Andrea Novelli

Dipartimento di Scienze della Salute Sezione di Farmacologia Clinica & Oncologia

Farmacologia dell’ospite critico: Terapia anti-infettiva nell’insufficienza d’organo e

cirtcolazione extra-corporea

• Angelini • Menarini Group • MSD

• Named • Valeas • Zambon Group

Transparency Declaration Honoraria or grant support received from:

Approach to the infected patient for the provision of optimal antibiotic therapy

Infected patient

Pharmacodynamic considerations

Time course of bacterial killing

Post-antibiotic effect Preclinical infection models

Pharmacokinetic considerations

Relative solubility

Protein binding Clearance mechanism

Clinical considerations

Site of infection Severity of illness Body composition Likely pathogens

Local resistance patterns

Clinical trial data

Modelling and simulation

Optimal antibiotic

dosing regimen

Onufrak NJ et al., Clin Ther, 2016

Causes of poor clinical evolution or therapeutic failure

• Lack of activity in empirical use (inappropriate treatment)

• Severe immunosuppression in the host • Concomitant diseases in advanced stage • Related adverse effects • Underdosing • Poor penetration in the infected tissue

Alvarez-Lerma F & Grau S, Drugs, 2012

PK-PD of antibacterials Antibacterial class Optimal PD

parameter PK-PD magnitude required for

efficacy

Theoretical Clinically needful Aminoglycosides Cmax/MIC 8 10 – 12 β-lactams (penicillins) T > MIC 50 > 70

Carbapenems T > MIC 30 > 45

Vancomycin AUC/MIC 400 > 400 Linezolid AUC/MIC 80 > 100 Daptomycin AUC/MIC 400-600 > 600

Tigecycline AUC/MIC ≥ 7 > 18

Adembri C & Novelli A, Clin Pharmacokinet 2009 ; Scaglione F & Paraboni L, Int J Antimicrob Agents, 2008; Alvarez-Lerma F & Grau S, Drugs, 2012; Canut A et al., Eur J Clin Microbiol Infect Dis, 2012.

• Low Vd • Predominant renal Cl • Low intracellular penetration

• High Vd • Predominant hepatic Cl • Good intracellular penetration

• ↑ Vd • Cl ↑ or ↓ dependent on renal

function

• Vd largely unchanged • Cl ↑ or ↓ dependent on

hepatic function

• β-lactams • Aminoglycosides • Glycopeptides • Daptomycin • Colistin

• Fluoroquinolones • Macrolides • Lincosamides • Linezolid •Tigecycline

General PK

Altered ICU PK

Examples

Hydrophilic antibiotics Lipophilic antibiotics

Roberts JA et al., Crit Care Med, 2009, mod

Severe sepsis/septic shock and impact on antibiotic pharmacokinetics

Cotta MO et al., Med Intensiva, 2015

Gentamicin

Relationship between albumin levels and volume of distribution or peak concentration

Bukkems LH et al., Int J Antomicrob Agents, 2018

Vd Cmax

Aminoglycoside use

Univariate analysis of risk associated with renal impairment

No renal impairment

n (%)

Renal impairment

n (%)

OR (95% CI)

P value

Sex Male

71 (57.3)

22 (73.3)

2.1 (0.85-4.97)

0.11

Nephrotoxic drug Yes

83 (66.9)

27 (90.0)

4.5 (1.27-15.52)

0.02

Prescription duration ≥ 3 days

26 (21.0)

16 (53.3)

4.3 (1.86-9.95)

< 0.001

N. administration per day > 1

9 (7.3)

6 (20.0)

3.2 (1.04-9.82)

0.04

Monitoring residual level Yes

45 (36.3)

17 (56.7)

2.3 (1.02-5.16)

0.04

Fraisse T et al., Age and Ageing 2014

• When prescribing aminoglycosides, it is important:

• to use the highest dose and the shortest as possible (<3 days) • limiting other nephrotoxic drugs • PK monitoring is to adapt aminoglycoside dose

β-lactam antibiotics

SHORT DOSING INTERVALS OR

CONTINUOUS INFUSION ?

