Voriconazole susceptibility of yeasts isolated from the mouths of patients with advanced cancer

Patients.

Yeast isolates were available from the mouths of 199 patients who were receiving palliative care for advanced cancer. These isolates had been collected and characterized during two previous studies of fungal carriage among randomly selected patients receiving palliative care in the UK (Bagg et al., 2003; Davies et al., 2002).

Whilst all of these patients had previously received the standard forms of treatment appropriate to their particular malignancies, for example surgery, radiotherapy and chemotherapy, such modalities were no longer being pursued with a view to curing their disease. They were receiving, instead, symptom control. Accordingly, all of the patients were under the care of one of three hospices, either as in-patients or day-stay patients, and all were suffering from very advanced disease with limited life expectancies. The most common diagnoses were carcinomas of the breast, bronchus, prostate and large bowel. This large group of patients was therefore demographically representative of those receiving treatment in a palliative medicine setting.

Whilst it would have been interesting to collect detailed information on previous antifungal drug use by each patient, this was not possible. The patients had been treated by a wide range of both primary-care and secondary-care health professionals in many settings during the course of their illness and the reliability of data that could be collected was poor.

Yeast isolates.

A collection of 270 clinical isolates of oral yeasts was available. These yeasts had been isolated from oral swabs and oral rinses of the 199 patients described above. The yeast isolates were identified by germ-tube tests and API 20C AUX and API ID 32C (bioMérieux) profiles. Isolates with indeterminate results were further investigated using species-specific PCR and 26S rRNA gene sequencing. All germ-tube-positive isolates were screened by PCR using Candida dubliniensis-specific and fungal universal primers (Donnelly et al., 1999).

Fluconazole and itraconazole susceptibility testing.

Susceptibility to fluconazole and itraconazole was determined for all isolates by means of a broth micro-dilution assay as described in the NCCLS standard M27A (NCCLS, 1997). For fluconazole, isolates with an MIC of ⩽ 8 μg ml⁻¹ were classified as susceptible, those with an MIC of 16–32 μg ml⁻¹ as susceptible dose-dependent and those with an MIC of ⩾ 64 μg ml⁻¹ as resistant. For itraconazole, isolates with an MIC of ⩽ 0.125 μg ml⁻¹ were classified as susceptible, those with an MIC of 0.25–0.5 μg ml⁻¹ as susceptible dose-dependent and those with an MIC of ⩾ 1 μg ml⁻¹ as resistant. In all cases, if results fell between categories, then the higher category applied (NCCLS, 1997).

Voriconazole susceptibility testing.

An adapted NCCLS method was used as described by Pelletier et al. (2002). Pfizer Global Research & Development kindly supplied voriconazole (UK-109496) as a 99.2 % pure powder. Stock solution (1600 μg ml⁻¹) was prepared by dissolving the powder in DMSO. A serial two-fold intermediary dilution was prepared by adding DMSO and RPMI 1640 to stock solution according to NCCLS M27-A procedures (NCCLS, 1997). After reconstitution with Candida suspension, the voriconazole concentrations ranged from 0.016 μg ml⁻¹ to 8 μg ml⁻¹, with a DMSO concentration of exactly 1 % in each well, including the growth control. The assays were performed in flat-bottomed micro-dilution plates. Individual Candida suspensions were prepared according to NCCLS M27-A methods. After incubation for 48 h at 35 °C, plates were read in a Dynatech MR 5000 microplate reader set at 405 nm. The MIC was defined as the lowest concentration of the drug that gave a 50 % reduction in optical density when compared with the turbidity of the growth control well (Rex et al., 2001).

Yeast isolates

The identities of the 270 isolates studied are shown in Table 1. There were a significant number of isolates of non-C. albicans yeasts, reflecting the heterogeneous oral mycological flora in patients with advanced cancer.

Fluconazole, itraconazole and voriconazole susceptibilities

The comparative in vitro susceptibilities of the yeast isolates to the three antifungal agents are shown in Table 2. Overall, 206 (76 %) of the isolates were susceptible to fluconazole (MIC ⩽ 8 μg ml⁻¹). Twenty-five isolates were susceptible dose-dependent and 39 isolates (14 %) were fully resistant to fluconazole (MIC ⩾ 64 μg ml⁻¹). The comparative in vitro susceptibilities of the 64 isolates with a fluconazole MIC of > 8 μg ml⁻¹ are shown in Table 3; seven were susceptible to itraconazole, 16 were susceptible dose-dependent and 41 were itraconazole resistant.

Overall, 160 (59 %) of the isolates were susceptible to itraconazole (MIC < 0.125 μg ml⁻¹). Sixty-one isolates were susceptible dose-dependent and 49 isolates (18 %) were fully resistant to itraconazole (MIC ⩾ 1 μg ml⁻¹). Of the 49 isolates that were resistant to itraconazole, 41 were also resistant to fluconazole and eight were susceptible.

