Toxicity of LY303366, an Echinocandin Antifungal, in Mice Pretreated with Glucocorticoids

Abstract

LY303366 is a semisynthetic derivative of the echinocandin class. During preclinical studies, lethal toxicity was observed in DBA/2 mice pretreated with a cortisone acetate dose followed by treatment with LY303366 at doses ranging from 12.5 to 50 mg/kg of body weight/day given intraperitoneally (i.p.). In the cortisone-treated, uninfected controls, 90% given LY303366 at 50 mg/kg died. Deaths occurred only in steroid-treated mice. In additional experiments, uninfected DBA/2 and CD-1 mice were pretreated with different glucocorticoids. Dosages were adjusted for comparative potency with cortisone and were given at one, two, or five times the equivalent cortisone dosage of 5 mg prior to treatment with LY303366 at 25 mg/kg/day given i.p. Lethal toxicity occurred in DBA/2 mice given hydrocortisone (1× or 2×), triamcinolone (1× or 5×), and cortisone. However, no mice pretreated with 1× or 5× dexamethasone died. In CD-1 mice, deaths occurred only in those given 5× triamcinolone; three of five died 2 days after the cessation of 10 days of LY303366 treatment. The causes of the deaths and why inbred DBA/2 mice are more sensitive than outbred CD-1 mice to the combined lethal effects of LY303366 and some glucocorticoids could not be determined histologically and remain unexplained. This is the first report of this toxicity of combination glucocorticoids and LY303366. Whether a similar toxicity might apply to the other compounds in the echinocandin class of antifungals and the species specificity require additional study. In addition, the clinical relevance of these observations in steroid-treated patients to the clinical safety of LY303366 and other echinocandins needs to be determined.

LY303366 (LY) is a semisynthetic lipopeptide of the echinocandin class of antifungal compounds. These compounds are noncompetitive inhibitors of 1,3-β-glucan synthesis. The compound has been reported to have both in vitro and in vivo activity against various fungi, but it has primary activity against Candida spp. and Aspergillus (1–3, 9–12). In vivo, LY has been reported to have efficacy against pulmonary aspergillosis and invasive aspergillosis in neutropenic rabbits and mice, respectively (8, 11).

Data published thus far, largely in abstract form, indicates that LY is safe for animals and humans. In one report, it was noted that rabbits treated with a dosage of 20 mg/kg of body weight/day had marked pulmonary edema not typical in the model and that survival was decreased, compared to the control animals (8). None of these reports have examined LY safety in the presence of steroid therapy. During the course of studies on the efficacy of LY for the treatment of pulmonary aspergillosis, we observed a deleterious interaction in mice pretreated with cortisone acetate followed by intraperitoneal (i.p.) treatment with LY. Because of the observed toxicity, we were unable to evaluate completely the therapeutic efficacy of LY in our model of pulmonary aspergillosis, which is studied in glucocorticoid-suppressed mice. The studies reported herein detail our findings on this observation of combination toxicity between glucocorticoids and LY.

MATERIALS AND METHODS

Preliminary studies (experiment 1).

The initial observations of possible toxicity were made during the course of studies in which a murine model of pulmonary aspergillosis with DBA/2 mice (Taconic, Germantown, N.Y.) was established to assess the efficacy of LY303366 (Eli Lilly, Indianapolis, Ind.) as a therapeutic agent. Male DBA/2 mice (8 weeks old) were given 5 mg of cortisone acetate (Cortone acetate; Merck Sharp and Dohme) subcutaneously (s.c.) 1 day prior to infection. Mice were infected intranasally with Aspergillus fumigatus isolate 10AF while under methoxyflurane anesthesia.

Therapy was initiated 1 day later for groups of 10 mice with LY diluents or 3.125, 6.25, 25, or 50 mg/kg/day given once a day (QD) i.p. (some groups were treated with amphotericin B or itraconazole). Mice that received no antifungal drug therapy served as untreated controls. Steroid-treated but uninfected mice were also treated for 13 days with LY at 50 mg/kg. All therapies were given on days 1 through 13 postinfection.

