Application of Pharmacokinetic and Pharmacodynamic ...• Establish a US database for antibiotic use and ... – Work-horse agents (pip/tazo, cefepime, meropenem) FDA-approved prior

Application of Pharmacokinetic and Pharmacodynamic Principles in

Antimicrobial Stewardship

Michael B. Kays, Pharm.D., FCCPAssociate Professor of Pharmacy Practice

Purdue University College of PharmacyAdjunct Associate Professor of Medicine

Indiana University School of MedicineClinical Specialist, Infectious Diseases

Indiana University Health, Methodist HospitalIndianapolis and West Lafayette, Indiana

Disclosures

• Serves on the advisory board and speaker’s bureau for Allergan and advisory board for Cubist (Merck).

Learning Objectives

• At the end of this lecture, the learner will be able to:– Classify antibiotics based upon pharmacodynamic

– efficacy characteristics– Devise 3 examples of dose optimization initiatives

based on PK/PD principles– Design and defend a beta-lactam dosing protocol

utilizing extended infusion

Seven Steps to Preserve the Miracle of Antibiotics

• Establish a US database for antibiotic use and resistance comparable to the EU

• Restrict use of antibiotics in agriculture• Prevent selected nosocomial infections• Aggressively promote antimicrobial stewardship• Promote use of new diagnostics with emphasis on

point-of-care molecular methods• Reduce FDA antibiotic barrier• Facilitate public-private partnerships for antibiotic

developmentBartlett JG, Gilbert DN, Spellberg B. Clin Infect Dis 2013;56:1445-1450.

Antimicrobial Stewardship• Primary goal

– To optimize clinical outcomes while minimizing unintended consequences of antimicrobial use, including toxicity, selection of pathogenic organisms, and emergence of resistance

– Includes appropriate antibiotic selection, dosing, route, and duration of therapy → judicious use

• Secondary goal– To reduce healthcare costs without adversely

impacting quality of careDellit TH, et al. Clin Infect Dis 2007;44:159-77.

Antimicrobial StewardshipCore Strategies

• Core strategies– Prospective audit of antimicrobial use with direct

feedback to the prescriber– Formulary restriction and preauthorization

• Additional components– Education– Evidence-based practice guidelines/clinical

pathways– Streamlining or de-escalation of therapy– Dose optimization based on PK/PD– Parenteral to oral conversion

Dellit TH, et al. Clin Infect Dis 2007;44:159-77.

Pharmacodynamic Properties of Different Antibacterial Classes

Drug/Class Bactericidal ActivityPAE (Gram-negatives) PD Parameter

β-lactams

Time-dependent(concentration-

dependent up to 4xMIC)

Little to none(exception:

carbapenems)fT>MIC

Fluoroquinolones Concentration-dependent YesAUC/MICPeak/MIC

Aminoglycosides Concentration-dependent Yes Peak/MIC

Vancomycin Time-dependent Yes AUC/MIC

Ambrose PG, et al. Clin Infect Dis 2007;44:79-86.

β-Lactam Antibiotics

• Commonly recommended PK/PD targets– Carbapenem – fT>MIC ≥ 40%– Penicillins – fT>MIC ≥ 50%– Cephalosporins – fT>MIC ≥ 60%– Aztreonam – fT>MIC ≥ 50-60%

• Other “proposed” or “evaluated” targets– fT>MIC ≥ 100%– fT>4xMIC ≥ 50%– fT>4xMIC ≥ 100% (Cmin/MIC ≥ 4)

Clinical Pharmacodynamics of β-Lactams in Different Patient Populations

• Cefepime, Ceftazidime– VAP: fT>MIC > 53% (microbiologic success)1

• Cefepime– Gram-negative infections: T>4.3xMIC (83-95% for

microbiologic success of 80-90%)2– P. aeruginosa: fT>MIC > 60% (microbiologic success)3

• Meropenem– LRTI: fT>MIC > 54%, fCmin/MIC > 5 (microbiologic

success)4– FN with bacteremia: T>MIC 83% (clinical response)5

1 McVane SH, et al. Antimicrob Agents Chemother 2014;58:1359-64; 2 Tam VH, et al. J Antimicrob Chemother2002;50:425-8; 3 Crandon JL, et al. Antimicrob Agents Chemother 2010;54:1111-6; 4 Li C, et al. AntimicrobAgents Chemother 2007;51:1725-30; 5 Ariano RE, et al. Ann Pharmacother 2005;39:32-8.

