Disposition of Caspofungin: Role of Distribution in Determining Pharmacokinetics in Plasma

Abstract

The disposition of caspofungin, a parenteral antifungal drug, was investigated. Following a single, 1-h, intravenous infusion of 70 mg (200 μCi) of [³H]caspofungin to healthy men, plasma, urine, and feces were collected over 27 days in study A (n = 6) and plasma was collected over 26 weeks in study B (n = 7). Supportive data were obtained from a single-dose [³H]caspofungin tissue distribution study in rats (n = 3 animals/time point). Over 27 days in humans, 75.4% of radioactivity was recovered in urine (40.7%) and feces (34.4%). A long terminal phase (t_1/2 = 14.6 days) characterized much of the plasma drug profile of radioactivity, which remained quantifiable to 22.3 weeks. Mass balance calculations indicated that radioactivity in tissues peaked at 1.5 to 2 days at ∼92% of the dose, and the rate of radioactivity excretion peaked at 6 to 7 days. Metabolism and excretion of caspofungin were very slow processes, and very little excretion or biotransformation occurred in the first 24 to 30 h postdose. Most of the area under the concentration-time curve of caspofungin was accounted for during this period, consistent with distribution-controlled clearance. The apparent distribution volume during this period indicated that this distribution process is uptake into tissue cells. Radioactivity was widely distributed in rats, with the highest concentrations in liver, kidney, lung, and spleen. Liver exhibited an extended uptake phase, peaking at 24 h with 35% of total dose in liver. The plasma profile of caspofungin is determined primarily by the rate of distribution of caspofungin from plasma into tissues.

Caspofungin (CANCIDAS; MK-0991) is a parenteral antifungal agent that inhibits 1,3-β-d-glucan synthesis, which forms a critical component of many fungal cell walls (4). Caspofungin is active against many clinically important fungal species, including Candida spp. and Aspergillus spp. (3, 5, 7, 14), and in clinical trials it has been shown to be efficacious in the treatment of esophageal candidiasis (1, 15, 16), invasive candidiasis (9), and invasive aspergillosis (J. Maertens, I. Raad, G. Petrikkos, et al., Abstr. 42nd Intersci. Conf. Antimicrob. Agents Chemother., abstr. M-868, 2002).

This paper describes results from two studies conducted in healthy human subjects to investigate the disposition of caspofungin following intravenous (i.v.) infusion of radiolabeled caspofungin and supportive studies of [³H]caspofungin tissue distribution in rats, in vitro metabolism, and in vitro binding and partitioning in human plasma and blood. The metabolites of caspofungin, a cyclic hexapeptide, in humans have been previously reported (2). Caspofungin is the major component of radioactivity in plasma and urine in the first 24 to 30 h postdose, with a ring-opened form of caspofungin, M0, comprising a minor component. At time points of ≥5 days, M0 was the major component in plasma, and urine radioactivity was largely comprised of the synthetic amino acid dihydoxyhomotyrosine (M1) and its N-acetyl derivative (M2). Caspofungin concentrations in plasma fall by more than an order of magnitude in the first day or two postdose, and most of the area under the concentration-time curve (AUC) is accounted for in this period (13). This suggests that caspofungin plasma pharmacokinetics are being determined primarily by processes other than metabolism, such as distribution or excretion, since only minor levels of metabolites are seen during this period. This paper will describe results which provide insight into the mechanism largely responsible for determining caspofungin plasma pharmacokinetics.

MATERIALS AND METHODS

Study A: radiolabeled disposition study

In this open-label study, six healthy male subjects (three white, one black, and two Hispanic; aged 31 to 45 years [average, 39 years]; weighing 66 to 89 kg [average, 77 kg]) received a single 70-mg (200-μCi) dose of [³H]caspofungin infused over 1 h. Subjects enrolled in this study (and the second study) were male nonsmokers in generally good health who refrained from taking other medications, had no history of drug or alcohol abuse, and had not enrolled in any prior caspofungin studies. Plasma samples were collected at 0 (predose), 0.5, 0.75, 1, 1.25, 1.5, 2, 3, 4, 6, 9, 12, 24, 48, 60, and 72 h following the start of drug infusion and then once daily on days 5 to 28 for assay of caspofungin and radioactivity. Urine was collected over the intervals 0 to 2, 2 to 4, 4 to 6, 6 to 8, 8 to 12, 12 to 24, 24 to 36, and 36 to 48 h postdose and then for 24-h intervals on days 3 to 28 for assay of caspofungin and radioactivity. Feces were collected over 24-h intervals on days 1 to 28 for assay of radioactivity.

