Lecture 1: Principles of antibiotic therapy






Russell E. Lewis
Associate Professor of Infectious Diseases
Department of Molecular Medicine
University of Padua


russelledward.lewis@unipd.it
https://github.com/Russlewisbo

Agenda


  • Key pharmacokinetic (PK) parameters of antibiotics
  • Pharmacodynamic (PD) patterns of antibiotics
    • How can PD be determined pre-clinically and optimised clinically?
  • How can PK/PD information individualise patient therapy?
  • Combination antibiotic therapy

Pharmacokinetics and Pharmacodynamics

G patient Patient factors PK Pharmacokinetics (PK) patient->PK PKPD PK/PD PK->PKPD drug Drug factors drug->PK intrinsic Intrinsic resistance PD Pharmacodynamics (PD) intrinsic->PD PD->PKPD acquired Acquired resistance acquired->PD inoculum Infection inoculum inoculum->PD microbiological Microbiological outcome PKPD->microbiological clinical Clinical outcome microbiological->clinical

Pharmacokinetics and Pharmacodynamics

G patient Patient factors PK Pharmacokinetics (PK) patient->PK PKPD PK/PD PK->PKPD drug Drug factors drug->PK intrinsic Intrinsic resistance PD Pharmacodynamics (PD) intrinsic->PD PD->PKPD acquired Acquired resistance acquired->PD inoculum Infection inoculum inoculum->PD microbiological Microbiological outcome PKPD->microbiological clinical Clinical outcome microbiological->clinical


Doses can be adjusted to improved the probability of microbiological outcomes!

Pharmacology of antimicrobials

G cluster_0 Pharmacokinetics (PK) cluster_1 Pharmacodynamics (PD) start Dosing regimen dosing Conc. vs. time in serum       Absorption Distribution Metabolism Elimination start->dosing tissue Conc. vs. time in tissue dosing->tissue infection Conc. vs. time at site of infection dosing->infection PD1 Pharmacological or   toxicological effect tissue->PD1 PD2 Antimicrobal effect vs. time infection->PD2

Antibiotic pharmacokinetics are described by concentration-time curves in serum

Definitions: Absorption-Bioavailibility


  • Bioavailability = AUCoral / AUCiv
    • The relative amount of drug that reaches the greater systemic circulation (i.e. following absorption & first-pass metabolism)
    • It is the proportion of drug that gets into the plasma vs. the same amount injected IV
  • In order to get to the plasma the drug goes through:
    • Ingestion
    • Transit through the GI tract
    • Absorption in small intestinal mucosa
    • First pass metabolism in small intestinal mucosa and liver
  • Bioavailability is used to determine:
    • The optimal route to achieve the therapeutic concentration
    • The amount of drug that needs to be given orally

Oral absorption in the GI tract



  • Stomach: Little absorption takes place for most drugs here (thick gastric mucous layer prevents significant drug absorption)
  • Small intestine: Transit time 4-6h; 6-7 m long x 2.5cm internal diameter
    • It has a larger surface area (30-35m2) than the stomach (~1m2), due to:
      • Jejunal plicae circulare (circular infoldings) - increases surface area 1.6x
      • Epithelial villi + microvilli: increases surface area 80-100x
  • Additional factors aiding absorption:
    • Peristalsis churns the intestinal contents, ensuring regular contact between drugs & epithelium
    • pH is 6-7 (weakly acidic) —> determines the ratio of drug molecules that exist in the ionized and unionized state

Factors affecting absorption

Transiting the GI epithelium into the bloodstream



Method of transit Substrate
Passive diffusion <500Da, lipophilic, not ionic
Facilitated diffusion <500Da, hydrophilic, ionic drugs
Active transport ionised drugs
Pinocytosis large molecules

Passive diffusion:
How most antibiotics are absorbed



  • Most common way for drugs to cross lipid membranes (esp. if <500 Da & not strongly ionic)

  • Lipophilic drugs (e.g. steroids) easily pass through lipid bilayers

    • The pKa/b of the weak acid/base determines their rate of uptake:

      • pKa = pH at which the acid is 50% ionised (A-) & 50% in the unionised form (AH)

      • pKb is the same for bases

    • Ionised drug can’t diffuse through a lipid bilayer

    • Unionised drug moves out of the gut lumen, (e.g., as an AH form)

      • Afterwards, ionic (A-) drug molecules combine with free H+ to become AH, which can then move across by passive diffusion

    • This re-equilibration between ionised & unionised forms continues along the entire length of the gut, with increasing amounts of the drug being gradually absorbed

A note about absorption and pH


  • Drugs are often ‘weak acids’ or ’weak bases,, and so their rate of passive uptake is determined by the ratio of ionised:unionised drug in the gut lumen:

    • Acidic drug (A): Can release a H+ ion, going from HA —> A- and H+

    • Basic drug (B): Can accept a H+ ion going from B —> BH+ (protonated base)

  • The pKa/b of the weak acid/base determines their rate of uptake:

    • pKa = pH at which the acid is 50% ionised (A-) & 50% in the unionised form (AH)

    • pKb is the same for bases

  • Ionised drug can’t diffuse through a lipid bilayer

  • Unionised drug moves out of the gut lumen, (e.g., as an AH form).