Goal of Dosing β-Lactam Antibiotics

• Ensure adequate PD exposures by maximizing or optimizing fT>MIC

• Why not use current FDA-approved doses?– Decreased susceptibility for many nosocomial pathogens– Work-horse agents (pip/tazo, cefepime, meropenem) FDA-

approved prior to clinical application of PK/PD principles– Limited number of dosing regimens studied for each

approved indication• “Therapeutically equivalent” if fT>MIC for

alternative regimen ≈ fT>MIC for FDA-approved regimen (surrogate marker)

Strategies to Optimize PD Exposures for β-Lactams

• Increase dose, same interval– 0.5 g q8h vs. 1 g q8h

• Same dose, shorter interval– 1 g q12h vs. 1 g q6h

• Prolonged infusions– Infuse dose over 3-4 hours

• Continuous infusion– Stability issues– Need dedicated IV line– Ideal concentration/MIC ratio? ≥ 4xMIC?

Increase Dose, Same Interval

Chart1

000

0.50.50.5

111

1.51.51.5

222

333

444

666

888

500 mg q8h

1 g q8h

MIC

Time (hours)

Concentration (mg/L)

1.76

3.52

8

33.5

67.01

8

18.4

36.8

8

12.41

24.82

8

9.42

18.83

8

6.39

12.77

8

4.73

9.46

8

2.83

5.66

8

1.78

3.56

8

Sheet1

Concentration500 mg q8h1 g q8hMIC

01.763.528

0.533.567.018

118.436.88

1.512.4124.828

29.4218.838

36.3912.778

44.739.468

62.835.668

81.783.568

To update the chart, enter data into this table. The data is automatically saved in the chart.

Same Dose, Shorter Interval

Chart1

000

0.50.50.5

111

1.51.51.5

222

333

444

666

86.56.5

1277

7.57.5

88

99

1010

1212

1 g q12h

1 g q6h

Time (hours)


0.86

6.31

8

61.89

69.47

8

33.34

38.98

8

23.02

26.75

8

18.19

20.54

8

12.94

14.12

8

9.54

10.54

8

5.22

6.36

8

2.86

69.47

8

0.86

38.98

8

26.75

8

20.54

8

14.12

8

10.54

8

6.36

8

Sheet1

Time1 g q12hTime21 g q6hMIC

00.8606.318

0.561.890.569.478

133.34138.988

1.523.021.526.758

218.19220.548

312.94314.128

49.54410.548

65.2266.368

82.866.569.478

120.86738.988

7.526.758

820.548

914.128

1010.548

126.368


Same Dose, Prolonged Infusion

Chart1

000

0.50.50.5

111

1.51.51.5

222

333

444

666

888

1 g q8h (0.5 h)

1 g q8h (4 h)

MIC

Time (hours)


3.52

5.49

8

67.01

12.36

8

36.8

15.79

8

24.82

18.04

8

18.83

19.76

8

12.77

22.4

8

9.15

24.34

8

5

11

8

2.8

5.49

8

Sheet1

Time1 g q8h (0.5 h)1 g q8h (4 h)MIC

03.525.498

0.567.0112.368

136.815.798

1.524.8218.048

218.8319.768

312.7722.48

49.1524.348

65118

82.85.498


National Survey on Continuous and Extended Infusions of Antibiotics

George JM, et al. Am J Health-Syst Pharm 2012;69:1895-904.

* Random sample of 1,000 acute care hospitals

29/26015/260

30/60

21/60

Chart1

Random Survey*Random Survey*

SIDP SurveySIDP Survey

Continuous Infusion

Extended Infusion

Hospitals (%)

11.2

5.8

50

35

Sheet1

Continuous InfusionExtended Infusion

Random Survey*11.25.8

SIDP Survey5035


Pharmacokinetics of Piperacillin, with Tazobactam, in Hospitalized Patients

1

10

100

0 1 2 3 4 5 6 7 8

Conc

entr

atio

n (m

g/L)

Time (hours)

Shea KM, et al. Int J Antimicrob Agents 2009;34:429-33.