Study B: long-term study

In this open-label study, seven subjects (one white, two black, and four Hispanic; aged 21 to 40 years (average, 31 years); weighing 61 to 93 kg [average, 78 kg]) received the same dose, route of administration, and radiolabeled form of caspofungin as used in the preceding study. Plasma samples were collected at the same time points as in study A up to 72 h and then on days 5, 6, 7, 10, 14, 21, and 28 and at weeks 6, 8, 10, 12, 14, 16, 18, 20, 24, and 26 for assay of caspofungin and radioactivity. No urine or feces were collected in this study.

The study protocols described in this report were approved by the Southern Institutional Review Board of Miami, Fla., and informed consent was obtained from all subjects.

Bioanalytical analysis

Plasma and urine samples were stored at −70°C, and feces were stored at −20°C until analysis. Plasma and urine concentrations of caspofungin were determined by high-pressure liquid chromatography with fluorescence detection as previously described (12). The plasma assay was modified slightly to allow for smaller sample volumes; 0.5 ml of plasma was used, with a resulting limit of quantitation (LOQ) of 25 ng/ml. The standard curve range was 25 to 2,000 ng/ml in the modified assay. In addition, for study B a column-switching procedure was employed as described in reference 13. Bovine serum albumin (2 ml of a 35% solution in saline per 200 ml of urine) was added to urine samples to prevent adsorption of the drug to the container. The LOQ for urine was 10 ng/ml.

Total radioactivity in plasma, urine, and feces was determined by liquid scintillation spectrometry with quench correction by external standardization. Aliquots of plasma and urine were added directly to liquid scintillant. Plasma and urine concentrations expressed as microgram equivalents (eq) per milliliter were calculated from the known specific activity of the [³H]caspofungin dosing solution (2.84 μCi/mg). Fecal samples were homogenized in ethanol-purified water (50:50 [vol/vol]), and aliquots were combusted and analyzed.

Irreversible binding in plasma was determined by exhaustive extraction of the protein pellets with a series of acidic and organic solvents (0.4 N trichloroacetic acid, 80% methanol in water, 1:3 [vol/vol] ether-ethanol) until the radioactivity in the supernatant reached below twofold of background. Radioactivity remaining in the pellets was determined by liquid scintillation counting, and protein concentrations were determined using a colorimetric method (8).

Pharmacokinetic analysis

The plasma terminal rate constants β and γ were calculated by weighted (1/y²) nonlinear regression of individual plasma drug concentration data using a biexponential (caspofungin data) or monoexponential (radioactivity data) decay function. The onset of the log-linear β-phase was determined by inspection for each subject and generally occurred at 6 to 9 h postdose for caspofungin and at 5 to 8 days postdose for radioactivity. Half-lives (t_1/2) were computed as the quotient of the ln(2) and the rate constant.

The AUC profile was calculated by the linear-log trapezoidal method with extrapolation of the AUC over the interval of last quantifiable point to infinity as the quotient of the last quantifiable concentration and the terminal rate constant. Plasma drug clearance was determined as the quotient of dose and AUC_0-∞. For several of the concentration-time profiles, the end of infusion did not coincide precisely with the actual C_{1 h} sampling time. In these instances, an estimated end of infusion concentration, determined by fitting the plasma drug concentration-time data to a three-compartment linear model, was used along with the plasma drug concentration data in the AUC calculations. Actual dose administered was calculated from the infusate radioactivity concentration and net volume administered. The concentration-based pharmacokinetic parameters AUC, C_{1 h}, and C_{24 h} were adjusted to a 70-mg dose equivalent using the ratio of 70 over the administered dose. Renal clearances were calculated as the quotient of urinary drug recovery and the corresponding AUC over the same time interval.

The distribution of caspofungin was investigated using compartmental modeling of the plasma caspofungin concentration-time profile (Fig. 1). The differential equations that compose the three-compartment, linear pharmacokinetic model are as follows:

where A denotes amount; the subscripts 1, 2, and 3 denote amounts in the central and two peripheral compartments, respectively; k₀ is the elimination rate constant; and k₁₂, k₂₁, k₁₃, and k₃₁ are the microconstants defining transfer between the central and peripheral compartments. The location of the elimination term, indicated by the (−k₀ · A_X) terms in the equations, was varied in the models evaluated between the central compartment (1) and the peripheral compartments (2 and 3). In each model, the elimination term location was limited to only one compartment (1, 2, or 3). The measured plasma drug concentrations were related to A₁ by the central volume of distribution (V₁). The software package ACSL TOX (Aegis Software, Huntsville, Ala.) was used for the modeling analysis. The model parameters were fit to weighted (1/y²) plasma data by maximization of the log-likelihood function using a generalized reduced gradient search algorithm. The apparent volume of distribution (V) at a given time point was calculated from the quotient of the total amount in the body (A₁ + A₂ + A₃) and the plasma drug concentration at that time.