PROLONGED INFUSION (3-4 h)

Betalactams

Study endpoints by treatment group

Endpoint Continuous infusion

Intermittent bolus

P

Plasma antibiotic concentration > MIC

18/22 (81.8%) 6/21 (28.6%) 0.001

Piperacillin/tazobactam 9/12 (75%) 4/11 (36%) Meropenem 8/8 (100%) 2/9 (22%) Clinical cure (test of cure date)

23/30 (76.7%)

13/30 (43.3%)

0.032

Dulhunty JM et al., Clin Infect Dis, 2013

Clinical cure rate for patients treated by prolonged-infusion or IV bolus dosing

* P < 0.05 prolonged-infusion vs IB dosing; n = number of patients in the subgroup

Abdul-Aziz MH et al., J Antimicrob Chemother, 2016

* P < 0.05 prolonged-infusion vs IB dosing; n = number of patients in the subgroup

Abdul-Aziz MH et al., J Antimicrob Chemother, 2016

30 day survival in patients treated by prolonged-infusion or IV bolus dosing

Piperacillin/tazobactam in severly ill patients

Meta-analysis of clinical cure according to infusion scheme

Rhodes NJ et al., Crit Care Med, 2018

>20%

<20%

Daptomycin

Mean pharmacokinetic parameters in healthy volunteers and severely ill patients

Parameter

6 mg/kg 8 mg/kg

Volunt.a

(6) Pts.c

(13) P* Volunt.b

(6) Pts.c

(7) P*

Cmax (mg/l) 86.4 55.7 < 0.01 106.2 85.1 = 0.05

t½ (h) 7.8 8.8 NS 7.3 8.6 NS

AUC (mg·h/l) 705 406.1 < 0.01 773.3 584.3 < 0.05

Cl (ml/h/kg) 8.6 18.0 < 0.05 10.1 20.4 < 0.05

Vd (l/kg) 0.096 0.22 < 0.01 0.102 0.25 < 0.01

( ) no. Cases * ANOVA test a Dvorchik BH et al., Antimicrob Agents Chemother, 2009 b Benvenuto M et al., Antimicrob Agents Chemother, 2006 c Falcone M, Venditti M, Novelli A, 21st ECCMID-27th ISC Milan, Italy, 2011

Cmax (mg/l) 86.4 55.7 < 0.01 106.2 85.1 = 0.05

AUC (mg·h/l) 705 406.1 < 0.01 773.3 584.3 < 0.05

Cl (ml/h/kg) 8.6 18.0 < 0.05 10.1 20.4 < 0.05 Vd (l/kg) 0.096 0.22 < 0.01 0.102 0.25 < 0.01

Considerations for Higher Doses of Daptomycin in Critically Ill Patients With Methicillin-Resistant S. aureus Bacteremia Falcone M, Russo A, Venditti M, Novelli A, Pai MP

• Methods. We evaluated the plasma pharmacokinetics (PK) and clinical outcomes of a cohort of critically ill patients treated with daptomycin 6–8 mg/kg/day for primarily Staphylococcus species–related infections. Data were modeled by population PK analyses, with Monte Carlo simulation to estimate the probabilities of effect and toxicity with standard and alternate dosing regimens.

• Results. Significantly lower daptomycin exposures were observed despite comparable

doses in a subset of patients (n = 13) with augmented clearance (CL). • This subset was significantly more likely to:

– be in severe sepsis or septic shock – have higher Sequential Organ Failure Assessment scores – have MRSA bacteremia.

In-hospital mortality was significantly higher (30.7% vs 10.8%) in patients with augmented daptomycin CL.

Use of an empiric fixed dose of 750 mg of daptomycin is predicted to achieve a comparable PTA with a lower probability of toxicity as compared to the use of 10 mg/kg in critically ill patients.

Clin Infect Dis, 57(11): 1568-1576, 2013

Tissue penetration of antimicrobial drugs and severity of organ failure

(AU

C tis

sue :

AU

C pl

asm

a,fre

e ra

tio)

1.4 1.2 1.0 0.8 06 0.4 0.2 0.0

SOFA Score 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Levofloxacin

Piperacillin

Zeitlinger Clin Pharmacokin 2007

Meropenem*

Concentration-time profile in plasma and ELF in 39 VAP patients

Parameter AUCplasma (mg·h/l)

AUCELF (mg·h/l)

Penetration ratio (%)

Mean 150.8 82.3 81.6

Median 130.9 35.0 25.42

SD 87.4 140.1 223.0

Percentile

10th 63.9 4.76 3.67

25th 90.14 12.52 9.0

50th 130.9 35.0 25.42

75th 189.3 92.1 70.14

90th 262.1 204.7 177.9

* 2g prolonged infusion Lodise TP et al., Antimicrob Agents Chemother, 2011

PK/PD of linezolid after intermittent or continuous infusion

0 6 12 18 24 30 36 42 48 54 60 66 72 780

5

10

15

20Intermittent Continuous

** **

*** ** **

*p< 0.05 **p< 0.01

* *

Time (h)

Con

cent

ratio

n (m

g/L)

4 mg/l = Susceptibility breakpoint

2 mg/l = for Staphylococcus spp. and Enterococcus spp.

