Abstract
Introduction: The optimal meropenem dosing regimens in critically ill patients receiving continuous renal replacement therapy (CRRT) based on pharmacokinetic and pharmacodynamic (PD) concepts are not well established. This study aimed to (1) gather the available published pharmacokinetic studies conducted in septic patients receiving CRRT and (2) to define the optimal meropenem dosing regimens in these populations via Monte Carlo simulations. Methods: We used Medical Subject Headings “meropenem,” “continuous renal replacement therapy,” and “pharmacokinetics” or related terms to identify studies for systematic review. A one-compartment pharmacokinetic model was conducted to predict meropenem levels for the initial 48 h of therapy. The PD targets were 40% of free drug above a threshold of 1 times the minimum inhibitory concentration (MIC) (40% fT > MIC), 4 times the MIC (40% fT > 4MIC), and an additional target of free drug level above 1 times MIC 100% of the time (fT > MIC). The dose that achieved at least 90% of the probability of target attainment (PTA) was defined as an optimal dose. Results: Twenty-one articles were included for our systematic review. The necessary pharmacokinetic parameters such as volume of distribution and CRRT clearance were cited in 90.5 and 71.4% of articles, respectively. None of the published studies reported completed necessary parameters. A regimen of 750 mg q 8 h was found to be the optimal dose for pre-dilution continuous venovenous hemofiltration and continuous venovenous hemodialysis modality using two effluent rates (25 and 35 mL/kg/h) which achieved the PD target of 40% fT > 4MIC. Conclusion: None of the published studies showed the necessary pharmacokinetic parameters. PD target significantly contributed to meropenem dosage regimens in these patients. Differing effluent rates and types of CRRT shared similar dosing regimens. Clinical validation of the recommendation is suggested.
Introduction
Optimally dosing drugs in critically ill patients who receive continuous renal replacement therapy (CRRT) is complicated. Alteration of pharmacokinetic parameters (e.g., the volume of distribution [Vd], protein binding and residual renal clearance) and amount of drug removed via CRRT machine should all be considered to design appropriate dosing regimens in these populations [1].
The calculation of CRRT clearance depends on the CRRT modality. Using CRRT clearance equations, the sieving coefficient (SC or SA) and CRRT intensity (ultrafiltration rate in continuous venovenous hemofiltration [CVVH] or dialyzate rate in continuous venovenous hemodialysis [CVVHD] or effluent rate in continuous venovenous hemodiafiltration [CVVHDF]) dictate CRRT clearance. Blood flow rates and serum hematocrit are the correcting factors used to adjust drug clearance using the pre-dilution CVVH modality [2, 3].
While meropenem is a hydrophilic drug, it has a low molecular weight and small Vd [4]. Due to these properties, meropenem tends to be removed via CRRT machines. The optimal meropenem dosing regimen in patients receiving CRRT based on pharmacokinetic and pharmacodynamic (PD) concepts is not well established. In clinical practice, meropenem dosing regimens from 500 mg to 1,000 mg every 8–12 h have been used, with these derived from pharmacokinetic studies [5, 7]. The variability of dosage recommendations can be explained by the differences in study characteristics, such as populations studied, CRRT modalities used, and PD targets selected. Therefore, to be able to effectively utilize recommendations from these studies for individualized patient care, necessary data should be or must be stated. The objectives of our study are (1) to determine the sufficient data reporting in the pharmacokinetic studies conducted in the septic patients receiving CRRT via systematic review and (2) to define optimal meropenem dosing regimens using pharmacokinetic parameters from anuric or oliguric critically ill patients receiving CRRT.
Materials and Methods
The current systematic review was performed according to the recommendations of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement guidelines. The PRISMA checklists are documented in online supplementary Table 1 (for all online suppl. material, see www.karger.com/doi/10.1159/000529694).
Search Strategies
The following databases were used to search for original research articles from November 1998 to May 2020: PubMed, Embase, and ScienceDirect. We used the Medical Subject Headings of “meropenem,” “continuous renal replacement therapy,” “pharmacokinetics,” “continuous venovenous hemofiltration,” “continuous venovenous hemodiafiltration,” “continuous venovenous hemodialysis.” We also manually searched in the reference lists of retrieved studies to identify additional studies for systematic reviews. The inclusion was limited to those under human study and those published in English. Detailed search strategies for databases are presented in online supplementary Table 2.
Study Selection
Articles were included if they were pharmacokinetic studies of meropenem and conducted in adult patients who received CRRT. The following criteria were excluded: (1) population pharmacokinetics, (2) intermittent hemodialysis or sustained low-efficiency dialysis pharmacokinetics studies, (3) concomitant use of extracorporeal membrane oxygenation with CRRT, (4) non-English language article or review articles.
Data Collection
Eligible titles and abstracts of articles identified by the systematic search were screened independently by 2 researchers (T.C. and P.K.). Then the full-text version of the articles was reviewed using inclusion and exclusion criteria. Disagreements between independent researchers were resolved by discussion with a third independent researcher (W.C.). The rationale for the exclusion of specific studies was recorded.
Data Extraction and Main Outcome Measurement
Two researchers (T.C. and P.K.) extracted data from the included studies independently. Included studies were assessed for sufficient data reporting in the septic patients receiving CRRT. The data extraction form included ideal datasets or parameters that were reported from the published study [2].
The datasets included (1) drug data: for the antibiotic assayed, specific PD target, and dose recommendation(s); (2) patient demographics: age, weight, severity of illness, number of patients in the study, residual renal function, and hepatic function; (3) basic pharmacokinetics: Vd, total clearance, CRRT clearance, non-CRRT clearance, and protein binding or serum albumin; and (4) specific CRRT clearance parameters that depended on CRRT modality:
pre-dilution CVVH: sieving coefficient (SC), ultrafiltration rate (Qf), blood flow rate (Qb), hematocrit (Hct), and pre-dilution replacement flow rate (Qrep)
post-dilution CVVH: sieving coefficient (SC) and ultrafiltration rate (Qf)
CVVHD: saturation coefficient (SA) and dialysis flow rate (Qd)
CVVHDF: SC or SA, ultrafiltration rate (Qf) and dialysis flow rate (Qd) or effluent flow rate (Qf+Qd)
The percentage of whether each variable was found was analyzed to describe the completion of provided data.