FIG. 1.

Schematic of the compartmental model used in the distribution analysis. The elimination term (k₀) is from either compartment 1, 2, or 3, depending on the model used.

Statistical analysis

Geometric means and associated 90% confidence intervals (CIs) for the recovery data, AUC_0-∞, C_{1 h}, C_{24 h}, and clearance were determined for studies A and B. For study B, two-sided 95% CIs for arithmetic mean times required for elimination from plasma of radioactivity and caspofungin were calculated using the between-subject standard deviation and referencing the t-distribution.

In vitro plasma protein binding and partitioning into red blood cells

Various concentrations of [³H]caspofungin (0.1, 1.0, 10, or 100 μg/ml) or [³H]M0 (5 μg/ml) were added to blank human plasma (Seratech Biologicals, New Brunswick, N.J.), incubated at 37°C for 15 min, and centrifuged at 37°C and 180,000 × g overnight using a high-speed centrifuge. The centrifuge was permitted to stop without the use of brakes, and aliquots (200 μl each) of the supernatant were pipetted from top to bottom in sequence. The sedimented protein was resuspended in 0.5 ml of water. The radioactivity in each fraction was determined by liquid scintillation counting. Fractions with the lowest radioactivity were considered to be protein free. The unbound fraction (Fu) was calculated as follows: Fu = (disintegrations per minute in protein-free aliquot)/(total disintegrations per minute) · 100%.

Various concentrations of [³H]caspofungin (0.1, 1.0, 10, or 100 μg/ml) were added to blank fresh human blood and incubated at 37°C in a water bath for 30 min. Aliquots (1.0 ml) were taken, and plasma was obtained by centrifugation. Blood and plasma (0.2 ml each) were combusted to ³H₂O in a sample oxidizer for radioactivity measurements to determine the distribution between blood cells and plasma. Samples were prepared in triplicate. In a similar experiment, the effect of time on partitioning was assessed using a concentration of 10 μg of [³H]caspofungin/ml and varying the incubation time (0, 15, 30, 60, 120, and 240 min).

Tissue distribution in rats

The distribution of radioactivity in rat tissues was determined after a 2.0-mg/kg body weight i.v. bolus injection of [³H]caspofungin, with animals sacrificed at 0.5, 2, 24, and 288 h postdose (three rats/time point). Blood was collected, and tissues were excised, rinsed with saline, blotted, and weighed. All tissues, with the exception of adrenal glands, skin, eye, and bladder, were homogenized with distilled water. Aliquots (1 ml) of the tissue homogenates were combusted to ³H₂O and counted by scintillation spectrometry. Plasma, red blood cells, adrenals, skin, eyes, and bladder were combusted without homogenization.

In vitro metabolism

Rat tissues (liver and kidney slices and homogenates and fresh blood) and cellular fractions (liver and kidney S9 fractions and liver microsomes) were prepared from male Sprague-Dawley rats. Human liver microsomes were obtained from IIAM (Exton, Pa.), and human hepatocytes were provided by the Human Cell Culture Center (Laurel, Md.). Liver and kidney slices (∼106 mg [wet weight]) were incubated with [³H]caspofungin (100 μM in 2 ml) for up to 27 h with 7-ethoxycoumarin used as a marker substrate to assess enzyme activity. Liver and kidney homogenates were incubated with [³H]caspofungin (50 μM in 0.15 M Tris buffer) at 37°C for 2 h with added cofactors [0.2 M UDP-glucuronic acid and 5 mM 3-phosphoadenosine 5-phosphosulfate]. Blood (hemolyzed and nonhemolyzed) was incubated with [³H]caspofungin (50 μM) at 37°C for 2 h. Liver microsomes (2 mg/ml) were incubated with [³H]caspofungin (50 μM) at 37°C for 3 h (rat) or 1 h (human). Human hepatocytes (9 × 10⁵ cells/ml) were incubated with [³H]caspofungin (10 μM) at 37°C for up to 24 h.

