Potential for Interactions between Caspofungin and Nelfinavir or Rifampin

The potential for interactions between caspofungin and nelfinaviror rifampin was evaluated in two parallel-panel studies. Instudy A, healthy subjects received a 14-day course of caspofunginalone (50 mg administered intravenously [IV] once daily) (n= 10) or with nelfinavir (1,250 mg administered orally twicedaily) (n = 9) or rifampin (600 mg administered orally oncedaily) (n = 10). In study B, 14 subjects received a 28-day courseof rifampin (600 mg administered orally once daily), with caspofungin(50 mg administered IV once daily) coadministered on the last14 days, and 12 subjects received a 14-day course of caspofunginalone (50 mg administered IV once daily). The coadministration/administrationalone geometric mean ratio for the caspofungin area under thetime-concentration profile calculated for the 24-h period followingdosing [AUC_0-24] was as follows (values in parentheses are 90%confidence intervals [CIs]): 1.08 (0.93-1.26) for nelfinavir,1.12 (0.97-1.30) for rifampin (study A), and 1.01 (0.91-1.11)for rifampin (study B). The shape of the caspofungin plasmaprofile was altered by rifampin, resulting in a 14 to 31% reductionin the trough concentration at 24 h after dosing (C_24h), consistentwith a net induction effect at steady state. Both the AUC andthe C_24h were elevated in the initial days of rifampin coadministrationin study A (61 and 170% elevations, respectively, on day 1)but not in study B, consistent with transient net inhibitionprior to full induction. The coadministration/administrationalone geometric mean ratio for the rifampin AUC_0-24 on day 14was 1.07 (90% CI, 0.83-1.38). Nelfinavir does not meaningfullyalter caspofungin pharmacokinetics. Rifampin both inhibits andinduces caspofungin disposition, resulting in a reduced C_24hat steady state. An increase in the caspofungin dose to 70 mg,administered daily, should be considered when the drug is coadministeredwith rifampin.

INTRODUCTION

Top

Introduction

Caspofungin (known by the trade name CANCIDAS; also known asMK-0991) is a parenteral antifungal agent that inhibits 1,3-ß-D-glucansynthesis, which forms a critical component of the cell wallof many fungal species (3). Caspofungin is active against manyclinically important fungal species, including Candida spp.and Aspergillus spp. (2, 4, 8, 15). In clinical trials, it hasbeen shown to be efficacious in the treatment of esophagealcandidiasis (1, 16, 17), invasive candidiasis (9), and invasiveaspergillosis (J. Maertens, I. Raad, G. Petrikkos, D. Selleslag,F. Petersen, C. Sable, N. Kartsonis, A. Ngai, A. Taylor, T.Patterson, D. Denning, and T. Walsh, Abstr. 42nd Intersci. Conf.Antimicrob. Agents Chemother., abstr. M-868, 2002). Caspofunginis approved for the treatment of invasive aspergillosis in patientsrefractory to or intolerant of standard therapy, invasive candidiasis,and esophageal candidiasis.

Metabolism and excretion of caspofungin are very slow processes(13). In contrast to the situation with many other drugs, neitherof these processes is the rate-controlling step that determinesthe clearance of caspofungin from plasma. Rather, plasma clearanceis determined primarily by the rate of distribution of caspofunginfrom plasma into hepatocytes and possibly other tissue cells(13). The uptake of caspofungin into tissue cells appears tobe mediated, at least in part, by an active transport process(13).

This paper describes results from two phase I studies of healthysubjects conducted to evaluate the potential for nelfinavir,a human immunodeficiency virus protease inhibitor, or rifampin,an RNA polymerase inhibitor active against tuberculosis andother bacteria, to alter the pharmacokinetics of caspofungin.These studies were undertaken because a drug-interaction screeninganalysis using population pharmacokinetic data from patientswith esophageal and/or oropharyngeal candidiasis had indicatedthat the use of inducers of drug clearance or nelfinavir mightresult in clinically meaningful reductions in the caspofunginarea under the time-concentration profile (AUC) and trough concentrationat 24 h postdosing (C₂₄_h) (J. Stone, S. Li, G. Winchell, S.Bi, P. Wickersham, M. Schwartz, N. Kartsonis, and C. Sable,Abstr. 43rd Intersci. Conf. Antimicrob. Agents Chemother., abstr.A-1571, 2003). Additionally, it is reasonable to expect thatthese drugs will be used in combination at times in clinicalpractice. Immunocompromised people are susceptible to developingactive tuberculosis, as well as invasive fungal infections.Patients with AIDS may develop fungal infections due to theirimmunocompromised state.

