INTRODUCTION
Coccidioidomycosis is a fungal infection that can occur in humans and nonhuman mammals and is caused by the dimorphic, saprobic fungal species
Coccidioides posadasii and
C. immitis (
1). Disseminated or progressive disease, including coccidioidal meningitis, develops in 1 to 3% of infected individuals, and risk factors for severe or disseminated infection include Filipino or African-American ethnicity, HIV/AIDS, pregnancy, cardiopulmonary disease, poorly controlled diabetes, and immunosuppression (
2–4). The overall mortality in such patients with coccidioidomycosis has been reported to be high (∼60%) (
5,
6). Fluconazole remains the preferred triazole for meningitis and central nervous system (CNS) disease due to its excellent absorption following oral administration, a relatively low adverse effect profile, penetration into the central nervous system, and relative affordability as a generic medication (
7). The disadvantage of this azole is its inability to clear the fungus from the host. In addition, there is concern for reduced
in vitro susceptibility of
Coccidioides isolates to fluconazole (
8,
9).
With the goal of reducing the potential for drug-drug interactions and other human toxicities, tetrazole-based fungal CYP51 inhibitors have been developed that are significantly more specific for inhibiting the target CYP51 compared with mammalian CYP enzymes (
10,
11). VT-1598 is a structurally distinct CYP51 inhibitor with a broad spectrum of activity against a variety of yeast and molds, as well as endemic fungi, including
Coccidioides species (
10,
12). The objective of this study was to evaluate the
in vivo efficacy of VT-1598 against CNS coccidioidomycosis in separate experimental models of disease caused by both
C. posadasii and
C. immitis.
RESULTS
In vitro susceptibility. VT-1598 demonstrated in vitro activity against both the Coccidioides isolates used to establish infection in the murine models. The VT-1598 MICs were 1 and 0.5 μg/ml against C. posadasii and C. immitis isolates, respectively. In contrast, the MIC of fluconazole was 16 μg/ml against both isolates.
Coccidioides posadasii fungal burden.
In mice infected with
C. posadasii and treated with VT-1598, significant reductions in fungal burden were observed in the brain 48 h posttreatment following 7 days of therapy (
Fig. 1A). In both the 4-mg/kg and the 20-mg/kg VT-1598 groups, the brain fungal burden (means of 2.64 and 1.34 log
10 CFU/g, respectively, at lower limit of quantification) was significantly lower than that observed in the vehicle control group (5.02 log
10 CFU/g;
P < 0.0001 for both comparisons). In the VT-1598 20-mg/kg group, the fungal burden was undetectable in 4 of the 10 mice prior to correction for the lower limit of quantification. Although the brain fungal burden was also significantly reduced in mice treated with the positive control VT-1161 (3.83 log
10 CFU/g;
P < 0.0001 versus vehicle control), the fungal burden observed in this group was significantly higher than that observed with either dose of VT-1598 (
P < 0.0001 for both comparisons).
Coccidioides immitis fungal burden.
In mice infected with
C. immitis, the brain tissue fungal burden was also significantly reduced. After 7 days of therapy, mean CFU counts ranged between 1.23 to 1.39 log
10 CFU/g in the VT-1598 groups and 1.37 log
10 CFU/g in mice treated with fluconazole, and the fungal burden in each of these groups was significantly lower than that observed with the vehicle control (5.56 log
10 CFU/g;
P < 0.001 for all comparisons) (
Fig. 1B).
Coccidioides posadasii survival.