Recombinant cytochrome P450 (CYP) isozymes (1A1, 1A2, 2C8, 2C9, 2C19, 3A4, 3A5, and 4A11) were purchased from Gentest (Woburn, Mass.), and hydrolytic enzymes (amidase, peptidase, aminopeptidase I, leucine aminopeptidase, carboxypeptidases B and Y, endopeptidase, prolidase, trypsin, ficin, and proteases V, XXVII, and XXIV) were obtained from Sigma Chemical (St. Louis, Mo.). The CYP isozymes were incubated with [³H]caspofungin (50 μM) at 37°C for up to 2 h. [³H]caspofungin was incubated with the hydrolytic enzymes at 37°C for up to 48 h with aliquots (0.1 ml) removed every 4 h, and the incubation mixture was supplemented with fresh enzymes. In addition, to select incubations, hepatic and renal S9 fractions were added to provide a broader metabolic system.

Metabolite formation was measured by high-performance liquid chromatography using in-line radioactivity flow detection.

RESULTS

Pharmacokinetics in plasma

Following a single dose of [³H]caspofungin, plasma concentrations of total radioactivity were roughly comparable to caspofungin concentrations for approximately the first day postdose, after which parent drug levels fell more rapidly than total radioactivity (Fig. 2). In study B, caspofungin plasma concentrations fell below the LOQ (0.025 μg/ml) at an arithmetic mean (95% CI) of 6.3 (4.5, 8.0) days, while total radioactivity fell below the LOQ (∼0.002 μg eq/ml) at an arithmetic mean (95% CI) of 22.3 (20.6, 24.1) weeks. Table 1 summarizes the noncompartmental plasma pharmacokinetic parameters for caspofungin and total radioactivity. Parent drug comprised 15 to 17% of the total radioactivity AUC.

FIG. 2.

Mean plasma drug concentration profiles of caspofungin and total radioactivity following single 1-h i.v. adminitration of 70 mg of [³H]caspofungin to healthy male subjects. (A) Within-study comparisons of caspofungin and radioactivity profiles (expanded scale in inset plots). (B) Across-study comparison of radioactivity profile plotted on semi-log scale.

TABLE 1.

Pharmacokinetics following single 70-mg doses of [³H]caspofungin

Parameter	Geometric mean value (90% CI)
	Study A (n = 6)	Study B (n = 7)
Caspofungin plasma pharmacokinetics
AUC(0-∞) (μg · h/ml)	110.98 (90.76, 135.71)	121.12 (104.07, 140.96)
C1 h (μg/ml)	11.65 (10.16, 13.36)	11.82 (10.75, 13.00)
C24 h (μg/ml)	1.25 (0.94, 1.65)	1.33 (1.05, 1.68)
Clearance (ml/min)	10.51 (8.60, 12.85)	9.63 (8.28, 11.21)
t1/2 β (h)	7.95 (1.19)†a	8.94 (1.82)†
t1/2 γ (h)	27.28 (4.69)†	26.68 (8.64)†
Total radioactivity plasma pharmacokinetics
AUC(0-∞) (μg eq · h/ml)	634.95 (577.54, 698.06)	814.44 (723.06, 917.36)
t1/2 (days)	11.98 (0.69)†	14.57 (1.54)†
AUC ratio (caspofungin/total radioactivity)	0.17 (0.15, 0.20)	0.15 (0.14, 0.16)

^a†, harmonic mean (jackknife standard deviation).

Plasma concentrations of caspofungin declined in a polyphasic manner following the 1-h infusion. A short α-phase was seen immediately postinfusion. A dominant β-phase characterized much of the profile and exhibited log-linear behavior from ∼6 to ∼48 h postdose. An additional γ-phase was evident at low concentrations after 48 h postdose. For total radioactivity, a long terminal phase characterized much of the plasma profile, with log-linear behavior evident from ∼7 days postdose (Fig. 2).

Recovery of caspofungin and radioactivity in urine and feces

In study A, only 1.44% of the dose was recovered as intact caspofungin in the urine. Renal clearance of caspofungin was 0.15 ml/min. The geometric mean (90% CI) recovery of total radioactivity from urine and feces over 27 days was 75.4% (69.85%, 81.42%) of the dose administered. The recovery was roughly evenly divided between urine (40.7% [39.0%, 42.6%]) and feces (34.4% [29.6%, 40.1%]).

