The success of antimicrobial therapy is determined by interactions between the administered drug, the host, and the infecting agent (1). The complexity of these interactions is responsible for the difficulty in selecting an appropriate antibiotic regimen (2).
Antimicrobial drug selection is based primarily on a static in vitro parameter, the minimum inhibitory concentration (MIC). Currently, we know that the serum antibiotic concentrations and the drug’s mode of administration are extremely important and should be considered (3, 4).
Antibiotics exert their effects not within the plasma compartment, but in defined target tissues to which these drugs must be distributed. Unfortunately, a complete and lasting balance between blood and tissue cannot always be taken for granted. This fact has been taken into account through local-regional strategies for drug applications, e.g., intra-arterial, intrathecal, intra-articular, or nebulized administration (2).
The pharmacodynamic (PD) effect is the result of the exposure of the bacteria to the unbound antibiotic drug fraction at the site of infection. This means that the overall serum concentration does not reflect the active fraction of the drug (3, 5). Moreover, altered pharmacokinetic (PK) parameters among intensive care unit (ICU) patients, e.g., decreased renal function, hyperdynamic state, and use of vasoactive drugs, have direct implications on the antimicrobial serum concentrations (6).
The antibiotic distribution process is characterized by high intertissue variabilities, and target site drug levels may substantially differ from corresponding plasma drug levels (6, 7). Inadequate target site concentrations may have important clinical implications that can explain therapeutic failures, primarily, for bacteria with high in vitro MICs (Table 1).
Pharmacokinetic-pharmacodynamic concepts are based on antimicrobial exposure-response relationships. These concepts were initially identified in the 1940s and 1950s by Dr. Harry Eagle through studies in rodents. Their studies were the first ones that differentiated time-dependent (penicillins) and concentration-dependent antibiotics (streptomycin). Despite this, only in the 1970s PK-PD concepts were rediscovered (8).
Based on in vitro and in vivo evidence, antimicrobial drugs are categorized on the basis of the PK-PD measure that is predictive of efficacy. The three PK-PD parameters are the duration of time a drug concentration remains above the MIC (T>MIC), the ratio of the maximal drug concentration to the MIC (Cmax/MIC), and the ratio of the area under the concentration time-curve at 24 h to the MIC (AUC0–24/MIC). Indeed, in more recent years, PK-PD parameters have had a critical role in drug development and antimicrobial therapy (8).
This article focuses on alterations in the PK-PD parameters that commonly occur among ICU patients. In addition, we describe the PK-PD parameters of the major classes of antimicrobial drugs used in ICUs.
Alterations in the PK parameters in ICU patients
It is very important to understand the effects of the pathophysiological changes that occur in critically ill patients because these changes can affect antimicrobial serum and tissue concentrations (9). The clearance rate (C) and volume of distribution (V) are the major PK parameters that influence drug levels. The half-life (T½) of a drug is related to C and V. The initial dosing is commonly based on the volume of distribution, whereas the subsequent dosing is based on the clearance rate. Alterations in these two parameters due to physiological changes during severe sepsis or septic shock are very common in ICU patients (Fig. 1).
Different antimicrobial classes have different PK-PD indices that are related to adequate antimicrobial activity (10). Thus, dosing regimens should take into account the specific PD parameters of each antimicrobial drug to achieve the appropriate antimicrobial effect. Moreover, understanding the physicochemical characteristics of antimicrobial classes is also very important, i.e., hydrophilic versus lipophilic drugs, and protein binding.
In septic patients, the production of exotoxins and endotoxins by microorganisms results in the production of several mediators that cause endothelial damage and greater capillary permeability (9, 11, 12). Because of these alterations, a fluid shift occurs from the intravascular space to the extravascular space (interstitial space), resulting in lower plasma and tissue concentrations of hydrophilic drugs. By contrast, the concentrations of lipophilic antimicrobials are less influenced by this fluid shift because of their extensive distribution (larger V) (9). In patients with septic shock, the impairment of microvascular perfusion is associated with much lower tissue antimicrobial concentrations than in patients with sepsis. Thus, greater plasma concentrations are required to achieve the target tissue concentrations. The potential site of infection is also very important and should be considered to achieve the proper concentration of antibiotic in that site. Zeitlinger et al. (13) described a clinical scoring system, TPPS (Tissue Penetration Prediction Score), which can be used at the bedside to analyze antibiotic penetration in tissues in patients with sepsis. The oxygen saturation, serum lactate concentration, and the dose per time unit of norepinephrine administered showed the best correlation with tissue penetration and were used in this score.