VT-1598 resulted in significant and dose-related improvements in survival in mice infected with
C. posadasii. As shown in
Fig. 2A, the median survival in each of the VT-1598 treatment groups (24, 34, and 39 days in the 3.2-, 8-, and 20-mg/kg groups, respectively) was significantly longer than in the vehicle control group (7.5 days;
P < 0.001 for all comparisons). Mice treated with VT-1598 began to succumb to infection between days 21 and 31 postinoculation, which corresponded to 6 to 16 days after therapy was stopped. The percent survival in mice treated with VT-1598 at 4 weeks posttreatment was 0% in the 3.2-mg/kg group, 11.1% in the 8-mg/kg group, and 33.3% in the 20-mg/kg group. All mice in the vehicle control group succumbed to infection by day 8. Nine of ten mice treated with the positive control VT-1161 survived to day 43, 28 days posttreatment, which was significantly longer than mice given VT-1598 (
P ≤ 0.01 for both comparisons). To compare with the
C. immitis study, survival was also evaluated at day 30 postinoculation, which was 15 days after therapy had stopped. When censored at day 30 in the
C. posadasii study, the median survival was >30 days in mice treated with VT-1598 at 8 and 20 mg/kg (
Fig. 2B). The percent survival at this time point was also significantly higher in the VT-1598 8- and 20-mg/kg groups (55.5 and 100%) compared to the vehicle control (0%;
P ≤ 0.003 for both comparisons).
In the
C. posadasii survival study, the fungal burden was also measured as a secondary endpoint at sacrifice, which for many mice was prior to scheduled study termination due to moribundity. In this study, mean fungal burdens in mice treated with VT-1598 (range, 4.91 to 5.36 log
10 CFU/g) and VT-1161 (4.36 log
10 CFU/g) were similar to control mice (5.29 log
10 CFU/g) (
Fig. 3A).
Coccidioides immitis survival.
Treatment with VT-1598 also resulted in significant improvements in survival in mice infected with
C. immitis, and the survival advantage observed was also dose proportional (
Fig. 2C). Median survival was >30 days in each of the VT-1598 treatment groups, which was significantly greater than that observed with the vehicle control (9 days;
P < 0.001 for all comparisons). The median survival was also significantly greater in mice treated with the positive-control fluconazole (
P < 0.001) compared to the vehicle control. The percentage of mice surviving to day 30, the endpoint used in the
C. immitis survival experiment, was higher in the VT-1598 treatment groups (70% in the 5-mg/kg group and 100% in the 15- and 45-mg/kg groups) compared to vehicle control (0%;
P ≤ 0.003 for all comparisons). There was also a trend toward a higher survival percentage in the fluconazole group, although this did not reach statistical significance (40%;
P = 0.087).
In the
C. immitis survival arm, fungal burdens were higher in mice treated with the lowest dose of VT-1598 (4.83 log
10 CFU/g) or fluconazole (5.73 log
10 CFU/g) (
Fig. 3B). However, mice treated with 15 and 45 mg/kg VT-1598 had mean fungal burdens of 2.19 and 2.58 log
10 CFU/g, respectively, values significantly lower than for the vehicle control (6.11 log
10 CFU/g;
P ≤ 0.002 for both comparisons).
VT-1598 plasma concentrations.
Plasma drug concentrations of VT-1598 and VT-1161 were determined for mice infected with C. posadasii. Two days after the last treatment, VT-1598 plasma concentrations in the fungal-burden study were 1.95 ± 0.63 μg/ml for the 4-mg/kg group and 17.6 ± 5.50 μg/ml for the 20-mg/kg group. Both were above the 1 μg/ml MIC for C. posadasii. VT-1161 concentrations (20 mg/kg dose) were 13.9 ± 3.35 μg/ml. The half-life of VT-1598 in mice is ∼24 h, which is reflected in the lack of any measurable VT-1598 at the time of sacrifice in the survival study. This likely explains the dose-dependent survival time of mice after stopping treatment with VT-1598, with animals failing after the plasma concentrations fell below the MIC. In contrast, the mean VT-1161 concentration in the survival study 28 days posttreatment was 1.34 ± 0.74 μg/ml, a finding consistent with its extended half-life of ∼14 days in mice.
Tolerability of VT-1598.
In both the C. posadasii and C. immitis models, VT-1598 was well tolerated. The weights of infected mice treated with VT-1598 in the fungal-burden arms of both the C. posadasii and the C. immitis models increased or remained the same over the 10-day period between inoculation and tissue collection (mean range of 0.05 to 2.9 g weight gain), whereas mice in the vehicle control groups lost weight over the same period (4.0 to 6.39 g weight loss). In the C. posadasii survival study, four mice in the VT-1598 groups died due to gavage injury (two in the 3.2-mg/kg group and one each in the 8- and 20-mg/kg groups). This may have been due to the prolonged period of oral dosing (14 days) in this model, although no mice were lost to gavage injury in the C. immitis model. These tolerability results are consistent with those of a 28-day study in rats, where the no-observed-adverse-effect level (NOAEL) for VT-1598 was 100 mg/kg when dosed once daily by oral gavage (E. Garvey and R. Schotzinger, unpublished data).