Because of the observations made in the initial experiment, a second small study (experiment 1A) was done with uninfected mice. Five mice (DBA/2) were given 5 mg of cortisone and five mice were not given cortisone. Treatment with LY at 50 mg/kg began 1 day later and was given for 10 days.

Experiment 2.

A dose escalation study followed, in which LY was administered to uninfected animals that either had received a single 5-mg dose of cortisone or had not been pretreated with cortisone. On day 0, 30 uninfected DBA/2 (8-week-old male) mice received 5 mg of cortisone acetate (Merck) s.c. One day later, therapy with LY began. There were three groups of 10 given cortisone, and three groups were given no cortisone. The groups were treated with either LY diluent (appropriate for the 50-mg/kg dose), LY at 25 mg/kg, or LY at 50 mg/kg. Treatments were given i.p. for 10 consecutive days. The diluent formulation for the LY used in this experiment and all subsequent experiments was a polysorbate 80 formulation, different from that used in experiments 1 and 1A, which was a micronized formulation of LY. The formulation change was made by the supplier of the drug, Eli Lilly & Co.

Experiment 3.

On the basis of the results obtained in experiment 2, in which animals given lower doses of LY showed no signs of toxicity, a second model of pulmonary aspergillosis was established in male DBA/2 mice as described above. The various treatments included LY given i.p. daily at 12.5 mg/kg. Both infected and uninfected mice received this dosage.

Glucocorticoid specificity of the interaction between glucocorticoid and LY, (experiment 4).

Forty 8-week-old female DBA/2 mice (Taconic) (mean weight of 18.9 g) were used. Uninfected animals were randomized into groups of five mice and given either no steroid treatment or 5 mg of cortisone acetate (Merck) s.c., hydrocortisone acetate (Steris Laboratories, Inc.) at 5 or 10 mg s.c., triamcinolone acetonide (Steris) at 1 or 5 mg s.c., or dexamethasone (Dexaject; Vetus Animal Health) at 0.17 or 0.85 mg s.c. Doses of other steroids were determined on the basis of relative cortisone activity with a 5-mg dose of cortisone being equal to 1 (7). Thus, hydrocortisone was given as 1× and 2× equivalents, and triamcinolone and dexamethasone were given as 1× and 5× equivalents. One group received no steroid treatment and served as a control. Two days after steroid treatment, all mice were started on LY at 25 mg/kg given i.p. once daily for 10 consecutive days. All mice were monitored daily for clinical signs and for deaths.

Surviving animals were euthanatized and necropsied, and tissues were sampled for histological analysis 1 day after the cessation of therapy. Tissues were placed in 10% buffered formalin and embedded in paraffin, and sections were stained with hematoxylin and eosin.

Specificity of the toxicity by mouse strain (experiment 5).

To determine whether the apparent deleterious drug interaction between LY and glucocorticoids was mouse strain specific, an experiment was done with 6-week-old female CD-1 mice (mean weight, 27 g) (Charles River Laboratories, Portage, Mich.). Mice were randomized into groups of five or six mice. The steroid treatment groups were the same as those described for experiment 4, as was the LY treatment (25 mg/kg given i.p. daily for 10 days).

RESULTS

Initial observations.

In our initial study on the efficacy of LY against pulmonary aspergillosis, some dose regimens of LY for infected mice seemed to be toxic. These toxic regimens consisted of LY at 25 or 50 mg/kg/day. No infected animals receiving LY at 50 mg/kg/day survived the treatment period, with deaths attributed to toxicity beginning on day 2. Lethality, presumed due to toxicity, was observed for LY at 25 mg/kg/day, with the death of four animals occurring prior to any deaths in control groups. Lethal toxicity was also observed in uninfected mice given LY at 50 mg/kg/day (Table 1).

TABLE 1.

