MIC Values of Voriconazole Are Predictive of Treatment Results in Murine Infections by Aspergillus terreus Species Complex

Abstract

We evaluated the efficacy of voriconazole against nine strains of Aspergillus terreus with different MICs (0.12 to 4 μg/ml) by using a murine model. Markers of efficacy included survival, tissue burden, galactomannan antigenemia, and drug serum levels. Voriconazole was especially effective in prolonging survival and reducing the fungal load in infections by strains that showed MICs that were less than or equal to the epidemiological cutoff value (1 μg/ml). In vitro data might be useful for predicting the outcome of A. terreus infections.

TEXT

Invasive aspergillosis is a frequently fatal infection that mainly affects immunosuppressed patients (1). Some species of the A. terreus complex have become increasingly important in recent years (2). Voriconazole is the recommended drug for treatment of invasive aspergillosis (3), although Aspergillus isolates resistant to voriconazole have been reported in association with clinical failure (4, 5). Epidemiological cutoff values (ECV) were proposed recently for several Aspergillus spp., including A. terreus (ECV, 1 μg/ml), and those isolates showing MICs higher than the ECV may have acquired resistance mechanisms (6, 7).

We evaluated the efficacy of voriconazole in the treatment of invasive murine infection by A. terreus species complex, testing nine clinical isolates with MICs ranging from 0.12 to 4 μg/ml (see Fig. 1 to 3, below) that were previously determined using a reference method (8). The aim was to assess if the in vitro data correlated with the in vivo antifungal drug efficacy.

Fig 1.

Mean survival time of mice infected with A. terreus. VRC 25, voriconazole administered at 25 mg/kg orally once a day. Error bars indicate standard errors of the means (n = 8 mice per group). Statistical significance values are indicated by lowercase letters (following strain names): a, P < 0.05 versus control; b, P < 0.05 versus FMR 8752; c, P < 0.05 versus FMR 8806; d, P < 0.05 versus UTHSC 07-3300; e, P < 0.05 versus UTHSC 08-3714.

Fig 3.

Galactomannan serum levels in mice infected with A. terreus as measured on day 5 of treatment. VRC 25, voriconazole administered at 25 mg/kg orally once a day. The horizontal line indicates the cutoff for positivity (GMI, > 1.5). Error bars represent standard errors of the means. Statistical significance: a, P < 0.05 versus control.

Male OF1 mice were used in this study. All animal care procedures were supervised and approved by the Universitat Rovira i Virgili Animal Welfare and Ethics Committee. Animals were immunosuppressed 1 day prior to infection by a single intraperitoneal injection of 200 mg of cyclophosphamide/kg of body weight plus a single intravenous injection of 150 mg/kg of 5-fluorouracil. Mice were challenged with 2 × 10⁵ CFU A. terreus species complex via the lateral tail vein. This inoculum was suitable to produce an acute infection, with 100% of the animals dying within 13 days (data not shown).

Voriconazole was administered at 25 mg/kg/dose once a day orally (9) during 7 days. From 3 days before infection, the mice were given grapefruit juice instead of water (10). All animals received ceftazidime at 5 mg/kg subcutaneously once daily. The efficacy of voriconazole was evaluated based on prolonged survival of mice, reduced tissue burden, and reduced galactomannan serum levels. Groups of 8 mice were randomly established for each strain. For tissue burden studies, animals were sacrificed on day 5 postinfection, and the numbers of CFU/g of kidney or brain tissue were calculated.

Additionally, before being sacrificed, approximately 1 ml of blood from each mouse belonging to one of the tissue burden groups was extracted by cardiac puncture. Pooled serum samples were used to determine the drug concentration by bioassay 4 h after the drug was administered (11, 12), and the galactomannan levels were determined by enzyme immunoassay (Platelia Aspergillus). Values were expressed as the galactomannan index (GMI), defined as the optical density of a sample divided by the optical density of a threshold serum sample provided in the test kit.

The Kaplan-Meier method and log rank test were used for survival studies. When multiple comparisons were carried out, the Bonferroni correction was used to avoid an increase in type I error. The tissue burden studies were analyzed using the Mann-Whitney U test. The Kolmogorov-Smirnov test was carried out to determine the normal distribution of the galactomannan serum levels and bioassay data, so that they could be analyzed using the t test.

