|Home | About | Journals | Submit | Contact Us | Français|
5-Fluorocytosine (5FC) is an oral antifungal that is currently used in combination with amphotericin B to treat Cryptococcus neoformans meningoencephalitis. The oral dosing of 5FC could be optimized by the use of a controlled-release (CR) formulation. The objective of the current study was to develop two prototype 5FC-CR formulations and evaluate the single-dose (1,500-mg) serum pharmacokinetic profiles of those formulations relative to the profile of the commercially available, immediate-release 5FC product (Ancobon) by the use of a phase 1, open-label, randomized, three-phase, crossover pharmacokinetic study design. Hydroxypropyl methylcellulose was utilized as the rate-controlling matrix to compound the 5FC-CR tablets. The two prototype 5FC-CR formulations demonstrated 80% release at 13.0 and 18.4 h, respectively, whereas the immediate-release product demonstrated 80% release at 0.28 h, as determined in vitro by the United States Pharmacopeia apparatus 2 dissolution method. Five subjects completed all three phases of the study without any adverse events. The mean maximum concentration, the area under the curve from time zero to 24 h, and the area under the curve from time zero to infinity were approximately 50% lower (P < 0.01) with the 5FC-CR formulations than with the immediate-release 5FC product. However, no statistically significant differences in the minimum concentrations at 24 h were noted between the formulations. The gastric absorption profile of 5FC-CR was well predicted by in vitro dissolution. Future exploration of a gastroretentive 5FC-CR formulation could overcome the marked lack of bioequivalence observed in the present study.
Cryptococcus neoformans is an opportunistic fungal pathogen that is associated with significant morbidity and mortality. A 27% mortality rate has been associated with Cryptococcus neoformans meningoencephalitis (CNME) among AIDS patients in South Africa (15). An induction antifungal regimen of amphotericin B (AMB) and 5-fluorocytosine (5FC) for 2 weeks, followed by long-term fluconazole (FLZ) maintenance therapy, is considered the standard of care for AIDS patients with CNME (20). While it is effective, the use of induction therapy has been too complex for implementation in developing countries (11). The use of FLZ monotherapy in AIDS patients with CNME has been utilized but is associated with treatment failure and the emergence of resistance (3, 18). Hence, access to a simple, easy-to-administer combination oral antifungal regimen could improve CNME-related outcomes in countries without sufficient resources. The use of oral 5FC and FLZ combination regimens have demonstrated promising clinical results (4, 8, 10, 14, 16).
However, currently recommended 5FC dosing regimens may not be optimal on the basis of pharmacokinetic (PK) and pharmacodynamic (PD) principles (7). For example, administration of 5FC at 25 mg/kg of body weight by mouth every 6 h yields steady-state maximum plasma concentrations (Cmaxs) of 60 to 80 μg/ml and minimum plasma concentrations (Cmins) of 5 to 20 μg/ml in patients with normal renal function (26). The PK-PD index predictive of the effects of 5FC against C. neoformans is most likely the time that the concentration remains above the MIC (T > MIC) (1, 13). The MIC90 of C. neoformans var. grubii is approximately 8 μg/ml on the basis of the findings of a survey of nearly 2,000 isolates collected worldwide (19). Given that 5FC has a low level (4%) of protein binding and concentrations in cerebrospinal fluid (CSF) that are 80% of the values in serum, a serum Cmin above 10 μg/ml should theoretically achieve a 100% T > MIC in CSF (26). An oral controlled-release (CR) 5FC (5FC-CR) formulation could be used to maintain a steady serum concentration above 10 μg/ml. Moreover, the use of lower 5FC doses could reduce the nausea, vomiting, and bone marrow suppression that have been associated with high plasma 5FC concentrations (26). The combined use of 5FC-CR and FLZ could also simplify oral administration (once daily) and improve outcomes among AIDS patients with CNME in sub-Saharan Africa.
This report outlines the development of two model CR formulations of 5FC. The PK profile of a single dose of immediate-release (IR) 5FC (Ancobon) capsules was compared to the PK profiles of two 5FC-CR formulations in a cohort of five subjects by the use of a three-phase crossover pharmacokinetic study design. The objectives of this pilot study were to (i) determine the feasibility of creating a 5FC-CR formulation, given its relatively low molecular weight and hydrophilic chemical structure and the high dose that is required and (ii) compare the serum PK profile of a single dose of 5FC-CR to that of the commercially available IR product in healthy volunteers.
