Single- and Multiple-Dose Pharmacokinetics of Caspofungin in Healthy Men

Abstract

Caspofungin, a glucan synthesis inhibitor, is being developed as a parenteral antifungal agent. The pharmacokinetics of caspofungin following 1-h intravenous infusions in healthy men was investigated in four phase I studies. In an alternating two-panel (six men each), rising-single-dose study, plasma drug concentrations increased proportionally with the dose following infusions of 5 to 100 mg. The β-phase half-life was 9 to 10 h. The plasma drug clearance rate averaged 10 to 12 ml/min. Renal clearance of unchanged drug was a minor pathway of elimination (∼2% of the dose). Multiple-dose pharmacokinetics were investigated in a 2-week, serial-panel (5 or 6 men per panel) study of doses of 15, 35, and 70 mg administered daily; a 3-week, single-panel (10 men) study of a dose of 70 mg administered daily; and a parallel panel study (8 men) of a dose of 50 mg administered daily with or without a 70-mg loading dose on day 1. Moderate accumulation was observed with daily dosing. The degree of drug accumulation and the time to steady state were somewhat dose dependent. Accumulation averaged 24% at 15 mg daily and ∼50% at 50 and 70 mg daily. Mean plasma drug concentrations were maintained above 1.0 μg/ml, a target selected to exceed the MIC at which 90% of the isolates of the most clinically relevant species of Candida were inhibited, throughout therapy with daily treatments of 70 or 50 mg plus the loading dose, while they fell below the target for the first 2 days of a daily treatment of 50 mg without the loading dose. Caspofungin infused intravenously as a single dose or as multiple doses was generally well tolerated. In conclusion, the pharmacokinetics of caspofungin supports the clinical evaluation of once-daily dosing regimens for efficacy against fungal infections.

Caspofungin (Cancidas; MK-0991) is an echinocandin that was recently approved by the U.S. Food and Drug Administration for patients with invasive aspergillosis who are refractory to or intolerant of standard therapy. Caspofungin inhibits the synthesis of 1,3-β-d-glucan, which forms a critical component of many fungal cell walls (3). It has been shown to have potent activity in vitro against many clinically important fungi, including Candida spp. and Aspergillus spp. (2, 7, 9; M. Del Poeta, W. A. Schell, and J. R. Perfect, Abstr. 36th Intersci. Conf. Antimicrob. Agents Chemother., abstr. F33, 1996). In in vivo studies with healthy and immunocompromised animals, prolonged survival in models of disseminated aspergillosis and candidiasis and clearance of Candida spp. from a target organ have been demonstrated with caspofungin treatment (1, 5, 6). In phase II clinical trials, caspofungin has been efficacious in the treatment of esophageal and oropharyngeal candidiasis (E. Arahoon, E. Gotuzzo, L. Noriega, J. Andrader, Y. S. Kim, C. A. Sable, and M. DeStafano, Abstr. 36th Annu. Meet. Infect. Dis. Soc. Am., abstr. 99, 1998; C. A. Sable, A. Villanueva, E. Arathon, E. Gotuzzo, G. Turcato, D. Uip, L. Noriega, C. Rivera, E. Rojas, V. Taylor, R. Berman, G. B. Calandra, and J. Chodakewitz, Abstr. 37th Intersci. Conf. Antimicrob. Agents Chemother., symposium S-74, 1997) and invasive aspergillosis (J. Maertens, I. Raad, C. A. Sable, A. Ngai, R. Berman, T. F. Patterson, D. Denning, and T. Walsh, Abstr. 40th Intersci. Conf. Antimicrob. Agents Chemother., abstr. 1103, p. 371, 2000).

Caspofungin has been developed as a parenteral agent due to its high molecular weight, unfavorable log P (partition coefficient), and extremely poor oral bioavailability in animals. In all the clinical studies, caspofungin has been administered as a constant-rate, 1-h intravenous (i.v.) infusion. This paper describes the results from four phase I studies conducted with healthy male subjects to investigate the single- and multiple-dose pharmacokinetics of caspofungin.

MATERIALS AND METHODS

Study A: initial single-dose study.