Monte Carlo Simulations and PD Target
Given the published population pharmacokinetic study, the characteristic of meropenem in CRRT patients was better described by a one-compartment linear model [8]. In addition, most of the previous traditional pharmacokinetic studies analyzed data using a one-compartment model [5, 7, 9]. Then, we developed a one-compartment pharmacokinetic model of the first 48-h meropenem concentration for acute AKI patients receiving CRRT. Pharmacokinetic equations are presented in online supplementary Table 3 [10]. From the previous 21 reviewed studies, the averages of necessary pharmacokinetic parameters, weight, Vd, nonrenal clearance (CLNR), and SC or SA, from 13 published meropenem pharmacokinetic studies conducted in anuric or oliguric patients and ongoing treatments with CRRT, were input into the model [5, 11, 22]. Blood flow rate of 200 mL/min and hematocrit of 30% were used in our pharmacokinetic model. The infusion time of the drug was 0.5 h. The correlation (r2) between body weight and Vd or CLNR was input into the model to generate realistic virtual patients. Based on Kidney Disease: Improving Global Outcomes (KDIGO) recommendation, ultrafiltration or dialyzate flow rates of 25 mL/kg/h were used in the pre-hemofilter dilution CVVH and CVVHD models, respectively [23]. Additionally, at high CRRT intensity, the ultrafiltration or dialyzate flow rate of 35 mL/kg/h, which represents the common effluent flow rate in clinical practice, was also used in the simulation [24].
The drug concentration-time profile of 5,000 virtual patients was created by Monte Carlo simulation (Oracle Crystal Ball Classroom). In this study, the PD targets were 40%, 100% of unbound fraction of drug level above the minimum inhibitory concentration (MIC) (40% fT > MIC, 100% fT > MIC), and 40% of unbound fraction of drug level above 4 times the MIC (40% fT > 4MIC) [6, 25]. Given that the most common Gram-negative pathogens causing nosocomial infections in critically ill patients were Pseudomonas aeruginosa, Acinetobacter baumannii, and Enterobacterales, we chose to apply an MIC of 2 mg/L for our models. According to Clinical Laboratory Standards Institute (CLSI), the susceptible MIC breakpoint for Enterobacterales is 1 mg/L and 2 mg/L for Pseudomonas aeruginosa and Acinetobacter baumannii. Consequently, the highest MIC breakpoint in a susceptible organism of that category was 2 mg/L [26]. The proportion of patients who attain the PD target was calculated as the probability of target attainment (PTA). The optimal dosage regimen was the lowest dose that achieved a PTA of at least 90%.
Results
A PRISMA flow diagram is demonstrated in Figure 1. A total of 421 articles were recruited from 3 scientific databases. Duplicated and irrelevant studies were excluded via screening of titles and abstracts. As a result, 44 articles were eligible for full-text assessment. Of these, 9 included participants undergoing intermittent hemodialysis or sustained low-efficiency dialysis, 9 were population pharmacokinetic studies, 3 were review articles, and 2 studied participants with extracorporeal membrane oxygenation and CRRT combination modalities. Finally, 21 full-text articles were included for the assessment of completed pharmacokinetic parameters. All included studies were published from June 22, 1998, through January 4, 2020. The key characteristics of the included studies were illustrated in Table 1[5, 7, 9, 11, 22, 27, 31].
A PRISMA flow diagram. CRRT, continuous renal replacement therapy; ECMO, extracorporeal membrane oxygenation; IHD, intermittent hemodialysis; POP-PK, Population pharmacokinetics study; SLED, sustained low-efficiency dialysis.
A PRISMA flow diagram. CRRT, continuous renal replacement therapy; ECMO, extracorporeal membrane oxygenation; IHD, intermittent hemodialysis; POP-PK, Population pharmacokinetics study; SLED, sustained low-efficiency dialysis.
Key characteristics of published pharmacokinetic studies of meropenem in critically ill patients undergoing CRRT
First author and year . | CRRT mode . | SC or SA . | Vd . | CLNR (mL/min) . | CLCRRT (mL/min) . | Prespecified PD target . | Dose recommendation . |
---|---|---|---|---|---|---|---|