DISCUSSION
Central nervous system involvement of disseminated coccidioidomycosis is a serious complication, and if left untreated coccidioidal meningitis is nearly always fatal (
13–15). The current recommendation for the treatment of patients with coccidioidal meningitis is fluconazole, and therapy is lifelong due to the suppressive rather than the curative effects of currently available antifungals in this disease (
7,
16,
17). Although fluconazole is well tolerated relative to other antifungals, including other triazoles, prolonged therapy can be problematic due to drug-drug interactions that are mediated by interactions with CYP 450 isoenzymes, including CYP3A4, which can become problematic at doses greater than 200 mg (
18). In addition, there is concern about a rise in the number of
Coccidioides isolates with reduced susceptibility to fluconazole (
8). Although the other triazoles have more potent
in vitro activity (
8), and there is clinical experience with itraconazole in the treatment of this
Coccidioides infections (
7), these agents have clinically significant CYP450 mediated drug-drug interactions, and issues with variable oral bioavailability, pharmacokinetics, and adverse effects may also limit their effectiveness (
19–24). Thus, there is a need for newer agents in the treatment of coccidioidal meningitis.
VT-1598 is a novel therapeutic antifungal candidate that has been designed to be a more specific inhibitor of fungal Cyp51 than mammalian CYP450 isoenzymes. This tetrazole-based agent has previously shown potent and nearly identical
in vitro activities against both
C. posadasii and
C. immitis (
12), which were confirmed here for the specific isolates used in the murine studies. In the present study, we demonstrate that VT-1598 is also similarly effective in murine models of CNS coccidioidomycosis caused by both species, with significant improvements in survival and reductions in brain tissue fungal burden.
The suppression of fungal burden within the brain tissue 48 h posttreatment, including undetectable fungal growth in some mice, was observed with VT-1598 plasma concentrations above the MIC for C. posadasii, demonstrating the in vivo potency of VT-1598 against Coccidioides in the brain. VT-1598 given at ≥8 mg/kg/day also significantly prolonged survival in ≥90% of mice infected with either species for 15 days beyond the last dose of drug. In the C. posadasii study, mice were observed out to 28 days posttreatment, where the length of survival became dose dependent, and mice began to succumb to infection between 21 and 31 days postinoculation or between 6 and 16 days posttreatment. Presumably, deaths of mice that occurred after stopping treatment were related to the progression of fungal burden following reduction of circulating drug and not due to any drug-related intolerability, since VT-1598 was well tolerated, and overall survival was dose dependent. With ongoing treatment, which is the current standard of care in patients with coccidioidal meningitis, VT-1598 would be expected to continue to potently suppress fungal burden.
Some limitations in this study are noted as follows. The studies were conducted independently at two different institutions, using different sexes and strains of mice, and the same doses of VT-1598 were not used throughout. Thus, the exact threshold dose associated with in vivo efficacy is unknown, although the positive response observed over the range of doses that were evaluated is encouraging. This limitation may also be a strength insofar as the efficacy was reproducible with differing conditions and operators. As previously noted, a residual fungal burden was observed in the survival arm of the C. posadasii model where mice were monitored off therapy for 28 days. Since this survival endpoint was an additional 13 days after therapy had stopped relative to the C. immitis study, it is unknown whether such a rebound would have been observed with VT-1598 in the C. immitis model. In addition, VT-1598 plasma concentrations were not measured in the C. immitis model, so it is unknown how these may have compared to those measured in mice infected with C. posadasii. Furthermore, because mice were treated for only a relatively short period of time (14 days) in the survival arm where the subsequent rebound in fungal burden was observed in the C. posadasii study, it is unknown whether this translates into VT-1598 being only suppressive rather than curative for the treatment of CNS coccidioidomycosis. Thus, additional in vivo studies could give valuable insight into these and other questions. The in vivo antifungal activity of VT-1598 demonstrated in these studies lays the foundation for clinical investigation of this novel antifungal for the treatment of coccidioidomycosis.