Study A was a double-blind, randomized, placebo-controlled, rising-single-dose, three-period study conducted to investigate the safety, tolerability, and pharmacokinetics of caspofungin in two panels with nine healthy male subjects in each panel. In each period, six subjects per panel received single i.v. doses of caspofungin and three subjects per panel received a placebo. Single i.v. doses of 5, 10, 20, 40, 70, or 100 mg or a placebo in a volume of 200 ml were infused over 1 h in alternating panels. Subjects in panel A received 5, 20, or 70 mg of caspofungin or the placebo, and subjects in panel B received 10, 40, or 100 mg of caspofungin or the placebo. Administrations of the single doses for each subject were separated by at least 2 weeks. We proceeded to the next higher dose within a panel after we determined from a review of the safety data that the previous lower dose was generally well tolerated in the nine subjects from the previous panel. At least 1 week separated administrations of the single doses for panels A and B. Subjects enrolled in this study (and the three other studies described here) were nonsmoking males in generally good health who had not taken prescription medications within 14 days or nonprescription medications within 7 days of the start of the study, had no history of drug or alcohol abuse, and had not enrolled in any prior caspofungin studies. Blood samples were drawn and plasma was collected at 0 h (predose); at 30, 45, 60, 75, and 90 min; and at 2, 3, 4, 6, 9, 12, 24, 48, 72, 96, 144, and 216 h following the start of drug infusion in each study period (periods 1 to 3) for the caspofungin assay. Urine samples were also collected during the intervals 0 to 2, 2 to 6, 6 to 12, and 12 to 24 h in each study period (periods 1 to 3). In this study, as well as in the other three studies, safety and tolerability were assessed throughout the study by clinical evaluation (including evaluations of a 12-lead electrocardiogram, blood pressure, heart rate, and orthostatic blood pressure and a physical examination) and laboratory safety tests. Local i.v. tolerability measurements were taken at specified time points for each day of dosing. Liver function tests were performed on specified days during follow-up. The average age of the 20 healthy men enrolled was 29 years (range, 21 to 39 years), and their average weight was 75 kg (range, 60 to 88 kg). The group consisted of 11 whites, 6 blacks, 2 Hispanics, and 1 South Pacific Islander.

Study B: initial multiple-dose study.

Study B was a double-blind, placebo-controlled, serial-panel, rising-multiple-dose study of 24 healthy male subjects to investigate the safety, tolerability, and pharmacokinetics of caspofungin following daily doses of 15, 35, or 70 mg (panels A, B, and C, respectively). In each panel, six subjects received the active drug and two subjects received a placebo (0.9% saline diluent) in a randomized fashion. Subjects, who had fasted, were administered 14 doses of the study drug infused i.v. in a volume of 200 ml given over 60 min. There was an interval of at least 2 weeks between panels. Plasma drug profiles defined by concentrations at 0 h (predose) and at 0.5, 0.75, 1, 1.25, 1.5, 2, 4, 8, 12, and 24 h postdose were obtained on days 1 and 14. Intervening trough (24-h postdose) concentrations (C_24h) were obtained prior to dosing on days 3, 4, 7, and 10. Plasma drug washout profile samples were obtained on days 16, 17, 19, 21, 24, and 28. Urine was collected during the intervals 0 to 2, 2 to 6, 6 to 12, and 12 to 24 h on days 1 and 14. Subjects initially enrolled in panel B (35 mg daily) were released following 4 days of treatment due to the potential instability of the drug caused by intermittent elevations in the temperature of the freezer storing the drug supplies for this panel. These subjects were replaced, and the panel was restarted with new drug supplies that had been stored at the appropriate temperature. The average age of the 32 healthy men enrolled was 29 years (range, 19 to 45 years), and their average weight was 77 kg (range, 60 to 96 kg). The group consisted of 20 whites, 3 blacks, 7 Hispanics, and 1 Asian.

Study C: 3-week multiple-dose study.

Study C was a double-blind, placebo-controlled, single-panel study of 12 healthy men. Subjects were randomized to receive 70 mg of caspofungin daily (10 subjects) or to receive a matched placebo (2 subjects) administered i.v. in an infusion volume of 200 ml over 60 min for 21 consecutive days. Plasma drug profiles defined by concentrations at 0, 0.5, 0.75, 1, 1.25, 1.5, 2, 4, 8, 12, and 24 h postdose were obtained on days 1, 14, and 21. Intervening C_24h were obtained prior to dosing on days 3, 4, 6, 8, 10, 12, 16, 18, 19, and 20. Plasma drug washout profile samples were obtained on days 23, 24, 26, 28, 31, 42, and 49. No urine was collected. The average age of the 12 healthy men enrolled was 32 years (range, 23 to 45 years), and their average weight was 84 kg (range, 69 to 92 kg). The group consisted of 10 whites, 1 black, and 1 Hispanic.

Study D: loading-dose study.

Study D was a placebo-controlled, randomized, parallel-panel study consisting of five treatment regimens (one treatment regimen per panel) administered daily over 14 days (eight men per panel). This report addresses only the results from two of the five treatment panels: that in which 50 mg of caspofungin was administered daily on days 1 through 14 and that in which 70 mg of caspofungin was administered on day 1 followed by 50 mg of caspofungin on days 2 through 14. Results from the other three panels of this study, which addressed gender effects and the potential for drug interactions between caspofungin and itraconazole, will be addressed in a separate publication. In the two panels discussed in this paper, plasma samples were collected for the caspofungin assay by using the same sampling scheme used in study B. No urine was collected. The average age of the 19 healthy men enrolled in the two panels addressed in this report was 31 years (range, 20 to 43 years), and their average weight was 79 kg (range, 64 to 103 kg). The group consisted of 7 whites and 12 blacks.