Thalhammer et al. [5] 1998 | CVVH | 1.09±0.1 | 29.5±2.7 L | NS | 49.7±8.3 | Cmidpoint > 12 mg/L (40–50% T > MIC, MIC = 8 mg/L) | 1,000 mg q 8 h |
Krueger et al. [11] 1998 | CVVHDF | 1.06 | 0.26±0.09 L/kg | NS | 30.4±2.0 | NS | 1,000 mg q 12 h |
Meyer et al. [12] 1999 | CVVHDF | 1.13±0.22 | 0.63 L/kg | NS | 20±7.64 | NS | 1,000 mg q 12 h |
Tegeder et al. [13] 1999 | CVVH | 1.17±0.11 | 12.4±1.8 L | 29.9±5.4 | 22±4.7 | Css, av of 12 mg/L | mean, 898.0±145.3 mg/d range, 690–1,197 mg/d |
Ververs et al. [15] 2000 | CVVH | 0.63±0.25 | 0.37±0.15 L/kg | 59±17.67 | 17.17±7.0 | NS | 500 mg q 12 h |
Giles et al. [14] 2000 | CVVH, CVVHDF | 0.93±0.06 | 0.35±0.1 L/kg | 37.83±18 | 32.67±11.17 | Cat 8 h > 4 mg/L | 1,000 mg q 12 h |
Valtonen et al. [27] 2000 | CVVH, CVVHDF | NS | NS | NS | NS | NS | 500 mg q 8 h (CVVH)1,000 mg q 12 h (CVVHDF) |
Robatel 2003 [16] | CVVHDF | 0.65±0.18 | 33.2±7.53 L | NS | 27.82±7.07 | 75% T > MIC,MIC = 4 mg/L | 750 mg q 8 h, 1,500 mg q 12 h |
Krueger et al. [28] 2003 | CVVH | 0.91±0.10 | 0.28±0.07 L/kg | NS | 24.42±8.03 | 40% T > MIC,MIC = 4 and 8 mg/L | 500 mg q 12 h |
Isla et al. [17] 2005 | CVVH, CVVHD | 0.80±0.12 | 0.57±0.29 L/kg | NS | 27.01±6.90 | NS | NS |
Langgartner et al. [29] 2008 | CVVHDF | 0.97 (0.87, 1.05)* | 32.3 (28.9, 40.7)* | NS | NS | T > MIC,MIC = 4 and 8 mg/L | 1,000 mg q 12 h (IB)500 mg LD then 2 g IV CI over 24 h |
Bilgrami et al. [18] 2010 | CVVH | 0.93 | 0.26 L/kg | 40 (26.67, 50)* | 58.17 | NS | 1,000 mg q 8 h |
Seyler et al. [6] 2011 | CVVH, CVVHDF | NS | 0.45 L/kg | NS | NS | 40% T> 4MIC,MIC = 2 mg/L | 1,000 mg q 8 h |
Afshatous et al. [9] 2014 | CVVHD | NS | 29.6 L | NS | 32.4±9.8** | 40% T> 4MIC,MIC = 2 mg/L | NS |
Beumier et al. [7] 2014 | CVVH, CVVHDF | NS | 0.39 L/kg | NS | NS | 40% T> 4MIC,MIC = 2 mg/L | 1,000 mg q 8 h |
Jamal et al. [19] 2015 | CVVH | 1.10 (IB) | 0.43 L/kg | 33 (IB) | 34.8 (IB) | IB: 40% T> 4MIC,MIC = 2 mg/L | 2g LD then 1g q 8 h (IB)1g LD then 3 g IV CI over 24 h |
1.02 (CI) | 36 (CI) | 39.2 (CI) | CI: 100% T> 5MIC,MIC = 2 mg/L | ||||
Kawano et al. [20] 2015 | CVVHD | NS | 17.5±5.6 L | NS | NS | 40% T > MIC,MIC = 2, 4, 8, 16 mg/L | 250 mg q 24 h for MIC = 2 mg/L |
250 mg q 12 h for MIC = 4 mg/L | |||||||
250 mg q 8 h for MIC = 8 mg/L | |||||||
500 mg q 8 h for MIC = 16 mg/L | |||||||
Varghese et al. [21] 2015 | CVVHDF | 1.08 | 0.35 L/kg | NS | 48.33 | NS | 500 mg q 8 h |
Singhan et al. [22] 2019 | CVVH | 0.94 | 26.66±9.89 L | 39.17 | 21.67 | NS | NS |
Le Noble et al. [30] 2019 | CVVH | 0.45–0.96 | 19.9 L | NS | 0.45 mL/kg/min | Css > 4MIC,MIC = 2 mg/L | 2,000–3,000 mg CI over 24 h |
Nowak-Kózka et al. [31] 2020 | CVVHD | NS | NS | NS | NS | NS | NS |
First author and year . | CRRT mode . | SC or SA . | Vd . | CLNR (mL/min) . | CLCRRT (mL/min) . | Prespecified PD target . | Dose recommendation . |
---|---|---|---|---|---|---|---|
Thalhammer et al. [5] 1998 | CVVH | 1.09±0.1 | 29.5±2.7 L | NS | 49.7±8.3 | Cmidpoint > 12 mg/L (40–50% T > MIC, MIC = 8 mg/L) | 1,000 mg q 8 h |
Krueger et al. [11] 1998 | CVVHDF | 1.06 | 0.26±0.09 L/kg | NS | 30.4±2.0 | NS | 1,000 mg q 12 h |
Meyer et al. [12] 1999 | CVVHDF | 1.13±0.22 | 0.63 L/kg | NS | 20±7.64 | NS | 1,000 mg q 12 h |
Tegeder et al. [13] 1999 | CVVH | 1.17±0.11 | 12.4±1.8 L | 29.9±5.4 | 22±4.7 | Css, av of 12 mg/L | mean, 898.0±145.3 mg/d range, 690–1,197 mg/d |
Ververs et al. [15] 2000 | CVVH | 0.63±0.25 | 0.37±0.15 L/kg | 59±17.67 | 17.17±7.0 | NS | 500 mg q 12 h |
Giles et al. [14] 2000 | CVVH, CVVHDF | 0.93±0.06 | 0.35±0.1 L/kg | 37.83±18 | 32.67±11.17 | Cat 8 h > 4 mg/L | 1,000 mg q 12 h |
Valtonen et al. [27] 2000 | CVVH, CVVHDF | NS | NS | NS | NS | NS | 500 mg q 8 h (CVVH)1,000 mg q 12 h (CVVHDF) |
Robatel 2003 [16] | CVVHDF | 0.65±0.18 | 33.2±7.53 L | NS | 27.82±7.07 | 75% T > MIC,MIC = 4 mg/L | 750 mg q 8 h, 1,500 mg q 12 h |
Krueger et al. [28] 2003 | CVVH | 0.91±0.10 | 0.28±0.07 L/kg | NS | 24.42±8.03 | 40% T > MIC,MIC = 4 and 8 mg/L | 500 mg q 12 h |
Isla et al. [17] 2005 | CVVH, CVVHD | 0.80±0.12 | 0.57±0.29 L/kg | NS | 27.01±6.90 | NS | NS |
Langgartner et al. [29] 2008 | CVVHDF | 0.97 (0.87, 1.05)* | 32.3 (28.9, 40.7)* | NS | NS | T > MIC,MIC = 4 and 8 mg/L | 1,000 mg q 12 h (IB)500 mg LD then 2 g IV CI over 24 h |
Bilgrami et al. [18] 2010 | CVVH | 0.93 | 0.26 L/kg | 40 (26.67, 50)* | 58.17 | NS | 1,000 mg q 8 h |
