.5 if ≤2 years 4.5 0.175*Age if >2 to ≤20 years 1 if >20 years Group Cmax (mg/L) AUC 012 (mg*h/L) Adult, 3000 mg 1,2 Sim: 29.5 (11.0) P=0.55 Obs: 26.8 (3.9), n=6 Sim: 96.2 (28.8) P=0.44 Obs: 91.0 (17.0), n=6 Adult, 500 mg 1,2 Sim: 8.4 (2.6) P=0.71 Obs: 8.8 (1.8), n=6 Sim: 24.4 (6.1) P=0.24 Obs: 25.0 (3.0), n=6 Child, 15 mg/kg 3 Sim: 6.6 (1.8) P=0.40 Obs: 6.9 (3.0), n=27 Sim: 25.7 (6.5) P=0.53 Obs: 24.9 (9.6), n=26 Child, 25 mg/kg 3 Sim: 10.7 (3.0) P=0.87 Obs: 10.6 (5.1), n=23 Sim: 42.6 (10.8) P=0.50 Obs: 44.1 (24.6), n=22 Table 2 Numerical predicEve check comparing simulated (Sim) vs. observed (Obs) maximum amox concentraEon (Cmax) and area under the EmeconcentraEon curve (AUC) for varying doses with a dose interval of 12 hours in adults and children. ObservaEons are taken from referenced studies, with the number (n) of parEcipants in each group indicated. SimulaEons for children were done using a weight of 12 kg and an age of 15 months. Parameter FiQed Published 1 Tlag (h) 0.04 (0.01 0.11) 0.09 Tabs (h) 2.59 (1.91 3.36) 1.72 Tabs Tlag (h) 2.51 (1.80 3.32) 1.60 Ksys (h 1 ) 1.10 (0.82 2.97) 1.31 Kma (mg) 864.39 (341.16 2286.17) 1077 Vma (mg/h) 1056.70 (841.00 1862.80) 1407 Vma/Kma (h 1 ) 0.98 (0.69 3.52) 1.78 V1s (L/kg) 0.11 (0.10 0.12) 0.15* K10s (kg * h 1 ) 5.86 (5.59 7.03) 4.48* K12 (h 1 ) 0.81 (0.51 1.04) 1.11 K21 (h 1 ) 0.92 (0.87 1.08) 1.31 Table 1 Median (interquarEle) parameter esEmates for the populaEon model compared with those in the published reference study of 6 adults. *Assuming weight of 70 kg and age ≥20 years, which were not reported. 0 2 4 6 8 10 12 0 10 20 30 40 Adults 3000 mg Time (h) Plasma amoxicillin (mg/L) 0.05 0.25 0.5 0.75 0.95 0 2 4 6 8 10 12 0 2 4 6 8 10 12 Adults 500 mg Time (h) Plasma amoxicillin (mg/L) 0.05 0.25 0.5 0.75 0.95 0.05 0.10 0.20 0.50 1.00 2.00 5.00 10.00 0.05 0.10 0.20 0.50 1.00 2.00 5.00 10.00 Plasma Amoxicillin (mg/L) Predicted (population) Observed R-squared = 0.82 Inter = -0.03 (95%CI -0.1 to 0.03) Slope = 0.99 (95%CI 0.9 to 1.08) 0.05 0.10 0.20 0.50 1.00 2.00 5.00 10.00 0.05 0.10 0.20 0.50 1.00 2.00 5.00 20.00 Plasma Amoxicillin (mg/L) Predicted (population) Observed R-squared = 0.9 Inter = 0.04 (95%CI -0.02 to 0.1) Slope = 0.97 (95%CI 0.91 to 1.03) 0.01 0.05 0.10 0.50 1.00 5.00 10.00 0.01 0.05 0.10 0.50 1.00 5.00 Plasma Amoxicillin (mg/L) Predicted (posterior) Observed R-squared = 0.98 Inter = 0.04 (95%CI 0.02 to 0.06) Slope = 0.98 (95%CI 0.96 to 1.01) 0.02 0.05 0.10 0.20 0.50 1.00 2.00 5.00 20.00 0.02 0.05 0.20 0.50 2.00 5.00 20.00 Plasma Amoxicillin (mg/L) Predicted (posterior) Observed R-squared = 0.96 Inter = 0.08 (95%CI 0.04 to 0.12) Slope = 0.94 (95%CI 0.9 to 0.98) 500 mg 3000 mg Figure 2 Observed vs. populaEon (top) or Bayesian posterior (middle) predicted plots. Visual predicEve check (boeom). 0 2 4 6 8 0 5 10 15 20 25 Approved: 40 mg/kg divided three times daily Time (h) Plasma amoxicillin (mg/L) 0.05 0.25 0.5 0.75 0.95 MIC=2 MIC=4 MIC=8 0 2 4 6 8 10 12 0 5 10 15 20 25 Approved: 45 mg/kg divided twice daily Time (h) Plasma amoxicillin (mg/L) 0.05 0.25 0.5 0.75 0.95 MIC=2 MIC=4 MIC=8 MIC T>MIC T>MIC≥50% T>MIC T>MIC≥50% T>MIC T>MIC≥50% T>MIC T>MIC≥50% 2 53 (15) % 64% 47 (11) % 39% 79 (14) % 99% 63 (12) % 91% 4 28 (14) % 4% 32 (10) % 3% 58 (15) % 75% 48 (11) % 41% 8 1 (3) % 0% 14 (9) % 0% 33 (15%) 10% 32 (10) % 3% 0 2 4 6 8 10 12 0 5 10 15 20 25 Off-label: 90 mg/kg divided twice daily Time (h) Plasma amoxicillin (mg/L) 0.05 0.25 0.5 0.75 0.95 MIC=2 MIC=4 MIC=8 0 2 4 6 8 0 5 10 15 20 25 