The study protocols described in this report were approved by the Institutional Review Board of Thomas Jefferson University (studies A and D) or by the Research Consultants Review Committee (studies B and C), and informed consent was obtained from all subjects.

Bioanalytical analysis.

All plasma and urine samples were stored at −70°C until they were analyzed. Plasma drug concentrations were determined by high-pressure liquid chromatography (HPLC) with fluorescence detection as previously described (8). The limit of quantification was 10 ng/ml. This assay was modified slightly for studies C and D. To allow for smaller sample volumes, 0.5 ml of plasma, rather than 1.0 ml, was used, with a resulting limit of quantitation of 25 ng/ml. The standard curve range in the modified assay was 25 to 2,000 ng/ml. In addition, for study D, a column-switching procedure was employed by using a Keystone Phenyl Guard cartridge, a rapid resolution Zorbax C₁₈ 4.6- by 75-mm column, and a Zorbax C₈ 4.6- by 150-mm column. HPLC separation was performed at a flow rate of 1.5 ml/min and a temperature of 40°C. This column-switching procedure was employed to avoid the potential appearance of an interfering peak that was seen in a single batch of control plasma.

Urine samples from the men in study A and the men of study B receiving 15- and 35-mg doses were not analyzed because significant losses of drug due to adsorption to plastic during the collection procedure were identified. Bovine serum albumin (2 ml of a 35% solution in saline per 200 ml of urine) was added to urine samples from the 70-mg group of study B to prevent adsorption of the drug. Urine samples from the 70-mg group were analyzed by HPLC with fluorescence detection as previously described (8). The limit of quantitation was 10 ng/ml.

Pharmacokinetic analysis.

The plasma drug terminal rate constants β and γ were calculated by weighted (1/y²) nonlinear regression of individual terminal plasma drug concentration data using a monoexponential or biexponential decay function. The onset of the log-linear phase was determined subjectively for each subject. Onset of the second disposition phase (β phase) generally occurred at 6 to 9 h postdose. Profiles with a single quantifiable point in the γ phase (typically occurring for single doses of 20 and 40 mg) were fit to monoexponential decay functions, with the final point being excluded. Half-lives at the β and γ phases (t_1/2β and t_1/2γ, respectively) were computed as the quotient of ln(2) and the rate constant. In the single-dose study (study A), the t_1/2γ could not be calculated with adequate precision because of insufficient data above the quantification limit.

The area under the concentration-time curve (AUC) was calculated during the interval from 0 h to the last quantifiable point or 24 h for the multiple-dose studies by the linear-log trapezoidal method. For the single-dose study, the AUC during the interval of the last quantifiable point to infinity was extrapolated as the quotient of the last quantifiable concentration and the terminal rate constant. Clearance was determined as the quotient of the dose and the AUC from 0 h to infinity (AUC_0-∞). Actual sampling times, as recorded by the investigator, were used for the calculation of β, γ, and AUCs. For several of the concentration-time profiles, the end of infusion did not coincide precisely with the actual sampling times for the drug concentration at 1 h (C_1h). 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. Incremental and cumulative renal clearances were calculated as the quotient of drug recovered from the urine and the corresponding AUC during the same interval.

On day 1 in study D, plasma samples at the 24-h sampling time point for many subjects were drawn substantially before the nominal time, with the earliest sample being drawn at 22.3 h. These early sampling times were projected to result in overestimations of the C_24h of up to 10% and underestimations of the AUC from 0 to 24 h (AUC_0-24) of up to 4%. Therefore, C_24h and AUC_0-24 on day 1 were corrected through extrapolation of the measured C_24h with the β-phase rate constant as follows: C_24h (corrected) = C_24h (measured) ·e^{−β(24−t_sampling}), where β(24−t_sampling) is the β times the number of hours prior to the 24-h time point at which the sample was taken.

Statistical analysis.

All subjects who completed the pharmacokinetic sampling for a study or period (study A) were included in the pharmacokinetic analysis. Unless otherwise noted, all tests were two sided and assumed a significance level, α, of 0.05.

For study A, a mixed analysis of variance (ANOVA) model was used to fit models for the natural log of AUC, C_1h, C_24h, and clearance and for untransformed rate constants. Independent variables were the panel, the dose nested within the panel (fixed effects), and the subject nested within the panel (random effect). Mean estimates and confidence limits were back-transformed (by taking antilogs), providing geometric means and 90% confidence intervals (CIs) for the geometric means. With regard to assessing dose proportionality, since the difference between slopes in the two panels was small (judged by the 90% CIs), a combined estimate of the slope was obtained by fitting the ln AUC, C_1h, and C_24h to a model of the panel and ln dose. Linear (intercept fixed at the origin, as well as fit), quadratic, and cubic (in ln dose) models were fit to assess dose proportionality for the AUC, C_1h, and C_24h. Dose proportionality corresponds to a first-order coefficient of 1 and coefficients for higher-order terms of zero. Type III estimates, P values, and 90% CIs were obtained for the coefficients of the ln dose. If the CIs for the higher-order coefficients were sufficiently narrow near 0 and the first-order coefficient was very close to 1, then the parameter was considered a linear effect of the dose.