Seyler et al. [6] 2011 | CVVH, CVVHDF | NS | 0.45 L/kg | NS | NS | 40% T> 4MIC,MIC = 2 mg/L | 1,000 mg q 8 h |
Afshatous et al. [9] 2014 | CVVHD | NS | 29.6 L | NS | 32.4±9.8** | 40% T> 4MIC,MIC = 2 mg/L | NS |
Beumier et al. [7] 2014 | CVVH, CVVHDF | NS | 0.39 L/kg | NS | NS | 40% T> 4MIC,MIC = 2 mg/L | 1,000 mg q 8 h |
Jamal et al. [19] 2015 | CVVH | 1.10 (IB) | 0.43 L/kg | 33 (IB) | 34.8 (IB) | IB: 40% T> 4MIC,MIC = 2 mg/L | 2g LD then 1g q 8 h (IB)1g LD then 3 g IV CI over 24 h |
1.02 (CI) | 36 (CI) | 39.2 (CI) | CI: 100% T> 5MIC,MIC = 2 mg/L | ||||
Kawano et al. [20] 2015 | CVVHD | NS | 17.5±5.6 L | NS | NS | 40% T > MIC,MIC = 2, 4, 8, 16 mg/L | 250 mg q 24 h for MIC = 2 mg/L |
250 mg q 12 h for MIC = 4 mg/L | |||||||
250 mg q 8 h for MIC = 8 mg/L | |||||||
500 mg q 8 h for MIC = 16 mg/L | |||||||
Varghese et al. [21] 2015 | CVVHDF | 1.08 | 0.35 L/kg | NS | 48.33 | NS | 500 mg q 8 h |
Singhan et al. [22] 2019 | CVVH | 0.94 | 26.66±9.89 L | 39.17 | 21.67 | NS | NS |
Le Noble et al. [30] 2019 | CVVH | 0.45–0.96 | 19.9 L | NS | 0.45 mL/kg/min | Css > 4MIC,MIC = 2 mg/L | 2,000–3,000 mg CI over 24 h |
Nowak-Kózka et al. [31] 2020 | CVVHD | NS | NS | NS | NS | NS | NS |
IB, intermittent bolus; CI, continuous infusion; NS, data not stated; LD, loading dose,
*Median (Q1, Q3).
**Reported as carbapenem CRRT clearance.
Main Outcome Measurement
None of the current studies reported the complete necessary parameters.
Drug Data
Meropenem concentration analysis methods were reported from all studies. Prespecified PD target and dose recommendation were reported in 57.1% and 76.2% of studies, respectively. The 40–100% T > MIC was most often used as a PD target with MIC varying from 2 to 8 mg/L (shown in Fig. 2a)
Percentage of studies reporting the required parameters for pharmacokinetic studies conducted in septic patients receiving CRRT. a Drug data. b Patient demographic. c Basic pharmacokinetic parameters. d Specific CRRT clearance parameters.
Percentage of studies reporting the required parameters for pharmacokinetic studies conducted in septic patients receiving CRRT. a Drug data. b Patient demographic. c Basic pharmacokinetic parameters. d Specific CRRT clearance parameters.
Patient Demographic
Age and sample size were reported in 100% of the studies, whereas the residual renal and hepatic functions were described in 76.2% and 4.8% of the reviewed studies. 90.5% of the studies provided individual or mean weight. Half of the studies mentioned the severity score, e.g., Acute Physiology and Chronic Health Evaluation (APACHE) II and the Sequential Organ Failure Assessment (SOFA) (shown in Fig. 2b).
Basic Pharmacokinetic Parameters
The most basic PK parameters reported were the Vd and the total clearance (90.5%). CRRT and non-CRRT clearance were noted in 71.4 and 28.6% of the studies, respectively. Protein binding affinity or serum albumin was reported in 33.3% of studies (shown in Fig. 2c).
Specific CRRT Clearance Parameters
Most studies used patients undergoing the CVVH modality (13 studies, 61.9%), whereas CVVHD and CVVHDF modality were used in 4 (19.1%) and 9 (42.9%) of studies, respectively. Dilution methods were stated in 66.7% of the studies. Pre-dilution modality was mentioned in 44.4% of both CVVH and CVVHDF. Blood flow rate and hematocrit, the parameters for calculating the correction factor in pre-dilution CVVH, were presented and accounted for 100% and 25%, respectively. All the studies conducted in CVVH, CVVHD, and CVVHDF models reported the ultrafiltration, dialyzate, and effluent flow rates. They reported SC or SA in 71.4% of the studies (shown in Fig. 2d).
Optimal Dosage Regimens of Meropenem
The PTAs of meropenem dosing regimens for Gram-negative bacilli infections with MIC =2 mg/L in pre-dilution CVVH and CVVHD modalities with 25 and 35 mL/kg/h effluent rates are shown in Tables 2 and 3. A regimen of 500 mg every 12 h was the optimal dose for the two CRRT modalities and two effluent rates with PD target of 40% fT > MIC. For the higher PD target, 40% fT > 4MIC and 100% fT > MIC, the regimen of 750 mg every 8 h and 2 g every 8 h was considered to achieve 90% of the PTA target in pre-dilution CVVH modality with two effluent flow rates.
