Efficacy of Delayed Therapy with Fosmanogepix (APX001) in a Murine Model of Candida auris Invasive Candidiasis

The emerging pathogenic yeast Candida auris is associated with antifungal resistance and high mortality. The novel antifungal agent manogepix (APX001A) inhibits glycosylphosphatidylinositol-anchored protein maturation and has demonstrated activity against numerous pathogenic fungi, including C. auris.

portion of isolates have also been reported to be resistant to all clinically available antifungal classes, including the polyene amphotericin B (13). Given the rapid spread of this emerging infection, the high mortality rates associated with invasive disease (1), and limited treatment options in the face of antifungal resistance, new therapeutic agents and treatment strategies are needed.
Manogepix (APX001A, formerly E1210; Amplyx Pharmaceuticals, Inc., San Diego, CA) is a novel antifungal agent that inhibits the inositol acyltransferase Gwt1 in the glycosylphosphatidylinositol (GPI) anchor biosynthesis pathway, thereby preventing GPI-anchored protein maturation (14). This agent is administered as the N-phosphonooxymethyl prodrug fosmanogepix (APX001, formerly E1211), which is rapidly converted to the active moiety following intravenous (i.v.) or oral administration. Studies have demonstrated that manogepix inhibits inositol acylation in fungi, including Candida and Aspergillus species, but not the human form of this enzyme (14). Thus, the potential for human toxicity is limited. Previous studies have demonstrated both in vitro and in vivo activity against various pathogenic yeasts, including Candida albicans, Candida glabrata, Cryptococcus neoformans, and Coccidioides posadasii, as well as filamentous fungi, including Aspergillus, Fusarium, and Scedosporium species (15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26), although it lacks in vitro activity against Candida krusei and some of the Mucorales (15,16). Two recent studies have also reported in vivo efficacy against invasive disease caused by C. auris (27,28). However, in both studies, therapy was started soon after intravenous inoculation (2 hours). Thus, it is unknown how effective fosmanogepix would be with delayed initiation of treatment. This may be of clinical importance given the challenges associated with the diagnosis of infections caused by this species (3,29), which may delay therapy. The objective of this study was to evaluate the in vivo efficacy of fosmanogepix in an experimental model of invasive candidiasis caused by C. auris, where the start of therapy was delayed until 24 hours after inoculation.

RESULTS
In vitro susceptibility. Manogepix demonstrated potent in vitro activity against C. auris isolates, with MICs ranging from Յ0.002 to 0.03 g/ml and MIC 50 and MIC 90 values of 0.03 g/ml and 0.125 g/ml, respectively. The geometric mean (GM) MIC was 0.013 g/ml. In contrast, the MIC range for fluconazole was 2 to Ͼ64 g/ml, with MIC 50 and MIC 90 values each at Ͼ64 g/ml. Caspofungin MIC values ranged from 0.06 to 0.5 g/ml, and the MIC 50 and MIC 90 values were 0.25 g/ml and 0.5 g/ml, respectively. The MIC values for manogepix, fluconazole, and caspofungin against the isolate used in the infection model were 0.03 g/ml, Ͼ64 g/ml, and 0.25 g/ml, respectively.
Pharmacokinetics and dose tolerability. The pharmacokinetic parameters determined in uninfected mice dosed with fosmanogepix in the preliminary pharmacokinetic/dose tolerability study are shown in Table 1. The overall exposures obtained after the last dose on day 7 (area under the concentration-time curve from 0 to infinite time [AUC 0-inf ]) ranged between 11.3 to 58.4 g·h/ml. Since the different cohorts were dosed with fosmanogepix twice a day (BID) or three times a day (TID), total daily exposures ranged from approximately 34 to 175 g·h/ml for the active component manogepix. Increases in manogepix exposure were greater than the proportional increases in the fosmanogepix doses administered. Manogepix, administered as the prodrug fosmanogepix, appeared to be well tolerated, with the exception of mice that received the 260 mg/kg of body weight TID dose, where 5 of the 21 mice were observed to be moribund prior to the study endpoint. Based on these results, the top dose of fosmanogepix used in the in vivo efficacy study was 260 mg/kg BID. Survival. A survival advantage was observed in mice treated with fosmanogepix at clinically relevant doses ( Fig. 1) (30,31). At each dose level, median survival (Ͼ21 days for each group) and percent survival (range, 90% to 100%) were both significantly greater than the vehicle control (5 days and 10%, respectively; P Ͻ 0.0001 for each comparison). Survival was also significantly improved in mice that received a dose of caspofungin that results in exposures in mice that are 3-to 6-fold greater exposures than the standard dose used clinically (Ͼ21 days and 100%; P Ͻ 0.0001 versus vehicle control) (32)(33)(34)(35). In contrast, there was no difference in median survival or percent survival compared with that of the control in fluconazole-treated mice (6 days and 10%). This finding is consistent with the in vitro fluconazole resistance observed for the clinical isolate used for infection.