Hypoalbuminemia is another common finding in critically ill patients and is often caused by increased capillary permeability and leakage into the extravascular compartment (14). Low plasma albumin levels lead to an increase in the size of the unbound fraction of a drug, and this free concentration results in a greater tissue distribution. However, the increased fluid loading utilized during the acute phase response in sepsis leads to a greater of amount interstitial fluid and reduced tissue concentrations of antibiotics (antimicrobial dilution). This reduced tissue concentration can be significant for highly protein-bound antimicrobials, e.g., ceftriaxone, ertapenem, and teicoplanin (15, 16).
The hyperdynamic state in sepsis is associated with increased renal clearance of drugs eliminated by glomerular filtration, i.e., β-lactams and glycopeptides. The use of vasoactive drugs and the excessive administration of fluids also can lead to an increase in cardiac output and higher glomerular filtration. It is noteworthy that only the unbound fraction of the antibiotic can be cleared by the organism. When end-organ dysfunction occurs (renal and/or hepatic), the decreased clearance and metabolism of drugs will result in the accumulation of drugs/metabolites, and dose reductions should be considered (9).
Application of PK-PD principles to clinical practice
Utilizing mathematical modeling is possible to apply PK-PD concepts to clinical practice. Monte Carlo simulation is used to integrate PK, PD, and MIC data to design antimicrobial drug regimens with a high probability of achieving the most adequate PD target (17). This technique encompasses the variability of PK values and the MIC distribution to predict antibiotic exposure. Briefly, by utilizing this program, one can calculate the probability of obtaining a target exposure that drives a microbiological effect for a range of MIC values (17). The question that arises is about the utilization of PK values from healthy volunteer studies and applying these values to hospitalized patients, especially critically ill patients. In practice, Monte Carlo simulation results have been very similar when utilizing PK data from hospitalized patients and when using data from healthy volunteers.
PK-PD parameters of the major classes of antimicrobial drugs
β-Lactam antibiotics have an important role in the treatment of infections, and they are the most prescribed class within the hospital environment. The clinical indications of these antimicrobiotics range from simple skin infections to difficult-to-treat infections, such as osteomyelitis (18).
This antimicrobial class shows hydrophilicity, renal excretion, a low volume of distribution, and moderate to low protein binding (the exception is ceftriaxone, which has a longer half-life and a high protein-binding rate). The major PD parameter is time-dependence (i.e., the antibiotic levels have to be above the MIC of the bacteria over the duration of exposure). Therefore, to achieve an optimal effect, the β-lactam serum levels should remain between 20% and 60% of the duration of exposure above the MIC (50% for penicillins, 60%–70% for cephalosporins, and 40% for carbapenems) (17).
The various methods to achieve this goal include more frequent dosing and the administration of higher doses. Increases in the dose are less effective, and the use of more frequent dosing has been more promising. There are two ways to do this: increasing the frequency of dosing and optimizing the time that the β-lactam serum concentration remains above the MIC of the bacteria through continuous infusion or extended infusion.
The continuous infusion of β-lactam antibiotics has the advantage of covering the entire treatment period and thus increasing the exposure of the pathogen to the antimicrobial drug. This model of infusion was initially adopted more than 60 years ago. During the 1940s, episodes of bacterial endocarditis were effectively treated with penicillin administered as a continuous infusion (19–21). More recent studies have shown that continuous infusion is a more effective strategy than intermittent infusion (22–25).
Continuous infusion is the best mode of administration of β-lactam agents, but these drugs have low physical and chemical stability at room temperature, especially carbapenems (19). For this reason, the extended/prolonged infusion of β-lactam antibiotics has been considered in recent years as a valuable alternative for achieving suitable serum levels of β-lactams. This type of infusion is interesting because the antibiotic plasma level does not necessarily need to be above the MIC for 100% of the interval.