MATERIALS AND METHODS
Fungal isolates.
For murine studies,
C. posadasii, strain Silveira (ATCC 28868, CBS 113859), for the murine studies was grown to maturity on glucose-yeast extract agar, and arthroconidia were harvested as previously described at the University of Arizona (
25,
26). Arthroconidia were enumerated by hemacytometer count, and viability was quantified by fungal culture to adjust concentration of infecting suspensions.
C. posadasii, strain Silveira, was also provided to the University of Texas Health Science Center at San Antonio for
in vitro susceptibility testing. A contemporary clinical isolate collected from a CNS shunt that was confirmed to be
C. immitis at the Fungus Testing Laboratory at the University of Texas Health Science Center San Antonio (UTHSCSA DI17-143) was also used for both animal studies and
in vitro testing (
27). At this institution, arthroconidia were harvested by the addition of sterile phosphate-buffered saline (PBS) to the potato dextrose agar (PDA) and gentle scraping of the agar surface with a sterile disposable loop to dislodge the arthroconidia. The arthroconidia suspension was centrifuged, the supernatant was discarded, and the arthroconidia were resuspended in PBS and filtered to remove hyphal fragments. The number of arthroconidia was adjusted to the desired inoculum concentration via dilution, which was verified by the use of a hemocytometer. The viability of the arthroconidia was evaluated by plating aliquots onto PDA and enumerating the number of CFU after incubation for 4 days at 35°C. All manipulations of
Coccidioides species were performed at biosafety level 3 (BSL3).
Antifungal agents.
VT-1598 and VT-1161 powders were provided by Viamet Pharmaceuticals, Inc. (Durham, NC), and were prepared for oral dosing using Cremophor EL (20% [vol/vol]). For the C. posadasii studies, the free base of VT-1598 was used, and for the C. immitis studies, a tosylate salt of VT-1598 was used, with the correction factor of 1.3 used for dose amount. Pharmacokinetics studies have demonstrated that exposures of the doses used in this study were not different between the two forms of VT-1598 (Viamet, unpublished data). Pharmaceutical-grade fluconazole was used for oral dosing in the C. immitis murine model. Pure powders of VT-1598 and fluconazole (Sigma, St. Louis, MO) were used for in vitro susceptibility testing.
In vitro susceptibility tests.
Antifungal susceptibility testing was performed by broth macrodilution according to Clinical and Laboratory Standards Institute standard M38-A2 (
28). Isolates were adjusted by a spectrophotometer to a starting inoculum of 1 × 10
4 to 5 × 10
4 arthroconidia/ml and then added to tubes containing serial 2-fold dilutions of VT-1598 (0.03 to 16 μg/ml) or fluconazole (0.125 to 64 μg/ml) in RPMI 1640 medium (0.165 M morpholinepropanesulfonic acid [pH 7.0] with
l-glutamine and without bicarbonate). The tubes were incubated at 35°C for 48 h, and the MIC was recorded as the lowest concentration that resulted in >80% inhibition of growth compared to the drug-free control.
Murine models of CNS coccidioidomycosis.
The C. posadasii murine model was conducted at the University of Arizona, and the C. immitis murine model was conducted at the University of Texas Health Science Center at San Antonio. All animals were housed and utilized according to PHS guidelines under approved institutional animal care and use protocols. The experiments were independently designed and performed at the two institutions. For the C. posadasii studies, 8-week-old female Swiss-Webster mice were used, and 4-week-old male ICR mice were used in the C. immitis studies. Both Swiss-Webster and ICR mice were purchased from Envigo (Indianapolis, IN). Throughout the course of the experiments, animals were monitored at least twice daily to prevent and minimize unnecessary pain and distress. Any animal that appeared moribund prior to the scheduled endpoint was euthanized.