In study B, separate ANOVA models for the log-transformed values for AUC_0-24, C_1h, and C_24h were utilized; dose was the only fixed effect to be entered into each ANOVA model. Mean least-squares estimates for each dose and the lower and upper confidence limits were then back-transformed to provide geometric means and 90% CIs. For drug accumulation, the difference in the log-transformed values for AUC_0-24, C_1h, and C_24h for the last (day 14) and first (day 1) doses was the dependent variable in the ANOVA model. These differences on the log scale were then exponentiated to obtain the geometric-mean ratios and 90% CIs for each dose. In addition, for AUC_0-24, C_1h, and C_24h the day 14/day 1 geometric-mean ratios at each dose were compared using the ANOVA model discussed above with appropriate pairwise contrasts among doses. Geometric means and 90% CIs based on the distribution over time were calculated for urinary data from study B.

In study C, caspofungin AUC_0-24, C_1h, and C_24h were compared between days 1 and 14 and days 1 and 21. For AUC_0-24 and C_1h, a pooled estimate of the within-subject variance was obtained by calculating the mean square error from an ANOVA model containing terms for subject and day (days 1, 14, and 21), whereas for C_24h, there was evidence of unequal variability among days 1, 14, and 21 (P = 0.014) and comparisons were made in a pairwise fashion (days 1 and 14, days 1 and 21). An assessment of whether steady state was achieved by day 14 in study C was conducted by comparing pharmacokinetic parameters at days 14 and 21 by the method for determining accumulation ratios.

For between-panel comparisons in study D, we performed a one-way ANOVA of log-transformed AUC_0-24, C_1h, and C_24h by using a factor for panel. The difference in least-square means from this model between panels of interest and the corresponding 90% CIs was then exponentiated to obtain the 90% CI for the geometric-mean ratio for the comparison.

Safety information was evaluated by tabulating adverse experiences and by clinical assessment of laboratory data. All subjects enrolled in these studies were included in the safety analyses.

RESULTS

Single-dose pharmacokinetics.

Following single 1-h i.v. infusions, concentrations of caspofungin in plasma declined in a polyphasic manner, as shown in the mean concentration values for the drug in plasma in Fig. 1. A short distribution phase was seen immediately postinfusion. A dominant β phase characterized much of the profile and exhibited clear log-linear behavior from ∼6 to ∼48 h postdose. At higher doses, an additional longer γ phase was evident 2 to 3 days postdose at low concentrations. Table 1 summarizes the pharmacokinetics following single doses. Clearance from the plasma was very slow, with geometric-mean clearance ranging from 9.85 to 12.43 ml/min, and remained approximately constant over the range of doses evaluated, suggesting pharmacokinetic linearity. t_1/2β values were generally independent of dose, further supporting linearity. The geometric-mean AUC, C_1h, and C_24h parameters, with their respective 90% CIs, plotted versus dose in Fig. 2, were well represented by linear regression lines identified in the analysis, which was consistent with dose proportionality.

FIG. 1.

Profiles of mean concentrations of caspofungin in plasma following single 1-h i.v. infusions of 5 to 100 mg in healthy male subjects.

TABLE 1.

Geometric-mean pharmacokinetic data following single doses of caspofungin^a

Dose (mg)b	No. of subjects	AUC0-∞ (μg · h/ml)	C1h (μg/ml)	C24h (μg/ml)	CLc (ml/min)	t1/2βd (h)
5	6	8.04 (7.06, 9.17)	0.78 (0.71, 0.86)	0.10 (0.08, 0.12)	10.36 (9.09, 11.81)	10.26 ± 2.31
10	6	15.13 (13.33, 17.17)	1.60 (1.46, 1.76)	0.18 (0.15, 0.21)	11.02 (9.71, 12.50)	9.56 ± 1.63
20	6	30.97 (27.17, 35.30)	3.06 (2.78, 3.37)	0.36 (0.30, 0.44)	10.77 (9.45, 12.27)	10.72 ± 0.73
40	6	55.60 (49.00, 63.10)	6.17 (5.61, 6.79)	0.57 (0.48, 0.68)	11.99 (10.56, 13.61)	9.64 ± 0.83
70	6	118.45 (103.92, 135.01)	12.04e (10.87, 13.33)	1.42 (1.18, 1.71)	9.85 (8.64, 11.23)	9.29 ± 1.96
100	5	134.11 (117.78, 152.71)	14.03 (12.69, 15.53)	1.47 (1.23, 1.76)	12.43 (10.91, 14.15)	8.61 ± 1.15

^aValues in parentheses are 90% CIs.