Fungal burden. Following 7 days of therapy, kidney fungal burden was significantly reduced in mice treated with 260 mg/kg BID fosmanogepix (mean, 3.86 log 10 CFU/g; P ϭ 0.0273) or with 10 mg/kg intraperitoneal (i.p.) once a day (QD) caspofungin (3.41 log 10 CFU/g; P ϭ 0.0033) compared with that of the vehicle control (5.61 log 10 CFU/g) ( Fig. 2A). In contrast, while fungal burden was numerically reduced in mice treated with fosmanogepix at doses of 104 or 130 mg/kg TID (range, 4.11 to 4.30 log 10 CFU/g), these differences were not significant. However, there was a trend toward significance in the 130 mg/kg TID fosmanogepix group compared with that of the vehicle control (P ϭ 0.076). No reduction in fungal burden was observed in the kidneys of mice treated with fluconazole (5.88 log 10 CFU/g). Similar results were also observed in the brain tissue in mice treated with fosmanogepix, as a significant reduction in CFU counts was observed in mice treated with the 260 mg/kg dose TID (2.99 log 10 CFU/g) compared with that of the vehicle control (4.40 log 10 CFU/g; P ϭ 0.0088) (Fig. 2B). In contrast, no significant reductions in fungal burden were observed between the vehicle control group and mice treated with the lower doses of fosmanogepix (range, 3.78 to 3.85 log 10 CFU/g), fluconazole (4.91 log 10 CFU/g), or caspofungin (4.36 log 10 CFU/g). Overall, the in vivo activity observed following 7 days of fosmanogepix therapy was static in nature since treatment with this investigational agent resulted in reductions in fungal burden of less than 1 log 10 CFU/g compared with that measured just prior to the start of therapy (range, Ϫ0.50 to Ϫ0.94 log 10 CFU/g) ( Table 2). Treatment with a high (10 mg/ kg) dose of caspofungin, which results in exposures 3 to 6 times that obtained in

Fosmanogepix Delayed Therapy versus Candida auris
Antimicrobial Agents and Chemotherapy humans at the clinical dose did result in cidal activity within the kidneys (Ϫ1.39 log 10 CFU/g) but not in the brain tissue (0.63 log 10 CFU/g) (32)(33)(34)(35). Fungal burden was also assessed in the survival arm as mice succumbed to infection or at the predetermined endpoint (day 21 postinoculation). Here, CFU counts within the kidneys of mice treated at each dose level of fosmanogepix (range, 4.47 to 4.55 log 10 CFU/g) were significantly lower than that observed in the vehicle control group (7.90 log 10 CFU/g; P Ͻ 0.0001 for all comparisons) (Fig. 3A). Similar to the day 8 fungal burden results and consistent with the survival curves, high dose caspofungin also lowered fungal burden within the kidneys in the survival arm. Significant reductions were also observed in brain tissue fungal burden in the survival arm with each dose of fosmanogepix and caspofungin (fosmanogepix range, 3.10 to 3.27 log 10 CFU/g; caspofungin, 2.70 log 10 CFU/g; P Ͻ 0.0001 for all comparisons). Interestingly, in contrast to the results of the fungal burden arm, each fosmanogepix dose level, as well as high  dose caspofungin, resulted in a reduction in fungal burden within the brains and kidneys of greater than 1 log 10 CFU/g compared with that measured just prior to the start of therapy. This reduction observed with fosmanogepix ranged between Ϫ1.50 and Ϫ1.57 log 10 CFU/g in the kidneys and between Ϫ1.29 and Ϫ1.44 log 10 CFU/g in the brains. Consistent with the survival results and the day 8 fungal burden results, no reductions in CFU counts were observed within the kidneys or brain tissue of mice treated with fluconazole.