Rifampin was chosen as a model inducer to test the hypothesisthat inducers could alter caspofungin pharmacokinetics, sinceit is a potent inducer that affects a broad spectrum of drug-dispositionprocesses. These phase I studies provide a more definitive evaluationof the potential for drug interactions with nelfinavir or rifampinthan could be obtained with population pharmacokinetics datain view of their prospective study designs, the larger numbersof subjects receiving the combinations, more extensive pharmacokineticsevaluation, exclusion of other concomitant therapies, and thelack of major underlying illnesses in participating subjects.

MATERIALS AND METHODS

Top

Materials and Methods

Study A. Study A was an open-label, randomized, parallel-panel studyof healthy male subjects to investigate the potential for nelfinavirand rifampin to alter the pharmacokinetics of caspofungin. Subjectsreceived one of three treatment regimens in which drugs wereadministered daily over 14 days, as follows: panel 1, 50 mgof caspofungin administered intravenously (IV) once daily (n= 10); panel 2, 50 mg of caspofungin administered IV once dailyand 600 mg of rifampin administered orally once daily (n = 10);and panel 3, 50 mg of caspofungin administered IV once dailyand 1,250 mg of nelfinavir administered orally every 12 h (n= 9). IV doses of caspofungin were administered as a constant-rateinfusion over 1 h. Subjects in each panel ingested a fixed moderate-fatmeal prior to dosing. Rifampin and the morning dose of nelfinavirwere administered just prior to the start of the caspofungininfusion on coadministration days. Subjects enrolled in thisstudy (and the second study described below) were male, nonsmokersin generally good health, who had no history of drug or alcoholabuse and who did not take prescription medications, St. John'swort, or other herbal remedies within 14 days or nonprescriptionmedications or grapefruit products within 7 days of the startof the study. Blood samples were drawn and plasma was collectedfor assay at 0 (predose) and 0.5, 0.75, 1, 1.25, 1.5, 2, 3,4, 6, 9, 12, and 24 h after dosing on days 1 and 14. To obtainintervening trough (24-h postdose) concentrations, blood sampleswere also drawn just prior to dosing on days 2 to 13. The 29healthy men enrolled had an average age of 31 years (range,18-44), an average weight of 76 kg (range, 62-93), and the followingracial distribution: 14 white subjects, 12 black subjects, 2Hispanic subjects, and 1 multiracial subject.

Study B. This was an open-label, randomized, parallel-panel study ofhealthy male subjects to further investigate the potential forinteractions between caspofungin and rifampin. Subjects receivedone of two treatment regimens, as follows: panel 1, 50 mg ofcaspofungin administered IV once daily for 14 days (n = 12);and panel 2, 600 mg of rifampin administered orally once dailyfor 28 days, with 50 mg of caspofungin coadministered IV oncedaily for the final 14 days (n = 14). IV doses of caspofunginwere administered as a constant-rate infusion over 1 h. Subjectsin each panel ingested a fixed moderate-fat meal prior to dosing.Rifampin was administered just prior to the start of the caspofungininfusion on coadministration days. To characterize caspofunginpharmacokinetics, blood samples were drawn and plasma was collectedfor assay at 0 (predose) and 0.5, 0.75, 1, 1.25, 1.5, 2, 3,4, 6, 9, 12, and 24 h after dosing on days 1 and 14 of caspofungindosing; blood samples (to obtain intervening trough concentrations)were also drawn just prior to dosing on days 2 to 13. For panel2, plasma samples were collected at 0 (predose) and 1, 2, 4,6, 9, 12, and 24 h after dosing on day 14 (prior to coadministrationof caspofungin), day 15 (day 1 of coadministration), and day28 (day 14 of coadministration) of rifampin dosing to characterizerifampin pharmacokinetics. The 26 healthy men enrolled had anaverage age of 35 years (range, 18-44), an average weight of76 kg (range, 58-92), and the following racial distribution:4 white subjects, 3 black subjects, and 19 Hispanic subjects.

