Are Antibiotics Appropriately Dosed in Critically Ill Patients with Augmented Renal Clearance? A Narrative Review

Aims Augmented renal clearance (ARC), which is commonly defined as increased renal clearance above 130 ml/min/1.73 m2, is a common phenomenon among critically ill patients. The increased elimination rate of drugs through the kidneys in patients with ARC can increase the risk of treatment failure due to the exposure to subtherapeutic serum concentrations of medications and affect the optimal management of infections, length of hospital stay, and outcomes. The main goal of this review article is to summarize the recommendations for appropriate dosing of antibiotics in patients with ARC. Methods This article is a narrative review of the articles that evaluated different dosing regimens of antibiotics in patients with ARC. The keywords “Augmented Renal Clearance,” “Critically ill patients,” “Drug dosing,” “Serum concentration,” “Beta-lactams,” “Meropenem,” “Imipenem,” “Glycopeptide,” “Vancomycin,” “Teicoplanin,” “Linezolid,” “Colistin,” “Aminoglycosides,” “Amikacin,” “Gentamycin,” “Fluoroquinolones,” “Ciprofloxacin,” and “Levofloxacin” were searched in Scopus, Medline, PubMed, and Google Scholar databases, and pediatric, nonhuman, and non-English studies were excluded. Results PK properties of antibiotics including lipophilicity or hydrophilicity, protein binding, the volume of distribution, and elimination rate that affect drug concentration should be considered along with PD parameters for drug dosing in critically ill patients with ARC. Conclusion This review recommends a dosing protocol for some antibiotics to help the appropriate dosing of antibiotics in ARC and decrease the risk of subtherapeutic exposure that may be observed while receiving conventional dosing regimens in critically ill patients with ARC.


What Is Known?
Augmented renal clearance (ARC) is a common phenomenon in critical care settings. e incidence of ARC was reported between 14 and 85% depending on the study population and the cutoff value of creatinine clearance (CrCl). e CrCl ≥130 ml/min/1.73 m 2 has been considered the ARC phenomenon in most studies, although different values have been suggested as well. e elimination rate of drugs, especially hydrophilic antibiotics that are mainly eliminated through the kidney, increased in ARC. According to the effect of ARC on the optimal management of infections, length of hospital stay, and clinical outcomes, determination of the ARC phenomenon is necessary for adjusting the optimal treatment to reduce the risk of subtherapeutic exposure that may be observed while receiving conventional dosing regimens in critically ill patients with ARC.

What Is New?
e higher rate of renal elimination of the medications in ARC impacts the dosing regimens of antibiotics to achieve target pharmacokinetic/pharmacodynamic (PK/PD) indices. ere are some review articles aimed to recommend dosing regimens for antibiotics in patients with ARC while reviewing the related articles, but each contains only several antibiotics, particularly B-lactams and vancomycin. We tried to prepare an almost complete article that involved most antibiotics with renal excretion.
To the best of our knowledge, there is no published guideline for drug dosing in patients with ARC. e main goal of this narrative review article is to summarize the recommendations for appropriate dosing of antibiotics in ARC to reduce the risk of subtherapeutic exposure that may be observed while receiving conventional dosing regimens in critically ill patients with ARC.

Quality of the Studies Included.
All studies except systematic reviews and meta-analyses and case reports were independently rated for quality by two reviewers using the National Institutes of Health (NIH) Quality Assessment Tools [21]. e studies were assessed with questions appropriate to the study design. We graded the quality of the study as good (G) if its rating was at least 70%, fair (F) if its rating was at least 50%, and poor (P) if its rating was less than 50% (Table 1).

Results
Studies that described the impact of ARC on the dosing regimens of antibiotics, based on achieving target pharmacokinetic/pharmacodynamic (PK/PD) indices, have increased in recent years.
ere are some review articles aimed to recommend dosing regimens for antibiotics in patients with ARC, but each contains only a few antibiotics, particularly B-lactams and vancomycin [4,8,15,57,58]. We tried to prepare an almost complete article summarizing the dosing recommendations for most antibiotics with renal excretion.

