Posaconazole and Amphotericin B Combination Therapy against Cryptococcus neoformans Infection

To investigate the effects of posaconazole (POS) and amphotericinB (AMB) combination therapy in cryptococcal infection, we establishedan experimental model of systemic cryptococcosis in CD1 miceby intravenous injection of three distinct clinical isolatesof Cryptococcus neoformans. Therapy was started 24 h after theinfection and continued for 10 consecutive days. POS was givenat 3 and 10 mg/kg of body weight/day, while AMB was given at0.3 mg/kg/day. Combination therapy consisted of POS given ata low (combo 3) or at a high (combo 10) dose plus AMB. Survivalstudies showed that combo 3 was significantly more effectivethan POS at 3 mg/kg for two isolates tested (P value, $≤$ 0.001),while combo 10 was significantly more effective than POS at10 mg/kg for all three isolates (P values ranging from <0.001to 0.005). However, neither combination regimen was more effectivethan AMB alone. For two isolates, combination therapy was significantlymore effective than each single drug at reducing the fungalburden in the brain (P values ranging from 0.001 to 0.015) butnot in the lungs. This study demonstrates that the major impactof POS and AMB combination therapy is on brain fungal burdenrather than on survival.

INTRODUCTION

Top

Introduction

Cryptococcus neoformans is the cause of the most common life-threateningopportunistic fungal infection in patients with AIDS (13). Althoughthe occurrence of cryptococcosis among this group of patientshas decreased in recent years due to the introduction of triplehuman immunodeficiency virus therapy, its incidence remainshigh, particularly in developing countries (6). The treatmentof choice for cryptococcosis remains amphotericin B (AMB), withor without flucytosine. For suppression therapy, a triazole,such as fluconazole, is the agent of choice (20). The toxicityof AMB combined with flucytosine and the increasing isolationof fluconazole-resistant strains of C. neoformans underlinethe need for improved treatments and the use of combinationstrategies to overcome the emergence of resistance (1, 2, 4,5, 7, 8, 16). The combination of a polyene and an azole hasalways been questioned because of a potential for antagonism(10). However, recent experimental and clinical data have demonstratedthat the effects of an azole antifungal agent on the efficacyof AMB are either drug or fungus specific (3, 12, 15, 17-19,21-23). Little is known about the interactions between azolesand AMB against C. neoformans (3, 9, 12, 17, 18). Thus, in thisstudy, we evaluated a combination of AMB and the new triazoleposaconazole (POS) in a murine model of systemic cryptococcosis.

MATERIALS AND METHODS

Top

Materials and Methods

Isolates. Three clinical isolates of C. neoformans obtained from AIDSpatients were used in this study (Table 1). They had been previouslytested in vitro for MIC determinations (3), and they were selectedbased on checkerboard test results (Table 1).

View this table:
[in this window]
[in a new window]
TABLE 1. Characteristics of C. neoformans isolates used in this study

Antifungal agents. A stock solution of POS (Schering Plough Research Institute,Kenilworth, N.J.) was prepared with polyethylene glycol 200(Janssen Chimica, Geel, Belgium). AMB was purchased as Fungizonefrom Bristol-Myers Squibb S.p.A. (Latina, Italy) and preparedaccording to the manufacturer's instructions.

Animal studies. A murine model of systemic cryptococcosis was established withCD1 male mice (weight, 25 g; Charles River Laboratories, Calco,Italy) by injection into the lateral tail vein of viable yeastcells grown overnight in brain heart infusion broth. Animalexperiments were conducted with the approval of the Universityof Ancona ethics committee. AMB was given intraperitoneallyat 0.3 mg/kg of body weight/day, while POS was administeredby oral gavage at concentrations of 3 and 10 mg/kg/day. Twocombination therapies were also used: AMB plus POS at 3 mg/kg/day(combo 3) and AMB plus POS at 10 mg/kg/day (combo 10). Drugswere administered 2 h apart. Therapy was started 24 h afterinfection and continued for 10 consecutive days.

For survival studies, the mice were observed through days 40or 80, and deaths were recorded daily. Moribund mice were sacrificed,and their deaths were recorded as occurring on the next day.For tissue burden studies, the mice were sacrificed on days1, 7, 14, and 21 after the end of therapy, and the number ofCFU per gram of brain and lungs was determined by quantitativeplating of organ homogenates on Sabouraud dextrose agar plates.There were 10 to 13 mice per group in survival studies and 7mice per group in tissue burden studies.