  • Afterwards, ionic (A-) drug molecules combine with free H+ to become AH, which can then move across by passive diffusion

  • This re-equilibration between ionised & unionised forms continues along the entire length of the gut, with increasing amounts of the drug being gradually absorbed

Facilitated diffusion





  • This is how lipophobic drugs cross cell membranes

  • This passively happens across an electrochemical concentration gradient.

  • Facilitated by solute carrier transporters (SLCs), which are present in the GIT, liver & kidney.

  • A note on SLCs:

    • SLCs are a group (>350 types) of transmembrane proteins that facilitate absorption of organic charged molecules up to 500 Daltons.
    • 2 types:
AT Organic Anion Transporter Carry deprotonated acids A-
OCT Organic Cation Transporter Carry protonated bases BH+

Active transport





- Primary AT: Mostly efflux of drugs from cells/specific body compartments

  • More important in limiting drug uptake

  • Secondary AT: This is how ionised drugs are absorbed

    • Carried out by SLCs (both OATs & OCTs) across pre-existing electrochemical gradients.
    • Drugs are co-transported e.g. with Na if moving into the intracellular space, or K if moving into the extracellular space
    • e.g. H+ and beta-lactam antibiotics

  • Pinocytosis: This is used to deliver large molecules across tissue barriers (e.g. blood-brain barrier)

    • Endocytosis: the cell membrane invaginates to non-specifically capture molecules present at the cell surface, pinching them off into a vesicle
    • Exocytosis is the opposite; vesicles are ejected from the cell

First pass metabolism





  • A small amount occurs in the gut epithelia before entering the portal vein. Most occurs in the liver.
  • As the drug passes through the liver some enters the hepatocytes and is metabolised by the CYP450 enzymes.





Oral is the new IV!

Bioavailability examples

Mechanism of Action Antimicrobial Bioavailability (%)
Cell wall
Penicillin V 60
Amoxicillin 80-85
Amoxicillin-clavulanate 80-85/64
Flucloxacillin 60
Pivmecillinam 75
Cephalexin (1g) 90
Cefuroxime axetil (2g) 40
Cefexime (3g) 40
Protein synthesis
Tetracycline >90
Doxycycline 80-90
Chloramphenicol 85
Erythromycin 20-25
Clarithromycin 55
Azithromycin 37
Clindamycin 90
Minocycline 95
Linezolid 100
DNA Synthesis
Ciprofloxacin 70
Levofloxacin 100
Moxifloxacin 90
Trimethoprim >90
Sulfamethoxazole >90
Misc.
Fosfomycin 33-44
Nitrofurantoin >95
Metronidazole >95

Caveats to bioavailibility



  • Amoxicillin: Absorption is saturable near 750 mg

    • higher doses are unlikely to be absorbed

  • Cephalexin: Maximum amount absorbed is ~4g/d

  • Macrolides: Not well-absorbed, but concentrate intracellularly

  • ‘UTI drugs’: Fosfomycin, trimethoprim, nitrofurantoin, pivmecillinam:

    • Concentrate in urine preferentially —> useful for UTI

Co-administration with food



Timing with food Antibiotic
“Empty Stomach”
(1h before or 2h after meal)		 Penicillin,
Flucloxacillin,
Azithromycin,
Fosfomycin
With food Amoxicillin-clavulanate (start of meal)
Erythromycin (just before or with),
Metronidazole (during or after),
Nitrofurantoin
No recommendation Amoxicillin,
Pivmecillinam,
Cephalexin,
Tetracyclines,
Clarithromycin,
Clindamycin,
Linezolid,
Co-trimoxazole,
Trimethoprim,
Quinolones

Pharmacology of antimicrobials

G cluster_0 Pharmacokinetics (PK) cluster_1 Pharmacodynamics (PD) start Dosing regimen dosing Conc. vs. time in serum       Absorption Distribution Metabolism Elimination start->dosing tissue Conc. vs. time in tissue dosing->tissue infection Conc. vs. time at site of infection dosing->infection PD1 Pharmacological or   toxicological effect tissue->PD1 PD2 Antimicrobal effect vs. time infection->PD2