^bWe evaluated doses of 5, 20, and 70 mg in the men in panel A and doses of 10, 40, and 100 mg in the men in panel B. At each dose level, three of nine subjects in each panel received a placebo in a randomized fashion, such that results within panel are partially, but not completely, crossed over.

^cCL, geometric-mean clearance.

^dHarmonic mean ± jack-knife standard deviation reported for t_1/2β.

^eThe number of C_1h obtained was 5 at this dose due to a missing value.

FIG. 2.

Geometric-mean AUC_0-∞, C_1h, and C_24h (with 90% CIs) versus dose following single 1-h i.v. infusions of caspofungin in healthy male subjects. The linear regression lines were obtained by taking the antilog of the estimated constant from the model with only the linear term included.

In the early clinical development program for caspofungin, a target concentration of 1 μg/ml was used as a guide for concentrations likely to be efficacious. This target was determined from in vitro susceptibility testing results and was selected as a concentration that exceeded the MIC at which 90% of most clinically meaningful isolates of Candida spp. were inhibited (2). In the initial single-dose study, the estimated geometric means and CIs for C_24h of the 70- and 100-mg doses exceeded the target concentration of 1 μg/ml.

Multiple-dose pharmacokinetics.

Figure 3 shows the mean values from plasma drug profiles following multiple daily 1-h i.v. infusions of caspofungin into healthy men in study B. Table 2 contains summary statistics of plasma pharmacokinetic parameters and the associated accumulation ratios from the three multiple-dose studies. The harmonic mean t_1/2β and t_1/2γ values following the last dose in these studies are provided in Table 2. The drug washout profiles demonstrated polyphasic behavior, similar to that seen following single doses, with a short distribution phase immediately postinfusion, a β phase dominating much of the profile, and a longer γ phase evident after 48 to 72 h.

FIG. 3.

Profiles of mean concentrations of caspofungin in plasma following multiple daily 1-h i.v. infusions of 15 to 70 mg in healthy male subjects (study B).

TABLE 2.

Geometric-mean pharmacokinetics data following multiple daily doses of caspofungin^a

Dose (mg/day)	Study	No. of subjects	AUC0-24 (μg · h/ml)					C1h (μg/ml)					C24h (μg/ml)					t1/2β (h) ± SD	t1/2γ (h) ± SD
			Day 1	Day 14	Day 14/day 1 ratio	Day 21	Day 21/day 1 ratio	Day 1	Day 14	Day 14/day 1 ratio	Day 21	Day 21/day 1 ratio	Day 1	Day 14	Day 14/day 1 ratio	Day 21	Day 21/day 1 ratio
15	B	5	19.66 (17.94, 21.55)	24.37 (21.45, 27.69)	1.24 (1.15, 1.33)			2.47 (2.20, 2.78)	2.79 (2.51, 3.10)	1.13 (1.04, 1.23)			0.29 (0.24, 0.34)	0.38 (0.31, 0.47)	1.31 (1.17, 1.46)			8.60 ± 0.89	40.90 ± 2.91b
35	B	5	41.18 (37.57, 45.14)	54.90 (48.32, 62.38)	1.33 (1.24, 1.43)			5.22 (4.65, 5.88)	5.97 (5.38, 6.63)	1.14 (1.05, 1.24)			0.59 (0.50, 0.70)	0.92 (0.75, 1.13)	1.56 (1.40, 1.74)			9.51 ± 1.21	46.73 ± 5.66
50	D	8	57.57 (51.85, 63.91)	86.90 (76.06, 99.29)	1.51 (1.43, 1.59)			7.64 (7.02, 8.33)	8.74 (7.92, 9.63)	1.14 (1.07, 1.22)			0.76 (0.60, 0.96)	1.59 (1.30, 1.95)	2.09 (1.82, 2.40)			10.09 ± 1.58	42.13 ± 10.38
70/50c	D	8	97.63 (87.94, 108.39)	100.47 (87.93, 114.79)	1.03 (0.98, 1.08)			12.09 (11.10, 13.16)	9.94 (9.02, 10.96)	0.82 (0.77, 0.88)			1.41 (1.12, 1.78)	1.77 (1.45, 2.17)	1.26 (1.09, 1.44)			10.58 ± 1.13	46.08 ± 8.18
70	B	6	96.01 (88.29, 104.40)	144.27 (128.40, 162.10)	1.50 (1.41, 1.60)			12.31 (11.06, 13.70)	15.42 (14.00, 16.97)	1.25 (1.16, 1.35)			1.33 (1.14, 1.56)	2.61 (2.15, 3.15)	1.96 (1.77, 2.17)			10.70 ± 1.22	50.16 ± 6.54
70	C	10	93.47 (86.34, 101.20)	129.61 (117.91, 142.47)	1.39 (1.35, 1.43)	137.29 (126.03, 149.55)	1.47 (1.43, 1.51)	12.14 (11.50, 12.81)	14.03 (13.26, 14.86)	1.16 (1.12, 1.19)	14.26 (13.47, 15.10)	1.17 (1.14, 1.21)	1.30 (1.11, 1.51)	2.37 (2.07, 2.72)	1.83 (1.77, 1.89)	2.57 (2.23, 2.96)	1.98 (1.85, 2.12)	11.16 ± 1.20	45.50 ± 6.30

