Preclinical Safety Evaluation of Avasimibe in Beagle Dogs: An ACAT Inhibitor with Minimal Adrenal Effects

Donald G. Robertson*,1, Michael A. Breider* and Mark A. Milad{dagger}

* Drug Safety Evaluation and {dagger} Pharmacokinetics, Dynamics, and Metabolism, Pfizer Global Research and Development, 2800 Plymouth Rd., Ann Arbor, Michigan 48106–1047

Received August 24, 2000; accepted October 13, 2000


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Avasimibe, a novel inhibitor of acyl coenzyme A:cholesterol acyltransferase (ACAT), is currently being developed as an antiatherosclerotic agent. The preclinical safety and toxicokinetics of the compound were assessed in beagle dogs in an escalating-dose study and in repeated-dose studies of 2-, 13-, and 52-week duration. Oral (capsule) doses up to 1000 mg/kg b.i.d. were assessed in the escalating dose study and once-a-day doses up to 300 mg/kg, 1000 mg/kg, and 1000 mg/kg were assessed in the 2-, 13-, and 52-week studies, respectively. Avasimibe was found to be a substrate and inducer of hepatic CYP 3A, producing pronounced decreases in plasma drug concentrations subsequent to Day 1. Plasma drug concentrations plateaued markedly at doses above 100 mg/kg. Significant toxicologic findings were restricted to the higher doses (>=300 mg/kg) and included emesis, fecal consistency changes, salivation, body weight loss, microscopic and clinical pathologic evidence of hepatic toxicity, and red blood cell (RBC) morphology changes. Mortality occurred at 1000 mg/kg due to hepatic toxicity. Toxicity was more closely associated with the exaggerated pharmacodynamic effects of the compound (e.g., marked serum cholesterol decreases) seen at the high doses of avasimibe used in these studies rather than with measures of systemic exposure (Cmax or AUC). Adrenal effects were noted only in the 52-week study and consisted of minimal to mild cortical cytoplasmic vacuolization and fibrosis at doses >=300 mg/kg, with no change in adrenal weight. In conclusion, avasimibe is an ACAT inhibitor that has minimal adrenal effects in dogs, with dose-limiting toxicity defined by readily monitored and reversible changes in hepatic function.

Key Words: avasimibe; ACAT inhibitor; dog toxicity; adrenal toxicity.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Avasimibe (2,6-bis(1-methylethyl)phenyl [[2,4,6-tris(1-methylethyl)phenyl]acetyl] sulfamate) also known as CI-1011 or PD 148515 (Fig. 1Go), is an inhibitor of acyl coenzyme A:cholesterol acyltransferase (ACAT), which is currently used in clinical trials to assess its potential as an anti-atherosclerotic agent. ACAT catalyzes the formation of intracellular cholesteryl esters from free cholesterol and is believed to play a critical role in cholesterol absorption and storage, lipoprotein production, and secretion and steroid hormone synthesis (Suckling and Strange, 1985). ACAT inhibition has long been recognized as a potential antiatherosclerotic target (Bocan et al., 1991Go; Krause et al., 1993Go; Sliskovic and Trivedi, 1994Go) with the emphasis for therapeutic intervention shifting from inhibition of cholesterol absorption to direct effects on developing atherosclerotic lesions (Roth, 1998Go). Adrenal gland toxicity, characterized by zonal atrophy or ablation of the zona fasciculata and zona reticularis with associated decreases in adrenal weights, was identified as a potential class-related effect in dogs, and was subsequently identified in guinea pigs and WHHL rabbits (Dominick et al., 1993aGo,bGo; Matsuo et al., 1996Go; Wolfgang et al., 1995Go). In early in vivo screening studies, avasimibe was identified as a novel ACAT inhibitor that lacked significant adrenal toxicity (Lee et al., 1996Go; Roth et al., 1998). The objectives of the studies reported here were to evaluate the preclinical toxicity of avasimibe in beagle dogs in support of clinical evaluation of the compound.