Statistical analysis. The log rank test was used to determine the difference betweensurvival groups (statistical significance for multiple comparisonswas set at a P value of $≤$ 0.0125), while the Mann-Whitney U testwas used to determine significance in tissue burden studies(statistical significance for multiple comparisons was set ata P value of $≤$ 0.0160).

RESULTS

Top

Results

To investigate the effects of AMB combined with POS againstC. neoformans, we performed in vivo experiments with three clinicalisolates for which the in vitro interactions between these drugswere previously determined (3). According to the consensus terminology(10) for describing results of combination testing, synergywas observed for C. neoformans 486, while indifference (or Loeweadditivity) characterized drug interactions against C. neoformans491 and 2337. Survival results are shown in Fig. 1. Mice wereinfected with 1.7 x 10⁵, 2.0 x 10⁵, and 2.1 x 10⁵ CFU/mousein experiments with C. neoformans 486, 491, and 2337, respectively.For C. neoformans 486, all treatment regimens were effectivein prolonging survival relative to the controls (P < 0.001).Both combo 3 and combo 10 were more effective than the respectivemonotherapies with POS (P = 0.001 and 0.005, respectively) butwere not more effective than AMB alone. For C. neoformans 491,only AMB (P = 0.008) and combo 10 (P < 0.001) were effectivein prolonging survival relative to the controls. No differencewas observed between these two regimens. Of note, combo 3 abolishedthe efficacy of the polyene. For C. neoformans 2337, all treatmentregimens, with the exception of POS at 3 mg/kg, were effectiverelative to the controls (P values ranging from <0.001 to0.012). For this strain, both combo 3 and combo 10 were moreeffective than the respective monotherapies with POS (P <0.001). Although the survival of mice treated with both combinationtherapies was not significantly different from that of micetreated with the polyene (for AMB versus combo 3, the P valuewas 0.070; for AMB versus combo 10, the P value was 0.039),there was a trend toward reduced activity of AMB in the combinationtherapies.

View larger version (27K):
[in this window]
[in a new window]
FIG. 1. Survival of mice infected intravenously with cells of C. neoformans 486, 491, or 2337 and treated for 10 consecutive days with POS at 3 mg/kg, POS at 10 mg/kg, AMB at 3 mg/kg/day, combo 3, or combo 10.

The second set of experiments consisted of tissue burden studies.In these experiments, a total of four groups were considered:control, AMB at 0.3 mg/kg/day, POS at 10 mg/kg/day, and combination.Mice were infected with 1.7 x 10⁵, 2.5 x 10⁵, and 2.6 x 10⁵CFU/mouse in experiments with C. neoformans 486, 491, and 2337,respectively. The animals were treated for 10 days and sacrified24 h after the end of therapy (early tissue burden). The resultsare shown in Table 2. For C. neoformans 486, all regimens weresignificantly superior to controls at reducing brain fungalburdens (P values ranging from 0.001 to 0.002), whereas onlythe triazole and the combination were similarly effective atreducing lung fungal burdens (P = 0.001). In the lungs, combinationtherapy and POS were also more effective than AMB alone (P =0.002). For C. neoformans 491, all regimens, with the exceptionof AMB in the lungs, were effective at reducing fungal burdensrelative to controls in both organs (P values ranging from 0.0006to 0.009). Also, combination and POS treatments were more activethan AMB in both organs (P values ranging from 0.001 to 0.009).Moreover, combination therapy was significantly more effectivethan each single drug at reducing fungal counts in the brain(P $≤$ 0.002). For C. neoformans 2337, all drug regimens were significantlysuperior to controls in both organs (P values ranging from 0.001to 0.002); the triazole was more effective than the polyenein brain tissue (P = 0.0002). Again, the combination therapywas more effective than both monotherapies at reducing brainfungal burdens (P = 0.015).

View this table:
[in this window]
[in a new window]
TABLE 2. Effects of combination therapy on early tissue burden in mice infected with isolates of C. neoformans^a

To investigate the effects of combination therapy along withthe progression of infection, we performed additional tissueburden studies with C. neoformans 2337. The mice were infectedwith 2.5 x 10³ CFU/mouse, treated for 10 days, and sacrificedon days 7, 14, and 21 after the end of therapy (late tissueburden). The results are shown in Table 3. The results on day7 were similar to those on day 1 after the end of therapy. Allregimens, with the exception of AMB in the brain (P = 0.019),were effective relative to controls at reducing the fungal countsin both organs (P values ranging from 0.002 to 0.009). Again,the combination therapy was significantly more active than eachsingle therapy in the brain (P $≤$ 0.009).