PTA of all recommended meropenem dosing regimens for Gram-negative bacilli infections in pre-dilution CVVH modality with 25 and 35 mL/kg/h effluent rates
Dosage regimens . | PTA over 48 h, MIC of 2 mg/L . | |||||
---|---|---|---|---|---|---|
effluent rates 25 mL/kg/h . | effluent rates 35 mL/kg/h . | |||||
40% fT > MIC . | 40% fT > 4MIC . | 100% fT > MIC . | 40% fT > MIC . | 40% fT > 4MIC . | 100% fT > MIC . | |
500 mg q 12 h | 0.9796 | 0.4270 | 0.0914 | 0.9744 | 0.3362 | 0.0664 |
500 mg q 8 h | 0.9988 | 0.7710 | 0.1626 | 0.9984 | 0.7086 | 0.1466 |
1,000 mg q 12 h | 0.9942 | 0.8940 | 0.5170 | 0.9900 | 0.8644 | 0.4832 |
1,000 mg LD 500 mg q 8 h | 0.9982 | 0.8538 | 0.5838 | 0.9972 | 0.8170 | 0.5596 |
750 mg q 8 h | 0.9998 | 0.9362 | 0.4216 | 1.00 | 0.9228 | 0.4150 |
1,250 mg LD 1,000 mg q 12 h | 0.9944 | 0.9032 | 0.6376 | 0.9904 | 0.8742 | 0.5924 |
1,000 mg q 8 h | 0.9996 | 0.9798 | 0.6392 | 1.00 | 0.9754 | 0.6298 |
1,500 mg q 12 h | 0.9956 | 0.9638 | 0.7534 | 0.9948 | 0.9484 | 0.7118 |
2,000 mg q 8 h | 0.9998 | 0.9968 | 0.9322 | 0.9998 | 0.9982 | 0.9206 |
Dosage regimens . | PTA over 48 h, MIC of 2 mg/L . | |||||
---|---|---|---|---|---|---|
effluent rates 25 mL/kg/h . | effluent rates 35 mL/kg/h . | |||||
40% fT > MIC . | 40% fT > 4MIC . | 100% fT > MIC . | 40% fT > MIC . | 40% fT > 4MIC . | 100% fT > MIC . | |
500 mg q 12 h | 0.9796 | 0.4270 | 0.0914 | 0.9744 | 0.3362 | 0.0664 |
500 mg q 8 h | 0.9988 | 0.7710 | 0.1626 | 0.9984 | 0.7086 | 0.1466 |
1,000 mg q 12 h | 0.9942 | 0.8940 | 0.5170 | 0.9900 | 0.8644 | 0.4832 |
1,000 mg LD 500 mg q 8 h | 0.9982 | 0.8538 | 0.5838 | 0.9972 | 0.8170 | 0.5596 |
750 mg q 8 h | 0.9998 | 0.9362 | 0.4216 | 1.00 | 0.9228 | 0.4150 |
1,250 mg LD 1,000 mg q 12 h | 0.9944 | 0.9032 | 0.6376 | 0.9904 | 0.8742 | 0.5924 |
1,000 mg q 8 h | 0.9996 | 0.9798 | 0.6392 | 1.00 | 0.9754 | 0.6298 |
1,500 mg q 12 h | 0.9956 | 0.9638 | 0.7534 | 0.9948 | 0.9484 | 0.7118 |
2,000 mg q 8 h | 0.9998 | 0.9968 | 0.9322 | 0.9998 | 0.9982 | 0.9206 |
The optimal dosing regimens are noted in bold and italic.
PTA, probability of target attainment; CVVH, continuous venovenous hemofiltration; LD, loading dose.
PTA of all recommended meropenem dosing regimens for Gram-negative bacilli infections in CVVHD modality with 25 and 35 mL/kg/h effluent rates
Dosage regimens . | PTA over 48 h, MIC of 2 mg/L . | |||||
---|---|---|---|---|---|---|
effluent rates 25 mL/kg/h . | effluent rates 35 mL/kg/h . | |||||
40% fT > MIC . | 40% fT > 4MIC . | 100% fT > MIC . | 40% fT > MIC . | 40% fT > 4MIC . | 100% fT > MIC . | |
500 mg q 12 h | 0.9820 | 0.4636 | 0.1048 | 0.9774 | 0.3394 | 0.0700 |
500 mg q 8 h | 0.9982 | 0.7772 | 0.1646 | 0.9988 | 0.7316 | 0.1556 |
1,000 mg q 12 h | 0.9952 | 0.8818 | 0.5188 | 0.9946 | 0.8530 | 0.4680 |
1,000 mg LD 500 mg q 8 h | 0.9984 | 0.8570 | 0.6022 | 0.9986 | 0.8142 | 0.5760 |
1,250 mg LD 1,000 mg q 12 h | 0.9968 | 0.9092 | 0.6506 | 0.9940 | 0.8728 | 0.6000 |
750 mg q 8 h | 0.9996 | 0.9406 | 0.4206 | 0.9990 | 0.9216 | 0.4076 |
1,000 mg q 8 h | 0.9990 | 0.9778 | 0.6318 | 0.9998 | 0.9726 | 0.6252 |
1,500 mg q 12 h | 0.9964 | 0.9640 | 0.7668 | 0.9962 | 0.9514 | 0.7242 |
1,750 mg q 8 h | 0.9998 | 0.9984 | 0.9096 | 1.00 | 0.9970 | 0.8906 |
2,750 mg q 12 h | 0.9984 | 0.9908 | 0.9112 | 0.9982 | 0.9874 | 0.8922 |
2,000 mg q 8 h | 1.00 | 0.9992 | 0.9320 | 1.00 | 0.9986 | 0.9220 |
Dosage regimens . | PTA over 48 h, MIC of 2 mg/L . | |||||
---|---|---|---|---|---|---|
effluent rates 25 mL/kg/h . | effluent rates 35 mL/kg/h . | |||||
40% fT > MIC . | 40% fT > 4MIC . | 100% fT > MIC . | 40% fT > MIC . | 40% fT > 4MIC . | 100% fT > MIC . | |
500 mg q 12 h | 0.9820 | 0.4636 | 0.1048 | 0.9774 | 0.3394 | 0.0700 |
500 mg q 8 h | 0.9982 | 0.7772 | 0.1646 | 0.9988 | 0.7316 | 0.1556 |
1,000 mg q 12 h | 0.9952 | 0.8818 | 0.5188 | 0.9946 | 0.8530 | 0.4680 |
1,000 mg LD 500 mg q 8 h | 0.9984 | 0.8570 | 0.6022 | 0.9986 | 0.8142 | 0.5760 |
1,250 mg LD 1,000 mg q 12 h | 0.9968 | 0.9092 | 0.6506 | 0.9940 | 0.8728 | 0.6000 |
750 mg q 8 h | 0.9996 | 0.9406 | 0.4206 | 0.9990 | 0.9216 | 0.4076 |
1,000 mg q 8 h | 0.9990 | 0.9778 | 0.6318 | 0.9998 | 0.9726 | 0.6252 |
1,500 mg q 12 h | 0.9964 | 0.9640 | 0.7668 | 0.9962 | 0.9514 | 0.7242 |
1,750 mg q 8 h | 0.9998 | 0.9984 | 0.9096 | 1.00 | 0.9970 | 0.8906 |
2,750 mg q 12 h | 0.9984 | 0.9908 | 0.9112 | 0.9982 | 0.9874 | 0.8922 |
2,000 mg q 8 h | 1.00 | 0.9992 | 0.9320 | 1.00 | 0.9986 | 0.9220 |
The optimal dosing regimens are noted in bold and italic.
