Residency · Residency · Infectiousdisease

Principles of Antimicrobial Pharmacology - PK/PD

Introduction and Framework

Why PK/PD Matters in Infectious Disease

The era of prescribing antibiotics based solely on minimum inhibitory concentration data has long passed. Contemporary infectious disease practice demands a sophisticated understanding of how pharmacokinetic and pharmacodynamic principles intersect to determine clinical outcomes. The relationship between drug exposure and microbiologic effect is the cornerstone upon which rational antimicrobial dosing is built, and ignoring these principles leads directly to treatment failure, excess mortality, and the emergence of resistant organisms.

Pharmacokinetic/pharmacodynamic optimization is particularly critical in critically ill patients, who represent the population most vulnerable to subtherapeutic dosing and most likely to harbor resistant pathogens. Patients in the intensive care unit frequently exhibit profoundly altered pharmacokinetics due to augmented renal clearance, massive fluid shifts with third-spacing, hypoalbuminemia, and the use of extracorporeal circuits such as ECMO and continuous renal replacement therapy. In these patients, standard dosing regimens derived from healthy volunteer studies are often grossly inadequate, and a failure to appreciate this reality results in measurably worse outcomes.

Core Definitions

Pharmacokinetics describes what the body does to the drug and encompasses the four classical processes of absorption, distribution, metabolism, and elimination. These processes collectively determine the concentration-time profile of a drug in the body and at the site of infection. Pharmacodynamics, by contrast, describes what the drug does to the organism. This encompasses kill kinetics, the post-antibiotic effect (the duration of persistent suppression of bacterial growth after drug concentrations fall below the MIC), and the mutant prevention concentration, which defines the threshold above which even first-step resistant mutants are suppressed.

The minimum inhibitory concentration, or MIC, is the lowest concentration of an antibiotic that prevents visible bacterial growth at 18 to 24 hours. While the MIC is the most widely reported susceptibility parameter, it is fundamentally a single-point estimate with inherent variability of plus or minus one doubling dilution. The MIC provides no information about kill kinetics, does not account for protein binding, and cannot predict clinical outcomes without integration into the broader PK/PD framework.

Pharmacokinetic Parameters in Detail

Absorption and Bioavailability

Oral bioavailability varies dramatically across antibiotic classes and is a critical consideration for intravenous-to-oral step-down strategies. Fluoroquinolones achieve oral bioavailability of approximately 90 to 100 percent, making them ideal candidates for early oral conversion. Linezolid is essentially 100 percent bioavailable orally, rendering the intravenous formulation pharmacokinetically unnecessary except when the gastrointestinal tract is non-functional. Amoxicillin achieves 75 to 90 percent bioavailability, while oral cephalosporins generally achieve only around 50 percent, which limits their utility for serious infections requiring high serum concentrations.

The criteria for a safe and effective intravenous-to-oral switch include a functioning gastrointestinal tract, hemodynamic stability, documented susceptibility of the causative organism, and the availability of an oral agent with adequate bioavailability for the infection being treated. Drug interactions affecting absorption deserve careful attention. Fluoroquinolones are chelated by divalent cations including calcium, magnesium, aluminum, and iron, leading to dramatic reductions in absorption. Posaconazole suspension requires an acidic gastric environment for adequate absorption, meaning that concurrent acid suppression therapy substantially reduces drug levels.

Volume of Distribution (Vd)

The volume of distribution is a theoretical parameter that relates the total amount of drug in the body to its plasma concentration. Hydrophilic agents, which include beta-lactams, aminoglycosides, and vancomycin, exhibit low volumes of distribution in the range of 0.1 to 0.3 L/kg. These drugs distribute predominantly into the extracellular fluid compartment and are therefore profoundly affected by changes in fluid status. In sepsis, the volume of distribution for hydrophilic drugs increases by 20 to 100 percent due to capillary leak and fluid resuscitation, leading to subtherapeutic plasma and tissue concentrations when standard doses are administered.