Figure 3 illustrates the distribution and excretion profiles for radioactivity. In this figure, the amount of radioactivity excreted into urine or feces was obtained from the cumulative recovery data, while the amount in plasma was estimated from the concentrations of radioactivity measured in plasma by assuming a plasma volume of 3 liters. The profile for amount in the tissues was then calculated from the mass balance. At the end of the 1-h infusion, ∼44% of the radioactive dose was found in plasma. Over the next 2 days, very little radioactivity was excreted (∼4% of dose), but there was extensive distribution of radioactivity into tissues. The mass balance calculation indicated that radioactivity in tissues peaked at 1.5 to 2 days postdose at ∼92% of dose, at which point only 4 to 5% of dose remained in plasma. Thereafter, radioactivity in tissues and plasma declined slowly due to continuing excretion of radioactivity into urine and feces. It was estimated that plasma contained 1.3% of the dose and tissues contained 23.0% of the dose at 27 days postdose. The rate of radioactivity excretion peaked at 6 to 7 days postdose for both urine and feces, and radioactivity continued to be excreted at a slow rate in both urine and feces at 27 days postdose (Fig. 4).

FIG. 3.

Mean distribution and excretion profiles of radioactivity following administration of a single 70-mg (200-μCi) dose of [³H]caspofungin to healthy male subjects (study A; n = 6).

FIG. 4.

Mean daily incremental recovery of radioactivity following administration of a single 70-mg (200-μCi) dose of [³H]caspofungin (study A; n = 6).

Irreversible binding to plasma proteins

At later time points (≥5 days), approximately 50% of the radioactivity in a given plasma sample was not extractable. This suggests that part of the caspofungin-derived radioactivity in plasma was irreversibly bound to proteins. The percentage of radioactivity in a given sample associated with the plasma proteins was fairly constant across time, while the amount of irreversible binding decreased with time. In study A, the irreversible binding averaged 6.94 pmol/mg of protein on day 5 to 6 and decreased to 3.32 pmol/mg of protein on day 19 to 20. In study B, the amount of irreversible binding decreased with time from an average of 6.71 pmol/mg of protein on day 14 to below the LOQ at week 12.

Model-based characterization of caspofungin distribution

The recovery data indicated that little elimination of caspofungin or its metabolites occurred during the first few days postdose. Furthermore, based on previously published data, little biotransformation of caspofungin to metabolites occurs in the first 24 to 30 h postdose (2). These findings indicate that the substantial decline in plasma concentrations during this period is largely due to distribution. Given this prominent role of distribution in controlling caspofungin plasma concentrations, it was of interest to investigate this distribution process.

Compartmental modeling of the plasma profile of caspofungin was used to characterize the distribution behavior of caspofungin. The plasma profile of caspofungin was well represented by a linear, three-compartment model with clearance from the central compartment or with clearance from the fast or slow peripheral compartments. This is illustrated in Fig. 5 for clearance from the central compartment, but essentially identical fits were obtained with clearance in either peripheral compartment. It is common for the location of the clearance term to be indistinguishable in compartmental modeling. Mean parameter estimates for all three models are in Table 2. The model parameters included the initial volume of distribution (V₁), and estimates for this parameter were quite small for all models (4 to 4.5 liters). The time courses of the apparent V for caspofungin indicated that over the first few hours, V increased from the initial distribution volume to 8 to 10 liters, depending on the model (Fig. 6). After this initial relatively rapid increase in volume, V continued to increase more slowly over the next 3 to 4 days until it reached a plateau at 23 to 160 liters, depending on the model used. The apparent V is known to overestimate the terminal phase volume of distribution due to the influence of elimination on the balance between plasma and tissue levels during this phase. However, these results can be expected to be reasonably accurate for the first 24 to 30 h postdose, since excretion and biotransformation are minor pathways for caspofungin during this period.

FIG. 5.

Fit of three-compartment linear model (clearance from central compartment) to individual plasma drug concentration profiles of caspofungin following administration of a single 70-mg (200-μCi) dose of [³H]caspofungin in study A (an expanded scale is shown in inset plots).

TABLE 2.

Model parameter estimates from fit of three-compartment, linear models to individual plasma drug concentration profiles of caspofungin obtained following administration of a single 70-mg dose

Parameter	Clearance from:
	Central compartment (1)		Fast peripheral compartment (2)		Slow peripheral compartment (3)
	Mean	SD	Mean	SD	Mean	SD
k0 (h−1)	0.151	0.028	0.145	0.019	0.0283	0.0049
k12 (h−1)	0.832	0.274	0.892	0.322	0.832	0.274
k13 (h−1)	0.0192	0.0057	0.0185	0.0049	0.171	0.033
K2a	1.17	0.23	0.816	0.120	1.17	0.23
K3b	1.71	0.28	1.77	0.34	0.0211	0.0042
VC (liters)	4.09	0.80	4.57	0.48	4.48	0.64

^aK₂ = k₂₁/k₁₂.

^bK₃ = k₃₁/k₁₃.

FIG. 6.

Time course of mean apparent volume of distribution of caspofungin following a single 70-mg dose of caspofungin as estimated by three-compartment linear modeling of plasma concentration-time data from study A (expanded scale is shown in the inset plot).