The study protocols described in this report were approved bythe Institutional Review Board of Thomas Jefferson University(study A) or by the Southern Institutional Review Board of Miami,Florida (study B), and informed consent was obtained from allsubjects.

Bioanalytical analysis. Plasma samples for determination of caspofungin concentrationswere stored at [minus]70°C until analysis. Plasma concentrationsof caspofungin were determined by high-pressure liquid chromatography(HPLC) with fluorescence detection as previously described (11).The plasma assay was modified slightly to allow for smallersample volumes; 0.1 ml of plasma was used, with a resultinglimit of quantitation of 125 ng/ml. The standard curve rangewas 125 to 10,000 ng/ml in the modified assay. The intradayprecision of the assay, as measured by the percent coefficientof variation (%CV) of the peak height ratios, was better (i.e.,less) than 10% (1.2 to 9.4%) at all points of the standard line.The interday variability of the assay, as assessed by the %CVof the quality control samples, ranged from 3.3 to 6.4%. Inaddition, a column-switching procedure was employed as describedin reference 12. No interference of rifampin with the caspofunginassay was seen.

Plasma samples were analyzed for the presence of rifampin byMDS Pharma Services, Sunnyvale, Calif., using HPLC with ultravioletlight detection. Approximately 1 mg of ascorbic acid/ml wasadded to all plasma samples for the rifampin assay prior tofreezing for improved stability. Rifampin and the internal standard,papaverine, were extracted from buffered plasma with chloroform.After evaporation and reconstitution, the extract was injectedonto a C₁₈ HPLC column and the rifampin was measured by absorbanceat 340 nm. The limit of quantitation was 0.1 µg/ml. Thestandard curve range was 0.1 to 25 µg/ml. The intradayprecision of the assay, as measured by the %CV, was better than10% (3.8 to 8.9%) at all points of the standard line. The interdayvariability of the assay, as assessed by the %CV of the qualitycontrol samples, ranged from 5.7 to 10.8%. No interference ofcaspofungin with the rifampin assay was seen.

Pharmacokinetic analysis. The AUC over the 24-h interval following dosing [AUC_0-24] wascalculated by the linear-log trapezoidal method. Actual samplingtimes, as recorded by the investigator, were used for calculationof the AUC_0-24. For several caspofungin concentration-time profiles,the timing of the end of infusion and the C_1h plasma samplingdiffered by more than 1 min, and in one instance the C_1h plasmasample was not obtained. In these instances, an estimated end-of-infusionconcentration, determined by fitting the plasma concentrationtime data to a three-compartment linear model, was used alongwith the plasma concentration data in the AUC calculations.For rifampin plasma profiles, the maximum concentration of drugin serum (C_max) and time to maximum concentration of drug inserum (T_max) were determined by inspection.

A pharmacokinetic model of the effect of rifampin on caspofunginpharmacokinetics was developed (Fig. 1). Caspofungin pharmacokineticswere represented by a three-compartment model with eliminationfrom the third compartment. This model structure is consistentwith disposition data which indicate that uptake of caspofungininto tissues is the primary mechanism controlling plasma pharmacokineticsand that the elimination processes (metabolism and excretion)are slow and appear to occur largely subsequent to the tissueuptake process (13). In the model, both the induction and inhibitioneffects of rifampin on caspofungin pharmacokinetics act on thetissue uptake rate constant (k₁₃). The induction process ismodeled by a scaling factor, IND, which represents the foldincrease in intrinsic k₁₃ at maximal (steady-state) induction.The inhibition effect is modeled as a function of rifampin concentration(C_RIF) using a Hill function 1 – C_RIF/(IC₅₀ + C_RIF), whereIC₅₀ is the 50% inhibitory concentration and the superscript ${gamma}$ is the sigmoidicity.

View larger version (13K):
[in this window]
[in a new window]
FIG. 1. Schematic of compartmental model to describe the effect of rifampin on caspofungin pharmacokinetics.

