Clinical Pharmacokinetics and Pharmacodynamics of Micafungin

The PK in healthy subjects has been investigated in both single- and repeated-dose studies [19‐24]. Table 1 shows the PK parameters and weighted averages. After a 100 mg single dose infused in 60 min, typical values for maximum plasma concentration (C _max) of 9.1 mg/L, trough plasma concentration 24 h after dosing (C _24h) of approximately 2 mg/L, area under the concentration–time curve (AUC) from zero to infinity (AUC_∞) of 133 mg h/L, clearance (CL) of 10.4 mL/h/kg, apparent volume of distribution (V _d) of 0.2 L/kg, and terminal half-life (t _½) of 15.4 h were found. Additionally, after a daily 150 mg dose infused in 60 min, C _max values after a single dose and at steady state of 12 and 16 mg/L were achieved, respectively; C _24h after a single dose and at steady state was approximately 2.5 and 4.5 mg/L, respectively; and AUC from zero to 24 h (AUC₂₄) after a single dose and at steady state of 116 and 181.5 mg h/L, respectively, were reported.

In a repeated-dose study in healthy volunteers, HIV patients and hematology patients receiving micafungin 50–150 mg daily, the drug was found to show accumulation in the body over time, with a ratio of approximately 1.5 [23, 25, 26]. This accumulation factor describes the elimination of micafungin from the body in relation to the dose interval, meaning that for micafungin administered daily, the AUC₂₄ at steady state was approximately 1.5-fold greater than the AUC₂₄ after a single dose.

In HIV patients with esophageal candidiasis treated with an intravenous dose of micafungin ranging from 50 to 150 mg/day, the V _d was approximately 0.4 L/kg, with the V _d after a single dose being similar to the V _d at steady state, indicating that equilibrium between plasma and tissue is rapidly reached [15]. Steady state is reached after approximately 4–5 days. Micafungin is highly protein bound in plasma (99.8%), mainly to albumin and α-1-acid glycoprotein, which is concentration-independent over a range of 10–100 mg/L [9]. At clinically relevant concentrations, binding to albumin was shown to be non-competitive, did not displace albumin-bound bilirubin, and showed no interaction with other protein-bound medication [9]. Micafungin is not significantly taken up by red blood cells, with a cell/plasma ratio of 0.7. Intracellular concentrations of micafungin in peripheral blood mononuclear cells and polymorphonuclear leukocytes were approximately tenfold higher compared with corresponding plasma concentrations [27].

Distribution into tissues throughout the body has not been extensively investigated and data mainly come from a limited number of case reports [28‐38]. Furthermore, these data are often difficult to interpret because homogenate samples may show incorrect concentrations due to differences between intra- and extracellular concentrations. Also, an unusually high number of mononuclear cells due to a local infection could lead to an overestimation of extracellular micafungin concentration [39]. However, some more easily accessible compartments can give a good estimation of tissue penetration, such as measurements in cerebrospinal fluid (CSF) and epithelial lining fluid (ELF). Tissue penetration of micafungin in relation to other antifungal agents has been thoroughly reviewed previously [39]. Penetration differs per organ system and a detailed description of available data is provided below.

The penetration of micafungin in burn eschar tissue has been reported in several publications [40‐43]. Asensio et al. reported an average tissue (T)/C _max ratio of 0.08, 1–3 h after administration of 100 mg on day 5 of therapy [40], while two investigations by Sasaki et al. examined tissue 24 h after a 200–300 mg dose and reported tissue/plasma (T/P) ratios ranging from 2.2 to 6.5 in one patient and from 0.76 to 2.32 in three other patients [42, 43].

Penetration in the lung, specifically the ELF and alveolar cells (AC) was investigated in two prospective studies, with a total of 35 volunteers receiving micafungin 150 mg [44, 45]. The T/P ratios changed over time and ranged from 0.01 to 1.1 for ELF/P and 0.61 to 7.62 for AC/P. By modeling and simulation, the authors predicted a mean AUC ratio from days 1 to 14 of 1.3 for ELF/P and 3.5 for AC/P [44, 45]. In two other patients, micafungin reached lower concentrations in pleural fluid, with an average T/P of 0.14 at steady state 2–4 h after dosing [46]. In another patient, T/P ratios of 0.57 (day 29) and 0.67 (day 43) were seen 22 h after the last dose [32].