Patients received 4.5 g q8h (4 h)Free concentrations (PB 30%)CLcr 83 ± 42 ml/min (23-148)

Pharmacokinetics of Piperacillin, with Tazobactam, in Hospitalized Patients

1

10

100

1000

0 2 4 6 8

Conc

entr

atio

n (m

g/L)

Time (hours)

Patient 1

Patient 2

Patient 3

Patient 4

Patient 5

Patient 6

Patient 7

Patient 8

Patient 9

Patient 10

Patient 11

Patient 12

Patient 13

Shea KM, et al. Int J Antimicrob Agents 2009;34:429-33.

Patients received 4.5 g q8h, infused over 4 hFree concentrations (PB 30%)

Probability of Target Attainment of Piperacillin in Hospitalized Patients

0

20

40

60

80

100

1 2 4 8 16 32 64

PTA

(%)

MIC (mg/L)

3.375 grams II q6h

4.5 grams II q8h

4.5 grams II q6h

3.375 grams PI q8h

4.5 grams PI q8h

Shea KM, et al. Ann Pharmacother 2009;43:1747-54.PD target: fT>MIC ≥ 50%

Pharmacokinetics of PI Cefepime in Hospitalized Patients

1

10

100

0 1 2 3 4 5 6 7 8

Conc

entr

atio

n (m

g/L)

Time (hours)

Cheatham SC, et al. Int J Antimicrob Agents 2011;37:46-50.

Patients received 1 g q8h (4 h)Free concentrations (PB 20%)CLcr 86 ± 31 ml/min (50-137)

Pharmacokinetics of PI Cefepime in Hospitalized Patients

1

10

100

0 2 4 6 8

Patient 1Patient 2Patient 3Patient 4Patient 5Patient 6Patient 7Patient 8Patient 9

Cheatham SC, et al. Int J Antimicrob Agents 2011;37:46-50.

Patients received 1 g q8h, infused over 4 hFree concentrations (PB 20%)

Probability of Target Attainment of Cefepime in Hospitalized Patients

0

20

40

60

80

100

0.5 1 2 4 8 16 32

PTA

(%)

MIC (mg/L)

1 g q8h (4 h infusion) 2 g q8h (4 h infusion) 1 q q12h (4 h infusion)2 g q12h (4 h infusion) 1 g q6h (3 h infusion) 2 g q6h (3 h infusion)

Cheatham SC, et al. Int J Antimicrob Agents 2011;37:46-50.PD target: fT>MIC ≥ 60%

Cumulative Fraction of Response of Cefepime in Hospitalized Patients

Regimen E. coli K. pneumoniae P. aeruginosa

1 g q12h (4 h) 96.9 88.6 73.8

2 g q12h (4 h) 97.8 91.1 87.1

1 g q8h (4 h) 97.7 90.9 88.6

2 g q8h (4 h) 98.9 95.4 96.2

1 g q6h (3 h) 98.2 92.6 92.7

2 g q6h (3 h) 99.4 97.5 98.2

Cheatham SC, et al. Int J Antimicrob Agents 2011;37:46-50.PD target: fT>MIC ≥ 60%

Potential Reasons for Not Adopting Prolonged Infusion Dosing

• “This dosing is not approved by the FDA.”– Do you only use drugs/doses as approved by FDA?– Treatment for HAP/VAP? No drugs FDA-approved

• “I want to see randomized, prospective, clinical studies first.”– Funding source?– Sample size needed to detect a significant difference?– Target less susceptible bacteria?– Ethical considerations?

• “It’s not better than what we are currently doing.”– Does it have to be “better”?

Piperacillin/Tazobactam Extended Infusion Dosing for P. aeruginosa

APACHE II < 17(n=115)

APACHE II ≥ 17(n=79)

14-day mortality 5.2% 21.5% (p=0.001)

Median LOS 18 days 27.5 days (p=0.02)

Intermittent(n=54)

Extended(n=61)

Intermittent(n=38)

Extended(n=41)

14-day mortality 3.7% 6.6% 31.6% 12.2%

p=0.5 p=0.04

Median LOS 18 days 18 days 38 days 21 days

p=0.5 p=0.02

Lodise TP, et al. Clin Infect Dis 2007;44:357-63.