The distribution of the MICs to the three antifungal drugs for the 206 fluconazole-sensitive strains is illustrated in Fig. 1(a). The median values (and interquartile ranges) for fluconazole, itraconazole and voriconazole, respectively, were 0.5 μg ml⁻¹ (0.25–2 μg ml⁻¹), 0.125 μg ml⁻¹ (0.06–0.25 μg ml⁻¹) and 0.016 μg ml⁻¹ (0.016–0.125 μg ml⁻¹). Fig. 1(b) shows the distribution of the MICs to the three antifungal drugs for the 64 fluconazole-resistant strains. The median values and interquartile ranges for fluconazole, itraconazole and voriconazole respectively were 64 μg ml⁻¹ (32–64 μg ml⁻¹), 1 μg ml⁻¹ (0.5–8 μg ml⁻¹) and 1 μg ml⁻¹ (0.125–8 μg ml⁻¹).

In total, 41 (15 %) of the isolates were fully resistant to both fluconazole and itraconazole. Of these, 23 were C. glabrata, 12 were C. albicans, four were Saccharomyces cerevisiae, one was Candida parapsilosis and one was Candida tropicalis. The MICs of voriconazole for these isolates are shown in Fig. 2.

Antifungals in oral fungal infections

Antifungal drugs are essential tools in the management of oral candidosis among debilitated patients. Fluconazole has been widely used to good effect. It is found in saliva at concentrations comparable to those in blood after single or multiple doses (Debruyne, 1997). Itraconazole, both as capsules and as the cyclodextrin solution, has also been shown to be effective in the treatment of oral candidosis (Cross et al., 2000) and the oral solution generates effective levels in both plasma and saliva (Reynes et al., 1997). However, the increasing incidence of colonization and infection with non-C. albicans yeasts displaying reduced susceptibility to the first generation azoles (fluconazole and itraconazole) has raised concerns for the future (Johnson et al., 1995). In the present study of a group of patients who we believe to be representative of those receiving palliative care for advanced cancer, 24 % of the isolates were not susceptible to fluconazole at standard doses. Consideration of the possible role of newer antifungal drugs in this group of patients is therefore important.

The data presented in this paper indicate that voriconazole is more potent than either fluconazole or itraconazole against yeasts isolated from the mouths of patients with advanced cancer. This finding is in keeping with published work on yeasts from a range of other patient groups and body sites (Cuenca-Estrella et al., 1999; Pelletier et al., 2002; Laverdiere et al., 2002). Of particular relevance is the voriconazole susceptibility of the isolates that were resistant to fluconazole and itraconazole. Many of these strains displayed low MICs to voriconazole, confirming reports by others that this agent is of value in treating some infections involving fluconazole-resistant Candida species (Pelletier et al., 2002; Laverdiere et al., 2002). However, the MICs of voriconazole for fluconazole-resistant isolates were higher than those for fluconazole-susceptible isolates, as reported by others (Cuenca-Estrella et al., 1999).

Whilst no interpretive breakpoints have yet been agreed for voriconazole, the pharmacokinetics of the drug suggest that adequate serum levels are available to treat infecting organisms with MICs of ⩽ 2 μg ml⁻¹ (Pelletier et al., 2002). Forty-four of the fluconazole-resistant strains had voriconazole MICs of ⩽ 2 μg ml⁻¹, but 18 of the remaining 20 strains had MICs of 8 μg ml⁻¹. The pharmacokinetics of voriconazole appear suitable for the management of oral candidosis. A recent study showed that the pharmacokinetic profiles for saliva followed a pattern similar to those observed for plasma, with a highly significant correlation between plasma and saliva voriconazole concentrations (Purkins et al., 2002).

C. glabrata in oral fungal infections

The frequent isolation of C. glabrata is worthy of note. Of the 51 isolates of C. glabrata, 34 were fluconazole-resistant, with correspondingly high MICs to voriconazole. C. glabrata therefore constituted 53 % of the 64 fluconazole-resistant yeasts isolated. Although the numbers were small, one other group has reported a relatively lower success rate for voriconazole treatment of C. glabrata infections (Perfect et al., 2003). These data indicate the value of accurate identification of isolates to species level, ideally accompanied by antifungal susceptibility testing, when managing florid oral fungal infections in those with advanced disease.

Biofilms in oral fungal infections

One further issue relevant to the management of oral fungal infections is that the yeasts are present in the form of a biofilm. It is well recognized that antimicrobial agents are often far less effective against sessile, as opposed to planktonic, organisms (Ramage & López-Ribot, 2004). These concepts are now being considered in relation to antifungal drug susceptibility testing (Ramage et al., 2001) and are to be examined in more detail in future work by our group.

Conclusions

In summary, 75 % of the oral isolates were susceptible to fluconazole and this drug is likely to continue as the mainstay of treatment for oral fungal infections in the terminally ill for the foreseeable future. A significant proportion of yeasts showed reduced susceptibility to fluconazole and itraconazole, many of which were inhibited by low concentrations of voriconazole. By contrast, some of the fluconazole-resistant isolates, in particular C. glabrata, demonstrated reduced susceptibility to voriconazole. These data suggest a potential role for voriconazole in management of some refractory oral fungal infections in those with advanced cancer. However, more comprehensive studies of voriconazole are required to determine the correlation between in vitro activity and clinical outcome in vivo.