Population plasma and urine pharmacokinetics and the probability of target attainment of fosfomycin in healthy male volunteers

Using plasma and urine samples, our study characterised the fosfomycin population PK and subject characteristics affecting the PK of fosfomycin, which serves as initial data for developing further nomograms for fosfomycin. The present finding suggests the sufficiency of intravenous fosfomycin standard dosing regimens of 4 g every 8 h as an intermittent infusion using the target of AUC24h/MIC 83 for 1-log bacterial reductions with a MIC of 4 mg/L, which is the current epidemiological cutoff values (ECOFF) of fosfomycin for E. coli. However, the ECOFF for fosfomycin is much lower than the EUCAST clinical breakpoint of 32 mg/L [12]. For the current clinical breakpoint of 32 mg/L, the currently recommended standard dosing regimen of 4 g every 8 h might not be sufficient, and a higher dosing of 8 g every 8 h as an intermittent infusion or 0.75 g/h as a continuous infusion might be required to reach the AUC24h/MIC 83 target.

Our results seem to be in line with the EUCAST current clinical breakpoint as no simulated dosing regimen is considered adequate to attain AUC24h/MIC target for the purpose of killing resistant bacteria (MIC > 32 mg/L). To exhibit the bacteriostatic effect of fosfomycin with the target of AUC24h/MIC 25 against E. coli with MIC 32 mg/L, 4 g every 12 h may be adequate. Thus, combining fosfomycin with other antibiotics should be considered as the main alternative to demonstrate the bactericidal effect against fosfomycin-resistant E. coli and to prevent resistance in general. A combination of fosfomycin and cefixime exhibited synergistic to kill resistant E. coli based on a previous in vitro study [36].

As concentration at the local site of infection is critical, simulations analysing the urine PK/PD target (AUC24h/MIC 3994) were performed. The present evaluation indicated that the standard daily dose schedule of intravenous fosfomycin (4 g every 8 h) can be used to treat urinary tract infections caused by susceptible E. coli with a MIC of 8 mg/L. However, it may not be effective against E. coli with a MIC of 16 mg/L meaning that if the dosing regimen attains the plasma target, the urinary target is not necessarily attained. This supports the results of the established in vitro dynamic bladder infection model, which also proved that E. coli was hardly killed with a MIC of 16 mg/L [35]. This result also suggests that the current breakpoint might be too high for the urine target of AUC24h/MIC 3994. Nevertheless, using only urine samples is generally not considered the best choice for therapeutic drug monitoring (TDM), since the collection of urine samples can be burdensome for some patients and thus may not be feasible in the majority of cases. Our study offers the feasibility of predicting fosfomycin exposure in urine by measuring plasma concentration and using a urine-plasma exposure ratio of 21,2. Yet, the link between urine and plasma concentration should be investigated more comprehensively to optimise the plasma PK/PD target that is strongly related to efficacy in the urinary tract.

The PK/PD index related to the time-killing activity of fosfomycin against E. coli remains disputed, as the target of %T>MIC was less robust than the AUC/MIC target in an earlier in vivo study [33]. Yet, a previous in vitro pharmacodynamic study proved that three out of the five E. coli strains investigated in the study seemed to be more time-dependent and two others were concentration-dependent [37]. To treat E. coli strains that are more time-dependent, our simulation proposed continuous infusion rather than intermittent infusion since the concentration remains constantly above the MIC. This is in line with al Jalali et al.’s noncompartmental analysis demonstrating that fosfomycin as a continuous infusion resulted in an improvement in the attainment of PK/PD determinants [15]. Our simulation of PTA targeting 75%T>MIC can be used as a direction to treat E. coli, and these results need to be validated in clinical studies.

Although fosfomycin is considered a safe drug and is generally well tolerated, possible fosfomycin toxicity was observed in this study. Two subjects experienced thrombophlebitis after continuous infusion administration [15]. A previous literature review and analysis of the reporting system database demonstrated that peripheral phlebitis was one of the most frequent adverse events of parenteral fosfomycin [14], which is in line with the result of the current study. Thrombophlebitis might be related to fosfomycin exposure, as these were observed in the two subjects with the highest maximum concentrations. Unfortunately, the evidence of fosfomycin toxicity, especially thrombophlebitis, and its relation to its exposure is lacking. No data on the threshold of fosfomycin toxicity are available.

