Table 1.

Summary of studies included in mini review of using PTA to optimize cefepime dosing.

Study	Simulated Population	tvCL in Model	Simulated Doses	PD Target and Definition of Success	TD Target	Assumed Fraction Unbound	Simulation of PK Parameters	Simulated Time Intervals	Final Dose Recommendation
Álvarez	2500 patients with hematologic malignancy	13.6 L/h scaled to sCR of 0.47	4–8 G/day as 30-min infusions, extended infusions or continuous infusions	fT 60% or 100% > MIC = 4 mg/L or 8 mg/L. PTA 90%	Toxicity not analyzed	0.8, fixed	Pop PK model developed from plasma samples of 15 patients	Not specified, first 24 h and at steady state	2 G q 8 with extended 4-h infusions achieves lenient PD target and 6 g/day CI achieves all targets.
Butterfield-Cowper	5000 patients	5.43 L/h scaled to CrCl of 80 mL/min	30-min and 5-min infusions of 1 G q 6 h and 2 G q 8 h	70% fT 70% > 1 × MIC at hour 24 of therapy	Toxicity not analyzed	0.8, fixed	2-cmpt pop PK model developed by Tam et al.	Not specified	Minimal difference in PK profile from 30-min to. 5-min infusion.
Chaijamorn	5000 anuric CKRT patients	1.46 L/h (patients assumed to be anuric)	1 to 2 g every 12 h to 2 g loading dose followed by 1 g every 8 h or 2 g every 12 h	≥70% fT > 4 × MIC = 8 mg/L in a 48-h time period. PTA 90%	Probability of trough >= 70 mg/L at end of 48-h interval	0.79, simulated with mean and SD	Log-normal distribution based on 1-compartment PK model developed via a literature search	Not specified, initial 48 h	1.75–2 G loading dose followed by 1.5–2 G q 8 h.
Delattre	1000 patients per group	4.5 L/h scaled to CrCl of 100 mL/min and weight of 70 kg	4 g or 6 g administered as a 0.5-h, 2-h or 3-h infusion every 8 h	70%T > 4 × MIC ≤8 mg/L within a dosing interval. PTA 90%	Toxicity not analyzed	1, fixed	Pop PK developed from 88 critically ill patients	Not stated	4 G loading dose infused over 3 h followed by 4 G q 6 h.
Huang	10,000 healthy patients	5.3 L/h corresponds to CrCl of 100–120 mL/min (need to follow up on Nye et al.)	1 g every 12 h (q 12 h), 1 g every 8 h (q 8 h), 2 g q 12 h, and 2 g q 8 h as an IV bolus (an assumption for the equation to generate %fT > MIC)	fT 50% ≥ MIC based on observed MIC distribution and 90% CFR defined as success	Toxicity not analyzed	0.8–0.9 uniformly distributed.	CL estimated from a study of healthy volunteers Nye et al., %fT generated from an equation.	Equation used, steady state	2 G q 8 h IV bolus PTA achieved > 90% to MIC 16 mg/L; however, adequate for non-esbls, not adequate for esbl based on CFR
Jang	10,000 patients receiving CRRT	1.46 L/h (assumed to be anuric)	Cefepime 1 and 2-g q 8 or q 12 h over 30-min infusion	fT 60% ≥ MIC of 8 mg/L (also 4XMIC tested) in 72-h time period. PTA 90%	Toxicity not analyzed	0.79, fixed	Log-normal distribution based on 1-compartment PK model developed via a literature search	Not stated, 72 h of initial therapy	No dose recommendation, 2 G q 8 achieved > 90% PTA in all simulated subgroups.
Lau	12,000 patients	2.29 L/h scaled to CrCl 60 mL/min with linear model	Per Australian dosing guidelines	Cmin > 32 mg/L. PTA 90%	49 mg/L derived via ROC analysis	Not specified, assumed to be 1	Via population PK model developed by Jonckheere et al.	Cmin at steady state	No recommendation, 89% of patients with CrCl > 50 mL/min would achieve PTA dosing of 2 G q 8 h.