Beta-Lactams (B-Lactams).
e pattern of B-lactams activity is time-dependent. e time that the free drug plasma concentration remains above the minimum inhibitory concentration (MIC) of the pathogen (fT > MIC) has an essential role in efficacy. In practice, depending on the pathogen and the type of B-lactam, fT > MIC for 40-70% of the dosing interval is considered an acceptable PD target [4,11]. In critically ill patients, fT > 4 × MIC has been suggested to improve clinical outcomes [4,8,13] (Table 2).
B-lactams are mainly eliminated through the kidneys, and renal function alteration could influence the elimination rate constant and consequently PK/PD parameters of these drugs [4]. Evaluation of the correlation between B-lactam concentration and creatinine clearance (CrCl) showed 55% and 36% subtherapeutic B-lactam levels in ARC patients with Pseudomonas aeruginosa and Enterobacter spp. infections, respectively [17]. Udy A.A. et al. also reported 82% and 72% of trough levels were less than the MIC and 4 × MIC, respectively, in critically ill patients with ARC treated with empirical doses of B-lactams [29].
Although Udy A.A. et al. reported no statistically significant differences in the outcome of the patients with ARC who received continuous B-lactam infusion compared with those who received intermittent infusion [59], some articles recommended increasing frequency of infusion as well as increasing the dosage to increase achieving the optimal fT > MIC in patients with ARC [13,15]. Continuous infusion (over 24 hr) of conventional doses can improve the optimal exposure to B-lactams in patients with ARC, compared to intermittent (over 30 min) or extended infusion (over 3-4 hr) [5]. Extended infusion of standard doses of B-lactams in patients with ARC resulted in 80% less than 100% fT > MIC and 37% less than 50%fT > MIC [1]. Reviewing the articles that compared prolonged infusion and intermittent infusion to achieve effective B-lactam exposure and maximal bacterial killing showed that even with prolonged infusion, effective exposure may not be achieved in critically ill patients with ARC [60]. Carrie et al. reported CrCl ≥ 170 ml/ min as a sensitive threshold (93%) to predict subexposure to B-lactams despite the continuous infusion of high doses [6]. Hobbs A.L.V. et al. had recommended a dosing nomogram for different antibiotics, including B-lactams, in critically ill patients with ARC according to their PK/PD targets and breakpoints that are included in Table 1 [8].    [10,12,30]. It has been shown that, in septic patients with ARC, 8-12 g meropenem daily is needed to obtain the PD target [12,30]. In patients with CrCl 60 to 90 ml/min/1.73 m 2 , 6 g meropenem daily has been recommended to achieve the PD target, whereas in patients with CrCl ≥ 90 ml/min/1.73 m 2 , in addition to increased dosage, increasing frequency or prolonging duration of infusion also has been reported [10]. In addition to doses and strategy of infusion, Agyeman A. A. et al. recommended combination therapy with meropenem and ciprofloxacin against P. aeruginosa isolates to increase synergistic killing and decrease resistance in patients with ARC [22].

Piperacillin-Tazobactam (PTZ)
. PK models of nine dosing regimens of PTZ (intermittent infusion, extended infusion (EI), and continuous infusion (CI) of 16 g, 12 g, and 8 g daily piperacillin) have been assessed as achieving the PD targets in septic patients with different CrCls and for different MICs. e MIC of 16 mg/L has been proposed as the clinical cutoff point of piperacillin for P. aeruginosa and the probability of target attainment (PTA) ≥ 90% for optimal piperacillin regimens. e results showed the PTA was ≥ 90% for 50% fT > MIC in all three CI regimens and EI of 12 and 16 g piperacillin daily, whereas the PTA for 100% fT > MIC was ≥ 90% only in all three CI regimens [9]. Treating the patients with 16 + 2 g/day and 20 + 2.5 g/day PTZ was also associated with 100% fT > MIC in 93% and 98% of patients with 130 ≤ CrCl < 200 ml/min and 80% and 90% of patients with CrCl ≥ 200 ml/min, respectively. e daily dose of 20 + 2.5 g PTZ was recommended for patients with CrCl ≥ 170 ml/min to reach the PD target with the highest probability [26]. No intoxication or supratherapeutic levels were reported by 20 + 2.5 g PTZ daily in critically ill patients with ARC [25,26]. In another PK model study, it is reported that even continuous infusion of higher doses of piperacillin (24 g daily) was insufficient to achieve the acceptable PD target in ARC patients with CrCl ≥ 90 ml/min [27].

Glycopeptides.
e antibacterial activity pattern of glycopeptides is both time-and concentration-dependent, and the area under the plasma concentration-time curve (AUC) for 24 hr, relative to the pathogen MIC (AUC0-24/ MIC) ratio, is considered the best parameter to predict their antibacterial activity [8,13] (Table 2). Glycopeptides are hydrophilic agents and are eliminated primarily through the  kidney [63], so the change in CrCl can affect the PK/PD parameters of these drugs.