Volume of distribution




  • The volume which appears to hold the drug if it was present in the body at the same concentration found in plasma

    • Estimated, not directly measured

  • Reported in litres (L) or litres per kilogram (L/kg)

  • Average plasma volume in adults is approximately 3L

\[ Vd=Dose/Cp \]

Visualizing the Vd





Clinical importance of Vd

Provides information on how much antibiotic is distributed in tissues vs. plasma → important for antibiotic selection

e.g., doxycycline, tigecycline do not achieve peak concentrations in bloodstream that surpass the MIC of many pathogens

Factors affecting Vd: Drug characteristics

  • Volume of distribution is affected by the physicochemical properties of the drug

  • Factors that favour low Vd: high water solubility, high protein binding, decreased tissue binding → converse is also true

Factors affecting volume of distribution


  • Ease of crossing phospholipid bilayers

  • Interactions with lipid tissue (lipophilic = more distribution into fat)

  • Interactions with protein

Examples of how drug factors affect Vd


Lipophilic vs. hydrophilic

  • More lipophilic antibiotics diffuse faster out of plasma into tissue; they favour distribution to lipid-rich tissues

  • If the drug has a net negative charge, it can still leave capillaries through endothelial fenestrations but further tissue penetration depends on:

    • Interstitial fluid pH

    • Drug pKa

    • Presence of OAT/OCT carriers

Plasma protein binding

There are several plasma protein that bind drugs with weak electrical polar bonds allowing

  • Storage/transport of “bound” drug

  • Quick release of the drug to become “free drug”

  • Only free drug distributes out of plasma and has pharmacological effect

Factors affecting volume of distribution, cont.



  • Tissue protein binding (e.g., muscle)

    • This ‘drains’ the drug from plasma

    • This will in turn reduce amount of plasma free drug

  • Tissue binding sites

    • Antibiotic example?

Factors affecting Vd: Vascular perfusion



  • Drugs diffuse from capillary beds—> interstitial —> intracellular space

  • More vascular tissue (e.g., heart, lung, kidney) have greater drug delivery vs. skin, bone, fat

  • Capillary leakiness:

    • Some capillary bed epithelial cells are fenestrated by 60-80nm diameter pores (e.g. intestinal, endocrine, pancreatic and kidney).

    • In others endothelial cells are separated by slit junctions or large intercellular gaps to allow large movements of molecular material (Liver, bone marrow, lymph nodes, spleen).

    • These ‘leak’ points facilitate access to the interstitial fluid.

Sepsis alters the Vd of antibiotics

Release of inflammatory mediators causes damage to the vascular endothelium, resulting in expansion of extravascular space (increased volume of distribution)

Factors affecting Vd:
Body size and chronic diseases

Anatomically privileged sites-1


Ventilator-associated pneumonia


Anatomically privileged sites-2


How is the drug metabolised,
concentrated, or excreted?

Which antibiotics are excreted in urine?

Pharmacology of antimicrobials

G cluster_0 Pharmacokinetics (PK) cluster_1 Pharmacodynamics (PD) start Dosing regimen dosing Conc. vs. time in serum       Absorption Distribution Metabolism Elimination start->dosing tissue Conc. vs. time in tissue dosing->tissue infection Conc. vs. time at site of infection dosing->infection PD1 Pharmacological or   toxicological effect tissue->PD1 PD2 Antimicrobal effect vs. time infection->PD2

Antibiotic elimination



  • The “drug out” part of pharmacokinetics

  • Metabolism of the drug determines how the drug is excreted

    • Not all drugs are metabolised before excretion (excreted unchanged)

  • Elimination starts as soon as drug first enters the liver; it is an ongoing process that does not ‘wait’ until the drug has finished distributing equally throughout the body

Metabolism


  • Chemical conversion of a drug into a form amenable to excretion

  • This is achieved through 2 main ways: Phase I & II reactions

Phase I Chemical reactions Effect

Oxidation/Reduction

Hydrolysis

 Introduce/unmask polar groups on drug: (-OH or -NH2)
Phase II Conjugation

Addition of polar groups to drug:

  • Glucuronate (glucuronidation)

  • Glutathione (Glutathione conjugation)

  • Glycine (Glycine conjugation)

  • Sulphate (Sulphation)

  • Acetyl (Acetylation)

  • Methyl (Methylation)

Metabolism



  • These reactions enhance the drug’s ionic charge, which makes them easier to excrete