(This work was presented in part at the 2007 American Association of Pharmaceutical Scientists Annual Meeting, San Diego, CA, 11 to 15 November 2007, and at the 2008 Controlled Release Society Annual Meeting, New York City, NY, 12 to 16 July 2008.)
The preparation of the 5FC-CR formulations included the use of 5FC (MP Biomedicals, Solon, OH) and the controlled-release matrix hydroxypropyl methylcellulose (HPMC; Methocel K4M and K100M; Dow, Midland, MI). Additional excipients included microcrystalline cellulose (Avicel PH-302; FMC Biopolymer, Philadelphia, PA), colloidal silicon dioxide (Cab-O-sil M-5P; Cabot, Tuscola, IL), and magnesium stearate (Spectrum, Gardena, CA). The 5FC and the excipients (excluding magnesium stearate) were mixed in a V-Shell blender for 5 min, sieved through a 40-mesh screen, and mixed again for 5 min. The blend was again sieved through a 40-mesh screen. The magnesium stearate was also sieved through a 40-mesh screen, added to the V-Shell blender, and mixed with the other materials for 3 min. The powder was compressed into oblong tablets by using a Stokes model 511-7 single-punch-station press to produce tablets containing 250 mg of 5FC. The tablet compression force, the depth of fill of the die, and the type of rate-controlling matrix material were varied in order to produce two 5FC-CR formulations. In addition, a small amount of Avicel PH-302 was added in order to improve the compressibility of the powder. The target release rates were 20 mg/h and 40 mg/h with a sustained release rate of at least 12 h during in vitro dissolution testing. These rates were targeted to achieve a steady serum concentration above 10 μg/ml and to theoretically achieve a 100% T > MIC in CSF.
The rates of 5FC release from the CR tablets (n = 6) and the commercially available IR capsules (Ancobon; n = 3) were determined by the United States Pharmacopeia (USP) apparatus 2 method (the paddle method, 100 rpm). The tablets and capsules were tested in deionized water (900 ml) heated to 37°C. Samples (3 ml each) were withdrawn from the apparatus at the baseline and at specific times up to 24 h after the introduction of the 5FC-CR dosage forms into the dissolution vessels. Samples (3 ml each) were withdrawn from the apparatus at the baseline and at specific times up to 4 h for the IR capsules, given the expected rapid dissolution of this formulation. The samples were assayed for 5FC by using a Shimadzu UV-2100 UV-visible spectrophotometer at an absorbance wavelength of 276 nm. These procedures were utilized to evaluate the CR formulations, and two of these formulations (5FC-CR-A and 5FC-CR-B) were selected for use in the subsequent clinical study. The in vitro drug-release profiles of the two CR formulations were compared by using the similarity factor (f2), a factor commonly used to evaluate dissolution profiles (21). This method includes the calculation of f2 by use of the following equation:
where Rt and Tt are the cumulative percentage dissolved at each of the selected n time points (t) of the two formulations, respectively. An f2 value of 50 to 100 represents similarity between formulations, whereas an f2 value of <50 indicates a difference.
This study was approved by the Human Research Review Committee (HRRC) at the University of New Mexico Health Sciences Center (UNMHSC). The study was conducted with a randomized, three-phase, open-label, single-dose, pharmacokinetics crossover design and with a washout period between phases. Twelve healthy volunteers who met the appropriate criteria were to be enrolled after the provision of informed consent. Male and female subjects were eligible to participate on the basis of the following criteria: (i) the subjects had to be between 18 and 40 years of age, (ii) they had to have a normal body mass index of 18 to 25 kg/m2, (iii) they had to be nonsmoking volunteers, and (iv) female subjects of childbearing potential had to be either surgically sterilized or using an effective method of contraception during the entire study period and 1 week following the study period. Subjects were not eligible to participate if they exhibited any of the following criteria: (i) a clinically significant chronic medical condition requiring the use of pharmacological or nonpharmacological management; (ii) a history of allergy or adverse reactions to 5FC; (iii) blood donation or participation in another study within 60 days of this study; (iv) pregnancy or breastfeeding for females; (v) a lack of willingness to abstain from alcohol, cigarette smoking, and caffeine intake during the study period (21 days); (vi) a clinically significant abnormal physical examination, defined as a physical finding requiring further workup or management with a pharmacological or nonpharmacological intervention; (vii) a significantly abnormal laboratory result (complete blood count, serum chemistry, transaminases, and total bilirubin); and (viii) a calculated creatinine clearance of <50 ml/min.