View larger version (11K):
[in this window]
[in a new window]
 
FIG. 1. Structure of avasimibe.

 

    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Dogs.
Beagle dogs were obtained from Marshall Farms USA, Inc. (North Rose, NY) and ranged in age from 9 to 28 months and in weight from 8 to 14 kg at study initiation. Dogs were housed individually in stainless steel cages in clean, climate-controlled rooms with a 12-h light:dark cycle. Water and food (Purina certified canine chow 5007TM) were available ad libitum.

Compound administration.
Avasimibe was prepared in-house and had nuclear magnetic resonance, and infrared and mass spectra consistent with the structure of the compound. The potency of each lot of compound used in these studies was determined by HPLC and elemental analysis to be >97%. The compound was administered orally as bulk drug in gelatin capsules on a mg/kg body weight basis. Control animals received empty gelatin capsules equal in number to those given to the high-dose group for each study.

Experimental Design
In all studies, clinical signs were monitored daily and body weights were taken weekly.

Escalating-dose study.
The 2-week repeated-dose study was the first study conducted with avasimibe in dogs (see below). The doses were selected based on experience with a previous ACAT inhibitor (Wolfgang et al., 1995Go). However, the study failed to define a dose-limiting toxicity. Therefore, an escalating dose study was conducted in order to establish a maximum tolerated dose. In the study, 2 male dogs were administered 100 mg/kg on Days 1–9, 1000 mg/kg once a day on Days 10–16 and 1000 mg/kg b.i.d. on Days 17–23. The b.i.d. doses were administered 8 h apart. Plasma drug concentrations were determined pre-dose, 1.5, 4, 8, 12, and 24 h post-dose on Days 9 and 16, and pre-dose, 1.5, 4, 8, 9.5, 12, 16, 24, and 32 h post-dose (first dose) on Day 23. Hematological and serum chemistry parameters were measured pre-test and on Days 8, 15, and 22. The animals were not euthanized at the end of the study.

Repeated-dose studies.
The experimental designs for the 2-, 13- and 52-week repeated-dose studies are summarized in Table 1Go. A full battery of standard clinical chemistry, hematology, and urinalysis assays was conducted at the times indicated. In addition to standard tests, serum lipoproteins were measured by an HPLC procedure described by Kieft et al. (1991). In the 13-week and 52-week studies, a selected battery of serum lipid markers and indices of liver toxicity were monitored at more frequent intervals, as indicated in Table 1Go. In the 52-week study, starting in Week 40, 1000 mg/kg males were given a 50% increased food ration due to clinical indications of thinness.


View this table:
[in this window]
[in a new window]
 
TABLE 1 Study Design
 
Pathology
At termination of the 2-, 13- and 52-week studies, dogs were euthanized by administration of an overdose of 4% Surital and subsequent exsanguination. Complete necropsies were performed and representative samples of all major organs were taken and processed for light microscopy. All tissues from control and high-dose animals were evaluated microscopically. Tissues identified as targets in the high-dose group were evaluated in intermediate-dose groups. Pathology slides were reviewed by a second pathologist for accuracy and interpretation.

Tissue Biochemistry
For the 2-, 13- and 52-week studies, hepatic microsomal fractions were prepared from all animals at study termination. In the 2-week study, total CYP (spectral), microsomal protein, methoxyresorufin-O-demethylase (MROD), ethoxyresorufin-O-deethylase (EROD), pentoxyresorufin-O-dealkylase (PROD) and erythromycin-N-demethylase (END) activities of microsomal fractions were determined. Additionally, microsomal CYP 1A, 2B, and 3A were evaluated by Western blot. In the 13- and 52-week studies only total CYP, protein, and END activities were determined on hepatic microsomal fractions. Microsomal assays were performed using procedures described previously (Robertson et al., 1995Go). Adrenal and liver total cholesterol, esterified cholesterol and free cholesterol were measured at termination in the 2-, 13- and 52-week studies using procedures described by Bocan and Guyton (1985).