  • This is essential for lipophilic drugs, which will otherwise simply diffuse back across the renal cell membranes, then back into the plasma. These reactions are responsible for the “First-Pass metabolism

  • The main sites of metabolism are liver, kidneys & lung, but these enzymes are present at various levels in most tissues

  • Not all drugs undergo both phases: e.g., some drugs undergo Phase II directly, depending if they already possess -OH, -NH or COOH groups

Phase I: Cytochrome P450 enzymes



  • Definition: A large family of >50 hepatic enzymes that primarily oxidise drugs

  • Site: mostly on hepatocyte endoplasmic reticulum

  • They are responsible for most Phase I metabolism

  • Main CYPs (~90% drug metabolism) are: 3A4, 2D6, then 1A2, 2C9, & 2C19. 3A4 metabolises 50% of drugs

  • Nomenclature:

CYP speed vs generalism


  • CYP enzymes are “generalists”

    • Each isoenzyme has a wide array of drug ‘targets’ ( drugs are metabolised by several different enzymes)

  • This “generalisability” comes at the expense of speed; compared to one specialised enzyme that could break down one specific molecule, CYP enzymes are much slower

  • This versatility contributes to the variation in elimination half-lives seen for drugs

Factors affecting CYP enzyme activity



Factor Explanation
Age Declines with age
Sex Males have larger livers —> more rapid metabolism of drugs
Genetics CYP isoenzymes exhibit high rates of genetic polymorphism, which can affect the efficiency with which the individual can process the enzyme’s drug targets.
Cardiac output Less cardiac output —> less drug delivered to the liver per minute —> slower rate of metabolism |
Drug interactions See next slide

CYP enzyme induction/ inhibition


  • Certain drugs can affect expression of CYPs through:

    • Induction (Increased transcription/translation, or slower degradation)

    • Inhibition (Increased degradation)

    • This will lead to increased/decreased metabolism of the CYP enzyme’s substrate drugs, respectively

  • CYP enzyme inducers

    • 200 drugs known to be inducers

    • Induction takes 1-2 weeks on average

      • NB: Rifampicin induction starts within 24h, and has full induction (of CYP3A4) at 72h
      • E.g. PC BRAS drugs: Phenytoin, Carbamazepine, Barbiturates, Rifampicin, Alcohol (Chronic use), Sulfonylureas

  • CYP enzyme inhibitors {.smaller}


    • These increase the plasma level of substrate (victim) drugs due to binding or modification of CYPs that are bound by inhibitors (perpetrators)

    • Occurs much more quickly (~1-3 days)

    • How long will inhibition last once perpetrator is removed?

      • Competetive inhibitor: Determined by the half life the perpetrator drug-i.e. should abate after 3-5 half lives
      • Non-competitive inhibitor: Determined by turnover of the enzyme (e.g., 20-50h)

Excretion



  • Drug metabolism has 3 outcomes:

    • Drug is metabolised to inactive metabolite

    • Drug is metabolised to metabolite with therapeutic activity (e.g., metronidazole’s metabolite is still active)

    • ‘Prodrug’ is metabolised to active form (e.g. codeine to morphine)

  • Regardless of what outcome occurs, the drug is ionised, making excretion easier

  • Main sites of excretion:

    • Kidney (main site; covered below)

    • GI tract

    • Lung (in breath)

    • Skin (sweat, tears)

Renal excretion



Occurs in 3 stages

  • Glomerular filtration

    • ~20% Renal blood flow serves the glomerulus

    • Free drug & metabolites passively diffuse across into Bowman’s capsule.

  • Proximal tubular secretion

    • ~80% Renal blood flow serves the rest of the nephron

    • OAT and OACTs actively transport polar molecules into urine

  • Distal tubular reabsorption

    • As water is reabsorbed along the tubule, urine drug/metabolite concentration increases

    • Lipophilic molecules will then diffuse down their concentration gradient, across the uroepithelium back into the plasma

Clearance: A key pharmacokinetic parameter


  • Drug elimination from the body

    • Described by volume of blood removed of drug unit per time

      • For example: If drug A as a clearance of 39 L/h or 650 ml/min, then every minute, the equivalent of 0.65L of the Vd can be “cleansed” of drug A

    • Unit of measure mL/min or L/hr

    • Clearance is affected by

      • Patient disease, organ function, genetic, interactions with other drugs

Example: Meropenem in septic patients

Integrating Volume of distribution and clearance


  • Vd and CL are physiologically-based

    • A change in fluid status or distribution can affect Vd

    • A change in kidney of liver function can affect drug CL

  • However, these parameters do not directly interact with each other

    • A change in Vd does not change clearance and vice versa

Why is this distinction important?