The differential equations that compose the caspofungin modelare as follows:

(1)

(2)

(3)
where A denotes amount; the subscripts 1, 2,and 3 denote the central, extracellular fluid, and tissue compartments,respectively; k₁₂, k₂₁, k₁₃, and k₃₁ are the microconstantsdefining transfer between the compartments; k₀ is the eliminationrate constant; and IR is the infusion rate (IND, IC₅₀, and ${gamma}$ are as described above). The measured concentrations in plasmaare related to A₁ by the central volume of distribution (V_C).

Several variations on this model were considered during modeldevelopment, including addition of a rifampin link model whichallowed rifampin concentrations in an effect compartment tolag behind plasma rifampin concentrations. This addition didnot improve the model fit to data, and the fit values obtaineddid not suggest a time delay between plasma concentrations andeffective concentrations. A second model variation evaluatedinvolved allowing induction and inhibition by rifampin to actonly on a fraction of k₁₃, with the remaining fraction unaffectedby rifampin. Model fits to data with this variation estimatedthe fraction of k₁₃ affected by rifampin to be $~$ 100%, so thisaddition also was not included in the final model.

The software package ACSL (Aegis Software, Huntsville, Ala.)was used for the modeling analysis. The model parameters werefit to weighted (1/y², where y is the magnitude of the concentration)caspofungin plasma data by maximization of the log-likelihoodfunction using a generalized reduced gradient search algorithm.Because the model did not allow for variation in the inductionfactor (IND) with time, the model was only fit to data obtainedunder either no induction (IND = 1.0), i.e., caspofungin monotherapyor day 1 of rifampin administration, or stable maximal induction,i.e., day 28 of rifampin administration in study B. First, themodel was fit to study B data, i.e., panel 1 (caspofungin monotherapy)data on all study days and panel 2 (caspofungin plus rifampin)data on day 14 of coadministration. Rifampin plasma concentrationson all days of coadministration were obtained by linear interpolationof the individual day 28 rifampin concentration data. Second,the model was fit to the day 1 data from study A, with IND setto 1.0 for both monotherapy and combination therapy. Becauseno rifampin plasma concentration data were available for studyA, rifampin concentrations for this fit were obtained by linearinterpolation of the mean rifampin profile concentrations obtainedin study B. Because the study A data fit was limited to day1, it was not possible to estimate all the parameters for theunderlying caspofungin model. Therefore, k₀ and k₃₁ were setto the values obtained in the fit-to-study-B data. For bothdata sets, the model was fit to the pooled set of individualprofile data from combination and monotherapy panels to obtainedmean parameter estimates. Therefore, no estimates of intersubjectdispersion in the mean parameter values were obtained. Thiswas necessary, since any given individual had only combinationor monotherapy data available due to the parallel-panel designof these studies. In addition to the parameter estimates, standarddeviations characterizing the estimation precisions were determinedfor each fit using established routines in the ACSL softwareprogram.

Statistical analysis. All subjects who completed the pharmacokinetics sampling fora study were included in the pharmacokinetics analysis. Unlessotherwise noted, all tests were two-sided and assumed a significancelevel of 0.05.

To evaluate the effect of rifampin (studies A and B) or nelfinavir(study A) on caspofungin pharmacokinetics, a between-panel comparisonof the natural log-transformed pharmacokinetics data [AUC_0-24,C_1h, and C_24h] was conducted based upon a one-way analysis ofvariance model with treatment as a factor. The observed differencesin means and limits of the corresponding 90% confidence intervals(CIs) were exponentiated to obtain the geometric mean ratios(coadministration/administration alone) and the corresponding90% CIs.

To evaluate the effect of caspofungin on rifampin pharmacokineticsin study B, a within-group comparison of the natural log-transformedAUC and C_max from panel 2 was conducted. Mean differences andthe corresponding 90% CIs were exponentiated to obtain geometricmean ratios (coadministration/administration alone) and 90%CIs. In order to assess the effect on the rifampin T_max, theHodges-Lehmann estimate of median difference and the corresponding90% CIs were calculated.