DISCUSSION
Candida auris is an emerging pathogen of great clinical significance. This yeast has rapidly spread globally, with numerous outbreaks of infections and colonization now having been reported in many countries (1)(2)(3). There are several characteristics of this fungal species which make it particularly challenging to deal with, including, but not limited to high associated mortality rate, a high rate of antifungal resistance (1,2), and reported resistance to multiple classes of clinically available antifungal agents (1,3,10,11). In addition, it is known to form biofilms which make clinical eradication, of either infection or colonization, and removal from environmental surfaces extremely challenging (36). Currently, echinocandins are recommended for initial therapy for patients with C. auris infections (https://www.cdc.gov/fungal/candida-auris/c-auris-treatment .html). However, given the limitations of IV-only therapy, the emergence of strains resistant to echinocandins, and limited brain penetration, there is a need for new therapeutic candidates. Several new antifungals currently under development have demonstrated promising activity, in vitro and/or in vivo, against C. auris. These include the echinocandin rezafungin (formerly CD101) and the triterpenoid ibrexafungerp (formerly SCY-078), both of which inhibit the production of 1,3-␤-D-glucan within the fungal cell wall (36)(37)(38)(39)(40). In addition, the tetrazole VT-1598, which acts in a similar fashion to the triazoles by inhibiting the biosynthesis of ergosterol (41), has also demonstrated in vitro and in vivo activity against C. auris (42).
Manogepix, which has a novel mechanism of action different than that of the triazoles, echinocandins, and polyenes, has been shown to have broad in vitro activity against numerous pathogenic fungi, including Candida, Cryptococcus, Aspergillus, Fusarium, and Scedosporium species (15)(16)(17)(18)(19)24). This in vitro activity has translated into in vivo efficacy of fosmanogepix in experimental models of candidiasis, cryptococcosis, coccidioidomycosis, aspergillosis, and scedosporiosis (20-22, 24, 28, 43). This therapeutic candidate has also previously been reported to have in vitro and in vivo activity against C. auris. Using a neutropenic model, Hager et al. reported that significant improvements in both in vivo outcome measures (survival and fungal burden) were achieved in mice infected with C. auris (27). These results are similar to the ones we report in the current study. The in vitro results reported by Hager et al. for manogepix against 16 C. auris isolates are also similar to those observed in the current study (MIC 90 , 0.03 g/ml in both). Interestingly, fosmanogepix appeared to be just as effective when therapy was delayed by 24 hours, as done in our study, compared with 2 hours postinoculation, as performed in the previous study.
In another study which evaluated the pharmacokinetics and pharmacodynamics (PK-PD) of APX001 against Candida species conducted by Zhao et al., fosmanogepix was also efficacious at reducing kidney fungal burden when assessed at 4 days in a mouse model that utilized a cyclophosphamide/cortisone acetate regimen to render mice neutropenic (28). Here, maximum reductions in C. auris fungal burden from the starting inoculum measured just prior to the start of therapy (2 hours after inoculation) ranged between 0.21 log 10 CFU/g to 1.02 log 10 CFU/g (mean, Ϫ0.69 log 10 CFU/g) and the PK-PD parameter most closely associated with efficacy was the area under the concentration-time curve for the free, unbound fraction of a drug (fAUC)/MIC ratio. In mouse PK-PD studies, a net stasis outcome (lack of change in burden over the treatment period) has been correlated with clinical efficacy in humans, especially for the echinocandins (44). In the current study, the doses utilized resulted in exposures at or below those previously associated with stasis. Zhao et al. reported the mean and median 24-hour total drug AUC (tAUC)/MIC values resulting in stasis, which were 7,336.4 and 5,864.2, respectively, for C. auris (28), whereas in this study, the 24-hour tAUC/MIC ranged between 1,130 to 3,520 in the three APX001 dose cohorts used to assess efficacy. However, fosmanogepix maintained efficacy, as is evident by both reductions in fungal burden and improvements in survival, especially at the 260-mg/kg BID dose where the 24-hour tAUC/MIC value was approximately half that reported for stasis in the previous PK-PD study (28).