Prolonged infusion has shown good results in clinical practice (26, 27). By maintaining optimal serum concentrations, extended infusion has also shown to result in prolonged antimicrobial exposure at infection sites such as the lungs (26). In addition, Lodise et al. (27) showed that the extended infusion (in 3 h) of piperacillin-tazobactam resulted in reduced mortality when compared with 0.5-h infusion for infections caused by Pseudomonas aeruginosa in critically ill patients (with APACHE II score >17).
The different changes in PK parameters among critically ill patients (creatinine clearance, volume of distribution, protein binding) can lead to excessive serum levels (e.g., cefepime neurotoxicity) (28) or low serum levels (for instance, in patients with creatinine clearance >120 mL/min). Moreover, Wells and Lipman (29) showed that the discrepancy between the estimated creatinine clearance and the measured clearance can be substantial. An alternative method for obtaining the correct dosage is utilizing therapeutic drug monitoring for β-lactams (as is utilized with aminoglycosides and vancomycin). Roberts and Ulldemolins (30) conducted a study involving β-lactam therapeutic drug monitoring in 236 patients. The authors found that 74.2% of the patients required dose adjustments, of which 50.4% required increased dosing.
Aminoglycosides have been used in clinical practice for over 50 years, although their use as monotherapies for the treatment of gram-negative infections declined considerably after the appearance of cephalosporins, carbapenems, and fluoroquinolones. Currently, the use of aminoglycosides has been restricted to combination therapies with β-lactams (31).
Aminoglycosides are hydrophilic molecules with low volumes of distribution, renal excretion, and short half-lives (approximately 2 h). These antibiotics exhibit concentration-dependent activity and prolonged postantibiotic effect (32, 33). This PD feature supports their use in a once-daily dose. The peak drug concentration (Cmax) is the PK parameter utilized for PK-PD analysis for this antimicrobial class. A Cmax/MIC ratio greater than 10 is predictive of microbiological success as demonstrated by previous clinical studies (33, 34).
Critically ill patients often have an increased volume of distribution, and this increased volume can result in a reduced Cmax for aminoglycosides. Thus, maximal weight-based doses (7 mg/kg per day for tobramycin/gentamicin and 20 mg/kg per day for amikacin) should be considered to obtain an ideal Cmax/MIC ratio (35).
Moreover, in recent years, the utilization of once-daily dosing has been demonstrated to be associated with a reduced likelihood of ototoxicity and nephrotoxicity (35–37).
Linezolid belongs to a new class of antibiotics, the oxazolidinones, which act by inhibiting the initiation of bacterial protein synthesis (38). This agent is characterized by moderate hydrophilicity, a moderate to low volume of distribution, and short half-life. Despite its hydrophilicity, linezolid penetrates well into tissues and reaches satisfactory levels in the epithelial lining fluid of patients with ventilator-associated pneumonia (39).
Linezolid is a time-dependent antibiotic with a persistent postantibiotic effect. The PD parameters that best characterize its activity are T>MIC and AUC/MIC. Based on these principles, Adembri et al. (40) recently examined the PK/PD profile of intermittent infusion versus continuous infusion in critically ill patients with sepsis. Continuous infusion was found to be superior with regard to the PK-PD parameters analyzed. However, further studies are needed to confirm the superiority of this mode of infusion.
The AUC/MIC ratio is the best PK-PD predictor of the response based on data from animal models and in vitro studies. Those studies suggest that an AUC/MIC ratio of at least 350 to 400 is essential for a good clinical outcome (17, 41).
The “MIC creep” (higher vancomycin MICs for Staphylococcus aureus in recent years) is very concerning because a higher MIC can reduce the AUC/MIC ratio contributing to a lower efficacy of vancomycin. Briefly, with an MIC of 0.5 mg/L, a standard dose of vancomycin (1 g every 12 h) is likely to reach the target attainment. However, when the MIC value increases to 1.0 or 2.0 mg/L, the probability of reaching the target attainment falls considerably to a suboptimal level (42). Moreover, higher vancomycin doses used to obtain better AUC/MIC rates have led to higher rates of nephrotoxicity (43).