RESULTS

Top

Results

Effect of nelfinavir on caspofungin pharmacokinetics. The effect of coadministration of nelfinavir on the pharmacokineticsof caspofungin was evaluated in study A. A modest differencein the concentration-time profiles of caspofungin was seen onday 1, with a shallower decline in caspofungin plasma concentrationsevident between 4 h and 24 h after dosing in the presence ofnelfinavir (Fig. 2). On day 14, little difference was observedin the plasma concentration profiles from the two panels. Alterationsin caspofungin pharmacokinetics with nelfinavir were at mostmodest (Table 1). There were statistically significant elevationsin AUC (P = 0.022) and C_24h (P = 0.001) of 16 and 58%, respectively,on day 1. After approximately day 2, the trough concentrationsclimbed more slowly in the nelfinavir combination panel thanin the control panel, such that, on day 14, there were no statisticallysignificant differences in any of the pharmacokinetics parametersin subjects receiving caspofungin alone and in combination withnelfinavir.

View larger version (19K):
[in this window]
[in a new window]
FIG. 2. Caspofungin plasma profiles following administration of caspofungin alone (50 mg administered IV once daily) or coadministration of caspofungin (as described above) and nelfinavir (1,250 mg administered orally twice daily) to healthy male subjects. Means and standard deviations (error bars) are indicated. Conc., concentration.

View this table:
[in this window]
[in a new window]
TABLE 1. Effect of nelfinavir on the pharmacokinetics of caspofungin (study A)^a

Effect of rifampin on caspofungin pharmacokinetics. The effect of coadministration of rifampin on the pharmacokineticsof caspofungin was evaluated in both studies A and B. In bothstudies, parallel panels received 14 days of caspofungin aloneor of caspofungin coadministered with rifampin. The primarydifference between the two study designs was in the timing ofthe initiation of the rifampin therapy. In study A, rifampinand caspofungin were initiated on the same study day. In studyB, rifampin was administered alone for 14 days prior to coadministrationwith caspofungin.

Mean plasma concentration-time profiles on days 1 and 14 ofcaspofungin therapy in subjects receiving caspofungin aloneand in combination with rifampin are shown in Fig. 3 (studyA) and 4 (study B). When caspofungin and rifampin were initiatedon the same day (study A), there were statistically significant(P < 0.001) elevations in AUC_0-24 and C_24h of 61 and 170%,respectively, on day 1, but not day 14, in the panel receivingcaspofungin with rifampin compared to the panel receiving caspofunginalone (Table 2). The caspofungin trough concentration data provideinsight into the time course of this transient elevation inplasma caspofungin concentrations with rifampin. Caspofungintrough concentrations in the rifampin combination panel of studyA reached an arithmetic mean peak of 2.49 µg/ml on day2 (Fig. 5). After day 2, the trough concentrations fell throughoutthe remainder of the study.

View larger version (19K):
[in this window]
[in a new window]
FIG. 3. Caspofungin plasma profiles following administration of caspofungin alone (50 mg administered IV once daily) or coadministration of caspofungin (as described above) and rifampin (600 mg administered orally once daily) when both drugs were initiated on the same study day (study A). Means and standard deviations (error bars) are indicated. Conc., concentration.

View this table:
[in this window]
[in a new window]
TABLE 2. Effect of rifampin on the pharmacokinetics of caspofungin^a

View larger version (24K):
[in this window]
[in a new window]
FIG. 5. Caspofungin trough concentrations following administration of caspofungin alone (50 mg administered IV once daily) or coadministration of caspofungin (as described above) and rifampin (600 mg administered orally once daily). Means and standard deviations (error bars) are indicated. Conc., concentration.

The mean plasma profile of caspofungin obtained on day 14 inthe rifampin combination panel was somewhat different in shapefrom the one obtained for the control panel (Fig. 3). Therewas a noticeable shoulder around 6 to 9 h after dosing in theprofile obtained with rifampin, indicating that caspofunginclearance is reduced by rifampin early in the dosing interval.Plasma concentrations in the interval 12 to 24 h after dosingappeared to fall faster in the combination panel than in controls,indicating that caspofungin clearance was increased by rifampinlater in the dosing interval. Consistent with the alterationin the profile shape observed, there is a trend towards a slightreduction in trough concentration on day 14 with rifampin instudy A. On day 1, the shoulder from 6 to 9 h after dosing wasalso prominent in the group receiving rifampin, but no subsequentaccelerated decline was observed during the 12- to 24-h postdoseinterval. This pattern suggests that only reduced caspofunginclearance is present on the first day of dosing with rifampin.

