Youjung Cho 1, Elin M. Svensson 1,2, Isabela Ribeiro 3, Thomas P.C. Dorlo 1
1 Department of Pharmacy, Uppsala University (Uppsala, Sweden), 2 Department of Pharmacy, Pharmacology and Toxicology, Radboud University Medical Center (Nijmegen, The Netherlands), 3 Drugs for Neglected Diseases initiative (Geneva, Switzerland)
Objectives:
Cryptococcal meningitis (CM) is a fungal infection of the central nervous system characterized by inflammation of the meninges. It is one of the leading causes of mortality among the HIV/AIDS or immunocompromised population, especially in sub-Saharan Africa.1 Immediate-release (IR) flucytosine (5FC), a key drug in CM combination therapy, is effective but limited by a short half-life, requiring four daily doses. Three new sustained-release (SR) 5FC pellet formulation prototypes (B-D) were developed to reduce dosing frequency and improve treatment accessibility. Currently, the pharmacokinetics (PK) of 5FC, variability in absorption, and food interaction are not well understood, which are essential for optimizing the treatment regimen.
The study aimed to compare the SR, absorption characteristics, and drug exposures of the three SR 5FC formulations by describing PK using phase 1 prandial clinical data through nonlinear mixed-effects modeling.
Methods:
Study design:
PK data from two phase 1 studies with fasted (n = 42) or fed (n = 36) were utilized. Both studies had rich sampling designs: 21 and 17 plasma samples, respectively, over 48 hours.2,3 The fasted study had a 4-way crossover design, and the fed study a 2-way crossover design with 7-14 day washout periods. In both studies, 1500 mg of oral IR 5FC was administered twice at 0 and 6 hours. In the fasted study, 3000 mg of oral SR 5FC prototypes B-D was administered once and in the fed study, 6000 mg of oral SR 5FC D was administered once.
Model Development:
Data handling was performed using RStudio (v. 4.4.3). Model development was conducted through NONMEM (v 7.5.1) and PsN (v. 5.3.1). Model selection was based on the objective function value (OFV) with a significant decrease defined as less than -3.84 (p < 0.05), in addition to visual predictive checks and Goodness-of-Fit plots. A 5FC PK model was initially developed using only IR 5FC data from the same Phase 1 individuals, and the resulting clearance and volume of distribution estimates were fixed during the development of SR models.4 Various SR absorption models were evaluated, such as i) zero-order (ZO) absorption, ii) dual ZO absorption, and iii) parallel ZO-first-order (FO) absorption. Food effects were evaluated on all absorption-related fixed-effect parameters. Inter-individual variability (IIV), inter-occasional variability (IOV) between formulations, and residual unexplained variability (RUV) were explored. All parameters were re-estimated after establishing structural and stochastic components of each formulation.
Results:
The PK of the SR formulations was best characterized by a two-compartment model with sequential ZO-FO, reflecting a slow dissolution, followed by a FO absorption process described by five transit compartments that split into fractions undergoing either a slow or a fast absorption, representing colonic absorption and small intestinal absorption, respectively.
SRDfed had the highest exposure with an AUC0-48 of 658.8 mg*h/L, followed by SRDfasted [261.2 mg*h/L], SRC [231.1 mg*h/L], and SRB [176.8 mg*h/L]. A relationship between the exposure and slow absorption was observed. SRB, which had the lowest exposure estimated, had the slowest absorption rate [Ka,slow,B: 0.00486 h-1], longest time to complete dissolution [TDISS,B: 3.36 h], and highest fraction slow absorption [FRACTslow,B: 73.4%]. This led to accumulation in the slow absorption compartment, reflecting limited colonic absorption and drug availability, resulting in lower systemic exposure. This was followed by SRC [Ka,slow,C: 0.0125 h-1, TDISS,C: 2.52 h, FRACTslow,C: 39.3%], then SRD [Ka,slow,D: 0.0231 h-1, TDISS,D: 2.05 h, FRACTslow,D: 25.2%]. In SRD, a food effect was identified on FRACTslow,D [-19.1%, dOFV: 19.14] and TDISS,D [+122%, dOFV: -139.872] where food consumption increased drug availability for small intestinal absorption, resulting in the highest AUC0-48 across all formulations and prandial states.
Conclusion:
The full model captured the distinct absorption profiles of the three SR 5FC formulations: SRB demonstrated the slowest absorption while SRD had the fastest. The estimated dissolution time of the various formulations ranged between 2-3 hours. A longer dissolution time was associated with a larger FRACTslow, leading to accumulation in the colon, where absorption is low, and consequently lowering systemic exposure. Food consumption prolonged TDISS and reduced FRACTslow, suggesting greater small intestinal absorption than colonic absorption under fed conditions. Data from a phase 2 trial with asymptomatic individuals will be applied to the final model to investigate potential differences in parameters in the asymptomatic population.
References:
1. Tugume, L., Ssebambulidde, K., Kasibante, J. et al. Cryptococcal meningitis. Nat Rev Dis Primers 9, 62 (2023). https://doi.org/10.1038/s41572-023-00472-z
2. Goyal, V., Krantz, E., Simon, F., Neven, A., Eriksson, J., Saayman, A., Ibnou Zekri Lassout, N., Louis, M., Robinson, S., Deshmukh, A., Antarkar, A., Ruffell, C., Victor, S., Chenel, M., Celebic, A., Caplain, H., Gillon, J. Y., & Ribeiro, I. (2024). Bioavailability of three novel oral, sustained-release pellets, relative to an immediate-release tablet containing 500 mg flucytosine: A randomized, open-label, crossover study in healthy volunteers. Clinical and translational science, 17(3), e13756. https://doi.org/10.1111/cts.13756
3. Ibnou Zekri Lassout, N., Goyal, V., Krantz, E., Simon, F., Neven, A., Eriksson, J., Saayman, A., Satam, V., Ruffell, C., Victor, S., Chenel, M., Celebic, A., Caplain, H., Gillon, J. Y., Deshmukh, A., Antarkar, A., Sjögren, E., & Ribeiro, I. (2024). Bioavailability of a novel sustained-release pellet formulation of 5-flucytosine in healthy-fed participants for use in patients with cryptococcal meningitis. Clinical and translational science, 17(9), e13908. https://doi.org/10.1111/cts.13908
4. Cho, Y. (2025). Population pharmacokinetics of the antifungal immediate-release oral flucytosine to characterize absorption and food effects. PAGE 33 (2025) Abstr 11374 [www.page-meeting.org/?abstract=11374]
Reference: PAGE 34 (2026) Abstr 12157 [www.page-meeting.org/?abstract=12157]
Poster: Drug/Disease Modelling - Absorption & PBPK