Lipophilic agents such as fluoroquinolones, macrolides, and tigecycline exhibit high volumes of distribution in the range of 1 to 5 L/kg. These agents distribute extensively into tissues and intracellular compartments and are therefore less affected by the fluid shifts characteristic of critical illness. This pharmacokinetic property explains why tigecycline, despite achieving relatively low serum concentrations, penetrates well into soft tissues and intra-abdominal compartments.

Protein Binding

Only the free, unbound fraction of a drug is pharmacologically active, capable of crossing membranes, reaching the site of infection, and interacting with bacterial targets. Highly protein-bound drugs include ceftriaxone (85 to 95 percent bound), daptomycin (92 percent), and ertapenem (95 percent). For these agents, the total drug concentration reported by the laboratory substantially overestimates the pharmacologically relevant exposure.

Hypoalbuminemia, defined as an albumin level below 2.5 g/dL, increases the free fraction of highly protein-bound drugs. This alteration affects both efficacy and toxicity, as a larger free fraction means more drug is available to exert pharmacologic effect but also more drug is available for clearance. When interpreting PK/PD targets, the relevant parameters are the free Cmax (fCmax) and free AUC (fAUC), not total drug concentrations. This distinction is particularly important for critically ill patients with hypoalbuminemia, where total drug levels may appear adequate while free drug exposure is actually subtherapeutic or, conversely, where modest total drug levels may be associated with toxicity due to elevated free fractions.

Clearance and Elimination

Understanding the primary route of elimination for each antibiotic is essential for dose adjustment in organ dysfunction. Aminoglycosides, most beta-lactams (with the notable exceptions of ceftriaxone and nafcillin), vancomycin, and fluconazole are predominantly renally cleared and require dose adjustment in renal impairment. Macrolides undergo hepatic metabolism primarily through CYP3A4, metronidazole is hepatically metabolized, and voriconazole is subject to CYP2C19 polymorphisms that create enormous interpatient variability in drug levels.

Augmented renal clearance, defined as a creatinine clearance exceeding 130 mL/min, is a frequently under-recognized phenomenon that is common in young trauma, burn, and sepsis patients. These patients clear renally eliminated drugs far more rapidly than expected, resulting in subtherapeutic drug levels despite what appear to be adequate or even generous standard doses. Failure to recognize augmented renal clearance is an increasingly identified cause of treatment failure in the ICU population.

<image>A detailed pharmacokinetic curve diagram showing a concentration-time profile after IV bolus administration. The diagram should label Cmax, Cmin (trough), AUC (shaded area under the curve), the MIC as a horizontal dashed line, and T>MIC as the time interval where the curve is above the MIC line. Include annotations showing the three PK/PD indices: peak/MIC ratio, AUC/MIC ratio (shaded differently), and %T>MIC. Use a clean, clinical textbook style with black axes, a blue concentration curve, and red MIC line.</image>

PK/PD Kill Indices

Time-Dependent Killing (T>MIC)

Time-dependent killing is the pharmacodynamic pattern exhibited by beta-lactams, carbapenems, cephalosporins, and macrolides (though macrolides are generally bacteriostatic). For these agents, the rate and extent of bacterial killing are determined not by how high the peak concentration rises above the MIC, but by the percentage of the dosing interval during which the free drug concentration exceeds the MIC. Once the concentration exceeds approximately four to five times the MIC, further increases in concentration provide no additional killing benefit.

The PK/PD targets for time-dependent agents vary by drug class. Penicillins require at least 50 percent fT>MIC for optimal bactericidal activity. Cephalosporins, which have somewhat slower bactericidal kinetics, require 60 to 70 percent fT>MIC. Carbapenems, which benefit from a modest post-antibiotic effect against gram-negative organisms, achieve adequate killing with approximately 40 percent fT>MIC.

The evidence base for extended infusions has matured considerably. The BLING III trial, published in 2022, was a landmark multicenter randomized controlled trial that evaluated continuous infusion of meropenem or piperacillin-tazobactam versus standard intermittent bolus dosing in ICU patients with sepsis. The trial demonstrated a 90-day mortality benefit with continuous infusion, with an absolute risk reduction of approximately 3.7 percent. This trial provides the strongest evidence to date supporting the routine use of continuous or extended-infusion beta-lactams in critically ill patients with sepsis.

