Optimizing the entire therapeutic regimen in septic critically ill patients should be based not only on improving antibiotic use but also on optimizing the entire therapeutic regimen by considering possible drug–drug or drug–nutrient interactions. There are reports of endocarditis, meningitis, peritonitis, or pneumonia associated with probiotics in critically ill patients. In addition, probiotics are associated with risk of the spread of antimicrobial resistance. The therapeutic drug monitoring (TDM)-guided method ensures a true optimization of antibiotic therapy, and particular efforts should be applied globally.
1. TDM-Guided Recommendations in Septic Critically Ill Patients
Antibiotic treatment should be adjusted based on the pharmacodynamic (PD) properties, the pathogens’ susceptibilities (as assessed by their minimal inhibitory concentrations (MICs)) and pharmacokinetics (PK) characteristics of antibiotics in the critically ill host. In clinical practice, MIC is the most used PD parameter. Knowledge of MIC is the foundation for selecting antimicrobial therapy against bacteria and fungi and helps to guide dosing needs in critically ill patients
[1]. Its main limitation is that knowing MIC does not account for the individual host defenses and does not provide sufficient information on the patterns of exposure to antimicrobial agents for an optimal therapeutic response
[2]. In addition, bacterial status (tolerance or persistence), bacterial inoculum size, or antibiotic concentrations may influence their antimicrobial activity
[3].
Pathological changes in critically ill patients in the volume of drug distribution, protein binding, and clearance led to a high intra- and inter-individual PK variability that significantly alters exposure to antibiotics with the risk of suboptimal doses or increased risk of toxicity. Depending on their PK/PD characteristics (
Table 1), the PK parameters (e.g., C
max, C
min or AUC) should be assessed in relation to MIC to establish the effectiveness and safety of antibiotic therapy
[4][5][6]. Recently, a new concept of the maximum tolerable dose (MTD) for beta-lactams in critically ill patients was proposed as the highest dose deemed safe for the patient, which has the goal of maximizing the kill of bacteria and minimizing the risk of antimicrobial resistance and toxicity, but there are a lack of data on the association of beta-lactam antibiotic levels and markers of toxicity
[7].
Table 1. PK/PD relationship of antibiotics
[5][6].
For this reason, model-informed precision dosing (MIPD), such as populational PK, PK/PD models in combination with therapeutic drug monitoring (TDM), Bayesian algorithms, and Monte Carlo simulations, are important tools for the individualization of treatment in critically ill patients. TDM quantifies drug concentrations (in plasma) using validated bioanalytical methods, while drug exposure may be directly correlated with the therapeutic target for PD response (MIC) or estimated through Bayesian methods. Dosing nomograms, clinician-based predictions, or dosing software (e.g., BestDose
®, Antibiotics kinetics
®, MwPharm++
®, TDMx
®, etc.) allow for further individual dosage adjustment
[8][9][10][11]. TDM is already successfully implemented on a large scale; Therapeutic Drug Monitoring and Early Appropriate infection Management in European ICUs (A-TEAMICU) reported that TDM was used in 61% of intensive care units (ICUs) surveilled
[12].
In 2020, experts from many worldwide associations on antibiotic treatment revised the TDM guidelines. Accordingly, TDM is routinely recommended for various beta-lactams, aminoglycosides, linezolid, teicoplanin, or vancomycin in critically ill patients
[13].
2. Beta-Lactams TDM-Guided Recommendations
Beta-lactams have a time-dependent bactericidal activity. The maximum bactericidal activity of beta-lactams is considered to occur when free fractions are maintained at least four times above the MIC over the entire dosing interval (i.e., 100%
fT > 4xMIC) with an increase up to 8xMIC in critically ill septic patients, while minimum plasma concentration (C
min) estimates antibiotics toxicity (e.g., the cut-off value for C
min—361 mg/L for penicillins or 20 mg/L for cephalosporins)
[14][15].
Cefiderocol is a novel siderophore cephalosporin approved for infections caused by multidrug-resistant aerobic Gram-negative organisms in adults with limited treatment options and
fC
min/MIC ratio is considered the optimal PK/PD parameter. In a case series of 13 patients with extended drug-resistant
Acinetobacter baumanii, microbiological failure was reported in 80% of patients with suboptimal
fC
min/MIC (<1) compared with 29% of those with optimal or quasi-optimal
fC
min/MIC ratio (≥4, and respectively 1–4)
[16].