Because of this, in 2009, a task force by the Infectious Diseases Society of America, the American Society of Health-System Pharmacists, and the Society of Infectious Diseases Pharmacists set new targets for serum vancomycin trough (15–20 mg/mL), aiming to increase the likelihood of achieving the PD target that could treat infections caused by agents with MIC greater than 0.5 mg/mL (44). Moreover, these recommendations also suggest a loading dose for vancomycin (25–30 mg/kg). Later, Kullar and Davis (45) demonstrated that patients exposed to higher serum trough concentrations (15–20 mg/mL) achieved higher clinical success rates than those exposed to lower concentrations (45% vs. 60%, P = 0.034).
Fluoroquinolones are fully synthetic compounds that exert their effect by interfering with DNA replication. These substances are lipophilic antibiotics and are less susceptible to the alterations in the clearance and volume of distribution commonly observed among critically ill patients.
Fluoroquinolones are a classic example of concentration-dependent bactericidal agents. The AUC/MIC ratio is the major PK-PD parameter, although the Cmax/MIC ratio has also been considered. The first study analyzing the PD of fluoroquinolones was a retrospective analysis of various studies using ciprofloxacin (46). A significant relationship was observed between the AUC/MIC ratio and the probability of a good microbiological outcome. The breakpoints for organism clearance were an AUC/MIC ratio of 125 to 250. It is noteworthy that despite the fact that the CLSI (Clinical and Laboratory Standards Institute) susceptibility breakpoint is 2 mgq/L, the probability that the maximum dose of ciprofloxacin (400 mg intravenously every 8 h) will achieve an AUC/MIC ratio of 125 was less than 90% for MIC values in excess of 0.5 mg/L (17).
In addition, Drusano et al. (47) performed clinical PD studies with levofloxacin in community-acquired infections and hospital-acquired pneumonia.
Regarding the clinical response, the peak/MIC ratio and the primary infection site were the major determinants of a successful outcome. A levofloxacin AUC/MIC ratio of 87 had a great effect on pathogen eradication among patients with nosocomial pneumonia. It is noteworthy that, in this study, the probability of attaining the breakpoint value of 87 remained greater than 90% only at MIC values of 0.5 mg/L or less for levofloxacin. These findings suggest that the CLSI sensitivity breakpoint values for gram-negative bacilli are probably set incorrectly and need to be revised.
In summary, we face increased levels of bacterial resistance worldwide, and few options of antibacterial drugs will be available in the future. Thus, optimizing current antimicrobial treatments is necessary. Understanding PK alterations among critically ill patients as well as PK-PD parameters associated with antimicrobial efficacy are essential for the improved treatment of bacterial infections in ICUs.
We think that the application of PK-PD knowledge at the patient’s bedside needs to be reinforced to contribute to reducing morbidity and mortality among critically ill patients.
1. Müller M, dela Peña A, Derendorf H: Issues in pharmacokinetics and pharmacodynamics of anti-infective agents: distribution in tissue. Antimicrob Agents Chemother
48 (5):1441–1453, 2004.
2. Joukhadar C, Frossard M, Mayer BX, Brunner M, Klein N, Siostrzonek P, Eichler HG, Müller M: Impaired target site penetration of beta-lactams may account for therapeutic failure in patients with septic shock. Crit Care Med
3. Roberts JA, Lipman J: Pharmacokinetic issues for antibiotics in the critically ill patient. Crit Care Med
4. Nicolau DP: Pharmacodynamic optimization of β-lactams in the patient care setting. Crit Care
12 (Suppl 4):S2, 2008.
5. Schmidt S, Röck K, Sahre M, Burkhardt O, Brunner M, Lobmeyer ML, Derendorf H: Effect of protein binding on the pharmacological activity of highly bound antibiotics. Antimicrob Agents Chemother
52 (11):3994–4000, 2008.