Plasma protein binding and partitioning into red blood cells

The unbound fraction of caspofungin in human plasma averaged 3.5% and was concentration independent over the range evaluated (0.1 to 100 μg/ml). The unbound fraction of M0 in human plasma averaged 2.0%. The human blood/plasma partitioning ratio for caspofungin averaged 0.74 and was independent of both incubation time and concentration.

Tissue distribution in rats

Radioactivity was widely distributed throughout the body in rats following a single i.v. dose of [³H]caspofungin (Table 3). Tissues with the highest concentrations of radioactivity at 0.5 and 2 h postdose were liver, kidney, lung, and spleen. In general, the concentration of radioactivity peaked within 2 h in most tissues. The radioactivity in liver continued to increase and peaked at 24 h, when 35% of the total dose was found in liver.

TABLE 3.

Mean concentrations and amounts of radioactivity in the tissues of rats receiving a 2.0-mg/kg i.v. bolus of [³H]caspofungin^a

Tissue	Mean ± SD concn (μg eq/ml or μg eq/g) (mean amt as % of dose) at:
	0.5 h	2.0 h	24 h	288 h
Plasma	11.0 ± 5.73 (24.6)	6.10 ± 0.596 (13.7)	1.74 ± 0.849 (3.91)	0.068 ± 0.040 (0.151)
Red blood cells	4.08 ± 2.57 (7.50)	1.93 ± 0.127 (3.54)	0.445 ± 0.219 (0.818)	0.132 ± 0.094 (0.243)
Skeletal muscle	0.305 ± 0.042 (7.63)	0.386 ± 0.096 (9.64)	0.103 ± 0.013 (2.58)	0.009 ± 0.002 (0.214)
Fat	0.431 ± 0.087 (0.862)	0.357 ± 0.073 (0.714)	0.358 ± 0.110 (0.715)	0.007 ± 0.000 (0.013)
Skin	1.60 ± 0.137 (14.0)	1.91 ± 0.043 (16.7)	0.959 ± 0.364 (8.38)	0.285 ± 0.128 (2.49)
Heart	2.31 ± 0.138 (0.382)	1.87 ± 0.045 (0.284)	0.642 ± 0.089 (0.097)	0.030 ± 0.004 (0.005)
Lung	5.12 ± 0.192 (1.10)	4.50 ± 0.601 (0.937)	2.44 ± 0.416 (0.531)	0.115 ± 0.022 (0.024)
Kidney	9.15 ± 1.30 (3.48)	10.6 ± 1.80 (3.91)	11.4 ± 1.64 (4.18)	0.789 ± 0.132 (0.282)
Spleen	4.37 ± 0.042 (0.606)	3.87 ± 0.491 (0.585)	3.62 ± 0.607 (0.472)	0.299 ± 0.035 (0.050)
Liver	5.03 ± 0.597 (8.37)	7.04 ± 1.34 (14.4)	22.2 ± 2.43 (35.2)	1.65 ± 0.530 (2.82)
Small intestine	3.94 ± 0.487 (2.42)	3.69 ± 0.076 (3.58)	2.27 ± 0.220 (1.30)	0.097 ± 0.004 (0.062)
Large intestine	2.25 ± 0.164 (0.402)	2.00 ± 0.388 (0.527)	1.38 ± 0.160 (0.216)	0.044 ± 0.002 (0.009)
Brain	0.127 ± 0.004 (0.045)	0.153 ± 0.021 (0.054)	0.164 ± 0.111 (0.057)	0.022 ± 0.002 (0.008)
Lymph nodes	1.93 ± 0.118 (0.544)	1.85 ± 0.527 (0.335)	1.56 ± 0.352 (0.256)	0.062 ± 0.008 (0.013)
Eye	0.516 ± 0.064 (0.025)	0.478 ± 0.081 (0.018)	0.295 ± 0.051 (0.012)	0.025 ± 0.002 (0.001)

^an = 3 rats per time interval.

In vitro metabolism

Over the wide range of in vitro systems tested, there was no conversion of [³H]caspofungin to the polar hydrolytic metabolites, M1 and M2, detected in vivo in urine. There was a low-level formation of M0 in some reaction mixtures involving relatively long incubation times; however, similar levels were also formed spontaneously in control incubations under the same conditions in the absence of enzymes.

