Combination of Amphotericin B with Flucytosine Is Active In Vitro against Flucytosine-Resistant Isolates of Cryptococcus neoformans

Abstract

The combination of flucytosine and amphotericin B was tested against 10 flucytosine-resistant isolates of Cryptococcus neoformans by checkerboard, killing curves, and Etest. Although differences were observed depending on the technique used, antagonism was never observed. The synergistic interaction was related to the mechanism of flucytosine resistance of the isolates.

The combination of amphotericin B (AMB) and flucytosine (FC) is the treatment of choice for cryptococcal meningitis (21). The higher efficacy of the combination compared to that of AMB alone has been demonstrated in human immunodeficiency virus-negative (2) and human immunodeficiency virus-positive (6, 27) patients. For infections with FC-resistant isolates, the combination might not add any effect (15) or could be deleterious. However, data to support this are scarce (12).

As several mechanisms of FC resistance are possible, one may wonder if synergy is achievable for all FC-resistant isolates. We thus used a panel of FC-resistant Cryptococcus neoformans clinical isolates and various techniques to evaluate the influence of the FC resistance mechanism on the in vitro interaction between AMB and FC.

Ten FC-resistant isolates were studied. Susceptibility to fluorouracil (FU) and FC alone was tested twice. Checkerboard studies of AMB plus FC were performed twice based on the CLSI (formerly NCCLS) M27-A2 methodology (19). All experiments included quality control strains. Endpoint definition is a critical issue for assessing checkerboard interactions (25). Thus, the same endpoint (50% inhibition) was used for both drugs alone and in combination. Fractional inhibitory concentration indices (FICI) were calculated (9), and interactions were defined according to current recommendations (1). Although different approaches could be used to analyze in vitro antimicrobial combinations, we chose the FICI model instead of a surface response modeling (11) due to its simplicity and its wide use.

Killing curve testing and result interpretation were done as described elsewhere previously (28). FC was tested at 64 μg/ml and AMB was tested at 0.125 μg/ml alone and in combination. For three isolates with various FU susceptibilities (isolate 550, FU susceptible; isolate 98.0673, FU intermediate; and isolate 533, FU resistant), additional AMB concentrations and time points were tested. Killing experiments for these three isolates were performed twice. Preliminary results showed that isolate 550 was less susceptible to AMB than isolates 533 and 98.0673. Similar antifungal activity was found for AMB at 0.19 μg/ml against isolate 550 and for AMB at 0.12 μg/ml against isolates 533 and 98.0673.

Etest studies were performed once as described previously (16, 28). Additional tests performed as described previously by White et al. (29) gave similar results.

The 3.2-kb PCR-amplified fragment (primers FCY2-5′ [5′-CCGAAGTGCCTTTTATTGCAGAG] and FCY2-3′ [3′-GATGCCTATGAACAACGGTGGT]) containing the whole KNH99 FCY2 gene was cloned into the NotI site of a plasmid containing a hygromycin resistance cassette (13). The resulting plasmid, pNE344, was SspI digested and used to transform C. neoformans by biolistic DNA delivery (26). Transformants were selected on hygromycin agar medium.

Susceptibilities to FC and AMB alone and in combination by checkerboard experiments are presented in Table 1. Using the same breakpoints for FU as those for FC, four isolates were considered to be resistant to FU (MICs ≥ 32 μg/ml), and two were considered to be intermediate (MICs = 8 to 16 μg/ml). Sixty percent of the interactions were synergistic. Antagonism was not observed. There was no strict relationship between synergism and FU susceptibility.

TABLE 1.

MICs of single drugs and interaction of AMB with FC evaluated by the checkerboard method against 10 FC-resistant C. neoformans isolates, according to their susceptibility to FU

Isolate	MIC (μg/ml) of drug alone			MICs (μg/ml) of drugs in combination		Lowest FICI in combination	Interactionb
	FC	FUa	AMB	FC	AMB
98.0176	64	0.008	0.5	0.5	0.12	0.258	SYN
18476	256	0.25	0.25	1	0.12	0.504	IND
285	32	4	1	0.25	0.5	0.508	IND
550	128	0.5	0.5	0.5	0.06	0.129	SYN
98.0673	64	16	1	16	0.25	0.500	SYN
99.1026	32	8	0.5	1	0.06	0.156	SYN
98.1121	256	256	0.5	0.25	0.5	1.001	IND
276	256	256	0.5	4	0.06	0.141	SYN
280	256	256	1	0.5	0.5	0.502	IND
533	256	256	0.5	0.5	0.12	0.252	SYN

^aFU was not tested in combination.

^bSYN, synergistic; IND, indifferent.

Figure 1A shows the killing curves for isolate 550. AMB at 0.19 μg/ml plus FC were significantly better than either drug alone and were synergistic. For isolate 98.0673, AMB at 0.125 μg/ml plus FC were significantly better than either drug alone (Fig. 1B), and the interaction was synergistic. For isolate 533, interactions were indifferent (Fig. 1C).

FIG. 1.