6. Scaglione F, Paraboni L: Pharmacokinetics/pharmacodynamics of antibacterials in the intensive care unit: setting appropriate dosing regimens. Int J Antimicrob Agents
7. Zeitlinger MA, Sauermann R, Traunmuller F, Georgopoulos A, Muller M, Joukhadar C: Impact of plasma protein binding on antimicrobial activity using time-killing curves. J Antimicrob Chemother
8. Ambrose PG, Bhavnani SM, Rubino CM, Louie A, Gumbo T, Forrest A, Drusano GL. Pharmacokinetics-pharmacodynamics of antimicrobial therapy: it’s not just for mice anymore. Clin Infect Dis
9. Varghese JM, Roberts JA, Lipman J: Antimicrobial pharmacokinetic and pharmacodynamic issues in the critically ill with severe sepsis and septic shock. Crit Care Clin
10. Craig WA: Pharmacokinetic/pharmacodynamic parameters: rationale for antibacterial dosing of mice and men. Clin Infect Dis
11. Bochud PY, Calandra T: Pathogenesis of sepsis: new concepts and implications for future treatment. BMJ
12. Opal SM, Cohen J: Clinical gram-positive sepsis: does it fundamentally differ from gram-negative bacterial sepsis? Crit Care Med
13. Zeitlinger BS, Zeitlinger M, Leitner I, Muller M, Joukhadar C: Clinical scoring system for the prediction of target site penetration of antimicrobials in patients with sepsis. Clin Pharmacokinet
14. Fleck A, Raines G, Hawker F, Trotter J, Wallace PI, Ledingham IM, Calman KC: Increased vascular permeability: a major cause of hypoalbuminemia in disease and injury. Lancet
1 (8432):781–784, 1985.
15. Burkhardt O, Kumar V, Katterwe D, Majcher-Peszynska J, Drewelow B, Derendorf H, Welte T: Ertapenem in critically ill patients with early-onset ventilator-associated pneumonia: pharmacokinetics with special consideration of free-drug concentration. J Antimicrob Chemother
16. Pea F, Viale P, Candoni A, Pavan F, Pagani L, Damiani D, Casini M, Furlanut M: Teicoplanin in patients with acute leukaemia and neutropenia: a special population benefiting from higher dosages. Clin Pharmacokinet
17. Lodise TP, Drusano GL: Pharmacokinetics and pharmacodynamics: optimal antimicrobial therapy in the intensive care unit. Crit Care Clin
18. Suárez C, Gudiol F: Antibióticos betalactámicos. Enferm Infecc Microbiol Clin
27 (2):116–129, 2009.
19. Nicolau DP: Pharmacodynamic optimization of β-lactams in the patient care setting. Crit Care
12 (Suppl 4):1–5, 2008.
20. Chain E, Florey HW, Gardner AD, Heatley NG, Jennings MA, Orr-Ewing J, Sanders AG: Penicillin as a chemotherapeutic agent. Lancet
21. Loewe AB: The combined use of penicillin and heparin in the treatment of subacute bacterial endocarditis. CMAJ
22. Crandon JL, Nicolau DP: Pharmacodynamic approach to optimizing β-lactam therapy. Crit Care Clin
23. Bodey GP, Ketchel SJ, Rodriguez V: A randomized study of carbenicillin plus cefamandole or tobramycin in the treatment of febrile episodes in cancer patients. Am J Med
24. Grant EM, Kuti JL, Nicolau DP, Nightingale C, Quintiliani R: Clinical efficacy and pharmacoeconomics of a continuous-infusion piperacillin-tazobactam program in a large community teaching hospital. Pharmacotherapy
25. Roberts JA, Kirkpatrik CMJ, Roberts MS, Robertson TA, Dalley AJ, Lipman J: Meropenem dosing in critically ill patients with sepsis and without renal dysfunction: intermittent bolus versus continuous administration? Monte Carlo dosing simulations and subcutaneous tissue distribution. J Antimicrob Chemother
26. Kuti JL, Moss KM, Nicolau DP, Knauft RF: Empiric treatment of multi-drug resistant Burkholderia cepacia
lung exacerbation in a patient with cystic fibrosis: application of pharmacodynamic concepts to meropenem therapy. Pharmacotherapy
27. Lodise TP, Lomaestro B, Drusano GL: Piperacillin-tazobactam for Pseudomonas aeruginosa
infection: clinical implications of an extended-infusion dosing strategy. Clin Infect Dis
28. Lamoth F, Buclin T, Pascual A, Vora S, Bolay S, Decosterd LA, Calandra T, Marchetti O: High cefepime plasma concentrations and neurological toxicity in febrile neutropenic patients with mild of renal function. Antimicrob Agents Chemother
54 (10):4360–4367, 2010.