3.375 g q4-6h, infused over 30 minutes3.375 g q8h, infused over 4 hours

Retrospective Cohort of Extended-Infusion Piperacillin/Tazobactam

Yost RJ, et al. Pharmacotherapy 2011;31:767-75.

p=0.02p=0.03

p=NS

p=NSp=NS

EI: 3.375 g q8h, infused over 4 hours

Chart1

EntireCohortEntireCohort

ICUAdmissionICUAdmission

APACHE II≥ 17APACHE II≥ 17

EI P/T vs.NEI P/TEI P/T vs.NEI P/T

EI P/T vs.Other β-lactamsEI P/T vs.Other β-lactams

Extended-Infusion Pip/Tazo(n=186)

Non-Extended-Infusion Comparator(n=173)

Mortality (%)

9.7

17.9

14.9

22.5

16

25.2

9.7

20.2

9.7

15.7

Sheet1

Extended-Infusion Pip/Tazo(n=186)Non-Extended-Infusion Comparator(n=173)

EntireCohort9.717.9

ICUAdmission14.922.5

APACHE II≥ 171625.2

EI P/T vs.NEI P/T9.720.2

EI P/T vs.Other β-lactams9.715.7


Extended-Infusion Cefepime in Patients with P. aeruginosa Infections

Outcome Intermittent(n=54)Extended

(n=33) P-Value

Mortality 11 (20) 1 (3) 0.03

LOS (days, median)

Hospital 14.5 11 0.36

Infection-related 12 10 0.45

ICU 18.5 8 0.04

Duration of mechanicalventilation (days) 14.5 10.5 0.42

Cost (US$)

Total hospital costs $ 51,231 $ 28,048 0.13

Infection-related hospital cost $ 15,322 $ 13,736 0.78

Bauer KA, et al. Antimicrob Agents Chemother 2013;57:2907-12.

2 g q8h, infused over 30 minutes; 2 g q8h, infused over 4 hours

Clinical Outcomes with PI Vs. Intermittent Infusions – Meta-Analysis

Study Group EI or CI Intermittent Risk Ratio(95% CI)

Mortality Events Total Events Total

PI vs. Intermittent 96 859 137 789 0.66 (0.53, 0.83)

Clinical Success


Teo J, et al. Int J Antimicrob Agents 2014;43:403-11.

• 29 studies with 2,206 patients included in the meta analysis• Carbapenems, cephalosporins, and penicillins evaluated• PI: EI ≥ 3 hours, CI 24 hours; intermittent 20-60 minutes

Antibiotic-related adverse events were comparable between groups.


Study Subgroups No. of Studies No. of Patients Risk Ratio(95% CI)

Mortality

RCTs 10 779 0.83 (0.57, 1.21)

Non-RCTs 9 841 0.57 (0.43, 0.76)

Penicillins 8 974 0.60 (0.45, 0.82)

Cephalosporins 5 191 0.92 (0.52, 1.63)

Carbapenems 4 274 0.74 (0.42, 1.28)

Equivalent Daily Dose 10 813 0.82 (0.56, 1.20)

APACHE II ≥ 15 10 861 0.63 (0.48, 0.81)

All Studies 19 1620 0.66 (0.53, 0.83)



Study Subgroups No. of Studies No. of Patients Risk Ratio(95% CI)

Clinical Success

RCTs 14 1125 1.05 (0.99, 1.12)

Non-RCTs 5 421 1.34 (1.02, 1.76)

Penicillins 6 491 1.08 (0.94, 1.25)

Cephalosporins 9 662 1.11 (0.98, 1.25)

Carbapenems 3 333 1.16 (0.93, 1.46)

Equivalent Daily Dose 10 934 1.22 (1.05, 1.43)

APACHE II ≥ 15 8 663 1.26 (1.06, 1.50)

All Studies 19 1546 1.12 (1.03, 1.21)


Clinical Outcomes with EI or CI Vs. Short-Term Infusions – Meta-Analysis

Study Group EI or CI Short-Term Risk Ratio(95% CI)

Mortality Deaths Total Deaths Total

EI vs. Short-term* 34 314 46 273 0.63 (0.41, 0.95)

CI vs. Short-term* 10 257 24 272 0.50 (0.26, 0.96)

Total* 44 571 70 545 0.59 (0.41, 0.83)

Falagas ME, et al. Clin Infect Dis 2013;56:272-82.