Toxicokinetics
Plasma avasimibe concentrations were determined by solid phase extraction followed by HPLC/UV detection. The limit of quantitation was 0.25 µg/ml for the dose-escalating and 2-week studies. The limit of quantitation was 0.08 µg/ml for the 13- and 52-week studies.

Maximum concentrations (Cmax) were recorded as observed. Area under concentration-time profile (AUC) values was estimated using the linear trapezoidal rule. AUC(0-tldc) values were calculated from time zero to the time for the last detectable concentration (tldc). Apparent elimination rate constant ({lambda}z) values were estimated as the absolute value of the slope of a linear regression of natural logarithm (ln) of concentration versus time during the terminal phase of the concentration-time profile. AUC(0-{infty}) values were calculated as the sum of corresponding AUC(0-tldc) and ldc/{lambda}z values. AUC(0-Tau) were estimated using linear trapezoidal rule, where Tau is the dosing interval. Pharmacokinetic parameter values were calculated using SAS (SAS Institute, Inc., Cary, NC) except in the 52-week study, where WinNonlin (Pharsight Corporation, Mountain View, CA) was used.

Statistics
Treatment comparisons were performed on rank-transformed data using a one-factor analysis of variance (ANOVA). A dose-trend test sequentially applied at the 2-tailed 1% significance level was utilized for quantitative clinical laboratory data. If the high-dose linear trend test was not significant, a test for trend reversal was performed. If the trend reversal was significant, then treated groups were compared individually to the control group using Dunnett's test. For the 300 mg/kg dose group added to the 52-week study, a t-test was used to determine statistical differences between the treated groups and their concurrent control.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Two-Week Study
No deaths occurred and no drug-related clinical or microscopic evidence of toxicity was observed. Clinical chemistry findings, relative to control, were restricted to anticipated dose-related effects on total serum cholesterol (decreased 18–25%) and low density lipoprotein cholesterol (LDL-C, decreased 60–95%) and slight (<2-fold at all doses) increases in alkaline phosphatase that remained within the normal range. No effects were noted on liver or adrenal cholesterol concentrations. Total hepatic microsomal CYP and END activity were increased in a dose-related fashion indicative of induction of CYP 3A (Table 2Go). Changes in hepatic microsomal MROD, EROD or PROD were less than 2-fold and were not attributed to CYP 1A or 2B induction. Western Blot analysis of hepatic microsomes from the 300 mg/kg dose group showed increases in immunoreactive CYP 3A protein but not CYP 1A or CYP 2B (Fig. 2Go). Toxicokinetic results were consistent with induction of CYP 3A, with marked decreases in Cmax and AUC between Day 1 and Day 9 at all doses (Table 2Go). Increases in plasma drug concentrations were less than dose proportional on both Days 1 and 9, with a 30-fold increase in dose producing no more than a 6-fold increase in either Cmax or AUC across all doses on Day 9.


View this table:
[in this window]
[in a new window]
 
TABLE 2 Two-Week Study, Toxicokinetic and Hepatic CYP Parameters (Mean)
 


View larger version (64K):
[in this window]
[in a new window]
 
FIG. 2. Western blot of hepatic microsomes obtained from control dogs (C) or dogs treated with 300 mg/kg avasimibe (H) for 2 weeks and probed for CYP 1A, 2B or 3A. Positive controls included microsomes from rats treated with either phenobarbital (PB) or ß-naphthoflavone (BNF). MW, molecular weight standards (58.1 K, top standard; 39.8 K, bottom standard).

 
Escalating-Dose Study
No deaths occurred. Mild clinical signs of toxicity including soft or mucoid feces, salivation, and decreased food consumption were noted occasionally in both dogs after escalation of the dose to 1000 mg/kg, and white material, presumably drug, was frequently noted in the feces. Clinical pathology findings were consistent with the 2-week repeated-dose studies with Day 22 total cholesterol and LDL-C reduced, relative to pretest, by 29% and 68%, respectively. The toxicokinetic data is summarized in Table 3Go. Increases in AUC and Cmax were less than dose-proportional when the dose was increased from 100 to 1000 mg/kg. As expected, b.i.d. dosing had little effect on Cmax and increased AUC by approximately 50%. This result, coupled with concerns over clinical tolerance of administration of large number of capsules required for each 1000 mg/kg dose (up to 12), led to selection of a single dose of 1000 mg/kg as the maximum feasible dose for repeated dose studies.