PTA, probability of target attainment; CVVHD continuous venovenous hemodialysis; LD, loading dose.
For CVVHD modality, the regimens of 1,250 mg LD then 1,000 mg every 12 h and 750 mg every 8 h were suggested to achieve the PD target of 40% fT > 4MIC in effluent flow rates of 25 and 35 mL/kg/h, respectively. The optimal doses to achieve 100% fT > MIC were 1,750 mg every 8 h (effluent flow rate 25 mL/kg/h) and 2,000 mg every 8 h (effluent flow rate 35 mL/kg/h).
All necessary variables for dosing the drug including pharmacokinetic parameters, CRRT setting, and patient characteristics are detailed in Table 4. The recommendations of meropenem dosing regimens for treating Gram-negative bacilli infections with various PD targets in critically ill patients undergoing CRRT from the simulations are summarized in Table 5. Figure 3 demonstrates the PTA results of meropenem dosing regimens at different MICs in pre-dilution CVVH with 25 mL/kg/h effluent rates for treatment of Gram-negative bacilli infection in virtual patients for the first 48 h.
Virtual patient characteristics and key pharmacokinetic parameters
Parameters . | Simulation-based values (mean±SD [range limits])(N = 5,000) . |
---|---|
Weight, kg | 86.60±29.20 (40–8) |
Vd, L/kg | 0.39±0.17 (0.08–1.07) |
CLNR, mL/min | 51.30±41.35 (0–251.55) |
r2 weight and Vd | 0.1215 |
r2 weight and CLNR | 0.0331 |
Residual renal clearance, mL/min | 0 |
SC | 0.99±0.19 (0.21–1.78) |
SA | 0.77±0.31 (0.15–1.13) |
Qb, L/h | 12 |
Hct | 30% |
Qf or Qd | 25 mL/kg/h |
35 mL/kg/h | |
Free fraction | 0.79±0.09 |
PD target | 40% fT > MIC |
40% fT > 4MIC | |
100% fT > MIC |
Parameters . | Simulation-based values (mean±SD [range limits])(N = 5,000) . |
---|---|
Weight, kg | 86.60±29.20 (40–8) |
Vd, L/kg | 0.39±0.17 (0.08–1.07) |
CLNR, mL/min | 51.30±41.35 (0–251.55) |
r2 weight and Vd | 0.1215 |
r2 weight and CLNR | 0.0331 |
Residual renal clearance, mL/min | 0 |
SC | 0.99±0.19 (0.21–1.78) |
SA | 0.77±0.31 (0.15–1.13) |
Qb, L/h | 12 |
Hct | 30% |
Qf or Qd | 25 mL/kg/h |
35 mL/kg/h | |
Free fraction | 0.79±0.09 |
PD target | 40% fT > MIC |
40% fT > 4MIC | |
100% fT > MIC |
Recommendations of meropenem dosing regimens for treating Gram-negative bacilli infections with various PD targets in critically ill patients undergoing CRRT
PD target . | Effluent rates, mL/kg/h . | CVVH (pre-hemofilter dilution) . | CVVHD . |
---|---|---|---|
40% fT > MIC | 25 mL/kg/h | 500 mg q 12 h | 500 mg q 12 h |
35 mL/kg/h | |||
40% fT > 4MIC | 25 mL/kg/h | 750 mg q 8 h | 750 mg q 8 h* |
35 mL/kg/h | 750 mg q 8 h | ||
100% fT > MIC | 25 mL/kg/h | 2,000 mg q 8 h | 1,750 mg q 8 h |
35 mL/kg/h | 2,000 mg q 8 h |
PD target . | Effluent rates, mL/kg/h . | CVVH (pre-hemofilter dilution) . | CVVHD . |
---|---|---|---|
40% fT > MIC | 25 mL/kg/h | 500 mg q 12 h | 500 mg q 12 h |
35 mL/kg/h | |||
40% fT > 4MIC | 25 mL/kg/h | 750 mg q 8 h | 750 mg q 8 h* |
35 mL/kg/h | 750 mg q 8 h | ||
100% fT > MIC | 25 mL/kg/h | 2,000 mg q 8 h | 1,750 mg q 8 h |
35 mL/kg/h | 2,000 mg q 8 h |
CVVH, continuous venovenous hemofiltration; CVVHD, continuous venovenous hemodialysis; LD, loading dose.
*The dosing recommendation of 750 mg q 8 h is suggested as it is a more practical option than a dose of 1,250 mg LD then 1,000 mg q 12 h.
