Search

The methyl-d(3) amide derivative of the polyene antibiotic amphotericin B was synthesized, assayed for biological activity, incorporated into mechanically aligned bilayers of dipalmitoylphosphatidylcholine (DPPC), and examined by deuterium and phosphorus NMR. The amide derivative has a lesser, but qualitatively similar, biological activity relative to amphotericin B. Incorporation of the amide derivative and ergosterol into aligned DPPC bilayers resulted in a single, stable bilayer phase, as shown by phosphorus NMR of the DPPC headgroups. Deuterium NMR spectra revealed one major (2)H quadrupolar splitting and one major (2)H-(1)H dipolar splitting in the liquid-crystalline phase, consistent with a high degree of alignment and a single, averaged physical state for amphotericin B methyl-d(3) amide in the bilayer. Variations of the quadrupolar and dipolar splittings as a function of macroscopic sample orientation and temperature indicated that the amide derivative undergoes fast rotation about a motional axis that is parallel to the bilayer normal.

Amphotericin B (AmB) with deoxycholate (Fungizone) and AmB incorporated into mixed micelles (AmB-mixMs) composed of egg lecithin with glycocholate, deoxycholate, or taurocholate were compared as treatments for murine infections. For mice infected with Candida albicans, treatment consisted of a single intravenous injection; for mice infected with Cryptococcus neoformans, treatment consisted of two intravenous injections. The maximal tolerated doses of AmB as Fungizone were 1.25 mg/kg of body weight in mice with candidiasis and 2.5 mg/kg of body weight in mice with cryptococcosis. The AmB-mixMs were nontoxic to mice at doses of 80 and 100 mg/kg of body weight and were therapeutically more active than the maximal tolerated dose of Fungizone in both models of infection. However, when Fungizone or AmB-mixMs were administered at equivalent doses of AmB, AmB-mixMs were more active in treating murine candidiasis, whereas Fungizone was more active in treating murine cryptococcosis.

The cellular activities of amphotericin B (AmB) used as Fungizone were compared with those of AmB complexed to either egg lecithin and glycocholate (Egam) or egg lecithin and deoxycholate (Edam). Under conditions in which leakage of K+ from erythrocytes and cultured L cells treated with Fungizone was almost complete, Egam and Edam containing concentrations of AmB severalfold greater than the concentration of AmB in Fungizone had no effect but retained the ability to decrease the level of retention of K+ in fungal cells. Analysis by absorption and circular dichroism spectroscopy demonstrated that when these formulations containing AmB at concentrations of less than 10(-5) M were added to buffer, the AmB dissociated slowly as monomers from Egam or Edam and dissociated rapidly as a mixture of monomers and self-associated species from Fungizone. We propose that in Egam and Edam, the absence of free AmB in the self-associated form reduces the toxicity of AmB to mammalian cells, whereas the presence of monomeric AmB results in the retention of the antifungal activities of these complexes.

The effects of four monoesters of sucrose with different acyl chain lengths (palmitate, C16; myristate, C14; laurate, C12; and caprate, C10) on the aggregation state of amphotericin B (AmB), its binding to cholesterol and ergosterol, its toxicity to cells, and its lethality to mice were determined. In solution, all four of these esters inhibited AmB binding to cholesterol more than to ergosterol; this effect correlated with the ester-induced shift from the mainly aggregated form of AmB to the mainly monomeric form. In experiments with cells, the esters inhibited the toxicity of AmB to mouse erythrocytes and cultured mouse fibroblast L-929 cells more than its toxicity to Candida albicans cells. When injected intravenously with AmB, these esters decreased AmB lethality to mice. In all of these assays, the ester with the shortest chain length (caprate) was much less potent than the other three esters. Our results indicate a correlation between in vitro and in vivo assays and suggest that the in vitro and in vivo selectivity of AmB may be enhanced by surface-active agents which modulate the aggregation state of AmB.

Mixed micelles prepared from egg lecithin and the sodium salt of glycocholic acid markedly inhibited amphotericin B toxicity to mammalian cells without significantly affecting the antifungal effects of the drug.

The antifungal activities of amphotericin B and two triazoles, Sch 39304 and fluconazole, were tested against Histoplasma capsulatum. In this study Sch 39304 compared favorably with amphotericin B in treating histoplasmosis in normal and leukopenic mice, whereas fluconazole was much less active. The differences in the efficacies of the triazoles appeared to be due to differences in their pharmacokinetics and the dosage schedule that was used. For amphotericin B there was a good correlation between in vitro and in vivo efficacy, but this was not true of the triazole derivatives. These results further demonstrate that, with the methods used in this study, in vitro susceptibility testing of triazoles may not be predictive of in vivo activity against isolates of H. capsulatum.