Vancomycin.
For vancomycin, the target of AUC0-24/MIC ≥ 400 has been proposed to have optimal clinical outcomes [8,35]. Studies showed the correlation of ARC with the augmented clearance of vancomycin, subtherapeutic serum concentration, and a higher risk of treatment failure [14,64]. It has been observed that using conventional doses of vancomycin (1000 mg every 12 hr) resulted in 62.9% trough concentrations <10 mg/L. e trough level remained less than 10 mg/L despite increasing the dosage of vancomycin to 1000 mg every 8 hr or 1500 mg every 12 hr [33]. A mean daily dose of 44 ± 9 mg/kg/day vancomycin in trauma patients with ARC was also associated with 54.2% therapeutic trough concentration <10 mg/L [32]. According to TDM data, there was a need for higher doses of vancomycin in patients with ARC compared with non-ARC patients (35.7 mg/kg/day vs. 27.1 mg/kg/day) [34]. A loading dose (LD) of 25-30 mg/kg vancomycin followed by a maintenance dose (MD) of 15-20 mg/kg every 8-12 hr or 45 mg/kg/day for vancomycin is recommended in critically ill patients with ARC [8,15]. Also, it has been proposed that an LD of 35 mg/kg vancomycin followed by continuous infusion of 35 mg/kg/day needs to keep the vancomycin trough concentration within the target therapeutic range in the CrCl equal to 100 ml/min/1.73 m 2 and higher CrCls need the larger MD to maintain the therapeutic exposure [35]. Fransson et al. reported a case with intracranial infection caused by Streptococcus intermedius stating that they did not achieve a stable vancomycin target level until they increased the dose to 1.5 g, four times a day [31]. Baptista et al. published a vancomycin dosing nomogram in septic patients with different CrCls to achieve an ideal trough level of vancomycin on the first day of treatment with 84% success [36].

Teicoplanin.
Teicoplanin has approximately 90-95% affinity to binding serum protein, and hypoalbuminemia may cause a low total trough level, whereas the unbound trough concentration is in the therapeutic range (1.5-4.5 mg/ L). So in settings of hypoalbuminemia, measuring total and unbound trough levels of teicoplanin is recommended [11,40,65]. Although a total trough level ≥ 15 mg/L is considered an optimal therapeutic concentration of teicoplanin for most infections, higher trough levels are suggested for deep and severe infections [40,66,67]. But it should be noted that the levels were kept <60 mg/L to avoid toxicity [40,42]. e negative association between the trough concentrations of teicoplanin and renal function has been reported [42]. Serum albumin and body weight are also other confounding factors to appropriate dosing of teicoplanin [40]. Evaluation of the impact of LD on the trough concentration of teicoplanin during 10-day treatment showed 25%, 38.9%, and 68.6% of patients who received no LD, low LD (<9 mg/kg), and high LD ( ≥ 9 mg/kg) achieved the total trough concentrations ≥ 20 mg/L, respectively [39]. e PTA on the 3 rd and 15 th days of treatment with 400, 600, 800, and 1000 mg teicoplanin every 12 hr for four consequent doses followed by the same doses every 24 hr for 12 days was evaluated in different CrCls. e results showed the need for 800 or 1000 mg teicoplanin as the LD to provide a trough concentration ≥ 15 mg/L on day 3 with a PTA ≥ 90%. But these doses may cause trough concentrations > 60 mg/L on day 15 even in CrCls >90 mL/min/1.73 m 2 [41]. e administration of 12 mg/kg teicoplanin every 12 hr for five consequent doses and then 12 mg/kg every 24 hr resulted in trough concentrations ≥10 mg/L in 62% of critically ill patients with CrCl > 50 mL/min/1.73 m 2 .
e clinical effectiveness rate was reported as 88% in these patients [43]. 12 mg/kg teicoplanin for both LD and MD resulted in median total and unbound trough concentrations of 15.9 and 3.7 mcg/ml, respectively, in critically ill patients with pneumonia [45].
Due to the limitations of clinical trials in the field of adequate dosing regimen for teicoplanin in patients with ARC, we also mentioned some studies that have suggestions for teicoplanin dosing in critically ill patients whose risk of ARC is high among them. e recommendations of these articles are summarized in Table 1. Although according to the effect of both ARC and hypoalbuminemia on teicoplanin dosing and high prevalence of the hypoalbuminemia as ARC in critically ill patients, these dosing recommendations may not be an exact guide for teicoplanin dosing in patients with ARC but can be helpful.
No association was reported between teicoplanin-induced nephrotoxicity and its trough concentrations [42]. Teicoplanin trough concentrations ≥20 mg/L compared with <20 mg/L showed no significant difference in the rate of adverse events in patients with MRSA infections [38]. Although no teicoplanin-induced renal impairment was reported in trough concentrations ≥10 mg/L, the hepatotoxicity rate was 10% [43].