• 14 studies with 1,229 patients included in the analysis• Carbapenems and piperacillin/tazobactam evaluated• EI ≥ 3 hours; CI 24 hours; short-term 20-60 minutes

* Carbapenems and piperacillin/tazobactam combinedPneumonia: RR 0.50 (95% CI, 0.26, 0.96)Unspecified infection: RR 0.63 (95% CI 0.41, 0.95)



Mortality Deaths Total Deaths Total

Carbapenems

EI or CI vs. Short-term 14 110 18 103 0.66 (0.34, 1.30)

Piperacillin/Tazo


Carbapenem or P/T


Total 44 571 70 545 0.59 (0.41, 0.83)




Clinical Cure Success Total Success Total

Carbapenems


Piperacillin/Tazo


Total 236 267 222 290 1.13 (0.99, 1.28)

Adverse Events* 20% (5/25)16.9% (22/130)24% (6/25)

13.6% (18/132)


* No adverse events reported in 3 of 5 studies with data regarding AEs

Clinical Outcomes with Alternative Dosing Strategies for Piperacillin/Tazobactam




Clinical Success


Bacteriologic Success


Yang H, et al. PLoS One 2015;10(1):e0116769. doi:10:1371/journal.pone.0116769.

• 14 studies with 1,786 patients included in the meta analysis• Intermittent: 2.25-4.5 g q6-8h, infused over 20-30 minutes• PI: EI ≥ 3 hours, CI 24 hours; dose range 6.75-13.5 g daily

Clinical Outcomes with PI and Intermittent Infusion in Hospitalized Patients




Clinical Cure


Tamma PD, et al. BMC Infectious Diseases 2011;11:181.

• 14 studies with 1,491 patients included in the meta analysis (1979-2008)• 8 of 14 studies with ≤ 50 patients• Carbapenems, cephalosporins, penicillins• PI: EI ≥ 3 hours, CI 24 hours; Intermittent package insert recommendations

Prolonged Infusions for Suspected GN Infections in the ICU

Outcome Intermittent (n=242) Prolonged (n=261) P-value

Treatment success

Clinically evaluable 137 (56.6%) 133 (51.0%) 0.204

Micro evaluable 58/105 (55.2%) 50/101 (49.5%) 0.486

14-day mortality 32 (13.2%) 47 (18.0%) 0.141

30-day mortality 57 (23.6%) 67 (25.7%) 0.582

In-hospital mortality 47 (19.4%) 60 (23.0%) 0.329

ICU LOS (days)* 9.3 10.8 0.138

Hospital LOS (days)* 17.0 15.6 0.281

Time from infection onset to:

ICU discharge* 7.8 8.4 0.293

Hospital discharge* 13.2 12.4 0.481

Death* 36 19 < 0.001Arnold HM, et al. Ann Pharmacother 2013;47:170-80.* Median

Applying Prolonged Infusion Dosing to Clinical Practice

• Where will greatest benefit be achieved?– Hospital wide vs. specific unit vs. specific patient type

• Ability to interpret PK/PD literature• Determine which antibiotics and dosing regimens

are optimal for your institution• Access to MIC data at your institution• Ability to work with a multidisciplinary team to

implement dosing protocol• Educate hospital staff to minimize errors and

maximize complianceMacVane SH, et al. Int J Antimicrob Agents 2014;43:105-13.

Target Population for Prolonged Infusion Dosing Regimens

• Critically ill patients (e.g., ICU, burn)– Younger patients– Augmented renal clearance or metabolism

• Creatinine clearance 120 ml/min? 150 ml/min? 200 ml/min?