Peritoneal fluid concentrations were assessed in 10 postsurgical patients receiving micafungin 100 mg daily. Samples were taken at seven time points on days 1 and 3. On both days, the AUC T/P ratio was between 0.3 and 0.4 [47] and the penetration was shown to be twice as high as previously reported in a patient, with an ascites/plasma ratio of 0.15 [46].

Similar to the other echinocandins, micafungin penetrates poorly into brain and CSF. The CSF/P ratio found in three patients 2–5 h after administration varied widely, with values ranging between 0.002 and 0.73 [36, 46]. In another patient, brain tissue was obtained which contained a T/P ratio for micafungin of 0.18 [30]. The most recent study measured micafungin concentrations in nine samples of three premature neonates after a 10 mg/kg dose and measured between 1 and 1.5 mg/L, corresponding with a CSF/C _24h ratio of 0.16, and CSF/C _max ratio of 0.04 [38, 48]. Although highly variable, these data might indicate that there is sufficient penetration of micafungin in these compartments for an antifungal effect, specifically in neonates with Candida meningoencephalitis treated with a high dose.

Low penetration of micafungin was seen intraocularly after intravenous infusion. Eight patients with Candida endophthalmitis were described, with mean aqueous and vitreous humor-to-plasma ratios of 0.0043 and 0.0046, respectively [33‐35]. In one patient, the tissue concentration of other structures in the eye were also measured and penetration was much better compared with intraocular penetration, with T/P ratios of 0.094 in the cornea, 0.86 in irises, 0.071 in retinas, and 0.34 in choroids [34].

Distribution of micafungin to the bladder is known to be poor, with <1% unchanged micafungin excreted in urine. Despite this finding, some cases are reported where micafungin was successfully used for the treatment of candiduria [28, 29, 37, 49]. This might be due to the excellent penetration of micafungin in the kidney. Investigations in rabbits showed that the concentration of micafungin in the kidney was similar to the concentration found in plasma [50]. On the other hand, the lack of reports describing therapy failure when using micafungin for candiduria could be the result of publication bias. High-quality evidence is still poor, and guidelines, such as the recent Infectious Diseases Society of America (IDSA) guideline, do not recommend the use of echinocandins for the treatment of urinary tract infections due to Candida species [6].

Other compartments where micafungin concentrations have been measured are wound tissue (T/P ratio of 0.46) [46], bile fluid (T/P ratio of 1.25) [31], and pancreatic pseudocyst fluid [30]. In the latter, only one sample was measured 24 h after a dose that contained 0.38 µg/mL, but no dose or plasma concentration was reported [30].

Micafungin undergoes metabolism to at least 11 compounds (M1–M11). It is the main circulating compound, but also M1 (catechol form), M2 (methoxy form of M1) and M5 (hydroxylation at the side chain) have been detected in plasma. Both M1 and M2 are pharmacologically active metabolites, with an exposure of up to 11 and 2%, relative to the parent compound [26]. The M1 metabolite is produced by arylsulfatase and is further metabolized to M2 by catechol-O-methyltransferase [51]. All other metabolites are thought to be inactive. The main inactive metabolite is M5, with an exposure of 9–14% relative to the parent compound; this metabolite is formed mainly by cytochrome P450 (CYP) 3A4 but also by several other CYP isoenzymes (CYP1A2, 2B6, and 2C). Like the parent drug, all metabolites show linear PK [26].

Systemic CL after infusion is approximately 12 mL/h/kg, which is 840 mL/h for a 70 kg adult individual. t _½ is 14–15 h and is independent of the dose. The main metabolite, M5, has a half-life of 32 h. Excretion of micafungin and its metabolites was investigated in two mass balance studies, each with six subjects, with a collection period of 169 h and 28 days, respectively. After 28 days, a total of 83% of the administered dose was collected, 71% in the feces and 12% in urine. The distribution of metabolites in these excretion fluids was not reported. The mean t _½ for all metabolites was estimated on 340 h [15].