For all strains, voriconazole significantly prolonged survival with respect to the control group. For the two strains with the lowest MICs (0.12 μg/ml), survival was 100%. For the strains with MICs of ≤1 μg/ml, the survival rate of animals treated ranged from 50 to 60%. With the strain with MICs of 2 μg/ml, 25% of the infected mice survived, but in those infected with the strain with the highest MIC (4 μg/ml), none of the mice survived (0%). In general, voriconazole was significantly more effective in mice infected with strains with MICs of ≤1 μg/ml than in those with MICs of ≥2 μg/ml (P < 0.05) (Fig. 1).

Voriconazole was able to reduce the fungal load significantly in both organs tested from the animals challenged with isolates with MICs of ≤1 μg/ml compared to untreated groups and, in general, with respect to the groups infected with strains with MICs of ≥2 μg/ml. For the strain with a MIC of 2 μg/ml, voriconazole only reduced the fungal load in kidneys, and for the strain with a MIC of 4 μg/ml, there were no significant differences with respect to the control group in any of the organs studied (Fig. 2).

Fig 2.

(A) Box plot of changes in fungal loads of mice infected with 2 × 10⁵ CFU of A. terreus relative to the respective control in kidneys. Statistical significance values are indicated by lowercase letters following the strain name: a, P < 0.05 versus control; b, P < 0.05 versus UTHSC 08-3714; c, P < 0.05 versus UTHSC 11-320; d, P < 0.05 versus UTHCS 07-3300; e, P < 0.05 versus UTHSC 10-3389; f, P < 0.05 versus FMR 8752; g, P < 0.05 versus UTHSC 11-53. (B) Similar box plot for changes in fungal loads in brains of mice treated with 25 mg/kg voriconazole orally once a day: a, P < 0.05 versus control; b, P < 0.05 versus UTHSC 07-3300; c, P < 0.05 versus UTHSC 08-3714; d, P < 0.05 versus FMR 8806; e, P < 0.05 versus FMR. Error bars represent standard errors of the means (n = 8 mice per group).

The serum concentration of voriconazole on day 5 of the experiment was 7.01 ± 2.82 μg/ml (mean ± standard error of the mean). All serum drug concentrations were higher than the corresponding MICs for the strains tested (data not shown). Galactomannan serum levels were significantly lower in mice treated with voriconazole than in controls, but the galactomannan serum levels were above the cutoff for positivity (GMI, >1.5) in all cases (13) (Fig. 3).

In the absence of clinical data, establishing the ECV for several Aspergillus species (6) together with the results of animal studies might contribute to the creation of clinical breakpoints for Aspergillus spp. and azoles (14). To our knowledge, this is the first study to explore the possible relationship between the MIC values for voriconazole and the outcomes of murine experimental infections by isolates of A. terreus species complex. This study demonstrated that although voriconazole is able to significantly improve the survival of animals infected with all strains tested, excellent survival rates (100%) were achieved in those mice infected with the strains having the lowest MICs (0.12 μg/ml). However, the efficacy of voriconazole was low in reducing the fungal load in brains of mice infected with the strains with MICs that were >1 μg/ml.

A possible limitation of the present study was the inclusion of only two isolates with MICs that were above the ECV, owing to the difficulty in finding isolates with higher MICs.

Our data on galactomannan serum levels showed that, in general, voriconazole worked well against all the strains of A. terreus tested, although in most cases the GMI was still above the threshold considered positive for invasive aspergillosis in mice (13). This could be due to the fact that voriconazole shows time-dependent fungicidal activity against some species of Aspergillus (15). In a previous study, galactomannan levels of mice infected with A. terreus and treated with posaconazole were negative at day 7 (16). Here, we only tested one dose of voriconazole, because a previous pharmacokinetic study demonstrated that 25 mg/kg produced plasma levels in mice higher than those produced in humans after administration of therapeutic doses (9).

In this study, voriconazole MICs less than or equal to the ECV were more predictive of in vivo success. Although only approximately 3% of clinical isolates have MICs greater than the ECV (17), this parameter might be important in predicting drug failure in clinical practice.