It was unclear from the study A data whether further reductionsin caspofungin concentrations would be observed following morethan 14 days of rifampin treatment. In study B, rifampin alonewas administered for 14 days prior to coadministration in orderto ensure that the induction effect reached steady state duringthe coadministration period.

When a 14-day pretreatment with rifampin alone was given priorto coadministration (study B), there were no statistically significantalterations in the caspofungin AUC_0-24 or C_1h on days 1 or 14with coadministration of rifampin. The mean caspofungin plasmaprofiles obtained on days 1 and 14 in the rifampin combinationpanel of study B had an alteration similar to that observedon day 14 in study A (Fig. 3 and 4), including a noticeableshoulder around the 6- to 9-h postdose interval and an accelerateddecline in the 12- to 24-h postdose interval. Consistent withthe alteration in the profile shape observed, statisticallysignificant reductions of 29% (P = 0.007) and 31% (P = 0.006)in caspofungin trough concentration with coadministration ofrifampin were observed on days 1 and 14, respectively, in studyB. The time course of trough concentrations indicates that thereduced trough concentrations with rifampin are maintained throughoutthe 14 days of coadministration, when the pretreatment withrifampin alone was given prior to coadministration (Fig. 5).

View larger version (19K):
[in this window]
[in a new window]
FIG. 4. Caspofungin plasma profiles following administration of caspofungin alone (50 mg administered IV once daily) or coadministration of caspofingin (as described above) and rifampin (600 mg administered orally once daily) when rifampin was initiated alone 14 days prior to coadministration with caspofungin (study B). Means and standard deviations (error bars) are indicated. Conc., concentration.

Effect of caspofungin on rifampin pharmacokinetics. The effect of coadministration of caspofungin on the pharmacokineticsof rifampin was investigated in study B. Mean plasma concentration-timeprofiles for rifampin from panel 2 subjects receiving rifampinalone (day 14 of rifampin alone) and in combination with caspofungin(days 1 and 14 of coadministration) are shown in Fig. 6. Table3 summarizes the effect of caspofungin on the pharmacokineticsof rifampin. There were no statistically significant differencesin the rifampin AUC_0-24 or C_max on day 1 or 14 of coadministrationrelative to day 14 of administration of rifampin alone. TheT_max was similar on the 3 days that the profiles were evaluated.

View larger version (25K):
[in this window]
[in a new window]
FIG. 6. Rifampin plasma profiles following administration of rifampin alone (600 mg administered orally once daily) or coadministration of rifampin (as described above) and caspofungin (50 mg administered IV once daily) (n = 14). Data are means. Conc., concentration.

View this table:
[in this window]
[in a new window]
TABLE 3. Effect of caspofungin on the pharmacokinetics of rifampin (study B)^a

Modeling analysis of pharmacokinetic interaction between caspofungin and rifampin. A pharmacokinetic modeling analysis was used to explore whethera combination of inhibitory and induction effects could accountfor the changes in the shape of the caspofungin plasma profilethat were observed with coadministration of rifampin. Figure7 illustrates the model fit to caspofungin plasma concentrationdata from day 1 of study A and day 14 of study B. The day 14data were well represented by the model incorporating both inhibitoryand induction effects, including accounting for the shoulderin the combination profile at 6 to 9 h after dosing and theaccelerated decline in the 12- to 24-h postdose interval. Similarly,the day 1 data from study A were well represented by the modelwith the induction factor (IND) set to 1.0 (no induction) forboth combination and monotherapy, including accounting for theshoulder in the combination profile at 6 to 9 h after dosingand the continued elevation in the combination profile through24 h after dosing. Table 4 provides the parameter estimatesobtained from fitting the model to the data sets. The parameterestimates are reasonably consistent across both sets of dataevaluated.

View larger version (19K):
[in this window]
[in a new window]
FIG. 7. Mean model fit to caspofungin concentration data from day 1 of coadministration in study A (caspofungin and rifampin initiated on same day) and from day 14 of coadministration in study B (pretreatment with rifampin alone). Conc., concentration.