The activities of fluconazole and amphotericin B against Histoplasma capsulatum were investigated. The minimum inhibitory concentrations ranged from 0.12 to 0.47 microgram/mL for amphotericin B and from 16 to 250 micrograms/mL for fluconazole. Fluconazole given orally twice a day for 6 consecutive days compared favorably with amphotericin B given intraperitoneally once every other day for a total of six doses in the treatment of histoplasmosis in normal and leukopenic mice.

Fluconazole (UK-49,858) was compared with amphotericin B in treating histoplasmosis in female AKR mice immunosuppressed with either cyclophosphamide or cortisone. Both drugs protected animals from a lethal challenge with Histoplasma capsulatum, but neither regimen resulted in cures since viable organisms were cultured from spleens of survivors.

Lysis of human erythrocytes induced by amphotericin B was retarded when the oxygen tension of the incubation mixture was reduced or when the antioxidant catalase was added; lysis was accelerated when cells were preincubated with the prooxidant ascorbate. In the atmosphere of reduced oxygen tension, the erythrocytes containing carboxyhemoglobin lysed at a slower rate than did the cells containing oxyhemoglobin. Consistent with a role for oxidative damage in lysis, the mixture of erythrocytes and amphotericin B showed an increase in malonyldialdehyde, the product of peroxidation, which paralleled the progression of hemolysis. In contrast, the permeabilizing effect of amphotericin B, measured as a decrease in intracellular K+, was not affected by changes in oxygen tension, catalase, or ascorbate treatment. These results imply that oxidant damage is involved in the lytic, but not in the permeabilizing, action of amphotericin B.

Amphotericin B (AmB) and filipin effects on L and HeLa cells were compared by monitoring drug-induced potassium leakage from cells, changes in radioactive uridine incorporation into cellular ribonucleic acid, protein leakage from cells, and cell viability. L cells were much more susceptible to both AmB and filipin than were HeLa cells, but the overall dose response was similar. For AmB, the various effects were easily separable. Potassium leakage occurred at the lowest concentrations of AmB and was reversible. Inhibition of uridine incorporation and loss of viability occurred at intermediate levels, and protein loss occurred at higher levels. In contrast, filipin was much more potent; its effects on potassium leakage were only minimally reversible, and the separation of the permeabilizing effects from complete cell lysis was possible only over a limited concentration range and for a short time.

We have found that amphotericin B is unstable in two commonly employed fungal culture media. This instability leads to inaccuracies in determining the actual level of susceptibility of slow-growing strains that require prolonged incubation for growth. To help compensate for this problem, we have described two rapid methods of susceptibility testing.

Amphotericin B (AmB) treatment before infection with the bacterium Listeria monocytogenes prolonged survival of AKR mice but shortened survival of C57BL/6 mice compared with survival of untreated infected controls. C57BL/6 mice were also more sensitive to the acute toxic effects of AmB than AKR mice, as were (C57BL/6 X AKR)F1 hybrid mice. Spleen cells and erythrocytes (RBCs) from the C57BL/6 and the F1 hybrid mice were both more sensitive to the lytic and lethal effects of AmB than corresponding cells from AKR mice. Biochemical analysis indicated that catalase levels in RBCs from C57BL/6 and F1 hybrid mice were about 60% of those found in RBCs from AKR mice. The lysis by AmB of RBCs from all these strains of mice was inhibited by catalase or incubation in a low-oxygen environment. These findings suggest that (i) the low catalase levels in C57BL/6 and F1 hybrid mice may limit the protection of cells from the oxidant damage involved in AmB action, and (ii) the toxicity which occurs at low concentrations of AmB in the mouse strains with low intracellular catalase levels may interfere with or ablate the AmB-induced increases in mouse resistance to L. monocytogenes infection.