– Increased volume of distribution– Increased body weight

• Patients likely to be infected with pathogens with elevated MICs

• Site of infection

Risk Factors for β-lactam Non-Target Attainment in Critically Ill Patients

All Patients(n=343)

Patient Not Achieving

50% fT>MIC(n=66)

Patients Achieving

50% fT>MIC(n=277)

P-value

Age (years) 60 (47-73) 52 (39-65) 63 (49-75) < 0.001

APACHE II 18 (13-24) 15 (8-21) 19 (14-25) 0.001

SOFA 5 (2-8) 3 (1-6) 6 (3-9) < 0.001

CLcr (ml/min) 91 (54-141) 119 (84-164) 82 (48-124) < 0.001

RRT 33 (9.6%) 1 (1.5%) 32 (11.6%) 0.013

EI or CI 86 (25.1%) 7 (10.6%) 79 (28.5%) 0.03

De Waele JJ, et al. Intensive Care Med 2014;40:1340-51.

Risk Factors for β-lactam Non-Target Attainment in Critically Ill Patients

All Patients(n=343)

Patient Not Achieving

100% fT>MIC(n=142)

Patients Achieving

100% fT>MIC(n=201)

P-value

Age (years) 60 (47-73) 55 (40-66) 64 (51-76) < 0.001

APACHE II 18 (13-24) 16 (10-23) 19 (15-25)

Implementation of a Prolonged Infusion Dosing Strategy

• Assess the institution’s available resources– How should the protocol be implemented?– Under what circumstances?– Who is ultimately responsible?– What is each department’s involvement and responsibility?

• Develop the prolonged infusion dosing protocol– Simple formatting– Easily accessible– Explicitly define patient populations– Employ appropriate antibiotic pharmacodynamics– Provide credible source information

MacVane SH, et al. Int J Antimicrob Agents 2014;43:105-13.


• Assess causative pathogens– Institution-specific and/or unit-based

• Incorporate local resistance rates– Utilize antibiograms to select appropriate agents– Incorporate MIC data (obtain from microbiology lab)

• Multidisciplinary approach– Providers, infectious diseases, AST, pharmacy, microbiology,

nursing, infection control, IT• Develop an implementation plan

– Standardized platform (CPOE)– Delegate “enforcer” of the program (AST, ID Pharm.D.)



• Education– Education and engagement of hospital staff– Combination of interventions

• In-services, presentations, newsletters, order alerts• Evaluate progress

– Prospective monitoring to assess compliance and identify barriers

– Routine surveillance to identify changes in local epidemiology and/or resistance patterns

– Clinical outcomesMacVane SH, et al. Int J Antimicrob Agents 2014;43:105-13.

Challenges and Solutions to Implementation

Challenge Strategic Solution

Lack of provider buy-in

Involve skeptical providers as members of protocol development team.Obtain approval/support from key opinion leaders at the institution.Generate CFR data using institution-specific MIC data.

Limited resources Provide evidence-based information on potential benefits (clinical and economic)Compliance issues Automatic substitution (P&T, MEC)

Inappropriate use

Explicitly define inclusion and exclusion criteria for use.Implement dosing strategies using CPOE-derived protocols


Challenges and Solutions to Implementation

Challenge Strategic Solution

Uncertainty/confusionwith procedures

Provide education up-front and develop initial roll-out plan with contact information for trouble-shooting or questions.

Medication errorsElectronic alerts, product labels, standardize volumes, “smart” pumps, CPOE

Decreasing utilization over time

Continue education sessions.Re-evaluate MIC data to determine need for refinement of drug selection or dosing.Collect outcome data to demonstrate improvements at the institution.


What About the First Dose?

• Most PK/PD analyses have focused on steady-state serum drug concentrations.

• Bacterial inoculum is largest prior to first dose.• Resistant subpopulations may be present.• Need to maximize bactericidal activity to

reduce bacterial load and decrease risk for selection of resistant strains.

• What is the PD target for the first dose?– 50% fT>MIC? Cmax/MIC > 4? Cmin/MIC > 4?

Martinez MN, et al. Antimicrob Agents Chemother 2012;56:2795-805.