The effects of intrinsic factors on the PK of micafungin have been investigated in several large studies. No significant differences were found between sex, race, and age, with, for the latter covariate, a group of 66- to 78-year-olds being compared with a group of 20- to 24-year-olds. No differences were found in C _max, AUC, t _½, V _d, CL, and percentage protein binding between these groups [1, 15, 17, 26, 52]. Weight was found to explain a large proportion of variability in CL in a group of 64 hematology patients receiving a dose ranging from 12.5 to 200 mg/day. Patients weighing less than 66.3 kg were found to have a higher average AUC₂₄ of 121 mg h/L, compared with 81 mg h/L in patients weighing more than 66.3 kg [53]. Furthermore, micafungin plasma concentrations in a 230 kg patient were approximately 50% lower compared with a group of hematology patients with a mean weight of 82.6 kg [18, 54].

In a study of three groups of healthy volunteers with body mass indexes (BMIs) of <25, 25–40, and >40 kg/m², the effect of weight on the CL of subjects >66 kg was described by the function CL (L/h) = 1.04 × (weight/66)^0.75 [55, 56]. This relation seems to conflict with the CL found in healthy subjects in other reports (see Table 1). For example, Hebert et al. reported a mean CL of 10 mL/h/kg in subjects with a mean weight of 71.7 kg, thus having a CL of 0.717 L/h [21]. According to the formula, these subjects should have had a mean CL of 1.10 L/h, which is an overestimation by more than 50%. This questions the validity of a general formula using weight above a certain cut-off, without taking into account physiological changes that are associated with obesity [57].

The plasma concentrations of micafungin and covariates influencing the concentration have been investigated in 15 population PK models (see Table 2 for details) [38, 41, 45, 47, 52, 53, 55, 58‐65]. In all cases, the plasma concentration was best described using a two-compartment model; additional compartments were used to explain tissue concentrations in two studies [41, 47]. In the majority of the models, weight was incorporated to explain variability in systemic CL, and, in approximately half of the models, weight was also able to explain variability in V _d of the central compartment. Interestingly enough, with the exception of BMI, no other weight-derived covariates such as fat-free mass, normalized fat mass, or lean body weight have been investigated to explain variability in CL or volume. Other covariates explaining variability in systemic CL were platelet count in Japanese adults and pediatric patients [52], alanine transferase and total bilirubin in pediatric patients [58], the ratio of aspartate transaminase and alanine transaminase in preterm neonates [38], and, finally, albumin and Sequential Organ Failure Assessment (SOFA) score in critically ill patients [61]. Variability in volume in distribution was explained by albumin concentrations in both postsurgical patients with peritonitis and critically ill patients [47, 61].

Micafungin is metabolized partly by the liver through various enzymatic systems, as described above [1]. Micafungin is not a substrate for P-glycoprotein [9]. It has been demonstrated in vitro that micafungin is a strong inhibitor of a wide variety of efflux pumps [66]. As a perpetrator drug, it has been demonstrated that micafungin influences the CL of the following drugs: sirolimus, nifedipine and itraconazole. A mechanistic basis for these interactions is not provided in the Summary of Product Characteristics (SmPC) but possibly due to inhibition of CYP3A4. As increases in AUCs of sirolimus, nifedipine, and itraconazole have been found to not be clinically relevant (increases of 21, 18, and 22%, respectively), they do not warrant dose adaptations of the victim drug [1, 18‐20, 22, 23, 67‐69]. Coadministration of micafungin with amphotericin B deoxycholate leads to an increase in amphotericin B exposure of 30%, accompanied by the occurrence of more side effects [70], while micafungin remains unaffected. Patients receiving this combination should be monitored closely for (renal) side effects.