High concentrations of amphotericin B (AmB) killed mouse L cells, but low concentrations increased plating efficiency and stimulated the incorporation of labeled precursors into DNA and RNA. Thus, there were two disparate effects of AmB on L cells, stimulatory and toxic, and they occurred in distinct dose-related stages. AmB also affected the permeability of L cells. In dose-response studies, increases in cell membrane permeability, measured as the loss of K+ ions, occurred along with the stimulation of [3H]uridine incorporation into RNA. In contrast, stimulation of [3H]thymidine incorporation into DNA was only observed in cells recuperating from AmB-induced permeability changes. When the K+ concentration in the medium was lowered to 0.5 from 4.5 mM, or when 1 mM ouabain was added to the cultures, cell killing was potentiated, but the stimulatory and permeabilizing effects of subtoxic concentrations of AmB were unaffected. Furthermore, etruscomycin, a polyene antibiotic without any permeabilizing effects, nevertheless induced an enhancement of plating efficiency and of incorporation of [3H]uridine into RNA and [3H]thymidine into DNA. Our results suggest that the dose-related stimulatory, permeabilizing, and toxic effects of AmB most probably have distinct mechanisms of action and may be independent of one another.

Based on the enhanced fluorescence of amphotericin B in acid solutions, a quantitative assay for this polyene antibiotic has been developed that is sensitive and linear in the range of 0.1 to 10.0 muM. The binding of amphotericin B to Saccharomyces cerevisiae was assayed under various conditions as the amount bound to cells in a dialysis chamber or after centrifugation. Two types of binding were defined: weak, reversible binding occurred at 0 C or higher temperatures and even in the presence of inhibitors of energy metabolism, whereas strong, irreversible binding did not occur at 0 C and was inhibited when energy metabolism was blocked. Only strong binding was correlated with cell killing. Weak binding probably involves the outer layer of the membrane; strong binding probably requires disruption of hydrophobic regions of the cell membrane.

Among the polyene antibiotics, many, like filipin, cannot be used clinically because they are toxic; amphotericin B, however, is useful in therapy of human fungal infections because it is less toxic. Both the toxicity of filipin and the therapeutic value of amphotericin B can be rationalized at the cellular and molecular level by the following observations: (i) these polyene antibiotics showed differential effects on cells; filipin was more potent in lysing human red blood cells, whereas amphotericin B was more potent in inhibiting yeast cell growth; and (ii) the effects of filipin were more efficiently inhibited by added cholesterol, the major membrane sterol in human cells, whereas the effects of amphotericin B were more efficiently inhibited by ergosterol, the major membrane sterol in yeast. The simplest inference is that the toxicity and effectiveness of polyenes are determined by their relative avidities for the predominant sterol in cell membranes.

High-titered antisera specific for amphotericin B (AmB) were induced by immunization with a protein conjugate of the D-lysyl AmB methyl ester. These polyclonal anti-AmB antibodies reacted preferentially with AmB or the AmB methyl ester and discriminated sharply between nystatin and AmB. A solid-phase radioimmunoassay was developed with radioiodinated immunoglobulin G fractions derived from the anti-AmB antisera. This assay was capable of detecting AmB in the sera in the same concentration range that is regularly achieved during AmB treatment of systemic fungal infections. This study demonstrated the feasibility of immunoassays in measuring the concentration of AmB in blood and tissue fluids.

LY121019 and amphotericin B were equally active in vitro against most clinical isolates of Candida albicans and Candida tropicalis. Higher concentrations of LY121019 were required for inhibition of Candida glabrata. Other Candida species were inhibited by amphotericin B but not by LY121019.

A laboratory-derived mutant of Candida albicans B311 (L) and a clinical isolate (C) of C. albicans, both lacking membrane ergosterol, were less susceptible to amphotericin B (AmB)-induced cell membrane permeability to K+ and lethality than was the wild-type laboratory strain (B311) which contained ergosterol. The resistance of L and C to AmB-induced killing was much greater than the level of resistance to AmB-induced cell membrane permeability. L and C were also less susceptible to killing by H2O2 than was B311, and when treated with menadione, they each produced less H2O2 than did B311. In addition, their levels of catalase activity were 3.8-fold (L) and 2-fold (C) higher than that of B311. The ergosterol deficiency in L and C probably impaired AmB binding to the cells, thereby lowering AmB effectiveness as measured by both cell membrane permeability and killing. Resistance of strains L and C to oxidation-dependent damage likely contributed to a diminished response to AmB-induced lethality.

A patient had cutaneous protothecosis because of the alga-like organism, Prototheca wickerhamii. In vitro sensitivity tests showed that the organism was sensitive to amphotericin B, and was treated successfully with this polyene antibiotic. As with treatment of some fungal infections, a clinical response was achieved when therapy with low doses of amphotericin B was given during a short period of time. The basis of the amphotericin B response may have been due to a combination of its immunostimulatory and antibiotic properties.