Concentration-Dependent Killing (Cmax/MIC)

Concentration-dependent killing is the pharmacodynamic pattern characteristic of aminoglycosides, daptomycin, and metronidazole. For these agents, the rate and extent of bacterial killing increase proportionally with the peak drug concentration relative to the MIC. The higher the peak, the faster and more complete the killing. The target for aminoglycosides is a Cmax/MIC ratio of at least 8 to 10.

The clinical strategy for concentration-dependent agents is to administer high doses at extended intervals, which maximizes the peak concentration while allowing prolonged drug-free intervals that minimize toxicity and exploit the post-antibiotic effect. This pharmacodynamic principle underpins the now-standard practice of once-daily aminoglycoside dosing, with gentamicin and tobramycin dosed at 5 to 7 mg/kg and amikacin at 15 to 20 mg/kg. The Hartford nomogram provides a practical framework for monitoring extended-interval aminoglycoside therapy by using a single random drug level drawn 6 to 14 hours after the dose to guide interval adjustment.

AUC/MIC-Dependent Killing

The AUC/MIC-dependent killing pattern applies to vancomycin, fluoroquinolones, linezolid, tigecycline, and daptomycin. For these agents, clinical outcomes correlate most closely with the ratio of the 24-hour area under the concentration-time curve to the MIC. This parameter integrates both the magnitude and duration of drug exposure relative to the organism's susceptibility.

For vancomycin, the 2020 IDSA/ASHP/SIDP consensus guidelines established an AUC/MIC target of 400 to 600, assuming an MIC of 1 mg/L by broth microdilution. This AUC-guided dosing approach has replaced the previous practice of trough-only monitoring, which targeted trough concentrations of 15 to 20 mcg/mL. The rationale for this paradigm shift is that trough-guided dosing led to systematic overdosing, with target troughs of 15 to 20 corresponding to AUC values in the range of 600 to 800, well above the therapeutic target and associated with significantly higher rates of acute kidney injury. For levofloxacin, an AUC/MIC ratio of at least 125 is targeted for gram-negative infections.

Bayesian estimation software, such as PrecisePK and DoseMeRx, is the preferred method for AUC estimation in clinical practice. These platforms use population pharmacokinetic models combined with individual patient data to estimate AUC from as few as one or two drug levels, allowing clinicians to optimize dosing without the logistical burden of collecting precisely timed peak and trough samples.

<image>A comparison infographic with three panels showing the three PK/PD killing patterns. Panel 1: "Time-Dependent" shows a concentration-time curve with the T>MIC region highlighted in green above the MIC line, with beta-lactam examples listed. Panel 2: "Concentration-Dependent" shows a curve emphasizing the peak height (Cmax/MIC ratio) highlighted in orange, with aminoglycoside examples. Panel 3: "AUC/MIC-Dependent" shows the entire AUC shaded in blue above the MIC line, with vancomycin and fluoroquinolone examples. Each panel should include the target numerical goals.</image>

Clinical Applications of PK/PD Optimization

Extended and Continuous Infusions of Beta-Lactams

The practical implementation of extended and continuous infusions requires attention to several agent-specific factors. Piperacillin-tazobactam can be administered as a standard 30-minute infusion, a 4-hour extended infusion, or as a continuous infusion, with progressively greater probability of achieving PK/PD targets at each step. Meropenem administered as a 3-hour extended infusion achieves greater than 90 percent probability of target attainment for organisms with an MIC at or below 2 mg/L, which encompasses the vast majority of susceptible Enterobacterales and Pseudomonas isolates.

Cefepime extended infusion over 4 hours similarly improves target attainment, though clinicians must remain vigilant about the neurotoxicity risk associated with elevated trough concentrations exceeding 20 mg/L, particularly in patients with renal impairment. Cefepime-associated neurotoxicity manifests as encephalopathy, myoclonus, and non-convulsive status epilepticus and is frequently under-recognized in the ICU setting. Practical considerations for extended and continuous infusions include drug stability (meropenem remains stable for only 6 to 8 hours at room temperature, which limits continuous infusion to systems with appropriate temperature control), the need for a dedicated intravenous line, and the importance of nursing education to ensure correct administration.