PTA results of meropenem dosing regimens at different MICs in pre-dilution CVVH with 25 mL/kg/h effluent rates for treatment of Gram-negative bacilli infection in virtual patients for the first 48 h. CLCRRT, CRRT clearance; CLNR, nonrenal clearance; CLTOT, total clearance; CRRT, continuous renal replacement therapy; CVVH, continuous venovenous hemofiltration; CVVHD, continuous venovenous hemodialysis; CVVHDF, continuous venovenous hemodiafiltration; Hct, hematocrit; PD, pharmacodynamic; Qb, blood flow rate; Qd, dialyzate flow rate; Qf, ultrafiltrate rate; SA, saturation coefficient; SC, sieving coefficient; Vd, volume of distribution.
PTA results of meropenem dosing regimens at different MICs in pre-dilution CVVH with 25 mL/kg/h effluent rates for treatment of Gram-negative bacilli infection in virtual patients for the first 48 h. CLCRRT, CRRT clearance; CLNR, nonrenal clearance; CLTOT, total clearance; CRRT, continuous renal replacement therapy; CVVH, continuous venovenous hemofiltration; CVVHD, continuous venovenous hemodialysis; CVVHDF, continuous venovenous hemodiafiltration; Hct, hematocrit; PD, pharmacodynamic; Qb, blood flow rate; Qd, dialyzate flow rate; Qf, ultrafiltrate rate; SA, saturation coefficient; SC, sieving coefficient; Vd, volume of distribution.
Discussion
This systematic review indicates that the current literature does not provide complete dataset parameters for drug dosing in patients undergoing CRRT. The CRRT modality and setting, pharmacokinetic parameters (e.g., Vd, CLNR, and residual renal clearance) as well as PD target should be considered to design the dosing regimen in these populations.
Effect of Modality and Filter Membrane on CRRT Clearance
Theoretically, extracorporeal drug clearance depends on drug molecular weight and CRRT modality. The diffusion technique is more effective for small molecular weight solutes than the convection technique [3]. Hence, meropenem, with a low molecular weight of 437.52 g/mol [4], should be lost more in CVVHD modality than CVVH modality. Owing to the combination techniques, the CVVHDF modality should therefore have had the highest extracorporeal clearance. However, drug clearance via CRRT is defined as the sieving coefficient multiplied by the effluent rate (Qf in CVVH, Qd in CVVHD, or Qf+Qd in CVVHDF). The key parameter that affects drug clearance is the effluent flow rate rather than the modality. Either of CVVH or CVVHD modality with the same effluent flow rate, drug clearances are similar. In addition, Shaw and Mueller [32] noted that there is little difference in CRRT clearance between CVVHD and CVVH or CVVHDF, especially for small molecules including meropenem [33].
Unfortunately, only 71.4% of the previous studies reported the SC or SA parameter. Giles et al. conducted the meropenem pharmacokinetic study in patients undergoing two modalities, CVVH (hemofiltration rate 1–2 L/h) and CVVHDF (hemofiltration rate 1–2 L/h; dialysis rate 1–1.5 L/h). Fractional drug clearance by the machine was reported in 52.6% and 42.3% in CVVHDF and CVVH, respectively, whereas sieving coefficients were similar [14]. Thalhammer et al. [5] and Verves et al. [15] reported the different SC values while using the different type of hemofilter, SC 1.09 ± 0.10 (high-flux polysulfone) compared with 0.63 ± 0.252 (high-flux polyacrylonitrile), respectively. Modality and CRRT setting of both studies were comparable, as well as severity and patient characteristics. This result demonstrated that the characteristics of dialyzer filter membrane, e.g., membrane composition, surface area, and pore diameter, might affect the extracorporeal drug clearance.
Drug loss via pre-dilution CVVH modality might be less than the post-dilution method due to the dilutional effect. Pre-dilution method requires additional parameters to calculate drug clearance, i.e., blood flow rate, hematocrit as well as replacement flow rate [2]. 66.7% of the modality with convection technique (CVVH and CVVHDF) stated the dilution method. Nearly one-half of them utilized pre-dilution method, but only 25% of the studies reported the hematocrit value. There were no studies that directly compared the pharmacokinetic meropenem parameters in pre-dilution or post-dilution CVVH.
Effect of Pharmacokinetic Parameters
Physiological change in critically ill patients can significantly affect the pharmacokinetics of drugs used in the critically ill patient population. Large Vd is frequently observed in these populations due to septic shock and alteration of protein binding. Blood concentration of hydrophilic antibiotics such as meropenem might be sub-therapeutic. 90.5% of the studies reported Vd parameters with the average of 0.39 ± 0.17 L/kg or 27.2 ± 11.9 L. These results show that the AKI critically ill patients had a higher Vd than normal patients or healthy volunteer population (0.23–0.35 L/kg or 15–20 L) [34].
CLNR is responsible for approximately 20% of total meropenem clearance in healthy subjects with normal renal function, total clearance 186 mL/min/1.73 m2, and CLNR 44 mL/min/1.73 m2, respectively. In patients with GFR between 5 and 29 mL/min/1.73 m2, the CLNR increases to 50% of the total clearance (CLNR 29 mL/min/1.73 m2) [35]. Consequently, CLNR plays a major role in the total clearance in renal impairment patients. Because of the difficulty in the direct quantitative assessment, CLNR was clearly stated in only 28.6% of the reviewed studies with the mean of 51.30 ± 41.35 mL/min. Nevertheless, we were able to obtain the CLNR from the indirect method via calculation, where CLNR can be calculated by the difference between total clearance and CRRT clearance in anuric patients. From the studies reviewed, we were able to calculate CLNR in this fashion in three studies [5, 16, 17].
Effect of PD Parameters
Animal study has shown that the bacteriostatic effects of meropenem are observed at 20% f T > MIC and bactericidal effects can be achieved at 40% f T > MIC [36]. However, a higher target as 4-fold MIC is associated with maximum bactericidal activity of ß-lactams [37]. Additionally, maintaining meropenem concentration above the MIC for at least 75% of the time contributed to a higher clinical response rate in the febrile neutropenic patients with bacteremia and the patients with severe lower respiratory tract infections [38, 39]. Thus, three PD targets were used in our simulations.