It is known that fosfomycin is eliminated primarily by the kidneys, as a result, the correlation between eGFR and clearance of fosfomycin was expected, which was also the case in the previous studies showing kidney function as a significant covariate [22, 36, 37]. It is certainly assumed, but not confirmed by this analysis, that there is a need for dose reduction in patients with renal insufficiency based on eGFR calculated with the CKD-EPI formula. We tested the effect of kidney function on the fosfomycin PK using three different markers, including serum creatinine, eGFR calculated with the CKD-EPI formula, eGFR normalised to 1.73 of body surface area, and CrCl calculated with the Cockcroft–Gault equation. In this analysis, eGFR (mL/min) showed a strong correlation with CL and is considered a better kidney function marker than CrCL by Cockcroft–Gault, especially for some populations where it could underpredict. This finding is not in line with the product information, where the standard dosing regimen of fosfomycin is adjusted based on CrCl [11]. Therefore, to achieve a more optimal dosage, this model suggests adjusting dosing based on eGFR instead of CrCl.

The model showed a good ability to characterise fosfomycin PK based on model diagnostics and validation. Most other recent studies also confirmed that the two-compartment model was the best fit to describe the fosfomycin plasma PK [17,18,19,20]. The second exponential decay suggests a distribution into deeper tissue, which leads to a slower release into plasma. It has been reported that fosfomycin penetrates extensively into the interstitial fluid of soft tissues [38,39,40]. Parameter estimates calculated in this study were comparable with those previous studies in healthy volunteers [15, 21,22,23, 41, 42].

Despite the successful PK model development using plasma and urine samples, this study has potential limitations. The data are from healthy males, and extrapolation to other groups like severely ill patients and women is limited. The PTAs of fosfomycin in patients, especially those with impaired renal function, can be higher than in healthy volunteers, and thus the dosing recommendation needs to be reduced. However, the dosing recommendation for patients with an eGFR of < 90 mL/min was not evaluated in this study. Therefore, a future external validation study should be conducted to assess the model’s fit for the target (extrapolated) population with a wider range of eGFR values, beyond those observed in our study participants. Another limitation is related to the urine data, which was modelled with high residual variability. The recovery in urine was over 100% in six subjects, which was possibly due to errors in the urine collections, which were done manually, or due to small assay errors. Fortunately, both the original and modified datasets were modelled, resulting in similar parameter estimates. Our model indicated that fosfomycin is likely to be close to 100% excreted via kidneys as the estimate of nonrenal clearance was very low. However, due to the limitation of our study, it is not conclusive, so more research is needed to confirm this finding. Meanwhile, Wenzler et al. evaluated the plasma and urine PK of a single dose of 8 g fosfomycin in healthy volunteers using noncompartmental analyses, resulting in a total clearance of 7.8 L/h and renal clearance of 6.3 L/h [41]. This means that approximately 20% of fosfomycin was eliminated via other routes besides the kidneys. However, the second elimination route was not described in Wenzler’s study. It is known that fosfomycin is not metabolised in the human body and is mainly excreted unchanged in urine through glomerular filtration [43]. In the case of patients with decreased renal function, other routes of fosfomycin elimination may occur, such as biliary excretion, as fosfomycin has been detected in the bile [34, 44, 45]. Still, the understanding of fosfomycin elimination is very limited as the data are scarce. Moreover, the presence of fosfomycin metabolites is hardly studied because there is no validated assay to analyse it.

In conclusion, our simulation suggests that the dose of 4 g every 8 h is probably optimal to treat E. coli isolates within the wild-type distribution (ECOFF) with MICs of ≤ 4 g/L for both systemic and urinary tract infections. For the above wild-type MIC and up to the current clinical breakpoint of 32 mg/L, the dosages of 8 g every 8 h and 0.75 g/h may be needed. Although no dosing regimen may be able to kill E. coli with MIC above the clinical breakpoint, all simulated dosing regimens can be used to inhibit E. coli from reproducing using the AUC24h/MIC 25 target. This PK model and simulation results are in line with the current clinical breakpoint. Dosing guidelines based on eGFR calculated with the CKD-EPI formula instead of CrCL by Cockcroft–Gault should be further developed and investigated to treat E. coli infections with a more optimal dose.

留言 (0)

沒有登入
gif