29. Wells M, Lipman J: Measurements of glomerular filtration in the intensive care unit are oly a rough guide to renal function. S Afr J Surg
35 (1):20–23, 1997.
30. Roberts J, Ulldemolins M, Roberts MS, McWhinney B, Ungerer J, Paterson DL, Lipman J: Therapeutic drug monitoring of beta-lactams: proof of concept. Int J Antimicrob Agents
36 (4):332–339, 2010.
31. Palomino J, Pachón J: Aminoglucósidos. Enferm Infecc Microbiol Clin
21 (2):105–115, 2003.
32. Turnidge J: Pharmacodynamics and dosing of aminoglycosides. Infect Dis Clin N Am
33. Craig WA: Optimizing aminoglycoside use. Crit Care Clin
34. Moore RD, Leitman PS, Smith CR: Clinical response to aminoglycoside therapy: importance of the ratio of peak concentration to minimal inhibitory concentration. J Infect Dis
35. Nicolau DP, Freeman CD, Belliveau PP, Nightingale CH, Ross JW, Quintiliani R: Experience with a once-daily aminoglycoside program administered to 2184 adult patients. Antimicrob Agents Chemother
39 (3):650–655, 1995.
36. Buijik SE, Mouton JW, Gyssens IC, Verbrugh HA, Bruining HA: Experience with a once-daily program of aminoglycosides in critically ill patients. Intensive Care Med
37. Olsen KM, Rudis MI, Rebuck JA, Hara J, Gelmont D, Mehdian R, Nelson C, Rupp ME: Effect of once-daily dosing vs. multiple daily dosing of tobramycin on enzyme markers of nephrotoxicity. Crit Care Med
32 (8):1678–1682, 2004.
38. Pigrau C: Oxazolidinonas y glucopéptidos. Enferm Infecc Microbiol Clin
21 (3):157–165, 2003.
39. Boselli E, Breilh D, Rimmelé T, Djabarouti S, Toutain J, Chassard D, Saux MC, Allaouchiche B: Pharmacokinetics and intrapulmonary concentrations of linezolid administered to critically ill patients with ventilator-associated pneumonia. Crit Care Med
33 (7):1529–1533, 2005.
40. Adembri C, Fallani S, Cassetta MI, Arriqucci S, Ottaviano A, Pecile P, Mazzei T, De Gaudio R, Novelli A: Linezolid pharmacokinetic/pharmacodynamic profile in critically ill septic patients: intermittent versus continuous infusion. Int J Antimicrob Agents
41. Moise-Broder PA, Forrest A, Birmingham MC, Schentag JJ: Pharmacodynamics of vancomycin and other antimicrobials in patients with Staphylococcus aureus
lower respiratory tract infections. Clin Pharmacokinet
42. Patel N, Pai MP, Rodvold KA, Lomaestro B, Drusano GL, Lodise TP: Vancomycin: We can’t get there from here. Clin Infect Dis
43. Lodise TP, Lomaestro B, Graves J, Drusano GL: Larger vancomycin doses (at least four grams per Day) are associated with an increased incidence of nephrotoxicity. Antimicrob Agents Chemother
44. Rybak MJ, Lomaestro B, Rotschafer JC, Moellering RC, Craig WA, Billeter M, Dalovisio JR, Levine DP: Vancomycin Therapeutic Guidelines: a summary of consensus recommendations of Infectious Diseases Society of America, the American Society of Health-System Pharmacists, and the Society of Infectious Diseases Pharmacists. Clin Infect Dis
45. Kullar R, Davis SL, Taylor TN, Kaye KS, Rybak MJ: Effects of targeting higher vancomycin trough levels on clinical outcomes and costs in a matched patient cohort. Pharmacotherapy
32 (3):195–201, 2012.
46. Forrest A, Nix DE, Ballow CH, Goss TF, Birmingham MC, Schentag JJ: Pharmacodynamics of intravenous ciprofloxacin in seriously ill patients. Antimicrob Agents Chemother
47. Drusano GL, Preston SL, Fowler C, Corrado M, Weisinger B, Kahn J: Relationship between fluoroquinolone area under the curve: minimum inhibitory concentration ratio and the probability of eradication of the infecting pathogen, in patients with nosocomial pneumonia. J Infect Dis