Ascorbic acid enhanced the lethal but not the permeabilizing effects of amphotericin B on Candida albicans and Cryptococcus neoformans cells. Two other ene-diol acids, D-erythorbate and dihydroxyfumarate, also enhanced the lethal action of amphotericin B on Can. albicans. Maleic acid and gulanolactone, compounds structurally related to ascorbic acid but not containing the ene-diol group, had no such effect. It is assumed that ascorbic acid and the two other ene-diol acids acting as pro-oxidants augmented the oxidation-dependent killing of fungal cells induced by amphotericin B.

The effect of amphotericin B (AmB) treatment on the mononuclear phagocyte system of mice was investigated. Peritoneal macrophages from mice that received AmB treatment showed a higher phagocytic and antibacterial activity than those from normal untreated mice. When the levels of macrophage precursor cells in bone marrow and spleen were followed in mice after AmB treatment, an eightfold increase in the splenic content of limited stem cells for both macrophages and granulocytes (colony-forming units in culture) and a threefold increase in the number of pluripotent hemopoietic stem cells (colony-forming units in spleen) were observed on day 4. These were also accompanied by a slight increase in the colony-forming units in spleen and in culture in femoral marrows. AmB was capable of inducing a large number of peritoneal colony-forming cells in the peritoneum, and caused a significant rise in the serum level of colony-stimulating factor. No significant change in the level of blood monocytes was noted, although a transient increase in the proportion of neutrophils was observed within 24 h after AmB treatment.

The antifungal activity of UK 49,858, a difluorophenyl bis-triazole derivative, was evaluated in vitro against seven strains of Histoplasma capsulatum and in vivo in AKR and C57BL/6 murine models of histoplasmosis. UK 49,858 had a lower toxicity for AKR and C57BL/6 mice than amphotericin B did. The therapeutic index of UK 49,858 was 4.3 for AKR mice and 7.1 for C57BL/6; with amphotericin B it was 2 for both mouse strains. Given orally, UK 49,858 compared favorably with amphotericin B given intraperitoneally in either AKR or C57BL/6 mice infected with H. capsulatum.

Amphotericin B (AmB) binds to the cholesterol in lipoproteins, as determined by comigration in density gradient ultracentrifugation and changes in the circular dichroic spectrum. The saturation curve and Scatchard plots obtained with circular dichroism suggest that four to 10 cholesterol molecules in low-density lipoproteins bind to one molecule of AmB. AmB interacts more rapidly with low- and very-low-density lipoproteins than with high-density lipoproteins, but the circular dichroic spectrum of the complexed species is the same in all three cases. AmB also binds to other proteins in blood, but much higher concentrations of these proteins than of lipoproteins are needed for comparable binding. Interaction with lipoproteins stabilizes the antifungal activity of AmB. Interaction with lipoproteins and with much higher concentrations of other proteins in blood can also inhibit the effects of AmB on red blood cells, which contain cholesterol in their plasma membranes, but not the effects on Candida albicans, whose membranes contain ergosterol. An appropriate inference is that, when used clinically, AmB circulates in blood bound to lipoproteins and other proteins. The toxic and therapeutic effects of AmB in clinical situations are thus contingent on competitive interactions between sterol-containing cellular membranes of the host and the parasite and components of blood, such as lipoproteins and proteins.

Amphotericin B in combination with rifampicin was more effective in the treatment of murine histoplasmosis and blastomycosis than either agent alone. The increased therapeutic effects were not accompanied by an increase in toxicity.

N-D-Ornithyl amphotericin B methyl ester (N-D-ornithyl AmE) has a lower toxicity for animals than does amphotericin B (AmB), and peak serum levels can be achieved that are fourfold higher than those obtained with an equivalent dose of AmB. However, N-D-ornithyl AmE has one-fourth the in vitro activity and between one-fifth and one-eighth the in vivo activity of AmB. N-D-ornithyl AmE and the corresponding lysyl derivative also lack the immunoadjuvant effects of AmB.