Linezolid.
e antimicrobial activity of linezolid is timeand concentration-dependent. Besides AUC0-24/MIC and %fT > MIC that are correlated with the clinical efficacy of linezolid [46,68], its trough concentration is also associated with clinical response and adverse events and suggested to be kept in the range of 2-10 mg/L to decrease adverse event incidence [47] (Table 2). Due to the amphiphilicity of linezolid and the limited impact of renal clearance on its excretion (approximately 30%), it does not seem ARC can cause subtherapeutic exposure to linezolid [46,49,50,63]. But some studies reported the significant impact of ARC on PK properties of linezolid [46,49]. e use of LD or continuous infusion of linezolid was recommended to improve the PTA and efficacy of linezolid in patients with ARC and severe sepsis [49]. e administration of conventional dosing of linezolid results in no PD target achievement in patients with ARC, whereas the PTA was 70% in patients with ARC who received a continuous infusion of 600 mg linezolid every 12 hr. According to Monte Carlo simulation, continuous infusion of a higher dose of linezolid (600 mg every 8 hr) could result in a PTA of 93% in patients with ARC with no linezolid-induced adverse effect [46]. e PTA of different regimens of linezolid was evaluated for the MIC of 0.5 to 4 mg/L. An optimal dosing regimen with adequate exposure was considered the one that its PTA was >90%. Although the continuous infusion of 1200 or 1800 mg/day linezolid was recommended in patients with ARC, it was the optimal dosing for MIC ≤ 0.5 mg/L, 2400 mg/day linezolid was optimal for MIC ≤ 1 mg/L, and none was optimal for MIC ≥ 2 mg/L, whether as an intermittent infusion or a continuous infusion. e trough concentrations >10 mg/L were detected by none of the dosing regimens [47]. e PTA values of 600 mg every 12 hr, 800 mg every 12 hr, and 900 mg every 12 hr of linezolid for the MIC of 2 mg/L in critically ill septic patients were 0.26%, 85.59%, and 98.81%, respectively. e dose of 800 mg linezolid every 12 hr was associated with 33.19% probability of thrombocytopenia, whereas this rate was 51.37% for 900 mg every 12 hr. According to the efficacy and safety data, 800 mg linezolid every 12 hr was recommended in septic patients [48].

Colistin.
Colistin is a cationic, lipoprotein, hydrophilic antibiotic [51]. e antibacterial activity of colistin is more concentration-dependent than time-dependent, and the AUC/MIC ratio is considered the best PD index to predict its antibacterial efficacy [15,63,69,70] (Table 2). Colistimitate sodium (CMS), the prodrug of colistin, is primarily eliminated through the kidney, and the increased elimination rate of CMS in ARC could alter the PK properties of colistin due to the reduced systemic bioavailability [51,63,70]. Studies showed the need for a longer duration therapy and consequently higher cumulative doses of colistin and higher doses of colistin than conventional regimens in patients with ARC in comparison with other patients [51,53]. ere were two cases (12.5%) with mild to moderate colistin-induced AKI, according to the Acute Kidney Injury Network (AKIN) classification, among patients with ARC who were treated with a longer duration of colistin [51]. e colistin-induced AKI was reported to be 44.3%, according to AKIN criteria, among ARC patients who receive higher doses of colistin, whereas none of them need renal replacement therapy or discontinuing the colistin therapy [53]. A dosing nomogram for colistin was published to achieve the appropriate antibacterial efficacy balanced with the risk of nephrotoxicity in different CrCls. But in CrCl >80 ml/min/1.73 m 2 even by the maximum daily recommended dose of colistin in the nomogram (360 mg or approximately 11 mIU daily), the PTA was <40%. So, combination therapy with colistin in addition to high-dose regimens was recommended for these patients [50,52].