View this table:
[in this window]
[in a new window]
TABLE 3. Effects of combination therapy on late tissue burden in mice infected with C. neoformans 2337^a

On day 14 after the end of therapy, none of the treatments waseffective in both organs. Actually, the number of CFU per gramin the treatment groups reached those in the controls (Table3). Similarly, no significant difference was detected in bothorgans on day 21 after the end of therapy (data not shown).

DISCUSSION

Top

Discussion

In this study, we evaluated the effects of combination therapywith POS and AMB against C. neoformans in a murine model ofsystemic cryptococcosis. The effects of this combination therapywere previously investigated in vitro against a large numberof clinical isolates of C. neoformans. In that study, the combinationregimen yielded 33 and 67% synergistic and indifferent (Loeweadditivity) interactions, respectively (3). Therefore, threeclinical isolates of C. neoformans for which either synergisticor indifferent interactions were observed in vitro were selectedfor the present study.

Several observations can be made from this study. First, neithersurvival nor tissue burden data correlated with the previousresults observed in vitro. This finding may be due to the factthat the checkerboard methodology measures drug effects at staticconcentrations only, while in vivo concentrations vary upondrug absorption and elimination rates and thereby influencethe overall efficacy of the combination (11, 14).

Second, survival studies showed that combination therapy wasgenerally more effective than the triazole alone but not moreeffective than the polyene alone. Although in this study weused a low dose of AMB to try to avoid its rapid fungicidalactivity, the polyene given at 0.3 mg/kg/day was still effectiveat prolonging survival with all three isolates. It is importantto note that the dose of the triazole can affect the efficacyof AMB. In fact, we found that POS given at 3 mg/kg/day abolishedthe efficacy of AMB in mice infected with C. neoformans 491.Since this phenomenon was observed only for one isolate, wespeculate that these variable effects are more likely to becorrelated with the strain itself rather than with pharmacokineticparameters. Due to the considerable degree of variation amongisolates of C. neoformans with respect to genetic background,the effects of such combination therapies may vary among theisolates. We hypothesize that a low dose of the triazole maycause subtle changes in sterol membrane composition for a givenstrain (e.g., changes in methylated sterol/demethylated sterolratios) and thereby may reduce the efficacy of AMB. On the contrary,a high dose of the triazole may not allow the fungus to modifyits membrane composition and thereby may conserve targets forAMB.

Third, our tissue burden data showed a reciprocal potentiationof the polyene and the azole drug in brain tissues. This findingwas observed in mice infected with two isolates of C. neoformans(491 and 2337). Since the central nervous system is the targetorgan for cryptococcosis, this finding is of particular interest.This effect was mainly evident in the early phase of infection,as shown in studies conducted with C. neoformans 2337, in whichthe efficacy of therapy persisted until 7 days after the endof treatment. The residual burden analyzed on days 14 and 21after the end of therapy showed a lack of effect of any of theregimens, including combination therapy. This finding can beeasily explained by considering the possibility of completedrug clearance from the infected organs at these times. Overall,our data correlate with those reported by other investigators(9, 17). Early studies with a ketoconazole-AMB combination showedthat the major impact of dual therapy was on fungal tissue burdenrather than on survival (17). Similarly, Barchiesi et al. recentlyobserved that a fluconazole-AMB combination was more effectivethan fluconazole alone but not more effective than AMB alonein prolonging survival, while a reciprocal potentiation wasoften detected for fungal burden (3).

In conclusion, a combination of the new triazole POS with AMBdid not show significant antagonism in an experimental murinemodel of systemic cryptococcosis. Although this therapeuticapproach does not seem to be more effective than AMB alone interms of survival, it may have a theoretical advantage at thetissue level. It is clear that additional studies are warrantedto further elucidate the potential benefits of this strategybefore considering it suitable for clinical practice.

ACKNOWLEDGMENTS

This work was supported in part by a grant from Istituto Superioredi Sanità, Rome, Italy (IV AIDS project, grant no. 50D.29).

FOOTNOTES

* Corresponding author. Mailing address: Università Politecnica delle Marche, Azienda Ospedaliera Umberto I, Via Conca, 60020 Torrette di Ancona, Ancona, Italy. Phone: 39.071.5963467. Fax: 39.0715963468. E-mail: l.infettive{at}ao-umbertoprimo.marche.it.

REFERENCES

Top