DISCUSSION

These single-dose studies of [³H]caspofungin administration to human subjects examined the distribution, metabolism, and elimination of caspofungin. A striking feature of the radioactivity recovery was the observation that little excretion of drug-related material occurred during the first few days postdose and that the rate of recovery did not peak until 6 to 7 days postdose for both urine and feces. This finding was consistent with there being a series of slow disposition steps occurring prior to the major pathway(s) of excretion. Furthermore, previously reported data on metabolic profiling of samples from study A indicated that parent drug was the major component of radioactivity in plasma and urine in the first 24 to 30 h postdose (2). The general comparability of the parent and radioactivity profiles for the first ∼24 h in Fig. 2 is consistent with this more definitive metabolic profiling data. Thus, the initial disposition steps appear to be distribution processes, since little excretion or biotransformation of caspofungin was seen at 24 to 30 h postdose. This indicates that distribution plays a prominent role in caspofungin plasma pharmacokinetics and is the rate-controlling step in both the α- and β-disposition phases.

The initial V is not much larger than plasma volume (∼3 liters), suggesting that caspofungin initially is largely confined to the plasma space. Albumin space (∼7.5 liters) is similar in size to the 8- to 10-liter volume obtained at the end of the α-phase (10). Caspofungin is extensively bound to albumin, which is in turn restricted to plasma and extracellular fluid, suggesting that the expansion of the V occurring during the α-phase is roughly from plasma space to extracellular fluid space. Distribution into red blood cells appears to be a minor component of this volume expansion, since in vitro results indicate that the blood/plasma partitioning ratio is low.

Further distribution from extracellular space into tissue appears to be the rate-controlling step during the prominent β-phase of the caspofungin plasma profile, since minimal biotransformation is evident for much of this phase. This is further supported by the mass balance, which indicates that radioactivity in tissues peaks at ∼92% of dose at 1.5 to 2 days postdose. Although the specific composition of this radioactivity (parent drug versus metabolites) is unknown, the slow biotransformation of caspofungin indicates that the radioactivity taken up into tissue for at least the first 24 to 30 h is predominantly composed of parent compound.

The metabolism, excretion, and pharmacokinetics of caspofungin in laboratory animals, including rats, were similar to those observed in humans (11). Therefore, animal tissue distribution results likely provide insight into the nature of tissue distribution in humans. In the rat slow, but extensive, distribution of radioactivity into liver was observed. This pattern of slow, but extensive, uptake into liver could be consistent with the tissue uptake process inferred above to be the predominant mechanism controlling the β-phase of the caspofungin plasma profile in humans. In situ perfusion experiments performed using rat livers suggest that hepatic uptake of caspofungin is a two-step process involving an initial rapid, reversible adsorption to the cell surface followed by a slow transport across the cell membrane (X. Xu, F. Deluna, M. Cartwright, et al., Abstr. Pharm. Res. 13:S454, 1996). In these [³H]caspofungin perfusion experiments, most (81%) of the liver-associated radioactivity following a short-term (1-h) perfusion could be removed by a subsequent washing step, but little (19%) liver-associated radioactivity could be removed by washing of livers obtained from rats dosed in vivo 24 h prior. These results are consistent with the uptake of caspofungin into hepatocytes, and possibly other tissue cells, being mediated at least in part by an active transport process. This allows caspofungin, a drug that does not readily cross cell membranes, as evidenced by its poor oral bioavailability, to extensively accumulate in hepatocytes. This interpretation is also consistent with the fact that most of the caspofungin plasma AUC is associated with the α- and β-distribution phases, even though little excretion or biotransformation of caspofungin occurs during these phases. The observed accumulation on multiple daily dosing of ∼50% for AUC (13) is reasonably consistent with the single-dose profile and does not suggest substantial additional accumulation associated with disposition phases subsequent to the β-phase. This is consistent with the interpretation that caspofungin taken up into tissue cells during the β-distribution phase generally does not return to circulation as parent compound in appreciable amounts. Active transport of caspofungin into hepatocytes may act as a mechanism of plasma clearance if the rate of metabolism of caspofungin in the hepatocytes were much faster than the rate of return of intact caspofungin to plasma.

Application of the standard noncompartmental approach to determine the steady-state volume of distribution (V_SS) for caspofungin based on the AUC and the area under the first moment of the concentration-time curve yields a value of ∼9.5 liters; however, it is likely that this value substantially underestimates the true V_SS for caspofungin. In the derivation of the V_SS equation (6), it is assumed that the drug concentrations acted upon by the elimination processes (metabolism and excretion) are equivalent to the concentrations at the site of measurement (plasma). For caspofungin this is unlikely to be true, since drug taken up into tissues, which subsequently undergoes metabolism and possibly biliary excretion, does not appear to freely equilibrate with plasma. Because this assumption is violated, V_SS values determined by the standard equation largely reflect the volume associated with caspofungin in plasma and extracellular fluid prior to tissue uptake. The true value for caspofungin V_SS cannot be determined from available data; however, estimates for V in the terminal phase based on the model with elimination from the slow peripheral compartment, the model most consistent with the mechanistic interpretation of the caspofungin disposition data, suggest that it is probably large.