Survival of DBA/2 mice treated with LY (experiment 1)

Treatment group	Infection	No. surviving/ total	Days of deatha
Untreated controls	Yes	6/10	5, 7, 8, 10
LY diluentb	Yes	4/10	6, 6, 8, 11, 12, 12
LY at 25 mg/kg, i.p.	Yes	3/10	2, 2, 3, 3, 3, 5, 6
LY at 50 mg/kg, i.p.	Yes	0/10	2, 2, 2, 2, 4, 5, 7, 7, 7, 9
LY at 50 mg/kg, i.p.	No	1/10	1, 2, 2, 2, 2, 3, 3, 9, 13

^aTreatment was initiated on day 1. This was 1 day after infection for those animals in the efficacy study and 2 days after cortisone treatment. 

^bThe diluent was that of the micronized formulation of LY. 

The deaths of infected mice given LY at 50 or 25 mg/kg/day were thought to be a result of toxicity, since the uninfected animals being treated with LY at 50 mg/kg/day also died (50% mortality by day 3, 90% mortality overall) and because mortality in the infected groups being treated with 50 or 25 mg/kg/day was greater than that in untreated controls. Because no deaths occurred in the infected LY diluent control group before day 6, we do not feel that the diluent carrier is the cause.

Experiment 1A.

The aim of experiment 1A was to determine whether the apparent toxicity observed in the initial study was due to a drug interaction of cortisone and LY by using uninfected animals. All five of the nonsuppressed animals survived through the 10 days of treatment, whereas four of the five cortisone-treated mice died before the end of therapy with LY. Thus, these results were indicative of a toxic drug interaction between LY and cortisone.

Experiment 2.

In experiment 2, we sought to determine whether the results obtained in experiment 1A were related to the dosage of LY administered. Again, no mice in the groups not receiving cortisone died. In the cortisone-treated groups, no LY diluent-treated mice died, one mouse in the 25-mg/kg/day group died on day 4, and three mice in the 50-mg/kg/day group died, one each on days 4, 8, and 10.

Experiment 3.

In the second model of pulmonary aspergillosis, some signs and symptoms of toxicity in both infected and uninfected LY-treated mice were again evident. The primary signs were ruffled fur, lethargy, and respiratory distress, similar to those in the previous experiment. Even though the highest dosage of LY had been lowered to 12.5 mg/kg/day, there was 30 to 40% mortality between days 2 and 4 of infection (2 to 4 doses) of infected and uninfected mice treated with 12.5 mg/kg/day (Table 2).

TABLE 2.

Survival of DBA/2 mice treated with LY (experiment 3)

Treatment group	Infection	No. surviving/ total	Days of deatha
Untreated controls	Yes	7/10	7, 9, 14
LY diluentb	Yes	3/10	3, 6, 6, 7, 7, 7, 8
LY at 12.5 mg/kg, i.p.	Yes	5/10	2, 3, 3, 4, 12
LY at 12.5 mg/kg, i.p.	No	6/10	2, 2, 3, 4

^aTreatment was initiated 1 day postinfection, which was 2 days after cortisone treatment 

^bThe diluent was that of the polysorbate 80 formulation of LY. 

Summary of signs and symptoms of toxic events in the LY-treated DBA/2 mice.

The primary signs of toxicity were ruffled fur, lethargy, respiratory distress, and a possible partial paresis of the hind limbs in both infected and uninfected animals. In conjunction with this was the observation of hemorrhage occurring in the nasal area of several, but not all, animals. Whether hemorrhage was from the nasal mucosa or was pulmonary in origin was not determined. Necropsies were performed on mice that died as an apparent consequence of the treatment, as well as on euthanatized mice. Upon necropsy, little to no gross tissue damage was observed in animals that died during the experiment, nor were any gross signs of damage noted in animals that were euthanatized at the end of an experiment. Samples of lung, liver, kidney, brain, and spleen (in some cases, pancreas and adrenal) were examined histologically by a pathologist (R.A.S.) blinded to the clinical outcome of the treatments. Some abscesses in the lungs of the infected mice were observed and considered to be an expected consequence of the experimental infection, but no other histological abnormalities were consistently identified in any mice.