Off-label: 90 mg/kg divided three times daily Time (h) Plasma amoxicillin (mg/L) 0.05 0.25 0.5 0.75 0.95 MIC=2 MIC=4 MIC=8 PRIMARY OUTCOME The first two graphs show the modelsimulated distribuEon of amox concentraEons for FDAapproved dosing, followed by two distribuEons for offlabel dosing at 90 mg/kg/day. Reference lines are drawn at nonmeningiEs MIC breakpoints for IV penicillin (≤2 suscepEble; 4 intermediate; ≥8 resistant) and 50% of the dosing interval. The resistance breakpoint for oral penicillin is ≥2. The table indicates the mean (SD) percentage T>MIC and percent of 1000 simulated profiles with T>MIC for at least half of the dosing interval.
Transcript of PAGE 2010 amox 2010 amox... · 2018. 6. 7. ·...
-
High-‐Dose Amoxicillin Pharmacokine5cs (PK) and Pharmacodynamics (PD) in Children
Michael Neely1, Michael Reed2, Roger Jelliffe11University of Southern California Laboratory of Applied PharmacokineEcs, 134-‐B Alcazar Ave, Los Angeles, CA 90033 USA
2The Rebecca D. Considine Clinical Research InsEtute, Akron Children's Hospital, One Perkins Square, Akron, OH 44308 USA
Objec5ves• We are unaware of any published PK data addressing oral amoxicillin (amox) doses >25 mg/kg in children.
• Since amox absorption is saturable,1,2 we wished to estimate the percentage of the dosing interval that the plasma amox concentration would exceed S. pneumoniae (pneumococcal) breakpoint minimum inhibitory concentrations (T>MIC), and the percent of children with the target T>MIC of at least half the dosing interval after various amox doses, up to the commonly used daily dose of 90 mg/kg.
#1802
Conclusions• We extended a published adult model of saturable oral amox absorption to children.• Simulation from the model suggests that amox plasma exposure in children is dose proportional at least to a dose of 45 mg/kg. At a median dose of 864 mg (Kma), bioavailability is expected to be 50% of maximal.
• A resistance breakpoint of ≥2 mg/L for oral amox is appropriate at approved doses (40-‐45 mg/kg/day).• With a higher daily dose of 90 mg/kg, by percentage target attainment of a T>MIC for at least half the dosing interval, more resistant organisms could likely be treated, especially with dosing three times daily.
• ConTirmation of this simulation study is desirable.
www.lapk.org
Methods• We adapted a published adult four-‐compartment model (Figure 1) with saturable, delayed, time-‐limited amox absorption through an absorptive compartment into a central compartment with linear elimination and transfer to/from a peripheral compartment.1,2 The original model was derived from 6 adult patients Tirst given a single dose of amox 500 mg IV, with intensive 10-‐hour post-‐dose PK sampling. A two-‐compartment model was Titted to the data. Each subject was then given single oral amox doses of 500 mg and 3000 mg with the same sampling after each dose as for the IV dose. The order of oral doses was random and separated by a week minimum. For the oral model, volume and inter-‐compartmental rate transfer constants were Tixed for each participant according to their estimates from the IV model.
• We log-‐transformed and allometrically scaled the reported model parameters as the initial ranges prior to Titting the data. No parameter in our model was Tixed, including volume.
• Individual data after oral dosing from the original study were used to Tit the model using non-‐parametric methods implemented in MM-‐USCPACK (available from www.lapk.org), with standard visual/numerical checks.