Liu	1000 patients per group, fixed at 70 kg and varied CrCl	5.65 L/h scaled to CrCL 120 mL/min and 70 kg	1–2 G q 8–12 as 2-, 5- or 30-min infusions	70%fT > MIC. CFR based on SENTRY database of MIC distributions. PTA 90%, CFR 90%	Toxicity not analyzed	0.8, fixed	Pop PK model developed from 70 patients and 604 cefepime concentrations	Not specified, evaluated 1st dose	IVP is not likely to be as good as intermittent infusion. No regimen meets the 90% threshold for MIC > 8 mg/L in patients with CrCl > 60, but CFR is > 90% for 2 G q 8 h based on MIC distributions.
Koomanachai	5000 patients	6.04 L/h scaled to CrCl 103.74 mL/min per equation in Tam et al.	2 g every 12 h (0.5-h infusion) or 2 g every 8 h (0.5-h and 3-h infusion)	≥50% fT > MIC. CFR >= 90% against observed MIC distribution	Toxicity not analyzed	Not stated	Simulated used Tam et al.	Not stated, steady state	2 g q 8 h infused over 3 h achieved CFR > 80–90%
Rhodes	10,000 patients with CrCl simulated range 108–220 mL/min	6.33 L/h scaled to CrCl of 120 m/min	3–8 G/day infused over 0.5–24 h q6–12 h or CI	≥68% fT > 1 × MIC in first 24 h of therapy. PTA 90%	Toxicity not analyzed	0.8, fixed	2-cmpt Pop PK model developed via cefepime concentration data from 9 patients	Simulated every 0.5 h, first 24 h of therapy	3–4 g/day as continuous infusions and doses of 2 g administered q 6 h (0.5-h infusion) to q 8 h (2-h infusion)
Sember	5000 anuric patients receiving CRRT	1.46 L/h (patients assumed to be anuric)	2-g loading dose (LD) infused over 0.5 h, followed by 1 or 2-g every 8 or 12 h with a 4-h extended-infusion.	≥60% fT > 4 × MIC = 8 mg/L in a 48-h time period. PTA 90%	Probability of trough >= 20 mg/L at end of 48-h interval	0.79, fixed	Log-normal distribution based on 1-compartment PK model developed via a literature search	Every 0.1 h for initial 48 h	2 G load followed by 2 G q 8 h
Shaw	5000 anuric patients receiving CRRT	1.49 L/h (patients assumed to be anuric)	1 to 2 g every 8–12 h to 2 g with or without load 2 G loading dose	≥60% fT > 1 × MIC or 4 × MIC = 8 mg/L in a 48-h time period. PTA 90%	Toxicity not analyzed	0.79, fixed	Log-normal distribution based on 1-compartment PK model developed via a literature search	Not specified, initial 72 h	No recommendation, but 2 G q 12 achieved 100% PTA in lenient target and 88.58% in strict target.
Thompson	10,000 patients with Cystic Fibrosis	8.47 L/h scaled to CrCl of 111.11 mL/min	2 g every 8 h (bolus and prolonged infusion)	≥60% or 100% fT > MIC against observed MIC distribution in CF patients (MIC50 = 16 mg/L). PTA 90%.	Toxicity not analyzed	0.8, fixed	Simulated via equations using steady state CL from Huls et al.	Equations used N/A	2 G CI achieves 66% PTA success and therefore is not adequate to cover resistant pseudomonal strains in CF population
Wang	5000 patients with CrCl >= 50 mL/min	9.18 L/h, which scales to a CrCl of 166.25 mL/min as calculated from Nicasio’s equation for CLT = 0.048 × CLCR + 1.2	1 g q 12 h or 2 g q 12 h as 30-min infusion or 2 g q 12 h as 3-h infusion	50% fT > MIC within dosing interval based on observed MIC distribution with CFR 90% defined as success	Toxicity not analyzed	0.85, fixed	Used Pop PK developed by Nicasio et al.	Not stated, evaluated at steady state	2 g q 12 h, 3 h; and cefepime 2 g q 12 h, 0.5 h had CFR of 80–90% which was considered suboptimal and therefore other antibiotics were recommended.