Experiment 4.

Because of the results obtained in the earlier experiments, we chose to examine the specificity of the interaction between glucocorticoids and LY by administering a single dose of LY to uninfected mice that had been pretreated with either a 1× or 5× cortisone equivalent dose of a different glucocorticoid. Clinical observation showed that, as early as the second day of dosing, mice in the cortisone-LY group and the 5× triamcinolone–LY group showed signs of toxicity. Two mice in the cortisone-LY group appeared very ill, and one animal had reddish urine (possibly hemoglobinuria, hematuria, or myoglobinuria). Similarly, four mice in the 5× triamcinolone–LY group appeared extremely ill. Animals in the hydrocortisone-LY groups also appeared ill. The general observations of these animals, as well as of any animal that died during the study, were lethargy, ruffled fur, labored respiration for about 1 day prior to death, ataxia and immobility, and possible dehydration. In contrast, animals in the LY control group and the dexamethasone-LY groups showed no overt clinical signs of illness. The number of animals dying in each group and the days of death are summarized in Table 3.

TABLE 3.

Survival of uninfected DBA/2 mice pretreated with glucocorticoids, followed by LY at 25 mg/kg given i.p. (experiment 4)

Treatment group	No. surviving/total	Days of death
LY control (no steroid)	5/5
Cortisone	1/5	3, 3, 3, 3
Hydrocortisone (1×)	1/5	3, 3, 3, 7
Hydrocortisone (2×)	2/5	3, 4, 7
Triamcinolone (1×)	2/5	3, 3, 10
Triamcinolone (5×)	0/5	3, 3, 3, 3, 10
Dexamethasone (1×)	5/5
Dexamethasone (5×)	5/5

General observations of necropsies were that both LY control- and dexamethasone-LY-treated animals had normal-appearing tissues. It should be noted that some spleen atrophy was observed in the dexamethasone-LY-treated animals. Animals from the other steroid-LY treatment groups all showed atrophied spleens, as would be expected, slightly smaller kidneys than normal, and pallor of the lungs. The lone surviving animal from the cortisone group had overt kidney damage. The histological analyses are summarized in Table 4.

TABLE 4.

Histological observations on tissues from DBA/2 mice from experiment 4

Treatment with LY at 25 mg/kg given i.p., plus:	Tissue observations
LY control (no steroid)	Normal brain, liver, kidney, spleen, and lung
Cortisone	Two acute infarcts in kidney with little associated inflammation, apparent heavy intravascular calcification associated with acute necrosis, embolic calcific material in an arteriole; other tissues normal; calcific material in vessel lumens not definite
Hydrocortisone (1×)	Small embolic crystalline material in the lungs with no surrounding inflammation, some dilation of the vessels noted because of the embolic material; other tissues normal
Hydrocortisone (2×)	Vacuolation (perhaps fatty degeneration) of hepatocytes; blue granular calcified material near a piece of the myocardium; otherwise normal
Triamcinolone (1×)	Mouse 1, small embolic crystalline material in the lungs with no surrounding inflammation, some dilation of the vessels noted because of the embolic material, other tissues normal; mouse 2, some lung congestion, otherwise normal
Triamcinolone (5×)	No mice surviving
Dexamethasone (1×)	Normal brain, liver, kidney, spleen, and lung
Dexamethasone (5×)	Normal brain, liver, kidney, spleen, and lung

Experiment 5.

In experiment 5, we assessed whether the apparent deleterious drug interaction was a result of the strain of mouse being used and chose to examine this possibility by repeating experiment 4 with CD-1 mice rather than DBA/2 mice. Clinical observations indicated that few, if any, animals appeared overtly ill or distressed during the course of this study. Some ruffling of the fur was noted in the LY–5× triamcinolone group. No mice died during the 10 days of treatment in this study. However, three animals in the LY–5× triamcinolone group died on day 12, 2 days after the cessation of therapy. Necropsies of animals from each group showed no overt tissue abnormalities, with the exception of the expected spleen atrophy. Histological analyses of various tissues taken 3 days after the cessation of LY treatment showed no remarkable pathological processes.