We developed a rapid, sensitive, high-pressure liquid chromatographic (HPLC) procedure which incorporates a commercially available internal standard, 1-amino-4-nitronaphthalene, to measure amphotericin B in serum. Recovery was quantitative (greater than or equal to 90%), and the standard curve was linear from 0.04 to at least 10.0 micrograms/ml. The reproducibility of the assay was good, with intrarun coefficients of variation from 2.0 to 6.8% and interrun coefficients of variation from 4.9 to 10.0%. Comparison by linear regression analysis of the HPLC assay with an agar well diffusion bioassay gave a correlation coefficient of 0.942, with the HPLC assay exhibiting greater precision and sensitivity. No interference was encountered from over 20 drugs and three amphotericin B analogs. However, serum specimens that contained high concentrations of conjugated bilirubin (greater than 3 mg/dl) produced interfering peaks in both this assay and other previously reported HPLC assays for amphotericin B. We also describe a solid-phase extraction procedure which effectively removes this interference and uses an alternative internal standard (N-acetyl amphotericin B).

Two nitrosoureas, 1-(2-chloroethyl)-3-cyclohexyl-1-nitrosourea (CCNU) and 1,3-bis-(2-chloroethyl)-1-nitrosourea (BCNU), with strong carbamoylating and weak alkylating activities, interfered with the induction of potassium leakage and lethal action of amphotericin B (AmB) on Candida albicans. 2-Cyclohexyl isocyanate, the product of decomposition of CCNU, and 2-chloroethyl isocyanate, the product of decomposition of BCNU, also interfered with the anticandidal actions of AmB. In contrast, two nitrosoureas with weak carbamoylating and strong alkylating activities, 1-(2-chloroethyl)-3-(2,6-dioxo-3-piperydyl)-1-nitrosourea and 2-[3-(2-chloroethyl)-3-nitrosoureido]-D-glucopyranose, did not affect AmB action against C. albicans. These results indicate that the inhibitory action of CCNU and BCNU on the anticandidal effects of AmB is associated with the carbamoylating activity of these nitrosoureas.

Amphotericin B was infused into normal rabbits or rabbits made hypercholesterolemic by diet. There was no difference in amphotericin B-induced toxicity between these two groups. Amphotericin B given in a mixture with human low-density lipoproteins was more toxic than when given without lipoproteins.

We compared the in vitro antifungal action of amphotericin B (AmB) used alone or in combination with a second polyene antibiotic or with miconazole or ketoconazole. When AmB was used in combination with either filipin or Etruscomycin (Farmitalia, Milan, Italy), antagonism or potentiation of the antifungal effect against Candida albicans resulted. Addition of AmB to Etruscomycin- or filipin-treated cultures resulted in antagonism. In contrast, potentiation occurred when Etruscomycin or filipin was added to cultures treated with AmB. The outcome of incubating C. albicans with combinations of AmB and either miconazole or ketoconazole depended on the duration of exposure of the cells to the drugs. Short-term incubations resulted in antagonism, whereas potentiation of antifungal effects occurred after prolonged exposure of cells to the antibiotics. In addition, supplementation of cultures with serum protein-potentiated AmB induced k+ leakage at low protein concentrations and inhibited K+ leakage at high protein concentrations.

The dimorphic fungus Histoplasma capsulatum exists in two phases: a unicellular yeast form at 37 C and a mycelium at 25 C. We have found that these two phases have selective drug susceptibilities. The mycelial form of H. capsulatum was much more susceptible to the polyene antibiotic amphotericin B than the yeast form; in contrast, the yeast form was more susceptible to the antibiotic actinomycin D. The changes in susceptibility occurred early in the transition between the two phases and permitted the transitions to be blocked by sublethal concentrations of the appropriate drugs.

Systemic yeast infections are a major cause of morbidity and mortality in severely immunocompromised patients. The in vitro susceptibility to amphotericin B of 29 yeasts causing fungemia was examined in 26 patients undergoing allogeneic or autologous bone marrow transplantation and/or myelosuppressive chemotherapy. The minimal inhibitory concentrations (MICs) of amphotericin B observed with blood isolates from these patients were significantly higher than those observed with blood, sputum, or skin isolates from non-immunocompromised patients (p less than 0.01). All episodes (10 of 10) of bloodstream infection in immunocompromised patients caused by isolates with MICs greater than 0.8 micrograms/ml were fatal, versus eight of 17 episodes of bloodstream infection caused by yeasts with MICs of 0.8 micrograms/ml or less (p = 0.04). Although 15 of 26 patients received empiric treatment with amphotericin B before laboratory evidence of fungemia developed, the amphotericin B susceptibilities of their isolates were not significantly different from those of patients who had not received empiric amphotericin B treatment. It is concluded that yeast fungemia in severely immunocompromised patients is often caused by organisms resistant to the usual concentrations of amphotericin B obtainable in vivo, and that this finding is clinically significant.