  • Vd is useful fo calculating the loading dose of a drug

  • CL is useful for calculating the maintenance dose of a drug

    • CL is not used to calculate the loading dose (or initial dose) because it is not affected by the volume of distribution

Key pharmacokinetic parameter: Elimination rate



  • What is kel?

    • Rate of drug removed per unit of time
    • Calculated parameter: Unit of measure: reciprocal time (hr-1)

\[ Kel =CL/Vd \]

Key pharmacokinetic parameter:
Half life


  • Time it takes for plasma concentration or amount in body to be reduced by 50%

  • It is a calculated parameter

    • Function of clearance and volume of distribution

      • Unit of measure = time (hours, minutes, days)

\[ t_{½}= \frac{0.693}{k_{el}} \] \[ t_{½}= \frac{0.693\ast Vd}{Cl} \]

Half-life examples



Drug Half life (h) Dosing regimen
Amoxicillin 1 8-hourly
Doxycycline 12 daily



Most of a drug will be eliminated without repeat dosing by 5-6 half-lives

β-lactam examples


Antimicrobial Metabolism Excretion Vd (L); Bold = from DrugBank Clearance (ml/min) Half life (h)
Penicillin G 16-30% Penicilloic acid; Small amounts to 6-aminopenicillanic acid & active metabolites Renal: Most Hepatic: some Faecal: some 35.4 560 0.4 - 0.9
Penicillin V 35-70% Penicilloic acid; Small amounts to 6-aminopenicillanic acid & active metabolites Renal: 25 35.4 N/A 0.5
Amoxicillin 7 metabolites (M1-7) Renal: 74 27.7 355 1
Nafcillin None Hepatic <30 28 N/A 0.5 - 1
Flucloxacillin Hepatic: mostly 13 0.75 - 1
Pivmecillinam Renal: 50% / 6h Hepatic: partially 14 - 28 1.2
Piperacillin Renal: Most Hepatic: some 14 - 21 32 - 41 0.6 - 1.2
Tazobactam M1 metabolite Renal: 80% / 20% M1 18.2 48 - 84 0.7 - 1.2
Cephalexin None Renal: 99%/24h 5.2 - 5.8 1
Cefazolin None Renal: 80%/24h 9 - 15 1.9
Cefuroxime axetil None; Axetil to Acetic acid + Aldehyde Mostly (unchanged) 50 125 - 148 1.3
Ceftriaxone Renal: 33 - 67 Hepatic: 33 - 67 5.8 - 13.5 5.8-8.7
Ceftazidime ~85%/24h 15 - 20 115 1.5 - 2.8
Cefepime <1% ~85% 18 120 2
Ceftolozane Nil ~100% 13.5 57 - 110 2.8 - 3.1
Ceftaroline fosamil By plasma phosphatase to active ceftaroline Renal: Mostly (unchanged) Faecal: <6% 20 1.6
Ertapenem 50% to inactive metabolite Renal: 80% Faecal: 10% 8.4 28 4
Meropenem Negligible 70% 17.5 - 24.5 1
Aztreonam ~11% to inactive compound 100%/12h 12.6 91 ~1.7

Non β-lactam antibiotics


Target Antimicrobial Metabolism Elimination Vd (L); Bold = from DrugBank Clearance (ml/min) Half life (h)
Cell wall Vancomycin Negligible 75% 28 - 70 68 6
Dalbavancin <25% to less active metabolites Renal: 33% unchanged; 12% metabolites Faecal: 20% 14 0.9 346
Protein Synthesis Gentamicin Negligible 70% 20 - 26 as per CrCl (~100) 1.25
Amikacin Negligible 94% 24 as per CrCl (~100) 2
Doxycycline Renal: 40% Hepatic: significant 52.5 58
Tigecycline 10% 490 - 630 27-43
Clarithromycin predominantly by 3A4 30% 191 - 306 3-4
Azithromycin Renal: 6% Hepatic: significant 2177 630 68
Clindamycin to 2 inactive metabolites, by 3A4 and 3A5 Renal: 10% Faeces: 3.6% 43 - 74 250 3
Linezolid to 2 inactive metabolites, unclear mechanism Renal: 30% unchanged, 50% metabolites Faeces: 6% 40 - 50 100-200 5-7
DNA Synthesis Ciprofloxacin ~20% to 4 metabolites Renal: 45% Faeces: 62% 140 - 213 350 - 630 4
Levofloxacin <5% as 2 metabolites Renal: 87% Faeces: 4% 89 - 112 143 - 227 6 - 8
Moxifloxacin 50%; Phase I conjugation/ sulphation Renal: 20% Faeces: 25% 119 - 189 200 11.5 - 15.6
Trimethoprim 90% to 2 inactive metabolites; by 2C9/3A4 Renal: ~55% Hepatic: ~15% 25.2 50 - 90 8 - 10
Sulfamethoxazole To inactive metabolites; CYP2C9; Renal: 85% 13 20 10
Other Daptomycin Some to inactive metabolites Renal: 80% Faecal: 6% 7 ~9.8 7.5 - 9
Colistin Tissue metabolism: 20% Renal: 80% 9 - 12 5
Fosfomycin Nil Renal: 100% 21 - 30 283 5.7
Nitrofurantoin ~1.3% to Aminofurantoin Renal: 90% 42 278 - 323 ~0.76
Metronidazole To 5 metabolites; 1 is active also Renal: ~70% Faeces: ~10% 36 - 77 10 6 - 10