^aValues in parentheses are 90% CIs.

^bThe number of t_1/2γ obtained was 4 at this dose.

^c70 mg on day 1 followed by 50 mg daily on days 2 through 14.

Accumulation of drug on multiple dosing is evident in Fig. 3 as increased plasma drug concentrations on day 14 over those on day 1. The geometric-mean ratios characterizing accumulation in Table 2 were significantly greater than 1 for AUC_0-24, C_1h, and C_24h at all doses. The degree of accumulation was greatest for C_24h and least for C_1h. In addition, the degree of accumulation for both AUC_0-24 and C_24h increased with dose in these studies. For AUC_0-24, accumulation increased from 24% for 15 mg daily to ∼50% for 50 and 70 mg daily. For C_24h, this disproportionate accumulation was even more pronounced, increasing from 31% and to ∼100% over the range of doses evaluated. The accumulation for C_1h demonstrated minimal dose dependency, with values ranging from 13 to 25%. Statistical comparison of the day 14 accumulation ratios from study B indicated that the ratios from the 15- and 70-mg doses and the 35- and 70-mg doses were significantly different (P < 0.05) for AUC_0-24 and C_24h, but that there were no significant differences among the ratios for C_1h.

Steady state.

Figure 4, which plots the mean C_24h over time in the three multiple-dose studies, illustrates a dose dependency in the time course of trough accumulation, as well as in the degree of accumulation. Trough concentrations reached a stable plateau by 3 to 4 days at 15 mg daily. At 35 mg daily, trough concentrations climbed slightly after 3 to 4 days, while at 50 or 70 mg daily, they climbed more substantially throughout the 14-day studies. These patterns of trough accumulation at the various doses were fairly consistent among individuals. It is unclear from the current data whether steady state was obtained by 14 days at 35 mg daily. The individual profiles at 35 mg daily suggest that most subjects were approaching steady state by day 14.

FIG. 4.

Mean trough concentrations (C_24h) of caspofungin following multiple daily 1-h i.v. infusions of 15 to 70 mg in healthy male subjects.

The 3-week study of individuals at 70 mg daily was conducted in part to further characterize the approach to steady state for this regimen. Pharmacokinetic parameters at days 14 and 21 were compared to determine whether steady state was achieved during the first 2 weeks of the study. The day 21/day 14 geometric-mean ratios (90% CIs) for AUC_0-24, C_1h, and C_24h were 1.06 (1.04, 1.08), 1.02 (1.00, 1.04), and 1.08 (1.03, 1.14), respectively. These results indicate that steady state was not achieved by day 14, since the geometric-mean ratios for AUC_0-24 and C_24h are significantly (P = 0.001 and 0.017, respectively) greater than 1. However, the accumulation occurring during the third week was slight and suggests that the subjects were approaching steady state during the third week.

Loading dose regimen.

As shown in Table 2, the estimated geometric means and CIs for C_24h at 70 mg daily exceeded the target concentration of 1 μg/ml after the first dose as well as after multiple doses. At 50 mg daily, geometric-mean C_24h fell below 1 μg/ml on day 1 but exceeded the target on day 14. Administration of a 70-mg loading dose on day 1 followed by 50 mg daily on days 2 to 14 resulted in day 1 pharmacokinetics in plasma similar to those obtained on day 14 (Table 2). The ability of the loading dose to generate higher trough drug concentrations in plasma during the initial days of therapy is illustrated in Fig. 5. Mean trough concentrations were maintained above the target concentration of 1 μg/ml throughout the treatment regimen with the loading dose, while they fell below the target concentration for the first 2 days of the regimen without the loading dose.

FIG. 5.

Mean trough concentrations in men receiving 50 mg of caspofungin daily with or without a 70-mg loading dose on day 1.

Summary statistics describing the effects of the loading dose on the pharmacokinetics of 50 mg of caspofungin daily are given in Table 3. The pharmacokinetics on day 14 observed in the two panels were compared to determine if there were persistent effects of the loading dose. The day 14 geometric-mean AUC_0-24 ratio (90% CIs) for the loading dose regimen relative to that without a loading dose was 1.16 (0.96, 1.40). This difference was not statistically significant. C_1h and C_24h exhibited similar nonsignificant trends toward slight increases with the loading dose regimen. No statistically significant differences in t_1/2β or t_1/2γ values were identified in an analysis of the rate constant data.

