The recommended dosage regimen for caspofungin in patients with higher body weight or hypoalbuminaemia will result in low exposure: Five years of data based on a population pharmacokinetic model and Monte-Carlo simulations

Background: To develop a population pharmacokinetic (PPK) model for caspofungin, identify parameters influencing caspofungin pharmacokinetics, and assess the required probability of target attainment (PTA) and cumulative fraction of response (CFR) for various dosing regimens of caspofungin in all patients and intensive care unit (ICU)-subgroup patients. Method: The general PPK model was developed based on data sets from all patients (299 patients). A ICU-subgroup PPK model based on data sets from 136 patients was then analyzed. The effects of demographics, clinical data, laboratory data, and concomitant medications were tested. Monte-Carlo simulations (MCS) were used to evaluate the effectiveness of different caspofungin dosage regimens. Results: One-compartment model best described the data of all patients and ICU patients. Clearances (CL) were 0.32 L/h and 0.40 L/h and volumes of distribution (V) were 13.31 L and 10.20 L for the general and ICU-subgroup PPK models, respectively. In the general model, CL and V were significantly associated with albumin (ALB) concentration and body weight (WT). In the ICU-subgroup model, CL was associated with WT. The simulated exposure in ICU patients was lower than that in all patients (p < 0.05). MCS indicated that higher caspofungin maintenance doses of 70–150 mg may achieve target CFR of >90% for patients with higher WT (>70 kg) or with C. albicans or C. parapsilosis infections, and especially for ICU patients with hypoalbuminaemia. Conclusion: The PPK model and MCS presented in the study demonstrated that the recommended dosage regimen for caspofungin in patients with higher body weight or hypoalbuminaemia will result in low exposure.

A caspofungin population pharmacokinetic (PPK) study in 9 ICU patients undergoing CRRT obtained the full concentration-time profiles, and the plasma samples were taken at 2, 4, 8, 12 and 24 h after the start of caspofungin infusion (Roger et al., 2017). Caspofungin steady-state trough (Cmin) and Cmax plasma levels were determined in two studies of caspofungin PK for an ICU patient with liver cirrhosis (Spriet et al., 2011) and a patient during extracorporeal membrane oxygenation (Spriet et al., 2009).
Plasma samples from one study of caspofungin low exposure in 20 ICU patients were taken on day 3 (±1 day) just before administration of caspofungin at 1, 2, 3, 4, 6, 8, 12, and 24 h after the start of the infusion (van der Elst et al., 2017). The last one research we brought into the study was a study about caspofungin PK in 6 ICU patients, caspofungin steady-state Cmin and Cmax plasma samples were collecter and determined (Sinnollareddy et al., 2015).

General PPK model
As indicated in Table 2, ALB, mycophenolate mofetil, cyclosporine (CYC), methylprednisolone, WT, CMT, SOP, SOT and ICU were found to have impacts on the caspofungin clearance (CL), while ALB, WT and CMT had impacts on the volume of distribution (V) according to the basic model. Based on 299 patients, we found that ALB, CYC, WT and CMT could be considered as significant covariates for CL, whereas ALB and WT were significant covariates for V in the full model.

Model validation
The mean values of the overall parameters of the final models were close to that of the bootstrap validation method. The internal verification results showed that the two final models were stable and robust, and the estimated parameter values were reliable. The mean values and 95% confidence intervals of the corresponding parameters are shown in Table 3.

Pharmacokinetics of caspofungin in all patients
Based on the Kruskal-Wallis test, the area under the plasma concentration-time curve (AUC) and CL were different among the four patient groups (Figure 3a, P < 0.05). What's more, the relative clinical data (WT, ALB) were different among the four patient groups (P < 0.05). As shown in Table 4, 86.3% (44/51) ICU patients had ALB concentrations < 35 g/L and 70.6% (36/51) of them had WT ≤ 70 kg.

Pharmacokinetics of caspofungin in ICU patients
Based on the Kruskal-Wallis test, the AUC and CL were different between these ICU patients and all patients from the general model (P < 0.05). The mean AUC in these ICU patients was 101.83 mg·h/L. The AUC of the ICU patients from the teaching hospital was a little higher than that of the ICU patients from other seven studies, and they all had low ALB concentrations.

Dosing simulations for all patients
The PK/PD targets for the various dosage regimens under different WT conditions were simulated with MCS. Table S1

Dosing simulations for ICU patients
There were big differences in the exposure (AUC24) of caspofungin between all the five dosage regimens and different WT conditions (P < 0.05) for all patients ( Figure 5a) and ICU patients ( Figure   5b). For ICU patients, as show in Table S2,  Caspofungin at 70/35 mg did not achieved the required CFR for C.albicans and C.parapsilosis for all ICU patients with different weight groups (WT ≤ 70 kg or >70 kg). Caspofungin at 70/50 mg did not achieved the required CFR for C.albicans and C.parapsilosis for patients with WT >70 kg.
Caspofungin at 100/100 mg could achieve the required CFR for C.parapsilosis for all ICU patients with different weight groups.