Qualifying subjects were randomized (permuted-block method) to the treatment sequence, namely, treatment with the commercially available product, 5FC-CR-A, and 5FC-CR-B, over three phases, and there was a minimum of a 5-day washout period between phases. The subjects were admitted to the University of New Mexico General Clinical Research Unit (GCRC) on the evening prior to dosing and were required to have no food intake for 10 h prior to dosing. A single dose of 1,500 mg 5FC (6 × 250 mg) was administered with water, and the subjects remained fasted for an additional 4 h postdosing. Blood (7 ml) was collected through peripheral intravenous access and placed in serum separator tubes predosing (0 h) and at 0.5, 1.0, 1.5, 2.0, 2.5, 3.0, 4.0, 6.0, 8.0, 12.0, and 24 h postdosing. Serum was harvested after centrifugation of the blood samples at 1,500 × g and 4°C for 10 min and was stored at −70°C until analysis. Similar procedures were used during each of the three study phases, and there was a minimum 5-day washout period between phases. All potential adverse events and protocol deviations were recorded and reported to the HRRC and GCRC. In addition, the data safety monitoring plan (DSMP) included provisions for analysis of the data after enrollment of 50% of the target sample size (n = 6).
Analysis of stored serum samples to determine 5FC and 5-fluorouracil (5FU) concentrations was performed with the samples in a single batch at the midpoint of the PK study. A previously validated high-performance liquid chromatography (HPLC) technique was used to measure the 5FC concentration (24). The assay required the purchase of 5-methylcytosine (5MC; Sigma-Aldrich, St. Louis, MO), 5-chlorouracil (5CU; Sigma-Aldrich), trichloroacetic acid (Mallinckrodt Chemicals, Hazelwood, MO), potassium hydroxide (EMD, Gibbstown, NJ), and potassium dihydrogen phosphate (EMD). Serum samples were spiked with 5-methylcytosine and 5-chlorouracil to serve as the internal standards for recovery determination. Serum samples were treated with 10% trichloroacetic acid solution in disposable 10-ml glass tubes under constant vortexing. The samples were centrifuged at 2,500 rpm for 10 min, followed by injection of a 50-μl aliquot onto a 5-μm-pore-size C8 column (4.6 by 250 mm; Alltech Associate, Inc., IL) with a C8 guard column. The mobile phase was run at 1 ml/min and consisted of a 40-mmol/liter potassium dihydrogen phosphate solution adjusted to pH 7.0 with 10% (wt/vol) potassium hydroxide. The retention times of 5FC, 5FU, 5MC, and 5CU were approximately 5, 6, 8, and 11 min, respectively, and detection was performed at a wavelength of 266 nm. The lower limit of quantification of the assay was 0.8 mg/liter for 5FC.
5FC concentration-versus-time data were analyzed by the use of WinNonLin pharmacokinetic software (Pharsight, Mountain View, CA) by noncompartmental methods. The values of the following pharmacokinetic parameters were calculated: Cmax; Cmin; the time to Cmax (Tmax); the mean residence time (MRT); the terminal elimination half-life (t1/2); the apparent volume of distribution during the terminal phase (Vz/F, where F represents bioavailability); the apparent clearance (CL/F); the area under the curve from time zero to 24 h (AUC0-24), including weight-normalized values (WNAUC0-24); and the area under the curve from time zero to infinity (AUC0-∞), including weight-normalized values (WNAUC0-∞). Repeated-measures analysis of variance with time as the repeated factor was the overall method of statistical analysis used for comparison of the values of the pharmacokinetic parameters. Post-hoc comparisons of these parameter values was performed by Student's paired t test according to Fisher's least-significant-difference method. The geometric mean ratio and 90% confidence interval (CI) of the estimated values of the PK parameter (AUC0-24, AUC0-∞, Cmax) for 5FC-CR-A and 5FC-CR-B relative to those for the commercial IR product were calculated to evaluate bioequivalence. All statistical analyses were performed with the Stata program (version 10; Stata Corp., College Station, TX).