TABLE 3.

Effect of a 70-mg day 1 loading dose on the pharmacokinetics of caspofungin at 50 mg daily

Day	Geometric-mean ratio (loading dose/no loading dose) (90% CIs)
	AUC0-24	C1h	C24h
1	1.70 (1.46, 1.97)	1.58 (1.40, 1.78)	1.86 (1.34, 2.57)
14	1.16 (0.96, 1.40)	1.14 (0.99, 1.31)	1.12 (0.84, 1.48)

Recovery in urine.

Summary statistics of the urinary pharmacokinetic parameters, including drug concentration, amount excreted (amount recovered), and renal clearance over timed intervals and over the 24-h dosing interval, are provided in Table 4 for the 70-mg-daily panel of study B. Renal clearance of caspofungin was very slow, and very little caspofungin was excreted unchanged in urine.

TABLE 4.

Geometric-mean urinary pharmacokinetics of caspofungin following multiple daily 1-h i.v. infusions of 70 mg to healthy male subjects^a

Collection interval on day 14 (h)	Drug concn in urine (μg/ml)	% of dose recovered in urine	Renal clearance (ml/min)
0 to 2	0.40 (0.22, 0.72)	0.13 (0.11, 0.14)	0.06 (0.05, 0.08)
2 to 6	1.48 (0.82, 2.69)	0.58 (0.54, 0.63)	0.18 (0.15, 0.21)
6 to 12	1.03 (0.61, 1.74)	0.57 (0.44, 0.74)	0.18 (0.14, 0.23)
12 to 24	0.66 (0.36, 1.21)	0.72 (0.52, 0.99)	0.19 (0.13, 0.28)
0 to 24		2.02 (1.67, 2.45)	0.16 (0.13, 0.21)

^aValues in parentheses are 90% CIs.

Safety.

Safety and tolerability were carefully monitored in these phase I studies of caspofungin. In studies A, B, and C and the two panels of study D addressed in this paper, 56 of the 98 subjects enrolled had at least one clinical adverse experience. Most of the clinical adverse experiences were mild in intensity and rarely led to discontinuation. The most frequently reported adverse experiences included headache and local problems with tolerability, including erythema, pruritus, rash, induration, pain, and tenderness. Headache was common for subjects on both caspofungin and placebo. The local-tolerability findings were more common for subjects receiving caspofungin than those receiving the placebo. The only laboratory-determined adverse experience judged (possibly) study drug related was increases in aspartate aminotransferase and alanine aminotransferase (both <2-fold greater than the normal upper limit) in one subject receiving 50 mg of caspofungin daily; this subject was released from the study. The only other subject released for an adverse event judged possibly drug related was a subject who experienced a sensation of chest tightness while receiving caspofungin. There was one serious adverse experience, trauma secondary to a motor vehicle accident, which was considered definitely not study drug related, and this subject withdrew from further participation in the study.

DISCUSSION

The rising-single-dose study was the first administration of caspofungin to humans. Single i.v. infusions of doses ranging from low doses up to those likely to be adequate for efficacy in treating fungal diseases were studied. The relatively long t_1/2β obtained in the single-dose study supported the investigation of daily dosing regimens, and the three subsequent studies of multiple daily doses in healthy subjects, described in this paper, were conducted. In addition to an evaluation of safety, pharmacokinetic data were obtained and compared to target fungicidal concentrations to guide the dosing regimen to be employed in studies of drug efficacy in patients with fungal infections.

Following single i.v. drug infusion, most of the AUC was accounted for by the β phase. The 9- to 10-h t_1/2 obtained for this phase is consistent with the moderate accumulation of caspofungin seen with the daily doses of 15 to 70 mg. Clearance from plasma was very slow across the doses studied, averaging 10 to 12 ml/min. As such, quantifiable drug levels were obtained up to 7 days postdose following the higher single doses.

The results following single doses of caspofungin are consistent with pharmacokinetic linearity. The concentration data (C_1h, C_24h, and AUC_0-∞) are dose proportional. Clearance and t_1/2β are relatively consistent across doses within and between subjects. However, the results following multiple doses indicate that caspofungin has a modest pharmacokinetic nonlinearity. This nonlinearity is most evident in the accumulation data, most notably in the trough accumulation behavior. The degrees of accumulation with both AUC_0-24 and C_24h, but not C_1h, increased with dose. The time course of trough accumulation in the period from days 3 to 14 also indicates a nonlinearity, since the time to steady state was dose dependent. In a drug with linear pharmacokinetics, the degree of accumulation and the time to steady state are independent of dose. In the single-dose study, no reliable data regarding the linearity of the γ phase are available, so it is possible that the nonlinear accumulation behavior arises from nonlinearities in this phase. However, t_1/2γ following the last dose of the multiple-dose regimens showed only a small trend toward increases with dose (from 40 to 50 h).