In this systematic review, the wide variability of dosing recommendation from 250 mg every 24 h to 1,000 mg every 8 h were recommended. Kawano et al. [20] recommended meropenem dosing regimens to achieve the PD target of more than 40% T > MIC. For patients infected with isolates that have MICs of 2, 4, 8, and 16 mg/L, the optimal doses were 250 mg every 24 h, 250 mg every 12 h, 250 mg every 8 h, and 500 mg every 8 h, respectively]. In comparison with the Seyler et al. [6] study, the dosing regimen of 1,000 mg every 8 h was recommended to achieve 40%T > 4MIC (MIC of 2 mg/L). Evidently, PD targets clearly impact drug dosing regimen and should always be considered as well as the CRRT setting, severity of illness, and pharmacokinetic changes in critically ill patients. In the situation that requires high PD target, e.g., neutropenia, critically ill patient, or high MIC pathogen infection, the higher dose of meropenem should be considered.
Effect of Patient Demographic
Most pharmacokinetic studies did not provide enough patient demographic data for drug dosing adaptation in clinical practice. Weight variation correlated with the Vd. Similarly, the severity of illness and protein binding affinity may also affect Vd and free drug concentration. In addition, approximately 70% of meropenem is excreted unchanged in the urine [40]. Therefore, residual renal function is one of the key factors concerning for drug dosing in clinical applications.
Optimal Dosage Regimens of Meropenem
Our simulation results suggest that the meropenem dosing regimen of 750 mg every 8 h was the optimal dose to attain 40% fT > 4MIC (MIC of 2 mg/L) in two modalities with the effluent flow rate of 35 mL/kg/h. This regimen was aligned with previously published reports; however, body weights and Vd were higher in our study [41]. The studies by Jamal et al. [19] and Seyler et al. [6] recommended the dosing regimens of 1 g every 8 h with and without 2 g loading, respectively. These recommended doses are higher than ours. This could be due to their reported Vd (0.43–0.45 L/kg) and SC (1.10) being slightly higher than our study.
In contrast to the results of several resources, recommended meropenem dosing regimens of 250–1,000 mg every 12–24 h for the patients receiving CRRT could not achieve the PK/PD targets in our models with virtually critically ill patients. The possible reasons for the lower doses than ours were the difference in prespecified PD target and most of them reported the Vd that was lower than ours (Table 1) [11, 13, 15, 20, 28, 29].
We developed a pharmacokinetic model for adult anuric AKI patients undergoing uninterrupted CRRT for at least 48 h. The virtual patients were constructed, using pharmacokinetic data from previous literature. Thus, patient characteristics such as body weight and residual renal function should be considered along with CRRT characteristics when determining the optimal meropenem dosage in real-life practice to avoid supra- or sub-therapeutic drug concentration. The average body weight used in our study was 86.60 ± 29.20 kg; therefore, the dose of 750 mg every 8 h is the optimal dose for Gram-negative bacilli infection with an MIC of 2 mg/L. According to the PTA which nearly achieves 90% (Tables 2,,3), the lower dose of 1,000 mg every 12 h or 1,000 mg LD, followed by 500 mg every 8 h can be used for patients with lower body weight than our model. In addition, the MIC breakpoint of 2 mg/L, the highest value cut-off for susceptible Gram-negative bacilli, was utilized in this study to ensure appropriate efficacy. This would be the worst scenario based on MIC breakpoint recommended by CLSI in our models. Dosing modification according to infection with different MIC value pathogen should be individually adjusted as illustrated in Figure 3.
Additionally, the results of our study recommended initial meropenem dosing regimens that are needed to achieve optimal PK/PD targets in the first 2 days. However, PK changes in critically ill patients are considerably varied and unpredictable. We recommend monitoring any physiological changes closely and adjusting doses based on individualized patient circumstances.
Limitation
Limitations of our results include: (1) the study search was limited to those published in English, (2) our dosing recommendations should be used for patients who share similar characteristics and CRRT settings to our model, and (3) these results need to be validated in a clinical study.
Conclusion
None of the current studies revealed all necessary pharmacokinetic parameters to individualize drug dosing. PD targets significantly contributed to meropenem dosage regimens in these patients. The current study suggests that 750 mg every 8 h was the optimal dose for achieving the PD target of 40% fT > 4MIC. Different effluent rates and types of CRRT shared similar dosing regimens. Further exploring clinical outcomes utilizing our recommendations is necessary.
Acknowledgments
We would like to thank Associate Professor Dr. Wibul Wongpoowarak for the helpful comments of our simulations in the study and Associate Professor Dr. Chalermsri Pummangura for all great advice and support.
Statement of Ethics
The ethics statement is not applicable because this study is based exclusively on the published literature. The study has been granted an exemption from requiring written informed consent by the Ethics Committee of Faculty of Pharmacy, Siam University (COA. No. SIAMPY-IRB 2018/001.01).
Conflict of Interest Statement
The authors have no conflicts of interest to declare.
Funding Sources
This study was supported by the funding from the Siam University. The funding source had no role in the study design; the collection, analysis, and interpretation of data; the writing of the article; and the decision to submit it for publication.
Author Contributions
Taniya Charoensareerat, Weerachai Chaijamorn, and Sutthiporn Pattharachayakul equally contributed to the conception and design of the research; Taniya Charoensareerat, Dhakrit Rungkitwattanakul, Weerachai Chaijamorn, and Sutthiporn Pattharachayakul drafted the manuscript; and Taniya Charoensareerat, Weerachai Chaijamorn, Pathakorn Kerdnimith, Nutsinee Kosumwisaisakul, Piyakamol Teeranaew, Dhakrit Rungkitwattanakul, Apinya Boonpeng, Nattachai Srisawat, and Sutthiporn Pattharachayakul equally contributed to the acquisition, analysis, and interpretation of the data; critically revised the manuscript; and read and approved the final manuscript.
Data Availability Statement
All data generated or analyzed during this study are included in this article and its online supplementary material. Further inquiries can be directed to the corresponding author.