Twelve 5FC-CR formulations were compounded and evaluated for their drug-release rates by dissolution testing. The compositions of the two candidate 5FC-CR formulations selected for in vivo testing were 5FC (43.4%), Methocel (36.2%), Avicel PH-302 (19.4%), Cab-O-sil M-5P (0.5%), and magnesium stearate (0.5%). The only difference between the two CR formulations (5FC-CR-A and 5FC-CR-B) was the composition of the Methocel used, namely, K4M (5FC-CR-A) and K100M (5FC-CR-B). The higher-molecular-weight K100M forms a gelatinous barrier thicker than that formed by K4M, and this thicker barrier results in slower drug release. The dissolution profiles of the IR and the two CR formulations are illustrated in Fig. Fig.1.1. The relationship between the percentage of 5FC released over time (t; in hours) was best fit by the function a·(1 − e−b·t), where a represents a constant value and b is the release rate constant. The relationships for the amount of 5FC released (mg) for the commercial IR product, 5FC-CR-A, and 5FC-CR-B were 89.9·(1 − e−7.79·t) (R2 = 0.996), 92.6·(1 − e−0.154·t) (R2 = 0.997), and 92.0·(1 − e−0.111·t) (R2 = 0.997), respectively. The times required to achieve the release of 80% of the 5FC were 0.28, 13.0, and 18.4 h for the commercial IR product, CR-A, and CR-B, respectively. An approximately 10-fold reduction in the rate of 5FC release was achieved through the use of a CR matrix tablet formulation. An f2 value of 31.2 was calculated when the drug-release profiles were compared, thus indicating a difference in the rates of release for the two CR formulations.
Six subjects were enrolled into the study prior to the initial data analysis, with data from all three phases being available for five subjects. The sixth subject withdrew for reasons unrelated to the study medication or procedures. On the basis of the data safety monitoring plan, the study was terminated early due to the marked lack of bioequivalence. The five subjects (three males, two females) included in this analysis had a mean ± standard deviation age, total body weight, and serum creatinine level of 26.2 ± 3.06 years, 85.6 ± 18.1 kg, and 0.91 ± 0.12 mg/dl, respectively. The race/ethnicity of the subjects were Caucasian (n = 3), African American (n = 1), and Hispanic (n = 1). No adverse events occurred during the study.
The mean ± standard error of the mean (SEM) 5FC concentration-versus-time profiles for the IR, CR-A, and CR-B formulations are illustrated in Fig. Fig.2.2. Serum 5FU concentrations were below the limit of detection. The values of the PK parameters for 5FC estimated for these three formulations are included in Table Table1.1. The mean Cmax of 5FC was approximately twofold lower (P < 0.001) with the 5FC-CR formulations than with the IR commercial product. Given that a fixed 5FC dose of 1,500 mg was administered, the median dose by total body weight was 16.0 mg/kg (minimum, 13.8 mg/kg; maximum, 25.4 mg/kg). The mean Tmax values were approximately twofold higher for both 5FC-CR formulations than for the IR product, although the difference was not statistically significant. However, the AUC0-24 and AUC0-∞ values were approximately 50% lower (P < 0.001) for both 5FC-CR formulations than for the IR commercial product. The geometric mean ratios of the values of AUC0-24, AUC0-∞, and Cmax for 5FC-CR-A and the IR commercial product were 0.465 (90% CI, 0.326, 0.604), 0.499 (90% CI, 0.379, 0.620), and 0.465 (90% CI, 0.346, 0.584), respectively. The geometric mean ratios of the values of AUC0-24, AUC0-∞, and Cmax for 5FC-CR-B and the IR commercial product were 0.376 (90% CI, 0.279, 0.473), 0.540 (90% CI, 0.295, 0.784), and 0.400 (90% CI, 0.233, 0.567), respectively. Although the difference was not statistically significant, the higher CL/F and Vz/F values reflect the lower bioavailability of the 5FC from the CR formulations relative to that of the 5FC from the IR product. Despite this marked lack of bioequivalence, the MRTs and Cmins were not significantly different between the three formulations.