Therapeutic Drug Monitoring (TDM)

Therapeutic drug monitoring is an established practice for several antimicrobial agents and is increasingly recognized as important for many more. For vancomycin, AUC-guided dosing is now the standard of care. The preferred approach uses either a two-level strategy, obtaining a trough before the fourth dose and a peak 1 to 2 hours post-infusion, or Bayesian estimation software that can derive an AUC estimate from fewer data points. For aminoglycosides, once-daily dosing is monitored using the Hartford nomogram or other extended-interval nomograms, typically with a single random level drawn 6 to 14 hours after the dose.

Voriconazole TDM is mandatory due to its erratic pharmacokinetics driven by CYP2C19 genetic polymorphisms. The target trough concentration is 1 to 5.5 mcg/mL. Poor metabolizers accumulate voriconazole to potentially toxic levels, while ultrarapid metabolizers may be persistently subtherapeutic despite standard dosing. Posaconazole trough monitoring targets levels above 1 mcg/mL for prophylaxis and above 1.25 mcg/mL for treatment of invasive fungal infections. For beta-lactams in the ICU, therapeutic drug monitoring is an emerging practice that is gaining traction, particularly for continuous infusions where real-time concentration data can inform dose adjustments to maintain adequate target attainment.

Dosing in Special Populations

Obesity presents unique dosing challenges that differ by drug class. Loading doses of hydrophilic drugs should generally be based on actual body weight to ensure adequate initial drug exposure in the expanded extracellular fluid volume. For aminoglycosides, however, adjusted body weight is used for maintenance dosing to avoid excessive peak concentrations.

Patients on continuous renal replacement therapy require careful attention to drug clearance, which depends on the specific modality employed (continuous venovenous hemofiltration, hemodialysis, or hemodiafiltration), the membrane type, and the effluent rate. Dosing references such as the comprehensive review by Heintz and colleagues provide empiric starting recommendations, but TDM is strongly encouraged whenever available. Patients on extracorporeal membrane oxygenation experience increased volumes of distribution and drug sequestration within the circuit tubing, with hydrophilic and highly protein-bound drugs being the most significantly affected. Burn patients present a unique constellation of increased volume of distribution, augmented renal clearance, and decreased protein binding that collectively necessitates substantially higher doses of many antimicrobial agents.

Post-Antibiotic Effect (PAE)

Definition and Clinical Relevance

The post-antibiotic effect is defined as the persistent suppression of bacterial growth that continues after drug concentrations fall below the MIC. This phenomenon has direct clinical implications for dosing interval selection. Aminoglycosides exhibit a prolonged post-antibiotic effect against gram-negative organisms lasting 2 to 4 hours, which pharmacodynamically supports the practice of once-daily dosing by providing continued bacteriostatic activity during the drug-free interval.

Beta-lactams, by contrast, exhibit minimal post-antibiotic effect against gram-negative organisms, typically less than 1 hour, which reinforces the need for frequent dosing or continuous infusion to maintain drug concentrations above the MIC for the maximum possible proportion of the dosing interval. Against gram-positive organisms, beta-lactams do show a modest PAE of 1 to 2 hours. Carbapenems occupy an intermediate position, with a PAE against gram-negative organisms of 1 to 3 hours, which partially explains why their fT>MIC targets (approximately 40 percent) are less demanding than those of penicillins and cephalosporins.

Mutant Prevention Concentration (MPC) and Resistance

The Mutant Selection Window

The mutant prevention concentration represents the drug concentration that prevents the growth of first-step resistant mutants. The mutant selection window is the concentration range between the MIC, which suppresses susceptible organisms, and the MPC, which suppresses all subpopulations including first-step resistant mutants. When drug concentrations remain within this window, susceptible organisms are killed while resistant subpopulations are preferentially selected and amplified, a phenomenon that drives the emergence of resistance during therapy.