Micafungin acts by selective inhibition of the fungal enzyme that produces the cell wall polymer 1,3-β-d-glucan synthase. As human cells do not contain this polymer, a favorable toxicity profile can be expected through the absence of a direct pharmacological effect. Indeed, no dose-limiting toxicity has been reported up to a daily dose of 8 mg/kg (896 mg) for 1–4 weeks in adults [71]. Furthermore, one patient was described as receiving 1400 mg every other week for 12 weeks without any side effects associated with micafungin [72]. In addition, a newborn was accidently treated with a single 16 mg/kg dose of micafungin without any adverse reactions [1]. The most common side effects associated with micafungin are diarrhea, nausea, vomiting, pyrexia, thrombocytopenia, and headache [15]. The European Medicines Agency, but not the US FDA, issued a black-box warning for possible development of foci of altered hepatocytes (FAHs) and hepatocellular tumors as preclinical data indicated these tumors developed in rats treated with high-dose micafungin for 13 weeks [1]. After treatment discontinuation, the rats recovered for 13 weeks but the FAHs were still present. At least a part of these foci was not reversible [9]. The relevance of the hepatocarcinogenic potential for use in humans is unknown. As of today, no cases have been published reporting this effect in humans.

The effect of hepatic impairment was investigated as part of the registration studies in eight volunteers with moderate hepatic impairment due to hepatitis C, primary biliary cirrhosis, or alcohol abuse, with Child–Pugh scores ranging between 7 and 9. Exposure after a single 100 mg dose was decreased to a mean of 98 mg h/L, versus 126 mg h/L in matched healthy volunteers [21]. Similar results were found in a study of eight subjects with severe hepatic impairment who had an exposure of 100 versus 142 mg h/L in eight matched healthy controls [24]. A possible explanation can be found in decreased levels of albumin, resulting in an increased free fraction of micafungin [21, 24]. This results in a lower total plasma concentration and explains the decrease in AUC. Nevertheless, this decreased AUC is not considered to be clinically relevant and no dose adjustments are recommended for patients with moderate or severe hepatic dysfunction. Similar results were observed in 34 liver transplant recipients reported in three studies, with one patient being a remarkable exception [73‐75]. This patient had a small-for-size graft liver with a volume of only 26% of a standard liver, and showed a normal half-life of 16 h after a 50 mg dose. However, after administration of a 100 mg dose, the half-life increased to 76 h and the AUC₁₂ of this patient increased from 79 to 601 mg h/L, corresponding to an AUC after a 1000 mg dose in a healthy subject [75].

Although micafungin is not cleared renally, patients who suffer from renal impairment might have altered PK due to alterations in albumin concentrations available for protein binding. The effect of renal impairment on the PK of micafungin was investigated in nine patients with a creatinine CL below 30 mL/min after a single dose of 100 mg, and compared with nine matched healthy subjects. No differences in AUC_∞, CL, V _d, half-life, C _max, and protein binding were observed between groups [15, 21]; therefore, no adjustments are necessary for patients with renal dysfunction.

The use of extracorporeal elimination techniques can influence the PK of drugs by increasing the V _d, direct elimination and adsorption to membranes and tubing material. Micafungin is a large molecule that is highly protein bound and not renally cleared. Indeed, no changes in PK were observed in ten critically ill patients treated with micafungin 100 mg daily during continuous venovenous hemofiltration (CVVH) using polyethersulfone or polysulfone hemofilters. Samples at seven timepoints pre- and postfilter were taken, and removal of micafungin was not observed. In addition, in samples taken from the ultrafiltrate, micafungin levels were below the limit of quantification [63]. In a study of four patients receiving CVVH using cellulose triacetate hollow fiber, the same results were observed [73]. Furthermore, a similar study observed no changes in pre- versus postfilter micafungin concentrations in four critically ill patients treated daily with 150–300 mg during continuous venovenous hemodiafiltration (CVVHDF) using a hollow-fiber membrane composed of polymethyl methacrylate. In addition, these patients were compared with nine critically ill patients not receiving CVVHDF. Although interindividual variability in CL was large throughout both groups, no indication of a difference in CL or V _d was observed [76]. Dose adjustments of micafungin are not indicated in these patients.