• Upon comparison of model-‐simulated proTiles with published Cmax and AUC ranges for doses of 15 and 25 mg/kg in children,3 volume and elimination terms were multiplied by population-‐constant age functions to accurately simulate the adult and pediatric data.
• Finally, single-‐dose T>MIC were simulated in 1000 representative 15 month, 12 kg children for varying amox doses and dose intervals.
Results
References1. Piotrovskij VK, Paintaud G, Alván G, Trnovec T. Modeling of the saturable time-‐constrained amoxicillin absorption in humans. Pharm. Res.
1994;11(9):1346-‐1351.2. Paintaud G, Alván G, Dahl ML, et al. Nonlinearity of amoxicillin absorption kinetics in human. Eur. J. Clin. Pharmacol. 1992;43(3):283-‐288.3. Fonseca W, Hoppu K, Rey LC, Amaral J, Qazi S. Comparing pharmacokinetics of amoxicillin given twice or three times per day to children
older than 3 months with pneumonia. Antimicrob. Agents Chemother. 2003;47(3):997-‐1001.
Presented at the 19th PopulaEon Approach Group Europe MeeEng, Berlin, Germany, June 8-‐12, 2010
Acknowledgements/Declara5onsSupport from NIH-‐NIAID 1 K23 AI076106 (MN) and NIH-‐NBIB, R01 EB005803 (MN and RJ) is gratefully acknowledged. None of the authors has a Tinancial or commercial conTlict of interest.
Ra
AgeMult1
K10 = K10s / (AgeMult1*kg0.25) dX(1)/dt = Ksys*X(4) -‐ (K12 + K10)*X(1) + K21*X(2) dX(2)/dt = K12*X(1) -‐ K21*X(2) dX(3)/dt = -‐Ra dX(4)/dt = Ra -‐ Ksys*X(4)
OUTPUT
AgeMult2
Y(1) = X(1) / (AgeMult2*V1s*kg)
3 4 V1 2K12
K21
KsysRa
Dose
Tlag
K10
Figure 1 -‐ Structural model and defining equaEons.
! 3.5 if ≤2 years3.5 -‐ 1.25*Age if >2 to ≤20 years 1 if >20 yearsX(3)*Vma/(Kma + X(3)) if Time < Tabs0 if Time ≥ Tabs!
! 4.5 if ≤2 years4.5 -‐ 0.175*Age if >2 to ≤20 years 1 if >20 years
Group Cmax (mg/L) AUC0-‐12 (mg*h/L)
Adult, 3000 mg 1,2 Sim: 29.5 (11.0) P=0.55Obs: 26.8 (3.9), n=6
Sim: 96.2 (28.8) P=0.44Obs: 91.0 (17.0), n=6
Adult, 500 mg 1,2 Sim: 8.4 (2.6) P=0.71Obs: 8.8 (1.8), n=6
Sim: 24.4 (6.1) P=0.24Obs: 25.0 (3.0), n=6
Child, 15 mg/kg 3 Sim: 6.6 (1.8) P=0.40Obs: 6.9 (3.0), n=27
Sim: 25.7 (6.5) P=0.53Obs: 24.9 (9.6), n=26
Child, 25 mg/kg 3 Sim: 10.7 (3.0) P=0.87Obs: 10.6 (5.1), n=23Sim: 42.6 (10.8) P=0.50Obs: 44.1 (24.6), n=22
Table 2 -‐ Numerical predicEve check comparing simulated (Sim) vs. observed (Obs) maximum amox concentraEon (Cmax) and area under the Eme-‐concentraEon curve (AUC) for varying doses with a dose interval of 12 hours in adults and children. ObservaEons are taken from referenced studies, with the number (n) of parEcipants in each group indicated. SimulaEons for children were done using a weight of 12 kg and an age of 15 months.
Parameter FiQed Published1
Tlag (h) 0.04 (0.01 -‐ 0.11) 0.09
Tabs (h) 2.59 (1.91 -‐ 3.36) 1.72
Tabs -‐ Tlag (h) 2.51 (1.80 -‐ 3.32) 1.60
Ksys (h-‐1) 1.10 (0.82 -‐ 2.97) 1.31
Kma (mg) 864.39 (341.16 -‐ 2286.17) 1077
Vma (mg/h) 1056.70 (841.00 -‐ 1862.80) 1407
Vma/Kma (h-‐1) 0.98 (0.69 -‐ 3.52) 1.78
V1s (L/kg) 0.11 (0.10 -‐ 0.12) 0.15*
K10s (kg * h-‐1) 5.86 (5.59 -‐ 7.03) 4.48*
K12 (h-‐1) 0.81 (0.51 -‐ 1.04) 1.11
K21 (h-‐1) 0.92 (0.87 -‐ 1.08) 1.31
Table 1 -‐ Median (interquarEle) parameter esEmates for the populaEon model compared with those in the published reference study of 6 adults. *Assuming weight of 70 kg and age ≥20 years, which were not reported.