The clinical relevance of this concept is most apparent with fluoroquinolones, where dosing regimens that allow drug concentrations to dwell within the mutant selection window promote resistance emergence. The strategic implication is that dosing should be aggressive enough to maintain drug levels above the MPC whenever possible, particularly for fluoroquinolones. This pharmacodynamic principle supports the use of high-dose, less frequent fluoroquinolone regimens rather than low-dose, more frequent ones.

<image>A diagram illustrating the mutant selection window concept. Show a horizontal axis representing drug concentration with three vertical dashed lines marking MIC99, MIC (of the susceptible population), and MPC. The region between MIC and MPC should be shaded in red and labeled "Mutant Selection Window." Below the MIC, label "No killing - susceptible organisms grow." Above the MPC, label "All organisms killed including first-step mutants." Within the red MSW zone, label "Resistant mutants selectively amplified." Include a small inset showing a concentration-time curve passing through the MSW.</image>

Tissue Penetration

Key Tissue-to-Serum Ratios

Antibiotic selection must account for penetration to the site of infection, as serum drug concentrations do not necessarily predict tissue concentrations. Cerebrospinal fluid penetration is poor for vancomycin (5 to 20 percent of serum levels even with inflamed meninges), moderate for ceftriaxone (5 to 15 percent) and meropenem (5 to 20 percent), and good for metronidazole, linezolid, and fluoroquinolones. These penetration characteristics dictate agent selection for CNS infections and explain why linezolid and fluoroquinolones are sometimes preferred for CNS infections caused by susceptible organisms despite not being first-line agents for systemic disease.

In the lung epithelial lining fluid, beta-lactams achieve concentrations that are 20 to 50 percent of serum levels, while fluoroquinolones achieve 100 to 300 percent (concentrating in the pulmonary epithelium), linezolid achieves approximately 100 percent, and vancomycin achieves 20 to 50 percent.

Bone penetration is excellent for fluoroquinolones and linezolid, making them valuable oral agents for osteomyelitis, while vancomycin penetration into bone is variable at 10 to 30 percent of serum levels. Abscess and biofilm penetration is uniformly reduced for most agents, underscoring the paramount importance of source control through drainage. Rifampin is exceptional in its ability to penetrate biofilm, which underpins its role as an adjunctive agent in prosthetic device infections.

Key Clinical Pearls

• AUC-guided vancomycin dosing is the current standard; trough-only monitoring is outdated and associated with nephrotoxicity
• Extended-infusion beta-lactams should be strongly considered in ICU patients and those with higher MIC organisms
• Augmented renal clearance is under-recognized and leads to treatment failure with standard dosing of renally cleared drugs
• Always think in terms of free (unbound) drug when interpreting PK/PD targets
• The BLING III trial provides the strongest evidence to date for continuous infusion beta-lactams in ICU sepsis
• Tissue penetration should guide agent selection: for CNS infections, choose agents with proven CSF penetration; for pneumonia, consider ELF concentrations

References

1. Roberts JA, Abdul-Aziz MH, Lipman J, et al. Individualised antibiotic dosing for patients who are critically ill: challenges and potential solutions. Lancet Infect Dis. 2014;14(6):498-509.
2. Dulhunty JM, Brett SJ, De Waele JJ, et al. Continuous vs intermittent beta-lactam antibiotic infusions in critically ill patients (BLING III). JAMA. 2022;328(18):1855-1865.
3. Rybak MJ, Le J, Lodise TP, et al. Therapeutic monitoring of vancomycin for serious methicillin-resistant Staphylococcus aureus infections: a revised consensus guideline. Am J Health Syst Pharm. 2020;77(11):835-864.
4. Drusano GL. Antimicrobial pharmacodynamics: critical interactions of 'bug and drug'. Nat Rev Microbiol. 2004;2(4):289-300.
5. Abdul-Aziz MH, Alffenaar JC, Bassetti M, et al. Antimicrobial therapeutic drug monitoring in critically ill adult patients: a position paper. Intensive Care Med. 2020;46(6):1127-1153.

Read this lecture as Markdown