Aminoglycosides (AGs).
AGs are hydrophilic antimicrobial agents with concentration-dependent killing characteristics. Although Cmax/MIC is considered the best PD index to reflect their bactericidal effect [13,50,63], the monitoring of trough concentrations of AGs is also recommended to avoid drug toxicity when the plan of treatment is more than 3-5 days [50] (Table 2). AGs are eliminated predominantly through the kidneys, and their doses need to adjust following the renal function alteration [63,71]. Some studies evaluated ARC as an effective factor in changing the PK characteristics of AGs [72,73], but due to the limitations of articles in the field of adequate dosing regimen for AGs in patients with ARC, we also mentioned some studies that have suggestions for AG dosing in critically ill patients whose risk of ARC is high among them.
Carrie et al. used the Monte Carlo simulation to report different dosing regimens of amikacin that are needed to achieve PD targets in different CrCls and for different MICs. An optimal regimen was considered the one that had fractional target attainment (FTA) >85%.
In critically ill patients, 30 mg/kg/day amikacin and 7-10 mg/kg/day gentamycin and tobramycin were recommended as initial empirical doses to achieve the optimal PD targets in severe infections [20,50].
%fT > MIC is another PD index that should be noted to decrease the risk of developing resistant pathogens against AGs. It was shown that fT > MIC for less than 60% of AG dosing intervals is associated with resistance development. ereafter, 12.5 mg/kg amikacin every 12 hr instead of a high-dose extended-interval dosage regimen was recommended to achieve the PD target in the empiric treatment of septic patients with minimizing resistance development. e safety of the two regimens was also compared using neutrophil gelatinase-associated lipocalin (NGAL) during 7 days of treatment, and any significant difference was not detected [55].

Fluoroquinolones (FQs).
Although the bacterial killing property of FQs is concentration-dependent, they also show some time-dependent effects [50]. e AUC0-24/MIC ratio is considered the best PK/PD parameter to predict their antibacterial efficacy and is recommended to be kept more than 125 for ciprofloxacin and more than 80 for levofloxacin to improve outcomes in critically ill patients with Gramnegative infections [8,13,50,63] (Table 2). Among FQs that belong to lipophilic drugs, levofloxacin is the most hydrophilic FQ, and its PK is more affected by renal function alteration [50,74]. In the comparison of four dosing regimens of levofloxacin to achieve the optimal PD target for different MICs in different CrCls, it was shown none of them was optimal (PTA ≥ 85%) for MIC ≥ 1 mg/L in CrCl ≥ 130 ml/min. But 1000 mg once daily was more effective than other dosing regimens, even 500 mg every 12 hr. Any of these regimens was associated with levofloxacin-related adverse events [56]. 400 mg every 8 hr or 600 mg every 12 hr and 750-1000 mg once daily or 500 mg every 12 hr were recommended as initial empirical doses for ciprofloxacin International Journal of Clinical Practice 13 and levofloxacin in critically ill patients with ARC to achieve optimal PD targets [8,15,20]. According to the correlation between AUC0-24/MIC <100 and the risk of developing resistance to FQs, the use of maximum recommended doses of FQs was suggested in critically ill patients to avoid subtherapeutic exposure [50].

Conclusion
PK properties of antibiotics including lipophilicity or hydrophilicity, protein binding, the volume of distribution, and elimination rate that affect drug concentration should be considered along with PD parameters for drug dosing in critically ill patients with ARC. is review recommends a dosing protocol for some antibiotics to help the appropriate dosing of antibiotics in ARC and decrease the risk of subtherapeutic exposure that may be observed while receiving conventional dosing regimens in critically ill patients with ARC (Table 3).
Data Availability e data supporting this narrative review are from previously reported studies and datasets, which have been cited. e processed data are available upon request from the corresponding author.

Disclosure
All authors including Mohammad Sistanizad, Rezvan Hassanpour, and Elham Pourheidar declare being employed by Shahid Beheshti University of Medical Sciences with a primary function of research and education, and nobody is an official representative or on behalf of the government.

Conflicts of Interest
e authors declare no conflicts of interest for the research, authorship, and/or publication of this article.

Authors' Contributions
EP and RH designed the study and drafted the article. EP was involved in acquisition of data and their interpretation. MS revised the article critically for important intellectual content. All three authors gave the final approval of the version to be submitted. 14 International Journal of Clinical Practice