Interestingly, we observed greater reductions in fungal burden in the survival arm of the current study, where mice received 7 days of therapy and were observed through day 21, versus the day 8 cohort. This may be due to the fact that a greater inoculum concentration was used to infect mice in the survival arm of this model than the fungal burden arm. Alternatively, the recovery of neutrophils during the 10-to 21-day postinoculation period following the administration of 5-fluorouracil may have contributed to greater reductions in fungal burden in the survival arm. Other studies that have evaluated APX001 in murine models of invasive candidiasis caused by C. auris did not measure fungal burden in a survival arm or did not assess survival (27,28).
The pharmacokinetics of manogepix after administration of the prodrug fosmanogepix in mice are different than those observed in humans, as this agent has been shown to undergo rapid metabolism in outbred strains (e.g., CD-1 or ICR) with half-lives ranging between 0.9 to 2.75 hours (26,28). In contrast, in humans the half-life has been reported to be ϳ2.5 days in healthy volunteers in phase 1 clinical trials (30,31). In order to slow the clearance and improve its overall exposure in mice, others have coadministered the cytochrome P450 inhibitor 1-aminobenzotriazole to inhibit the rapid metabolism of manogepix observed in murine models and allow for higher total daily exposures. This strategy has been successfully used in studies of invasive candidiasis caused by species other than C. auris, invasive aspergillosis, and invasive scedosporiosis (20,22,25,26,28). However, this was not done in the current study or in the others that evaluated the in vivo efficacy of fosmanogepix against invasive C. auris infection (27,28).
Overall, the results of the current study are encouraging and confirm those that have previously reported good in vivo efficacy for fosmanogepix against C. auris infections. However, further studies are warranted to determine the exact role of this investigational agent against invasive infections caused by C. auris. This may include, but is not limited to, work to determine if the in vivo activity may be further enhanced with higher exposures. Overall, the results of the current study suggest that manogepix, administered as the prodrug fosmanogepix, may be a future option for the treatment of C. auris infections. Candida auris isolates. For in vitro susceptibility testing, 10 C. auris isolates from the CDC FDA Antibiotic Resistance (AR) Bank (https://www.cdc.gov/drugresistance/resistance-bank/index.html) and 3 clinical isolates received and identified as C. auris through DNA sequence analysis of the internal transcribed spacer (ITS) ribosomal DNA (rDNA) region by the Fungus Testing Laboratory at the University of Texas Health Science Center were used (45)(46)(47). Prior to in vitro testing, all isolates were subcultured twice on Sabouraud dextrose agar (SDA). A clinical isolate of C. auris (UTHSCSA DI17-46) was used to establish the infection in mice. Prior to inoculation, the isolate was subcultured twice onto SDA, and colonies were taken from the second subculture and placed into brain heart infusion broth, which was grown overnight with shaking at 200 rpm at 37°C. Following centrifugation, the cells were collected and washed in sterile saline with 0.1% Tween 20. A hemocytometer was used to adjust the numbers of cells to the desired starting inocula, and viability was confirmed by serial dilution of aliquots onto SDA and counting the number of colonies following incubation at 37°C.

Antifungals
In vitro susceptibility. In vitro susceptibility testing was performed by broth microdilution in RPMI according the methods in the CLSI M27 standard (48). MICs were determined visually for manogepix, fluconazole, and caspofungin after 24 hours of incubation as the lowest drug that inhibited 50% growth. The concentration ranges evaluated were 0.002 to 1 g/ml for manogepix, 0.125 to 64 g/ml for fluconazole, and 0.015 to 8 g/ml for caspofungin.