First Dose Concentrations of Free Piperacillin in Hospitalized Patients

0102030405060708090

100

0 1 2 3 4

Conc

entr

atio

n (m

g/L)

Time (hours)

3.375 g 4.5 g 6.75 g

fCmax/BP 5.4

fCmax/BP 3.6

fCmax/BP 2.7

First Dose Concentrations of Free Piperacillin in Hospitalized Patients

0

20

40

60

80

100

120

0 0.5 1 1.5 2 2.5 3 3.5 4

Conc

entr

atio

n (m

g/L)

Time (hours)

3.375 g 4.5 g

fCmax/BP 6.7

fCmax/BP 5.0

Application of PD Principles to a Specific Patient (β-Lactam)

Intermittent Infusion

Prolonged Infusion

fT>MIC (%) = TINF −R0 / CL

R0 / CL − MIC

V

CLlnln x x x+ − ln (MIC)

100

DI

V

CL

R0

CL

where TINF is the infusion time; R0 is the infusion rate calculated as (dose x fraction unbound/TINF); CL is systemic clearance (L/h); V is volume of distribution; DI is dosing interval.

Where Do You Find the PK Data?

• Two-stage PK modeling– PK parameters are independent of each other– Individual patients can influence parameter estimates

• Population PK modeling– Estimates covariance between PK parameters and

interindividual/residual variability– Outliers have less influence on parameter estimates more robust estimates

– Can estimate PK parameters in patients with limited drug concentrations

Where Do You Find the PK Data?

• Is your patient represented in the published study?– ICU vs. non-ICU– Burn injury– Morbidly obese– Cystic fibrosis

Estimate Systemic Clearance from the Patient’s Creatinine Clearance

y = 0.065x + 3.191R² = 0.857p

Clinical Application

• 49 year old white male admitted to Methodist Hospital on January 8, 2016 following semi vs. tree MVA (ventilator)

• No significant past medical history• On 1/15, he develops fever with increased

purulent sputum production, hypoxemic.• Temperature 102.3°F; WBC 17,800/mm3• He is 70 inches tall, weighs 105 kg, BMI 33.2

kg/m2, and his estimated CLcr is 90 ml/min.


• Sputum gram-stain: gram-negative bacilli• He was started on piperacillin/tazobactam

3.375 g q8h (infused over 4 h), tobramycin 560 mg IV q24h, and vancomycin 1500 mg IV q12h.

• Sputum culture is positive for P. aeruginosa, susceptible to piperacillin/tazobactam (8μg/ml) and tobramycin (1 μg/ml).

• Are his doses appropriate for his infection?


• Estimate clearance (CL) and volume (Vd)– Non-obese study1

• CL = (0.0651 * 90) + 3.191 = 9.05 L/h• Vd = 0.28 ± 0.06 L/kg * 105 kg = 29.4 L

– Population PK study2• CL = 11.3 + [0.0646 * (CLcr – 105)] + [0.0579 * (BMI –

35)] = 10.2 L/h• Vd = 31.3 + [0.132 * (TBW – 120)] = 29.3 L

1Shea KM, et al. Int J Antimicrob Agents 2009;34:429-33.2Chung EK, et al. J Clin Pharmacol 2015;55:899-908.

Clinical ApplicationPiperacillin/Tazobactam

• 4 h infusion– TINF = 4 h– R0 = (3000 * 0.7)/4 = 525 mg/h– CL = 10.2 L/h– Vd = 29.3 L– fT>MIC ≈ 110%

• 0.5 h infusion– fT>MIC ≈ 105%

Clinical ApplicationPiperacillin/Tazobactam

• Estimated V 29.3 L, CL 10.2 L/h• If we wanted to give a continuous infusion:

– K0 (mg/h) = Css (mg/L) * CL (L/h)• MIC 8 mg/L• K0 (mg/h) = 32 mg/L * 10.2 L/h = 326.4 mg/h 7,834 mg/d 8 g of piperacillin/day as continuous infusion

Conclusions• Pharmacodynamic principles and optimal targets

have been scientifically validated in numerous in vitro, animal, and human studies.

• An important goal of all ASPs is dose optimization of all antibiotics, especially in patients at risk for subtherapeutic concentrations.

• Prolonged infusions of β-lactams increases PTA compared to standard infusions and may allow for reduction in total daily doses.