For carbapenems, the optimal clinical results in critically ill patients are 100%ft > MIC and 100%ft > 4xMIC or target through concentration >4–8xMIC, while a C
min of 64.2 mg/L is potentially neurotoxic
[17]. In patients with normal renal function, the usual recommended doses are often subtherapeutic and do not reach the target concentrations. In a meta-analysis including 35 studies, it can be noted that meropenem should be administered in doses up to 6 g /day (every 6 h, 8 h or 12 h) or by continuous infusion or prolonged infusion (up to 3 h) to achieve the PK/PD target
[18][19].
3. Beta-Lactam and Beta-Lactamase Inhibitor Combinations TDM-Guided Recommendations
As many authors have highlighted, the widely used antipseudomonal piperacillin/tazobactam had a wide PK variability in ICU, and varied target concentrations could not be achieved. TDM-guided dose modifications dramatically improved therapeutic exposure and defined target concentrations of 100 mg/L and 361 mg/L for nephrotoxicity and neurotoxicity, respectively
[17][20]. Nevertheless, the recently published TARGET international study findings are not promising. After evaluating 244 patients, the authors found that TDM-guided piperacillin/tazobactam medication had no beneficial impact on critically ill patients
[21].
Ceftazidime/avibactam possesses a high activity against many carbapenem-resistant
Enterobacterales and
Pseudomonas aeruginosa with a therapeutic target of 24–30 mg/L in plasma and 8–10 mg/L at the site of infection; dosage adjustment based on TDM led to therapeutic failure in 1% of cases
[22].
Ceftolozane/tazobactam was recently approved for treating intraabdominal, renal, and lower respiratory infection (including ventilation-acquired pneumonia) caused by multiresistant
Enterobacterales,
Pseudomonas aeruginosa or
Haemophilus influenzae. In 40 patients with multi- or extended drug-resistant
P. aeruginosa infections, TDM-recommended dose reductions were applied in 84.2% of cases receiving prolonged infusion and allowed the achievement of 100%ƒT ≥ MIC target even at lower doses
[23][24].
4. Aminoglycosides TDM-Guided Recommendations
Aminoglycosides are concentration-dependent antimicrobials with known toxicity and very high PK variability in critically ill patients. TDM methods are already implemented at a large scale.
Plazomicin is a new aminoglycoside antibiotic effective against
Enterobacterales (including carbapenemase-producing bacteria),
Pseudomonas aeruginosa, and
Staphylococcus spp., including methicillin-resistant strains. In patients with normal renal function, the recommended dosage of plazomicin is 15 mg/kg per day; TDM is recommended in patients with a creatinine clearance <90 mL/minute to maintain plasma trough concentrations below 3 g/mL
[25].
5. Oxazolidinones TDM-Guided Recommendations
From the oxazolidinones group, Linezolid is extensively used in ICUs to treat infections produced by Gram-positive multiresistant cocci. It has a narrow therapeutic index and a high PK variability in critically ill patients; the usual recommended doses of 1.2 g/day often led to subtherapeutic levels. The ratio of the area under the drug plasma concentration–time curve over 24 h to the MIC (AUC/MIC) or percentage time above the MIC (%T > MIC) are predictors of the therapeutic response, and concentrations in the range of 2–8 mcg/mL appear to define the optimal window for acute bacterial infections, while a cut-off value >8 mg/L is an indicator for thrombocytopenia. Thus, TDM based on PK models or continuous infusion over 6 h instead of intermittent infusion is recommended for the optimization of Linezolid treatment to achieve the target AUC/MIC
[6][26][27][28][29].
6. Polymyxins TDM-Guided Recommendations
Even though there is not a well-defined recommendation for polymyxins (polymyxin B, colistin) monitoring, a narrow therapeutic index related to nephrotoxicity and the ratio of the area under the unbound plasma concentration–time curve over a dosing interval to minimum inhibitory concentration (
fAUC:MIC) is considered the most predictive PK/PD parameter of colistin activity
[30][31]. Many authors consider polymyxin E as the subject of TDM in ICUs as an adaptive feedback control because of the following: (1) CMS and colistin prodrug concentrations are not correlated, and (2) in critically ill patients, kidney function and renal replacement therapy dramatically impact the PK of prodrug CMS and colistin with the risk of under or overdosing
[32][33][34].
However, these are only a few examples of TDM–guided recommendations for antibiotics usually used in critically ill patients, and many other methods are published in the literature.