In the 3-week study, the comparison of pharmacokinetic parameters on days 14 and 21 indicates that steady state was not achieved within the first 2 weeks at 70 mg daily; however, the accumulation occurring in the third week was slight. The comparison of pharmacokinetic parameters on days 14 and 21 identified increases of less than 10% for AUC_0-24, C_1h, and C_24h during the third week of dosing in most individuals. These results suggest that caspofungin concentrations were nearing steady state by day 21. In subsequent studies of patients receiving long-term (>28 days) therapy, continued slow accumulation of caspofungin was not seen following the first few weeks of dosing (Merck Research Laboratories, unpublished data). Finally, it should be noted that much of the accumulation of caspofungin following daily dosing occurs during the first week of dosing, as illustrated in Fig. 4, and that the slow accumulation occurring during the second and third weeks of dosing at higher doses is small relative to that seen during the first week.

The relationship between pharmacokinetic variables, in vitro measures of microorganism susceptibilities, and clinical outcomes have not been well elucidated for antifungal agents, and no correlations have been established, even for some drugs that have been available for many years (4). With this new class of antifungals, even further caution in projecting these correlations is required. Nonetheless, it is helpful to work from an initial hypothesis to examine the likely adequacy of the concentration attained with caspofungin in plasma. The target concentration selected for this program was 1.0 μg/ml, which generally exceeds the MIC at which 90% of the isolates of clinically relevant species of Candida were inhibited (2). The comparisons of C_24h of caspofungin in plasma with the target concentration support the investigation of once-daily dosing regimens of 50 or 70 mg of caspofungin, since mean plasma drug concentrations following multiple doses exceed the target concentration for these regimens. The generally good tolerability and the relatively long t_1/2β and t_1/2γ values also support investigation of once-daily dosing in studies of caspofungin efficacy.

The loading dose regimen was investigated in study D because pharmacokinetic model simulations had suggested that this would allow target concentrations of caspofungin to be reached sooner. The pharmacokinetic data obtained confirm that there may be an advantage to using the loading dose for life-threatening infections requiring early effective treatment. The loading dose regimen maintained mean concentrations above the 1-μg/ml target throughout the study, while concentrations were below target at trough during the first 2 days of the 50-mg-daily regimen without a loading dose. In the evaluation of the loading dose, there was a trend toward higher day 14 plasma drug concentrations in men receiving the loading dose, which could be consistent with the modest nonlinearity in caspofungin accumulation. In a linear system, there should be no persistent effect of the loading dose at steady state or near steady state. However, after the initial dose, mean trough concentrations in men receiving the loading dose approached, but never reached, those obtained in men receiving no loading dose, as illustrated in Fig. 5. Rather, the two trough profiles climbed in parallel after approximately day 4, suggesting a slight persistent elevation in caspofungin concentrations in men receiving the loading dose, although this difference was not statistically significant.

Renal clearance of the parent drug was shown to play a minor role in caspofungin disposition (∼2% of dose cleared renally as unchanged drug). Therefore, altered renal clearance of the parent drug in the setting of renal insufficiency would be expected to have little direct effect on the pharmacokinetics of caspofungin. However, an effect of renal insufficiency on the pharmacokinetics of caspofungin cannot be ruled out based on this finding. Potential mechanisms through which renal dysfunction could indirectly alter caspofungin disposition include accumulation of a metabolite or an endogenous substance that interferes with caspofungin disposition or plasma protein binding and alterations in nonrenal organ function associated with chronic renal disease. Further study of patients with renal insufficiency would be needed to determine whether caspofungin pharmacokinetics is altered through these indirect mechanisms.

Caspofungin was generally well tolerated. The most frequent drug-related nonserious clinical adverse experiences included headache and the local tolerability signs or symptoms of induration, erythema, pain, swelling, and edema. In most cases, these were mild in intensity. Mild-to-moderate headaches may have been due in part to the fasting and caffeine restrictions. There were few laboratory adverse experiences.

In conclusion, the pharmacokinetics of caspofungin supports the clinical evaluation of once-daily dosing regimens for efficacy against fungal infections. Caspofungin shows moderate accumulation after 14 to 21 days of repeated daily i.v. administration. While the pharmacokinetic data of caspofungin administered i.v. as single doses between 5 and 100 mg are consistent with linearity, the degree of accumulation of multiple daily doses increases with dose, suggesting modest pharmacokinetic nonlinearity. Administration of a 70-mg loading dose on day 1, followed by 50 mg of caspofungin daily, maintains mean plasma caspofungin concentrations above a 1-μg/ml target throughout treatment.