Mice (
n = 10/group) were infected according to a previously published protocol (
25,
26,
29). Briefly, mice were anesthetized and inoculated intracranially with a lethal dose of
Coccidioides arthroconidia (∼90 arthroconidia of
C. posadasii and ∼137 arthroconidia of
C. immitis). Both fungal-burden and survival arms were used to evaluate the
in vivo efficacy of VT-1598 against CNS coccidioidomycosis. Treatment with antifungals began 48 h postinoculation and continued for 7 days in the fungal-burden arm and for 14 days in the survival arm. In the
C. posadasii studies, VT-1598 doses of 4 and 20 mg/kg administered by oral gavage once daily were used in the short-term fungal-burden study, and doses of 3.2, 8, and 20 mg/kg were used in the survival study. Based on previously published data (
25), VT-1161 administered at 20 mg/kg once daily by oral gavage served as the positive control in the
C. posadasii studies. In the
C. immitis model, VT-1598 doses of 5, 15, and 45 mg/kg once daily by oral gavage were used in both the fungal-burden and survival studies, and fluconazole administered at 25 mg/kg twice daily by oral gavage served as the positive control. In both models, Cremophor EL (20% [vol/vol]) served as the vehicle control and was administered by oral gavage once daily.
Fungal burden.
The brain fungal burden was determined at the time of sacrifice, either 24 or 48 h after final treatment in the fungal-burden studies or when euthanized due to moribundity or study termination in the survival studies. In the C. posadasii survival study, termination occurred 28 days after the last treatment dose (43 days postinoculation), and in the C. immitis study, animals were terminated 15 days after final dosing (day 30 postinoculation). Brains were collected, weighed, and homogenized in sterile saline. Dilutions of the homogenates were prepared and plated onto PDA or glucose yeast extract (GYE) agar, and the number of CFU per gram of brain tissue (CFU/g) was determined after at least 48 h of growth at 37°C.
Drug plasma concentrations.
Blood was collected terminally under anesthesia in the
C. posadasii infected mice and plasma was frozen until analysis. The plasma concentration of VT-1598 and the positive control VT-1161 were determined by liquid chromatography/tandem mass spectrometry as previously described (
25).
Data analysis.
Survival was plotted by Kaplan-Meier analysis and differences in the median survival time and percent survival were analyzed by the log-rank test and chi-square test, respectively. Since Coccidioides CFU data may be nonnormally distributed among groups, the Kruskal-Wallis Test with Dunn's posttest for multiple comparisons was used to determine whether differences between non-normally distributed fungal-burden data were significant. For normally distributed fungal-burden data, analysis of variance with a Tukey's posttest was used for multiple comparisons. For mice in which no fungal growth was detected, the lower limit of quantification of each tissue (10 colonies/weight of the individual organ) was used for data analysis. A P value of <0.05 was considered statistically significant for all comparisons. Descriptive statistics were used for VT-1598 plasma concentration results.
ACKNOWLEDGMENTS
We thank OpAns, LLC, for analysis of plasma concentrations of VT-1598 and VT-1161.
This project utilized preclinical services funded by the National Institute of Allergy and Infectious Diseases (NIAID), National Institutes of Health, Department of Health and Human Services, under contracts HHSN272201100018I and HHSN272201000038I, Task Orders A28 and A98, respectively, to the University of Texas Health Science Center at San Antonio. The University of Arizona studies were funded by Viamet Pharmaceuticals, Inc., through a NIAID R21/R33 grant awarded to Viamet (AI 101497). VT-1598 and VT-1161 powder was provided by Viamet Pharmaceuticals, Inc.
N.P.W. has received research support to the UT Health San Antonio from Astellas, bioMérieux, Cidara, F2G, Merck, and Viamet and has served on advisory boards for Merck, Astellas, Toyama, and Viamet. T.F.P. has received research grants to UT Health San Antonio from Astellas, Merck, and Revolution Medicines and has served as a consultant for Astellas, Gilead, Merck, Pfizer, Revolution Medicines, Toyama, Viamet, and Scynexis. L.K.N. has received travel support from Viamet Pharmaceuticals, Inc. C.M.Y., R.J.S., and E.P.G. are employees of Viamet Pharmaceuticals, Inc.