Pharmacokinetics and dose tolerability. A preliminary dose tolerability and pharmacokinetic study was done in uninfected, immunosuppressed mice. Two days prior to the start of dosing with the prodrug fosmanogepix, mice were rendered neutropenic with a single 5-mg intravenous (i.v.) dose of 5-fluorouracil. Antibiotic prophylaxis with enrofloxacin (50 ppm in the drinking water) was also provided to prevent bacterial infections. Fosmanogepix was administered by intraperitoneal (i.p.) injection at doses of 78 mg/kg TID, 130 mg/kg TID, and 260 mg/kg BID or TID. As a result of the 1.3-fold difference in the molecular weight for the prodrug versus the active moiety, these prodrug doses corresponded to manogepix doses of 60 mg/kg, 100 mg/kg, and 200 mg/kg, respectively. After the first dose on day 7 of therapy, groups of 3 mice per dose group were humanely euthanized at 0.5, 1, 2, 4, 8, 12, and 24 hours, and blood was collected by cardiac puncture. The plasma was separated and stored frozen. Plasma was shipped to QPS, LLC, and manogepix plasma concentrations were measured by an established liquid chromatography-tandem mass spectrometry (LC/MS-MS) assay in mouse plasma, with a lower limit of quantitation of 50.0 ng/ml. Pharmacokinetic parameters for manogepix were calculated using the noncompartmental method of the pharmacokinetic software package Phoenix WinNonlin (Certara, St. Louis, MO), and the area under the concentration-time curve (AUC) was calculated using the linear trapezoidal method.
Infection model. As noted above, mice were rendered neutropenic with a single 5-mg i.v. dose of 5-fluorouracil given 1 day prior to inoculation. This immunosuppression regimen results in neutrophil counts of Ͻ100 cells/mm 3 for 10 days (49). Enrofloxacin was administered in the drinking water (50 ppm) to prevent bacterial superinfection. Mice were infected on day 0 with a 0.2-ml inoculum of the fluconazole-resistant strain C. auris (UTHSCSA DI17-46) via the lateral tail vein in order to establish infection. The target inocula were 1 ϫ 10 7 cells/mouse in the survival arm and 5 ϫ 10 6 cells/mouse in the fungal burden arm. Therapy began 24 hours postinoculation and was continued through day 7. Treatment groups consisted of vehicle control (D5W); fosmanogepix at doses of 104 mg/kg i.p. TID, 130 mg/kg i.p. TID, and 260 mg/kg i.p. BID; fluconazole at 20 mg/kg by oral (p.o.) gavage QD; and caspofungin at 10 mg/kg i.p. QD. The doses of fosmanogepix corresponded to doses of 80 mg/kg, 100 mg/kg, and 200 mg/kg of the active moiety manogepix. Mice were observed multiple times per day to minimize unnecessary pain and distress, and any animal that appeared moribund prior to the study endpoint was humanely euthanized. The animal protocol was approved by the University of Texas Health Science Center Institutional Animal Care and Use Committee.
Fungal burden and survival. Both fungal burden and survival arms were included. In the fungal burden arm, on day 8 postinoculation, mice were humanely euthanized, and the kidneys and brains were collected aseptically. The organs were homogenized, and further dilutions were made in sterile saline, which were plated onto SDA. Colonies were counted after 48 hours of incubation, and the number of CFUs per gram of tissue (CFU/g) were determined. In the survival arm, mice were followed for 14 days after therapy stopped until day 21 postinoculation. As noted previously, mice were monitored multiple times per day, and any animal that appeared moribund was humanely euthanized and death was recorded as occurring the next day. Fungal burden was also measured in the kidneys and brains on day 21 or as the mice became moribund as described above.
Data analysis. Descriptive statistics were used to evaluate the in vitro activity of manogepix, including MIC range, concentrations that inhibited 50% and 90% of the isolates tested (MIC 50 and MIC 90 , respectively), and geometric mean (GM) MICs. Kaplan-Meier analysis was used to plot survival, and the differences in median survival and percent survival were evaluated by the log-rank and Fisher's exact test, respectively. Analysis of variance (ANOVA) with Tukey's posttest for multiple comparisons was used to assess for differences in fungal burden versus the vehicle control group, and in vivo fungicidal activity was defined as a 1-log 10 CFU/g reduction in fungal burden compared with that of the 24-hour control group (i.e., fungal burden measured just prior to the start of therapy).