View this table:
[in this window]
[in a new window]
TABLE 4. Model parameter estimates from fit of three-compartment, linear models with caspofungin uptake into tissue compartment induced and inhibited by rifampin

DISCUSSION

Top

Discussion

The results of these studies indicate that rifampin has bothan inhibitory and an induction effect on caspofungin disposition,with a net effect of slight induction at steady state. In studyA, coadministration of rifampin with caspofungin resulted ina transient increase in caspofungin concentrations on day 1.The results are consistent with net inhibition of caspofungindisposition in the initial days of concomitant administrationwhen rifampin therapy is initiated at the same time as caspofungintherapy. The effect's being evident on day 1 during the ßdistribution phase of caspofungin suggests that rifampin inhibitsthe uptake of caspofungin into tissues. Previous results froma single-dose disposition study of [³H]caspofungin (13) indicatethat little metabolism or excretion occur during the first 24h after dosing and that uptake of drug into tissues is the mechanismcontrolling the decline of drug concentration observed duringthe dominant ß phase evident from approximately 6to 48 h after dosing.

By day 14 in study A and on both days 1 and 14 in study B, thecaspofungin AUC and end-of-infusion concentrations in the combinationpanel were similar to those obtained with controls. However,an alteration in the shape of the caspofungin plasma concentrationprofile with rifampin coadministration resulted in a 14 to 31%reduction in caspofungin trough concentrations relative to controlson those study days. The declining trough concentrations observedwith continued concomitant dosing with rifampin in study A areconsistent with induction of a caspofungin disposition process.The results from study B, including the reduced caspofungintrough concentrations throughout the coadministration period,are also consistent with induction of caspofungin dispositionby rifampin, since the pretreatment with rifampin alone allowedsubstantial induction of this disposition process to occur priorto the start of coadministration.

The caspofungin tissue uptake mechanism controlling the declinein plasma concentrations in the ß phase, presumablythe mechanism inhibited by rifampin, is unknown. Rifampin hasrecently been shown to inhibit the rat uptake transporters Oatp1and Oatp2 (5, 14), but it is not known if either of these transportersis involved in the disposition of caspofungin. The same uptakemechanism inhibited by rifampin in the initial days of therapymay also be induced by rifampin with continued dosing. It isnot unreasonable to postulate that rifampin could induce uptaketransporters, since rifampin induces a number of metabolic pathwaysand has recently been shown to induce the efflux transporterP-glycoprotein (6, 7). Induction of caspofungin metabolism isunlikely to be the mechanism generating these pharmacokineticalterations with longer-term rifampin administration. Uptakeof caspofungin into tissues is the rate-controlling step forplasma clearance of caspofungin and the initial biotransformationprocess appears to be chemical decomposition (13). Therefore,induction of metabolic pathways would be expected to have little,if any, effect on caspofungin plasma pharmacokinetics.

The modeling analysis suggests that a combination of inductionand inhibition effects of rifampin on caspofungin tissue uptakecan account for the caspofungin plasma profile alterations observed.The parameter estimates obtained suggest that induction by rifampinroughly doubles the rate of caspofungin uptake into tissues,although the presence of ongoing inhibition by rifampin resultsin more-modest alterations in the plasma pharmacokinetics ofcaspofungin. In general, the model represented the concentrationdata well. On both day 1 and day 14, the model appeared to underpredictthe degree of inhibition at 2 to 4 h after dosing. A possibleexplanation for this is that rifampin plasma concentrationsmay underrepresent hepatic concentrations during the rifampinabsorption phase and, thus, the model may not fully accountfor the degree of inhibition of caspofungin uptake into hepatocytesduring this period.