0 2 4 6 8 10 12
010
2030
40
Adults 3000 mg
Time (h)
Pla
sma
amox
icill
in (m
g/L)
0.05
0.25
0.5
0.75
0.95
0 2 4 6 8 10 12
02
46
810
12
Adults 500 mg
Time (h)
Pla
sma
amox
icill
in (m
g/L)
0.05
0.25
0.5
0.75
0.95
0.05 0.10 0.20 0.50 1.00 2.00 5.00 10.00
0.05
0.10
0.20
0.50
1.00
2.00
5.00
10.00
Plasma Amoxicillin (mg/L)
Predicted (population)
Observed
R-squared = 0.82Inter = -0.03 (95%CI -0.1 to 0.03)Slope = 0.99 (95%CI 0.9 to 1.08)
0.05 0.10 0.20 0.50 1.00 2.00 5.00 10.00
0.05
0.10
0.20
0.50
1.00
2.00
5.00
20.00
Plasma Amoxicillin (mg/L)
Predicted (population)
Observed
R-squared = 0.9Inter = 0.04 (95%CI -0.02 to 0.1)Slope = 0.97 (95%CI 0.91 to 1.03)
0.01 0.05 0.10 0.50 1.00 5.00 10.00
0.01
0.05
0.10
0.50
1.00
5.00
Plasma Amoxicillin (mg/L)
Predicted (posterior)
Observed
R-squared = 0.98Inter = 0.04 (95%CI 0.02 to 0.06)Slope = 0.98 (95%CI 0.96 to 1.01)
0.02 0.05 0.10 0.20 0.50 1.00 2.00 5.00 20.00
0.02
0.05
0.20
0.50
2.00
5.00
20.00
Plasma Amoxicillin (mg/L)
Predicted (posterior)
Observed
R-squared = 0.96Inter = 0.08 (95%CI 0.04 to 0.12)Slope = 0.94 (95%CI 0.9 to 0.98)
500 mg 3000 mg
Figure 2 -‐ Observed vs. populaEon (top) or Bayesian posterior (middle) predicted plots. Visual predicEve check (boeom).
0 2 4 6 8
05
1015
2025
Approved: 40 mg/kg divided three times daily
Time (h)
Pla
sma
amox
icill
in (m
g/L)
0.05
0.25
0.5
0.75
0.95
MIC=2
MIC=4
MIC=8
0 2 4 6 8 10 12
05
1015
2025
Approved: 45 mg/kg divided twice daily
Time (h)
Pla
sma
amox
icill
in (m
g/L)
0.05
0.25
0.5
0.75
0.95
MIC=2
MIC=4
MIC=8
MIC T>MIC T>MIC≥50% T>MIC T>MIC≥50% T>MIC T>MIC≥50% T>MIC T>MIC≥50%
2 53 (15) % 64% 47 (11) % 39% 79 (14) % 99% 63 (12) % 91%
4 28 (14) % 4% 32 (10) % 3% 58 (15) % 75% 48 (11) % 41%
8 1 (3) % 0% 14 (9) % 0% 33 (15%) 10% 32 (10) % 3%
0 2 4 6 8 10 12
05
1015
2025
Off-label: 90 mg/kg divided twice daily
Time (h)
Pla
sma
amox
icill
in (m
g/L)
0.05
0.25
0.5
0.75
0.95
MIC=2
MIC=4
MIC=8
0 2 4 6 8
05
1015
2025
Off-label: 90 mg/kg divided three times daily
Time (h)
Pla
sma
amox
icill
in (m
g/L)
0.05
0.25
0.5
0.75
0.95
MIC=2
MIC=4
MIC=8
PRIMARY OUTCOME -‐ The first two graphs show the model-‐simulated distribuEon of amox concentraEons for FDA-‐approved dosing, followed by two distribuEons for off-‐label dosing at 90 mg/kg/day. Reference lines are drawn at non-‐meningiEs MIC breakpoints for IV penicillin (≤2 suscepEble; 4 intermediate; ≥8 resistant) and 50% of the dosing interval. The resistance breakpoint for oral penicillin is ≥2. The table indicates the mean (SD) percentage T>MIC and percent of 1000 simulated profiles with T>MIC for at least half of the dosing interval.
http://www.lapk.orghttp://www.lapk.org