Key pharmacokinetic parameter:
Area under the curve (AUC)


  • Total drug exposure over time, expressed as mg∙h/L

  • Dependent on the dose administered and rate of elimination

  • Calculated by adding up or integrating the amounts of drug eliminated in discreet time intervals, from zero (time of the administration of the drug) to a defined time-e.g., 24 hours using trapezoidal sections

  • Can simplistically be interpreted as average dur exposure over a dosing interval:

    • e.g., an AUC0-24h of 48 mg∙h/L for a dosing regimen administered every 24 hours would represent an average concentration over the 24 hour dosing interval of 2 mg/L (48 mg∙h/L divided by 24 hr=2 mg∙L average over 24 hours).

Renal impairment



  • Well established dosing adjustments for chronic renal insufficiency, based on estimated renal function (Cockcroft-Gault, MDRD etc…)
  • However, accuracy of these equations are limited in severely ill, high-low body mass, prolonged hospitalization

\[ CrCl\text{ estimated}=\left(\frac{\left(140-age \right)\ast \left( Weight, kg \right)}{72\ast SeCr} \right)\ast 0.85 \text{ if female} \]


  • Drug specific guidance for patient on dialysis- consult drug references or primary literature

Be careful about prematurely reducing
antibiotic doses based on serum creatinine


  • Antibiotic renal dose adjustments in drug labels are based on patients with chronic kidney disease

  • Renal impairment is acute, not chronic, in up to 50% of patients with infection

  • Renal impairment frequently resolves within the first 48 hours

  • Creatine-based equations for estimates of CrCl are based on steady-state conditions, and not as accurate in acute kidney injury

  • Decreases in SeCr are delayed with respect to injury resolution

  • Renal dose reduction in the first 48 hours of therapy may result in underdosing of antibiotics

Augmented renal clearance





  • Common clinical scenarios:

    • “hyperdynamic state” with Gram-negative sepsis

    • vasoactive medications to support blood pressure

    • large-volume fluid resuscitation

  • Most common populations with augmented clearance:

    • Younger patients (i.e. trauma)

    • Patients with severe burns, preganancy, sepsis

  • Combined with increased Vd, often leads to inadequate antibiotic exposures


Pharmacology of antimicrobial therapy

G cluster_1 Pharmacodynamics (PD) cluster_0 Pharmacokinetics (PK) start Dosing regimen dosing Conc. vs. time in serum       Absorption Distribution Metabolism Elimination start->dosing tissue Conc. vs. time in tissue dosing->tissue infection Conc. vs. time at site of infection dosing->infection PD1 Pharmacological or   toxicological effect tissue->PD1 PD2 Antimicrobal effect vs. time infection->PD2

Laws of antimicrobial pharmacodynamics


  • The shape of the antibiotic concentration versus antimicrobial effect curve is important for dosing

How does PD differ from MICs?

  • Good indicators of potency

  • Tell us nothing about time course of antibiotic activity

  • Nothing about dose-response relationship




  • How does the rate and extent of bacterial killing by an antibiotic change at concentrations near and above the MIC?

  • The shape of the curve affects drug dosing strategies

How to define concentration-time-kill
effect curves?

Interpreting time-kill curves

Post antibiotic effect (PAE)

How do you translate?

Integration of PK/PD concepts

Common PK/PD indices

Mutant selection window (MSW)
and mutant prevention concentration (MPW)


Dose fractionization study

Dose fractionization study-interpretation

In vitro hollow-fiber models


  • Simulate antibiotic dosing regimens in vivo

  • Bacterial inocula is retained in dialysis cartridge.

    • Media containing drug is pumped through cartridge.