DISCUSSION

The results of our studies demonstrate that a deleterious drug interaction occurs between glucocorticoids and LY. This interaction was found to occur in both DBA/2 and CD-1 mice, but it was particularly severe in the DBA/2 mice. It should be noted that this observation has been made for both male and female mice. Although the nature of the interaction is unclear at this point, we found it to be reproducible. The lethal toxicity was dose dependent with respect to LY (deaths occurred at doses of 12.5 mg/kg/day or higher). In addition, the interaction was not specific to only cortisone and LY but also occurred after pretreatment with hydrocortisone or triamcinolone. However, no interaction was suggested when the animals were pretreated with dexamethasone; those animals showed no overt signs of toxicity. The toxicity was observed in DBA/2 mice treated with LY in two different formulations, and thus the toxicity appears related to LY and diluent independent.

We also addressed the possibility that this phenomenon was mouse strain specific. Our results from the study using CD-1 mice indicate that the toxicity is reduced in these animals, since deaths occurred only in a single group (i.e., the 5× triamcinolone group, with three of five dying). Even though the deaths occurred 2 days after the cessation of therapy, they were attributed to toxicity. Thus, the toxic interactions do not appear to be limited to one strain of mice. However, the acuteness and severity of the toxicity appear much greater in the DBA/2 mice.

Histology revealed no clear cause for the deaths of the animals. In DBA/2 mice, a calcific embolic material was noted in the tissue sections, particularly in the kidneys. These findings may or may not be important, as there are reports in the literature of spontaneous calcification in the heart in this strain of mice (4, 6). However, in our study only those survivors from groups given a steroid and LY in which lethal toxicity occurred showed these pathological abnormalities. Moreover, the most marked pathological findings may have been in the animals that did not survive the treatment and thus died before histopathological examination could be performed. Mice from the dexamethasone-treated groups and the LY control group had normal tissues. No remarkable histological findings were found in the tissues taken from the CD-1 mice in experiment 5.

Data published thus far, largely in abstract form, indicates that LY is safe for animals and humans. In a rabbit model of aspergillosis, pulmonary edema was attributed to LY rather than to infection (8). In addition, those authors found reduced survivorship at the highest dosage of LY (20 mg/kg/day) used in the study. No other reports of toxicity with LY have been uncovered through literature searches. Whether the toxicity we observed in the glucocorticoid-treated mice is a property of only LY or might also occur with other compounds in the echinocandin class is unknown at this time. It would be of interest and possibly very important to address this question with those compounds currently in development.

The relevance of our finding to the clinical use of this drug in humans remains to be determined. This is especially true for steroid-treated patients that have been on long-term suppressive therapy and have a greater proclivity toward developing fungal infections, such as aspergillosis or candidosis. The differences in steroid dosing done with patients versus those used in our studies with mice would need to be addressed. Although the steroid dose is high on a milligrams-per-kilogram basis compared to doses used for humans, species dose scaling is problematic because of intrinsic interspecies differences in pharmacodynamics (5) and may correlate better with measures other than the milligrams-per-kilogram dose, such as milligrams per unit of body surface area (for which the proportionality between mice and humans is much closer). The total steroid dose administered, even on a milligrams-per-kilogram basis, though administered as a single dose, is similar to that used for humans over, e.g., a 2- to 3-month course of immunosuppressive therapy (7), one that would put a patient at risk for an opportunistic mycosis. Perhaps of greatest biological relevance is that the steroid doses used in the model are those required to produce progressive infection and thus may be analogous to those doses which are associated with progressive opportunistic infection in patients. It would be important to perform additional studies with various strains of mice, as well as to do dose-finding studies on the glucocorticoids. In addition, the species specificity of the toxicity should be determined.