Caspofungin trough concentrations were reduced by 14 to 31%following multiple doses of rifampin. The clinical significanceof this alteration is unclear. In a dose-ranging study conductedin patients with esophageal and oropharyngeal candidiasis, patientstreated with 35 mg of caspofungin had a numerically lower favorableresponse rate than patients treated with doses of 50 or 70 mgof caspofungin, although this difference was not statisticallysignificant (1). The caspofungin dose of 35 mg produced troughconcentrations that were $~$ 70% of the value obtained with 50 mgin the efficacy study described above (J. Stone et al., 43rdICAAC). Although the critical caspofungin pharmacokinetic parameter(AUC, peak, or trough) for efficacy is unknown, the responsedata at 35 mg indicate a possibility that a trough reductionof 30% or more may be associated with reduced efficacy. In patientswith proven or potentially life-threatening fungal infections,caution indicates that dosing decisions should err on the sideof ensuring that effective drug concentrations are achieved.The 90% CIs for the geometric mean ratios (coadministration/administrationalone) of caspofungin trough concentrations fell below 0.7 onday 14 of study A and on days 1 and 14 of study B, indicatingthat the reduction in caspofungin trough concentrations withcoadministered rifampin may be clinically meaningful. Therefore,an increase in the daily maintenance dose of caspofungin from50 to 70 mg should be considered when caspofungin and rifampinare coadministered. This dose increase should generate troughconcentrations similar to those obtained when 50-mg doses ofcaspofungin are given without concomitant rifampin. While thisdose increase would also increase the AUC_0-24 and C_1h, theseparameter values should be similar to or only slightly in excessof those obtained with 70-mg doses of caspofungin given dailywithout rifampin. The 70-mg/day regimen has been generally welltolerated in clinical studies (10, 12).

Administration of rifampin with caspofungin produces a transientelevation in caspofungin plasma concentrations on day 1 of coadministrationwhen both drugs are initiated together, but not when caspofunginis added after administration of rifampin for the preceding14 days. A dose reduction is not considered to be necessaryfor the transient elevation in caspofungin plasma concentrationswhen rifampin and caspofungin are initiated on the same studyday. The geometric mean caspofungin AUC_0-24 of 90.44 µg· h/ml obtained in the coadministration panel on day1 in study A is similar to the AUC_0-24 values that were obtainedwith acceptable tolerability following a single 100-mg dose(113.92 µg · h/ml) and following multiple dosesof 70 mg (129.61 to 144.27 µg · h/ml) (12) andis much less than values obtained with acceptable tolerabilityfollowing single doses of 150 and 210 mg (279.66 and 374.92µg · h/ml, respectively) and multiple doses of100 mg (227.36 µg · h/ml) (J. Stone, E. Migoya,S. Li, P. Deutsch, G. Winchell, K. Ghosh, A. Miller, S. Bi,A. Bass, G. Mistry, and R. Dawkins, Abstr. 42nd Intersci. Conf.Antimicrob. Agents Chemother., abstract A-1389, 2002). Therefore,it is unlikely that this transient elevation would be clinicallymeaningful. In addition, these transient elevations will occuronly under conditions where rifampin is initiated at the sametime or after caspofungin is initiated. Under typical clinicalusage, it is more likely that caspofungin will be added to preexistingrifampin therapy, and under these conditions, no transient elevationwill occur, as verified in study B.

The effect of caspofungin on the pharmacokinetics of rifampinwas also investigated in study B. A comparison of rifampin pharmacokineticson the final day of pretreatment with rifampin alone and ondays 1 and 14 of coadministration indicates that caspofunginhas no effect on the pharmacokinetics of rifampin.

Finally, the results from study A also provide information onconcomitant use of nelfinavir. Caspofungin pharmacokineticson day 14 were unaltered by coadministration of nelfinavir.Although a slight elevation in caspofungin concentrations wasnoted on day 1 of coadministration, this alteration is unlikelyto be clinically meaningful. Given that this study providesa more definitive evaluation of the potential for drug interactionswith nelfinavir than could be obtained with population pharmacokineticdata, the lack of clinically meaningful alterations in caspofunginpharmacokinetics indicates that no dose adjustment is necessarywhen caspofungin and nelfinavir are administered together. Italso suggests that the association of a reduced AUC and troughconcentration with concomitant use of nelfinavir in the prioranalysis of the population pharmacokinetics data was a spuriousfinding.

In conclusion, nelfinavir has no clinically significant effecton the pharmacokinetics of caspofungin. No dose adjustment isnecessary when caspofungin is coadministered with nelfinavir.Caspofungin has no effect on the pharmacokinetics of rifampin.Rifampin both inhibits and induces caspofungin disposition,resulting in reduced trough concentrations at steady state.An increase in the caspofungin dose to 70 mg daily should beconsidered when the drug is coadministered with rifampin.

FOOTNOTES

* Correspondence author. Mailing address: WP75-100, Merck Research Laboratories, West Point, PA 19486. Phone: (215) 652-5705. Fax: (215) 993-3533. E-mail: julie_stone{at}merck.com.

REFERENCES

Top