The metabolic pathways for caspofungin involve peptide hydrolysis and N-acetylation and not oxidative metabolism (2). A ring-opened peptide, M0, is seen in plasma, and additional metabolism of caspofungin appears to involve the hydrolysis of this hexapeptide into its constitutive amino acids or their degradates. The metabolism of [³H]caspofungin was extensively studied in vitro with at most only low levels of M0 formed in these systems, apparently due to spontaneous degradation, since similar levels were obtained in control incubations. These results indicate that caspofungin is a poor substrate for the major CYP isozymes and the various hydrolytic enzymes tested. They also suggest that the circulating ring-opened peptide, M0, is formed by spontaneous chemical degradation.

Excretion of intact caspofungin in urine and probably also in bile constitute minor pathways of elimination. Since little biotransformation is seen during the first 24 to 30 h postdose, fecal recovery of radioactivity during this period is probably due to biliary excretion of caspofungin. Renal clearance of caspofungin was very low (0.15 ml/min). Based on the free fraction of caspofungin and a glomerular filtration rate of 120 ml/min, the rate of filtration of caspofungin would be expected to be ∼4.2 ml/min, which is much greater than the observed renal clearance. Therefore, there appears to be net tubular reabsorption of caspofungin in the kidneys, probably also via an active transport process. The potential for renal reabsorption of caspofungin may also contribute to the high levels of radioactivity noted in rat kidneys in the tissue distribution study, although the excretion of parent compound and metabolites in urine may have also contributed to the observed radioactivity in rat kidney.

The slow decline of radioactivity in plasma appears to be due to a slow release of radioactivity from the tissues, rather than slow clearance of irreversibly bound radioactivity. If clearance of irreversibly bound radioactivity were the rate-controlling step, then the percentage of irreversibly bound material in the plasma samples should approach 100% during the long log-linear phase, as the other components of radioactivity are more rapidly cleared. Once irreversibly bound material accounted for 100% of the radioactivity in plasma, an additional phase with a longer half-life, which reflected the rate of elimination of irreversibly bound radioactivity, would then be evident in the decline of total radioactivity in plasma. In fact, the level of irreversible binding in plasma was generally constant out to week 8 at around 50% of radioactivity in a given sample, and the total radioactivity profile in plasma showed no evidence of an additional phase beyond week 8. This suggests that the release of radioactivity from tissues is the rate-controlling step in the elimination of radioactivity. This is consistent with the series of slow disposition steps that was proposed above to account for the delay in the peak rate of radioactivity recovery until day 6 to 7.

For antiinfective drugs, the potential for efficacy at a given infection site is sometimes inferred from measurements of drug concentration in tissue homogenates obtained from animal tissue distribution studies or from tissue samples obtained from clinical cases. This is unlikely to be a meaningful approach for caspofungin, since caspofungin concentrations within the extracellular and cellular spaces of tissues could be very different, given the apparent slow equilibration of drug across these spaces. Concentrations in homogenates will only reflect average concentrations within the entire tissue. Nonbloodstream, infective, fungal organisms are generally found within the extracellular space of tissues, not within the cell membranes of intact tissue cells and, as such, in most cases plasma drug concentration is probably a better surrogate for caspofungin concentration at the site of infection than tissue homogenate drug concentration. An exception would be tissues with a potential barrier to penetration from plasma to extracellular space, such as the central nervous system or the eye. However, even for these tissues, tissue homogenate drug concentrations will be very difficult to interpret for the reasons stated above, as well as due to the potential for substantial differences in the extent of protein binding in the extracellular fluid at these sites. Therefore, the potential for efficacy of caspofungin against infections at such sites is better evaluated through animal treatment models for these sites of infection or through clinical experience with treating infections at these sites.

In conclusion, following administration of a single 70-mg i.v. dose of [³H]caspofungin, (i) approximately 75% of radioactivity was recovered in urine and feces over 27 days; (ii) distribution, rather than excretion or biotransformation, was the predominant mechanism controlling the α- and β-phases of the caspofungin plasma profile, and clearance appeared to be largely distribution rate limited; and (iii) renal clearance of unchanged caspofungin constituted a minor pathway of elimination, and biotransformation of caspofungin was slow.