The macrolide heptaene amphotericin B (AmB) induced concentration-dependent effects on Saccharomyces cerevisiae which were separable into two distinct stages. At low concentrations the drug inhibited the growth of the yeast and reversible changed cell permeability to Na+ and K+. At high levels it was lethal. The intracellular K+ concentration of cells with reversible damage (stage I) could be increased by addition of K+ to the medium, but cells irreversibly damaged (stage II) were not able to retain K+. The addition of K+ to the medium did not influence the growth-inhibitory or killing action of AmB. Addition of Mg2+ to cultures increased S. cerevisiae resistance to the killing effects of AmB. At low concentrations of AmB, growth inhibition was also decreased by extracellular Mg2+, but at higher concentration of AmB, growth inhibition was increased, probably because the prevention by Mg2+ of the lethal effect allowed expression of the inhibitory effect in a greater range. Simultaneous addition of K+ and Mg2+ markedly decreased both the inhibitory and lethal action of AmB at all concentrations. Filipin, a pentaene macrolide, had only lethal effects, which were unaffected when K+ was added to the medium but were diminished when medium was supplemented with Mg2+.

Suboptimal doses of amphotericin B in combination with either rifampin or 5-fluorocytosine were better than single-drug therapy in the treatment of disseminated Aspergillus fumigatus infection in mice. Despite the increased effectiveness of combination therapy, none of the therapeutic regimens we used completely eradicated infections in the mice when evaluated by mycological culture, even in long-term survivors.

Amphotericin B in combination with 5-fluorocytosine was synergistic against three clinical isolates of Aspergillus fumigatus and one of three clinical isolates of A. flavus. Amphotericin B in combination with rifampin was synergistic against all six clinical isolates of Aspergillus tested. The levels of 5-fluorocytosine and rifampin required for synergism were higher than clinically achievable concentrations when measurements of synergism were based on visual turbidity; but when the effects of the drugs were measured by inhibition of ribonucleic acid synthesis or dry-weight increase, much lower concentrations were effective.

Several different methods of performing susceptibility tests on six clinical isolates of Aspergillus are described. Some of the conditions that affected the level of susceptibility to drugs were: the type of media used, temperature and time of incubation, and the initial inoculum size. For amphotericin B susceptibility testing, the effectiveness of the polyene antibiotic as measured by visual growth was equivalent to the effectiveness as measured by inhibition of ribonucleic acid synthesis and dry-weight increase. For 5-fluorocytosine and rifampin, the visual-turbidity method gave minimum inhibitory concentrations that were much higher than those determined by effects on ribonucleic acid synthesis and dry weight. The reason for these discrepancies in susceptibility testing with 5-fluorocytosine and rifampin are unknown. We conclude that the most relevant test of this fungus to antifungal agents will have to be determined by the correlation of in vitro data with animal experiments and clinical results.

In a survey of 350 laboratories, 41 of 210 respondents indicated that they performed antifungal susceptibility tests. Two-thirds performed 20 or fewer tests per year, and most used a broth dilution method to test amphotericin B and flucytosine activity against Candida albicans. The broth dilution procedure of S. Shadomy and A. Espinel-Ingroff (p. 647-653, in E.H. Lennette, ed., Manual of Clinical Microbiology, 3rd ed., 1980) was the method most frequently cited, and therefore this method was used to test the susceptibility of five isolates of C. albicans and one of Saccharomyces cerevisiae to amphotericin B and flucytosine in seven research laboratories. Agreement among replicates performed on the same day by each laboratory was excellent for both drugs, all values being within 1 twofold drug dilution. Precision from week to week for each laboratory was also good, with 95 and 92% of values being within 1 drug dilution for amphotericin B and flucytosine, respectively. Interlaboratory precision, however, was poor. For amphotericin B, values varied 8- to 32-fold, and for flucytosine, they varied 32- to greater than 512-fold. We conclude that antifungal susceptibility testing is currently being performed in small volumes by numerous laboratories in the United States and that results from one laboratory may not agree with results from another. Improved standardization of fungal susceptibility tests is necessary before their results can be generally applied to clinical situations.

Amphotericin B was found to potentiate the antifungal effects of mycophenolic acid glucuronide, tetracycline, and actinomycin D by increasing the penetration of these antibiotics through the fungal cytoplasmic membrane. Because the fungi were over 100 times more susceptible than animal cells to these effects of amphotericin B, these observations might have clinical application.