Changes in micafungin PK in critically ill patients in the intensive care unit (ICU) have been investigated in two prospective studies totaling 119 ICU patients. The first study investigated micafungin concentrations over a 14-day period, with daily trough samples and intensive sampling at day 3 and limited sampling on day 7 in 20 ICU patients receiving micafungin 100 mg daily. At days 3 and 7, the AUC₂₄ was 79 versus 66 mg h/L (no significant difference) [77]. These exposures are much lower than the exposure found in healthy volunteers (Table 1; mean value = 133 mg h/L). Investigations in another study of 99 ICU patients confirmed this and found an AUC_∞ ranging between 65.5 and 99.5 mg h/L, depending on SOFA score, albumin concentration, and bodyweight as relevant covariates [61]. The lowest exposure of 65.5 mg h/L was found in patients with a SOFA score of <10 and an albumin of ≤25 g/L. This study also showed that micafungin PK were not influenced by using extracorporeal membrane oxygenation (ECMO), confirming a report in a previously described patient [78]. One of the reasons for lower exposure might be the availability of albumin for protein binding. A second reason is the influence of the SOFA score on micafungin PK, which may impact the metabolic routes of a drug. In the case of micafungin, an induction of arylsulfatase, catechol-O-methyltransferase or CYP isoenzymes, or a change in biliary excretion, may be anticipated [61, 77]. The lower AUC decreases the probability of target attainment (PTA) when the licensed micafungin 100 mg is used in ICU patients. Simulations show that only 62% of patients reach the MIC/AUC target for non-C. parapsilosis spp.; therefore, these patients could benefit from a dose escalation to micafungin 200 mg, as indicated in the manufacturer’s label information [62].

Critically ill patients with thermal injuries showed a lowered plasma concentration of micafungin after a daily 100 mg dose, with a mean C _24h of 0.9 mg/L compared with approximately 2 mg/L in healthy volunteers [40]. Two case series report that patients treated with micafungin 200–300 mg had comparable plasma concentrations compared with healthy volunteers receiving 75 mg [42, 43]. Factors causing lower exposure in this patient population are similar to those in general ICU patients. An additional factor for the lower exposure might be the hypermetabolic state, a phase occurring beyond 48 h after the injury period for up to another 48 h, and also seen with other antifungals in severely burned patients [79]. PK in burn patients were compared with PK in patients with complicated intra-abdominal infections, and data from both populations were used to build a population PK model. No differences in PK between these groups were observed, except for the rate constant describing the distribution of micafungin between blood plasma and tissue fluid. The authors concluded that these populations should not be dosed differently from each other. Simulations showed that a micafungin dose of 100–150 mg should be sufficient to achieve a PK/PD target in plasma for non-C. parapsilosis spp. and C. parapsilosis species, with MIC values up to 0.008 and 0.064 mg/L, respectively. A licensed dose of 200 mg was found to be sufficient to achieve the European Committee on Antimicrobial Susceptibility Testing (EUCAST) susceptibility target for C. albicans (0.016 mg/L), but not for C. glabrata (0.03 mg/L) [41].

Hematology patients show different PK of micafungin compared with healthy subjects. PK in hematology patients has been investigated in a dose-escalation study of 62 patients treated with micafungin 12.5–200 mg/day (see Table 3 for a summary of the PK parameters). AUC_∞ was 81.1 mg h/L after the first 100 mg dose, and seems lower compared with healthy subjects (Table 2; 133 mg h/L). CL was higher being 13.6 mL/h/kg versus 10.4 mL/h/kg in healthy subjects, but V _d seems to be in the same range as healthy subjects, both approximately 0.2 L/kg. The half-life of micafungin in hematology patients is short, 13 h versus 15.4 h in healthy subjects [18].

PK of micafungin in premature neonates with a weight above and below 1 kg has been investigated in four small studies, with doses ranging from 0.75 to 15 mg/kg [64, 80‐82]. Initial investigations started with single doses of 0.75 and 1 mg/kg, and showed that the average CL was up to tenfold higher in neonates <1 kg compared with healthy adults, i.e. 79.3–98 versus 10.4 mL/h/kg [80, 81]. The reported average half-life was 5.5 and 6.3 versus 15.4 h in healthy adults [80, 81]. Subsequently, higher doses were investigated at steady state in the <1 kg population, with a dose of 10 and 15 mg/kg. CL was approximately 36 mL/h/kg, still almost fourfold higher than CL in healthy adults [64, 82]. AUC₂₄ at steady state after a daily 10 mg/kg dose was 308 mg h/L, and proportionally higher after a daily 15 mg/kg dose, with an AUC₂₄ of 472 mg h/L; both doses were well-tolerated [64, 82]. The reported volumes of distribution in this select population have a wide range, with averages ranging between 0.51 and 0.81 L/kg, which is at least 2.5-fold higher than the 0.2 L/kg found in healthy adults [64, 80‐82]. The high variability in both V _d and CL between the reported populations investigating PK at a very low dose of 0.75–1 mg/kg might be due to non-linear kinetics in this patient population; however, data are very sparse and this could also be an artifact due to small sample sizes.