• Recently published meta-analyses have shown decreased mortality and increased clinical success with prolonged and continuous infusions.

Early Observations Regarding Drug Concentrations, Bacterial Killing &

Resistance“It is not difficult to make microbes resistant to penicillin in the laboratory by exposing them to concentrations not sufficient to kill them, and the same thing has occasionally happened in the body . . . . Moral: If you use penicillin, use enough.”

Alexander Fleming, 1945Nobel Prize acceptance speech

Application of Pharmacokinetic and Pharmacodynamic Principles in Antimicrobial Stewardship DisclosuresLearning ObjectivesSeven Steps to Preserve the Miracle of AntibioticsAntimicrobial StewardshipAntimicrobial Stewardship�Core StrategiesPharmacodynamic Properties of Different Antibacterial Classesβ-Lactam AntibioticsClinical Pharmacodynamics of β-Lactams in Different Patient PopulationsGoal of Dosing β-Lactam AntibioticsStrategies to Optimize PD Exposures for β-LactamsIncrease Dose, Same IntervalSame Dose, Shorter IntervalSame Dose, Prolonged InfusionNational Survey on Continuous and Extended Infusions of AntibioticsPharmacokinetics of Piperacillin, with Tazobactam, in Hospitalized PatientsPharmacokinetics of Piperacillin, with Tazobactam, in Hospitalized PatientsProbability of Target Attainment of Piperacillin in Hospitalized PatientsPharmacokinetics of PI Cefepime in Hospitalized PatientsPharmacokinetics of PI Cefepime in Hospitalized PatientsProbability of Target Attainment of Cefepime in Hospitalized PatientsCumulative Fraction of Response of Cefepime in Hospitalized PatientsPotential Reasons for Not Adopting Prolonged Infusion DosingPiperacillin/Tazobactam Extended Infusion Dosing for P. aeruginosaRetrospective Cohort of Extended-Infusion Piperacillin/TazobactamExtended-Infusion Cefepime in Patients with P. aeruginosa InfectionsClinical Outcomes with PI Vs. Intermittent Infusions – Meta-AnalysisClinical Outcomes with PI Vs. Intermittent Infusions – Meta-AnalysisClinical Outcomes with PI Vs. Intermittent Infusions – Meta-AnalysisClinical Outcomes with EI or CI Vs. Short-Term Infusions – Meta-AnalysisClinical Outcomes with EI or CI Vs. Short-Term Infusions – Meta-AnalysisClinical Outcomes with EI or CI Vs. Short-Term Infusions – Meta-AnalysisClinical Outcomes with Alternative Dosing Strategies for Piperacillin/TazobactamClinical Outcomes with PI and Intermittent Infusion in Hospitalized PatientsProlonged Infusions for Suspected GN Infections in the ICUApplying Prolonged Infusion Dosing to Clinical PracticeTarget Population for Prolonged Infusion Dosing RegimensRisk Factors for β-lactam Non-Target Attainment in Critically Ill PatientsRisk Factors for β-lactam Non-Target Attainment in Critically Ill PatientsImplementation of a Prolonged Infusion Dosing StrategyImplementation of a Prolonged Infusion Dosing StrategyImplementation of a Prolonged Infusion Dosing StrategyChallenges and Solutions to ImplementationChallenges and Solutions to ImplementationWhat About the First Dose?First Dose Concentrations of Free Piperacillin in Hospitalized PatientsFirst Dose Concentrations of Free Piperacillin in Hospitalized PatientsApplication of PD Principles to a Specific Patient (β-Lactam)Where Do You Find the PK Data?Where Do You Find the PK Data?Estimate Systemic Clearance from the Patient’s Creatinine ClearanceClinical ApplicationClinical ApplicationClinical ApplicationClinical Application�Piperacillin/TazobactamClinical Application�Piperacillin/TazobactamConclusionsEarly Observations Regarding Drug Concentrations, Bacterial Killing & Resistance

Documents

Application of Pharmacokinetic and Pharmacodynamic ...• Establish a US database for antibiotic use and ... – Work-horse agents (pip/tazo, cefepime, meropenem) FDA-approved prior