Parameter Group I Group C

AUC/MIC 92.2 ± 45.2a 103.03 ± 39.8a

AUC/MIC (2mg/l) ≥ 85

5/8 (62.5%) 7/8 (87.5%)

Tfree > 2 mg/L MIC (%) > 80%

3/8 (40%)* 8/8 (100%)*

Adembri C et al., Int J Antimicrob Agents, 2007

a mean ± SD * P < 0.05 Group I vs Group C

Optimization of linezolid therapy in the critically ill: the effect of adjusted infusion regimens Taubert M, Zander J, Frechen S, Scharf C, Frey L, Vogeser M, Fuhr U, Zoller M

• RESULTS: Best target attainment according to T>MIC was observed for continuous infusions, followed by q6h, q8h and q12h.

• A substantially reduced target attainment was observed in patients with acute respiratory distress syndrome (ARDS)

• In patients without ARDS a dose of 1400 mg/day administered q6h or by continuous infusions provided an acceptable target attainment (e.g. cumulative fraction of response with regards to T>MIC≥93%)

• CONCLUSIONS: Irrespective of the regimen, 1200 mg/day linezolid might be insufficient for the treatment of ICU patients. Patients without ARDS might particularly benefit from q6h infusions with increased daily doses (e.g. 1400mg/day)

J Antimicrob Chemother, 72: 2304–2310, 2017

Antimicrobial drugs

PK considerations and adaptation of dosing regimen

Infection site PK alteration Potential change to dosing regimen

Blood Expanded Vd, enhanced Cl Provision of LD, increase frequency

Lung Impaired permeability Increase dose Soft tissue Contingent on body composition Increase dose in obesity Bone Impaired permeability Increase dose, duration of

therapy

CNS Impaired permeability Maximal dose

Onufrak NJ et al., Clin Ther, 2016

Cefepime-induced neurotoxicity

A clinical picture

• Renal dysfunction • Critical illness • Altered BBB • Older age • Drug overdose

• Altered mental status • Reduced consciousness • Confusion • Myoclonus • Aphasia • Agitation • Seizures

Risk factors Signs and symptoms

Payne LE et al., Crical Care, 2017

Vancomycin in ICU patients

Distribution of trough concentration versus age

Scaling: the marker size represents the number of patients in the vicinity of the spot Qian X et al., Int J Infect Dis, 2017

Scaling: the marker size represents the number of patients in the vicinity of the spot Qian X et al., Int J Infect Dis, 2017

Vancomycin in ICU patients

Distribution of trough concentration versus estimated glomerular filtration rate (eGFR)

Altered PK in Critically ill Patients

Increased Vd Decreased Cl Increased Cl Variable changes in Vd and/or Cl

Hypoalbuminaemia, leading to increased unbound drug

Renal hypoperfusion Augmented renal clearance

Extracorporeal interventions (eg RRT, ECMO)

Capillary leakage Acute kidney injury

Fluid resuscitation Renal/hepatic dysfunction

Third space loss

Wong G et al., BMC Infect Dis, 2014

Antibiotic Dosing in Critically Ill Patients Receiving CRRT: Underdosing is Overprevalent Lewis SJ and Mueller BA

• In CRRT patients our too-careful practice of “starting low and going slow” with antibiotic dosing to avoid the risk of toxicity may lead to an “overprevalence” of antibiotic underdosing and call for clinicians to strike a more aggressive antibiotic prescription in the intensive care unit

Semin Dial, 27(5):441-5, 2014

Amikacin in critical ill patients on CRRT

PTA of efficacy (Cmax/MIC ratio ≥8) and toxicity (Cmin2.5 mg/l)

Weight = 80kg and pathogens with 4 mg/l MIC Roger C et al., Antimicrob Agents Chemother, 2016

CVVH

Roger C et al., Antimicrob Agents Chemother, 2016

Amikacin in critical ill patients on CRRT

PTA of efficacy (Cmax/MIC ratio ≥8) and toxicity (Cmin2.5 mg/l) CVVHDF

Weight = 80kg and pathogens with 4 mg/l MIC

Population pharmacokinetics of daptomycin in adult patients undergoing continuous renal replacement therapy Xu X, Khadzhynov D, Peters H, Chaves RL, Hamed K, Levi M, Corti N