    • The pump rate simulates antibiotic half-life

  • Study antibiotic performance over a wide range of inocula (important for resistance studies)

  • No impact of protein binding, immune system, etc

In vitro hollow-fiber models


Example of dose fractionization study results

Ciprofloxacin example


From mice to humans



PK/PD parameters predictive
in animals and humans


Antibiotic class Optimal PK/PD index PK/PD magnitude for bacterial killing PK/PD index for clinical efficacy PK/PD index for resistance suppression
Aminoglycosides AUC0-24/MIC AUC0-24/MIC 50-100 Cmax/MIC Cmax/MIC ≥ 20
Cmax/MIC - Cmax/MIC > 8 Cmax/MIC ≥ 30
Penicllins T>MIC; Cmin/MIC ≥ 40-50% T>MIC ≥ 40-50% T>MIC ≥ 40-50% T>MIC
Cephalosporins T>MIC ≥ 40% T>MIC T>MIC ≥ 40% T>MIC
tMSW ≤ tMSW 45%
Carbapenems T>MIC ≥ 40% T>MIC ≥ 40-50%T>MIC ≥ 40% T>MIC
Quinolones AUC0-24/MIC AUC0-24/MIC 30-200 AUC0-24/MIC 35-250 AUC0-24/MIC 100-200
Cmax/MIC Cmax/MIC ≥ 8 Cmax/MIC ≥ 8 ≤ tMSW 30%

PK/PD parameters predictive
in animals and humans, cont.


Antibiotic class Optimal PK/PD index PK/PD magnitude for bacterial killing PK/PD index for clinical efficacy PK/PD index for resistance suppression
Vancomycin AUC0-24/MIC AUC0-24/MIC 86-460 AUC0-24/MIC 400-600 AUC0-24/MIC 200
Linezolid AUC0-24/MIC AUC0-24/MIC 50-80 - -
T>MIC ≥ 40% T>MIC ≥ 85% T>MIC -
Daptomycin AUC0-24/MIC AUC0-24/MIC 388-537 AUC0-24/MIC 200
Cmax/MIC Cmax/MIC 59-94 - -
Fosfomycin %T>MIC m

%T>MIC> 70%;

AUC0-24/MIC ≥ 24

 - -
Colistin AUC0-24/MIC AUC0-24/MIC 50-65 - -

Laws of antimicrobial pharmacodynamics


  • The shape of the antibiotic concentration versus antimicrobial effect curve is important for dosing

  • Only free-drug (non-protein bound fraction) is microbiologically active

Only non-protein-bound drug is
microbiologically active

Laws of antimicrobial pharmacodynamics


  • The shape of the antibiotic concentration versus antimicrobial effect curve is important for dosing

  • Only free-drug (non-protein bound fraction) is microbiologically active

  • A higher MIC will diminish the effect of a fixed dose

Effect of increasing MIC



Laws of antimicrobial pharmacodynamics


  • The shape of the antibiotic concentration versus antimicrobial effect curve is important for dosing

  • Only free-drug (non-protein bound fraction) is microbiologically active

  • A higher MIC will diminish the effect of a fixed dose

  • Administering a fixed dose of drug to many patients (even on a mg/kg basis) results in wide variability in exposure

Variability of antibiotic pharmacokinetics


Which patients are in clinical trials?

Types of pharmacokinetic analysis

Types of pharmacokinetic analysis


Differences Non-compartmental analysis
(NCA)		 Compartmental model
Model independence NCA does not assume any specific physiological or anatomical model. It treats the body as a system with no distinct compartments Compartmental analysis assumes that the body can be represented by one or more compartments that correspond to different tissues or groups of tissues with similar blood flow and drug affinity
Simplicity Basic pharmacokinetic principles to calculate parameters such as area under the curve (AUC), clearance (CL), volume of distribution (Vd), and half-life (t1/2) This approach is more complex and involves fitting the concentration-time data to a predefined model (e.g., one-compartment, two-compartment models)
Data requirements Requires precisely times samples to calculate PK parameters Compartmental analysis initially requires more frequent sampling and more data points to accurately define the model parameters
Parameters Pharmacokinetic parameters derived from NCA are descriptive and do not describe the process of drug distribution and elimination mechanistically The parameters derived from compartmental analysis (such as rate constants for absorption and elimination) are mechanistic and describe how the drug moves between compartments and is eliminated from the body
Use cases Early drug development for preliminary PK assessment and in situations where detailed modeling is not feasible or necessary Compartmental analysis is used when a more detailed understanding of the pharmacokinetics of a drug is needed, such as for dose optimization, to understand drug-drug interactions, or to predict concentrations in various tissues





How can we address pharmacokinetic uncertainty?