Guidelines from the Infectious Diseases Society of America (IDSA) recommend the use of a combination of intravenous AMB (0.7 to 1.0 mg/kg/day) and oral 5FC (100 mg/kg/day) for 2 weeks followed by oral FLZ at 400 mg/day for a minimum of 8 weeks for the treatment of CNME (20, 25). Several studies have verified the importance of 5FC in combination antifungal therapy for this disease (10, 14, 16, 22). Unfortunately, the IDSA-recommended treatment approach has been too complex for application in sub-Saharan Africa (13). An observational study performed in 1998 and 1999 in Zambia highlighted this problem during the management of 230 AIDS patients with CNME (18). AMB and 5FC were not available, so FLZ was administered as a 400-mg loading dose and orally at 200 mg once daily. Due to limited supplies of FLZ, 56% of the patients received FLZ therapy, while 44% of the patients received palliative care only. The case-fatality rate was 100% by 7 weeks and 6 months in the untreated and FLZ-treated groups, respectively.
Documentation of the poor clinical outcomes and advocacy led to free access to FLZ through a donation program led by Pfizer, Inc. Free access to FLZ has now expanded to nearly 80 countries, where the prevalence of HIV infection exceeds 1% (28). The impact has led to the use of FLZ monotherapy for CNME in sub-Saharan Africa. Kisenge et al. evaluated 149 consecutive HIV-infected patients presenting with meningitis at a tertiary referral hospital in Moshi, Tanzania (9). Forty patients were positive for CSF cryptococcal antigen. These patients were treated with FLZ at 800 mg and were initiated or continued on antiretroviral therapy. The rate of survival with this therapeutic approach was 50% within the first 2 weeks of admission (P. R. Kisenge, personal communication). The use of FLZ monotherapy to treat CNME has also been associated with the emergence of FLZ resistance (3). These data strongly advocate the need for alternative oral pharmacologic approaches to the treatment of CNME in sub-Saharan Africa. The use of a high dose of FLZ (greater than 800 mg) for the management of CNME has been documented to be beneficial (12). Milefchik and colleagues have documented the benefits of the use of a combination of a high dose of FLZ (up to 2,000 mg) and 5FC (16). Furthermore, the use of oral 5FC at 100 mg/kg/day in combination with oral FLZ at 1,200 mg/day was associated with a significantly lower rate of mortality at 2 weeks compared to that which occurred with the use of oral FLZ monotherapy at 1,200 mg/day (10% and 37%, respectively) (8).
Hence, the current study was undertaken to develop a practical solution to the current problem of CNME management in countries with insufficient resources. The ability to give 5FC once or twice daily could be translated into a simple combined regimen with FLZ. Therefore, CR formulations of 5FC were developed by using HPMC, a hydrophilic polymeric matrix material commonly used in the pharmaceutical industry for such release-controlling applications (11, 17). When it is exposed to aqueous medium, HPMC forms a highly viscous gelatinous barrier that controls drug release (6). Although the current study failed to demonstrate bioequivalence, many important observations that could aid with the continued exploration of this approach were identified. Our initial data suggest that the rate of release of 5FC can be controlled by use of the HPMC matrix, despite the high water solubility and small molecular structure of 5FC. Limited drug absorption was demonstrated after the presumed gastric emptying time, i.e., 2.5 h and beyond. Two potential explanations for the reduced absorption include a lower rate of drug release at alkaline pH values and/or the increased gut microbial conversion of 5FC to its metabolite, 5-fluorouracil (27). The HPLC method used in the current study demonstrated plasma 5FU concentrations below detectable levels. However, 5FU is known to be poorly absorbed from the gastrointestinal tract (27). Hence, confirmation of gut microbial conversion may require the collection of feces for total drug recovery estimation in future studies.
Despite the lack of success of our initial approach, alternative strategies that may be used to overcome the present limitation exist. The development of a gastroretentive CR formulation of 5FC appears to the best next approach. The options for achieving gastroretention include the use of floating drug delivery systems, bioadhesive materials, or systems that expand in size when they are hydrated (23). The successful formulations of ciprofloxacin and acyclovir obtained by the use of gastroretentive technologies serve as relevant examples (2, 5). The continued development and study of a once- or twice-daily 5FC-CR formulation could provide a practical solution to the management of CNME when that formulation is combined with fluconazole in countries that cannot implement the currently recommended treatment paradigm.
This study was supported by the University of New Mexico Clinical Translational Science Center, funded in part through the National Institutes of Health (grant M01-0997).
M. P. Pai acknowledges C. Richard Lyons for his thoughtful mentorship and the Pharmaceutical Society of Zambia for inspiration.
Published ahead of print on 28 December 2009.