Time-kill study with AMB alone and in combination with FC on C. neoformans (A) isolate 550, (B) isolate 98.0673, and (C) isolate 533. Concentrations alone and in combination were 64 μg/ml for FC and 0.125 to 0.5 μg/ml for AMB. × (dashed line), control; + (solid line), FC; □ (solid line), AMB at 0.125 μg/ml; ▪ (dotted line), AMB at 0.125 μg/ml plus FC; ▵ (solid line), AMB 0.19 μg/ml; ▴ (dotted line), AMB at 0.19 μg/ml plus FC; ○ (solid line), AMB at 0.25 μg/ml; • (dotted line), AMB at 0.25 μg/ml plus FC; ⋄ (solid line), AMB at 0.5 μg/ml; ⧫ (dotted line), AMB at 0.5 μg/ml plus FC. The limit of detection was 1.3 log₁₀ CFU per milliliter. FC_R, FC resistant; FU_S, FU susceptible; FU_I, FU intermediate; FU_R, FU resistant.

Table 2 shows time-kill results for AMB at 0.125 μg/ml and FC alone or in combination for nine isolates. The combination exhibited synergy for four isolates and indifference for five isolates. The combination of AMB and FC against isolate 99.1026 (FU intermediate) was more active than the most active single drug, although the decrease of 1.8 log₁₀ CFU/ml did not comply with the synergy definition. Antagonism was never observed.

TABLE 2.

Log₁₀ CFU per milliliter after 72 h of incubation of FC at 64 μg/ml and AMB at 0.125μg/ml alone and in combination against FC-resistant isolates of C. neoformans evaluated by time-kill methodology^a

Isolate	Suscepti- bility		Log10 CFU/ml				Interaction
	FC	FU	Control	FC	AMB	FC + AMB
98.0176	R	S	7.8	7.4	6.5	3.7	SYN
18476	R	S	7.9	7.5	6.7	3.9	SYN
550	R	S	7.4	7.2	7.2b	4.7b	SYN
98.0673	R	I	7.5	7.4	7.2	4.6	SYN
99.1026	R	I	7.7	7.1	5.2	3.4	IND
98.1121	R	R	7.5	7.5	1.5	1.3	IND
276	R	R	7.3	7.3	7.3	7.2	IND
280	R	R	7.5	7.4	5.6	5.2	IND
533	R	R	7.6	7.5	7.4	6.5	IND

^aR, resistant; S, susceptible; I, intermediate; SYN, synergistic; IND, indifferent.

^bData presented are for a concentration of AMB of 0.19 μg/ml.

The interaction of AMB plus FC was evaluated by Etest for 10 FC-resistant and 5 FC-susceptible isolates (data not shown). FC MICs for all FC-resistant strains were >32 μg/ml. AMB MICs ranged from 0.094 to 0.5 μg/ml. All interactions were indifferent.

The genome of C. neoformans serotype D (17) was blasted with the FCY2 gene sequence of Saccharomyces cerevisiae, encoding the purine-cytosine permease (10). The closest homologue gene was named FCY2. The deduced sequence of 523 amino acids shared 68.5% similarity with the sequence of the known S. cerevisiae cytosine permease. Two isolates (FU-susceptible isolate 550 and FU-resistant isolate 280) were transformed in four independent experiments, and 10 transformants were obtained for each isolate.

Transformants of isolate 550 showed a significant decrease in the FC MIC (0.25 to 4 μg/ml). Transformant MICs of isolate 280 did not change. Cytosine permease enzymatic activity was not specifically measured. For the FU-susceptible isolate (isolate 550), transformation led to FC susceptibility, indicating permease activity after transformation.

In vitro primary FC resistance has been reported in up to 7% of C. neoformans isolates (5, 7, 20). It is thus of clinical interest to know whether AMB plus FC are possibly antagonistic against FC-resistant isolates.

The in vitro combination of AMB and FC against FC-resistant C. neoformans isolates has been evaluated for a few strains (12, 18, 23-25), and synergy and antagonism (12) have been reported. In mice infected with an FC-resistant isolate, it has been shown that AMB plus FC were more effective than monotherapies (22).Whether the FC resistance mechanism influences the potential synergy remains unknown.

FC uptake by yeasts occurs via cytosine permeases. FC is then deaminated by cytosine deaminases into FU. Although it is not known if FU enters the cell via a specific transporter in basidiomycetes, this is the case for Saccharomycetales (14). In the case of a cytosine permease or cytosine deaminase defect, cells are FC resistant but remain FU susceptible. Isolates that are FC resistant due to enzymatic defects downstream of the cytosine deaminase are also FU resistant. FC-resistant isolates used in this study included strains of various FU susceptibilities (3, 4), allowing us to assess if the interaction of AMB with FC is influenced by different genetic backgrounds.

As no standardized method for the testing of antifungal combinations in vitro exists, three different techniques were used. Indifference was observed with the Etest, possibly due to difficulties in testing C. neoformans with RPMI medium (8). Synergy by checkerboard was observed for FU-susceptible strains but also for some FU-resistant strains. In contrast, time-kill curves showed synergy only for strains that were not resistant to FU. The present results suggest that the synergy of AMB plus FC depends on the FC resistance mechanism.

These results could strengthen the current recommendation for induction treatment of cryptococcal meningitis by demonstrating that adding FC to AMB cannot harm and can even be beneficial even if the C. neoformans isolate is FC resistant.

Nucleotide sequence accession number.

The FCY2 sequence was deposited in GenBank under accession number DQ190470.