In premature neonates with a weight above 1 kg, a dose range of 0.75–15 mg/kg has been investigated and seems to show dose proportionality of AUC [64, 80‐82]. The CL of approximately 39 mL/h/kg found in this group was similar to neonates with a weight below 1 kg [64, 80, 82]. The V _d seems a bit lower in the >1 kg group, at 0.4 L/kg [64, 80, 82]. In the study by Kawada et al., the PK parameters reported in neonates >1 kg are conflicting with the above-stated parameters . Here, an augmented CL of 81 (versus 39 mL/h/kg) and a V _d of 0.72 L/kg (versus approximately 0.4 L/kg) was found; however, these differences between data were explained by the study design. Kawada et al. performed their investigations in a neonatal population with a gestational age of 24–34 weeks within 12–24 h after birth, while Heresi et al. studied the same population, but 3–8 weeks after birth [81].

Overall, neonates have a higher CL than healthy adults, which can be explained by the fraction of unbound micafungin. Yanni et al. found that the fraction of unbound micafungin can be eightfold higher in neonates compared with adults, suggesting an age-dependent serum protein binding [83]. This explanation would discourage a dose increase because it would mean that although the total exposure decreases, the concentration of unbound micafungin is not necessarily lowered. Lower total exposure can additionally be caused by increases in intrinsic CL due to, for example, maturation. Nevertheless, a dose of 10–15 mg/kg/day was well-tolerated, and reported hepatic toxicity was reversible and manageable by monitoring hepatic markers [38].

PK of micafungin in children and adolescents has been extensively investigated, with reports of a variety of age and dose ranges, both after a single dose and in steady-state conditions [51, 84‐88]. This heterogeneity in study designs complicates the comparison between studies. Table 4 summarizes the main PK parameters.

Both single dose and steady-state parameters show dose proportionality and linearity for C _24h, C _max, and AUC throughout a dose ranging 0.5–4.5 mg/kg [51, 86]. The main PK parameters, i.e. half-life (approximately 13 h), CL (approximately 19 mL/h/kg), and V _d (0.3 L/kg), did not change throughout dose cohorts or time. Compared with healthy adults, CL is almost twofold higher in children and adolescents compared with adults (19.2 versus 10.4 mL/h/kg), and the remarkably high CL seen in neonates seems to decrease with age. This has been confirmed with a pairwise comparison in a population ranging between 2 and 17 years of age, in children below and above 8 years of age. The group aged <8 years had a higher CL and V _d, i.e. 23.1 versus 17.1 mL/h/kg, and 0.35 versus 0.28 L/kg [86]. Albano et al. investigated these differences in more detail and found that children aged from 4 months to 5 years have a much higher CL at steady state of approximately 20 mL/h/kg, versus approximately 13 mL/h/kg in children aged 6–16 years [84]. V _d in this group was also higher (0.32 L/kg versus 0.26 L/kg) [84]. These changes in PK throughout the processes of maturation and growth urge the need for a higher weight-corrected dose compared with adults, especially in children less than 8 years of age. The role of protein binding remains unresolved as an explanation of this variability. Furthermore, critically ill children were shown to have an additionally increased CL of 41 mL/h/kg and an increased V _d of 0.64 L/kg, and should receive an even higher dose than the registered 1–4 mg/kg dose. After a 4 mg/kg dose, these children had an AUC of 117 mg h/L, which approximately corresponds to the AUC in a child after a 2.5 mg/kg dose. A dose increase of 2.5–5 mg/kg should be considered in critically ill children [89].