Monte Carlo Simulation


  • Invented by John Neumann and Stanislaw Ulam during WWII

  • Computation algorithms that rely on repeated sampling to obtain a numercial result

  • Technique used to determine probability that antibiotic dosing regimen will acheive PD target with maximum effect

    • Allows us to simulate spread of values (e.g., AUC/MIC, % Time>MIC, VD, CL) that would bea seen in a large population or special populations (e.g., critically-ill with sepsis, CRRT)

  • Makes it possible to explore and test antimicrobial regimens based on the probability of what is most likely to happen and to estimate the probability of achieving PK:PD targets.

Monte-Carlo Simulation: Data inputs

  • Pharmacokinetic parameters
  • Measure of central tendency and associated dispersion (variance and covariance)
  • Pharmacodynamic target: AUC/MIC, % T> MIC, Cmin/MIC
  • MIC distribution
  • Microbiologic surveillance data MIC range

Example of a β-lactam with 1-compartment pharmacokinetics

Model-informed precision dosing (MIPD)

How does this work?

Antibiotic dosing uncertainty-part of medicine?




Medicine is a science of uncertainty,

and an art of probability


-Sir William Osler, M.D. 

1849-1919

References


Unless otherwise cited, all figures were created using www.biorender.com


Abdul-Aziz M, Lipman J, Mouton J, Hope W, Roberts J. Applying Pharmacokinetic/Pharmacodynamic Principles in Critically Ill Patients: Optimizing Efficacy and Reducing Resistance Development. Seminars in Respiratory and Critical Care Medicine 2015;36:136–53. https://doi.org/10.1055/s-0034-1398490.
Craig WA. Pharmacokinetic/pharmacodynamic parameters: Rationale for antibacterial dosing of mice and men. Clinical Infectious Diseases 1998;26:1–10. https://doi.org/10.1086/516284.
Crass RL, Rodvold KA, Mueller BA, Pai MP. Renal Dosing of Antibiotics: Are We Jumping the Gun? Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America 2019;68:15961602. https://doi.org/10.1093/cid/ciy790.
Drusano GL. Antimicrobial pharmacodynamics: critical interactions of ’bug and drug’. Nature Reviews Microbiology 2004a;2:289–300. https://doi.org/10.1038/nrmicro862.
Drusano GL. Antimicrobial pharmacodynamics: critical interactions of ’bug and drug’. Nature Reviews Microbiology 2004b;2:289–300. https://doi.org/10.1038/nrmicro862.
Forrest A, Nix DE, Ballow CH, Goss TF, Birmingham MC, Schentag JJ. Pharmacodynamics of intravenous ciprofloxacin in seriously ill patients. Antimicrobial Agents and Chemotherapy 1993;37:10731081.
Gilbert DN. Urinary tract infections in patients with chronic renal insufficiency. Clinical journal of the American Society of Nephrology: CJASN 2006;1:327–31. https://doi.org/10.2215/CJN.01931105.
Huttner A, Bielicki J, Clements MN, Frimodt-Møller N, Muller AE, Paccaud J-P, et al. Oral amoxicillin and amoxicillin clavulanic acid: Properties, indications and usage. Clinical Microbiology and Infection 2020;26:871–9. https://doi.org/10.1016/j.cmi.2019.11.028.
Kothekar AT, Divatia JV, Myatra SN, Patil A, Nookala Krishnamurthy M, Maheshwarappa HM, et al. Clinical pharmacokinetics of 3-h extended infusion of meropenem in adult patients with severe sepsis and septic shock: implications for empirical therapy against Gram-negative bacteria. Annals of Intensive Care 2020;10:4. https://doi.org/10.1186/s13613-019-0622-8.
Lodise TP, Sorgel F, Melnick D, Mason B, Kinzig M, Drusano GL. Penetration of meropenem into epithelial lining fluid of patients with ventilator-associated pneumonia. Antimicrobial Agents and Chemotherapy 2011;55:1606–10. https://doi.org/10.1128/AAC.01330-10.
Mouton JW, Brown DFJ, Cantón R, Giske CG, Ivanova M, MacGowan AP, et al. The role of pharmacokinetics/pharmacodynamics in setting clinical MIC breakpoints: the EUCAST approach. Clinical Microbiology and Infection 2012;18:E37–45. https://doi.org/10.1111/j.1469-0691.2011.03752.x.
Theuretzbacher U. Pharmacokinetic and pharmacodynamic issues for antimicrobial therapy in patients with cancer. Clinical Infectious Diseases 2012;54:1785–92. https://doi.org/10.1093/cid/cis210.