• AIM: the objective of this population pharmacokinetic (PK) analysis was to provide guidance for the dosing interval of daptomycin in patients undergoing continuous renal replacement therapy (CRRT)

• CONCLUSIONS: Q24 h dosing of daptomycin up to 12 mg/kg provides comparable drug exposure in patients on CVVHD and in those with CrCl ≥ 30 ml/min. Daily daptomycin use up to 8 mg/kg doses are appropriate for patients on CVVHDF, but higher doses may increase the risk of toxicity

Br J Clin Pharmacol, 83: 498–509, 2017

Cefepime in CVVHDF* simulated patients

Shaw AR et al., Seminars in Dialysis, 2016

* 25 ml/kg/hour effluent flow rate for the first 48 hours

Effects of continuous renal replacement therapy on linezolid pharmacokinetic/ pharmacodynamics: a systematic review Villa G, Di Maggio P, De Gaudio AR, Novelli A, Antoniotti R, Fiaccadori E and Adembri C

• Major alterations in linezolid PK/PD parameters might be expected in critically ill septic patients CRRT

• The optimal AUC/MIC ratio was reached for pathogens with an MIC of 4 mg/L in one study only

• Wide variability in linezolid PK/PD parameters has been observed across critically ill septic patients treated with CRRT. Particular attention should be paid to linezolid therapy in order to avoid antibiotic failure in these patients. Strategies to improve the effectiveness of this antimicrobial therapy (such as routine use of target drug monitoring, increased posology or extended infusion) should be carefully evaluated, both in clinical and research settings

Villa G et al., Critical Care (2016)

Population Pharmacokinetics and Dose Optimization of Teicoplanin during Venoarterial Extracorporeal Membrane Oxygenation Wi J, Noh H, Min KL, Yang S, Jin BH, Hahn J, Bae SK, Kim J, Park MS, Choi D, Chang MJ

• The presence of ECMO was associated with a lower central volume of distribution, and continuous renal replacement therapy (CRRT) was associated with a higher peripheral volume of distribution

• For severe infections, an optimal dose was an LD of 1,000 mg and an MD of 800 mg for ECMO patients not receiving CRRT and an LD of 1,200 mg and an MD of 1,000 mg for those receiving CRRT

• In conclusion, doses higher than the standard doses are needed to achieve fast and appropriate teicoplanin exposure during ECMO

Antimicrobial Agents and Chemotherapy, 61 (9): e01015-17, 2017

Dzierba AL et al., Critical Care, 2017 1 increased α1‐acid glycoprotein and decreased albumin concentrations 2 mostly affecting hydrophilic drugs

Critically ill patients and ECMO

Hydrophilic drugs Lipophilic drugs Volume of distribution (Vd) Low High Primary mode of clearance Renal Hepatic Log P Low High Potential effect of critical illness on PK

Increased Vd No change in Vd

Effect of ECMO on PK Increased Vd No change in clearance

Increased Vd Increased clearance

Dzierba AL et al., Critical Care, 2017

Factors affecting drug pharmacokinetic in patients receiving ECMO

Ha MA & Sieg AC, Pharmacotherapy, 2017

Age-dependent impact of extracorporeal membrane oxygenation (ECMO) prime volume on native blood volume

Sherwin J et al., Clin Ther, 2016

ECMO

Drug sequestration based on octanol/water partition coefficient and protein binding

Octanol/water partition coefficient

Protein binding

< 30% 30 – 70% > 70%

< 1 Minimal Minimal to moderate Moderate

1 – 2 Minimal to moderate Moderate Moderate to high

> 2 Moderate Moderate to high High

Ha MA & Sieg AC, Pharmacotherapy, 2017

Drug sequestration Dose adjustment

Minimal Dose adjustment likely not required

Moderate Increased dose, frequency or infusion rate may be required

High Increased dose, frequency or infusion rate likely required

ECMO

Manteinance dose adjustment based on drug sequestration

Ha MA & Sieg AC, Pharmacotherapy, 2017

What can be done in the clinical setting Pharmacological tricks for resistant strains

• Know the target pharmacokinetic/pharmacodynamic parameter for the specific drug in use

• Select the most appropriate administration modality according to pharmacokinetic/pharmacodynamic parameters

• Remember that standard susceptibility breakpoints may be inaccurate for the clinical scenario

• Maximize dosing, especially in severely ill patients, according to renal function, but limit the duration of therapy when possible

• Assess serum antimicrobial concentrations whenever possible Adembri C and Novelli A, PK-PD and potential for providing dosing regimens that are less vulnerable to resistance Clin Pharmacokinet, 2009