View this table:
[in this window]
[in a new window]
 
TABLE 3 Escalating-Dose Study, Toxicokinetic Data in Male Dogs (n = 2)
 
Thirteen-Week Study
No deaths occurred. No drug-related effects were noted at the 10 mg/kg dose. Anticipated dose-related changes in serum lipid parameters were evident with combined-sex mean serum cholesterol reduced from pretest levels of 179 and 169 mg/dl to 114 and 71 mg/dl by Week 13 in the 100 and 1000 mg/kg groups, respectively. LDL-C was completely absent by Week 4 in dogs treated with 1000 mg/kg. Other effects at 100 mg/kg were restricted to changes in fecal consistency (soft stool and/or diarrhea) and a slight (up to 218 IU/l), reversible increase in mean serum alkaline phosphatase (AP) activity. Drug-related effects at 1000 mg/kg included emesis and salivation, which were related to the number of capsules the dogs received, and change in fecal consistency. In addition to clinical signs, changes in clinical chemistry parameters in both sexes at 1000 mg/kg included moderate (up to 6-fold) elevations in AP (up to 550 IU/l), alanine amino transferase (ALT, up to 218 IU/l) and aspartate aminotransferase (AST, up to 73 IU/l), and moderate decreases in serum protein (~20%, down to 5.0 mg/dl). In general, the changes were considered mild to moderate, were maximal by Week 8 with subsequent return toward control levels by Week 13, and rapidly reversed (within 1 week) after discontinuation of drug treatment. Liver cholesterol was unaffected by avasimibe treatment; however, total adrenal cholesterol was markedly reduced (>65%) at 1000 mg/kg, with the greatest decrease occurring in the esterified cholesterol fraction. Total adrenal cholesterol was reduced ~20% in both sexes at 100 mg/kg; however, this effect was not statistically significant. Dose-related increases in total CYP and END activity consistent with induction of CYP 3A, as noted in the 2-week study, were also evident (Table 4Go). No definitive drug-related pathology was noted at termination or in reversal animals. Toxicokinetic data for the study are presented in Table 4Go. Decreases in Cmax (50–80%) and AUC (73–86%), values, attributed to autoinduction, were noted at doses >=100 mg/kg between Day 1 and Week 12. As in the 2-week study, increases in Cmax and AUC values were less than dose-proportional with a 100-fold increase in dose producing less than a 7-fold increase in either parameter at Week 12.


View this table:
[in this window]
[in a new window]
 
TABLE 4 Thirteen-Week Study, Toxicokinetic and Hepatic CYP Parameters (Mean ± SD)
 
Fifty-Two-Week Study
Two 1000 mg/kg females were sacrificed in moribund condition during Week 9. Clinical signs prior to sacrifice included hypoactivity and emaciation in one dog and head tilt, tremors, lameness, and severe neck pain in the other. The neurologic clinical signs were consistent with Beagle Pain Syndrome (Hayes et al., 1989Go); however, microscopic lesions in the spinal cord were not noted. Additionally, the dog displaying the neurologic signs had a triglyceride level of 1 mg/dl within 5 days of sacrifice, which may have contributed to the unusual clinical condition of the animal. Since the effect was noted in only one dog early in the study, the CNS signs were not considered a direct drug-related effect. Clinical pathology findings in both animals included markedly elevated (>40-fold) AP and ALT, 68% to 89% decreases in serum cholesterol, relative to pretest, and abnormal red-blood-cell morphology. Hepatic lesions, including prominent single-cell necrosis and bile stasis, were believed to have contributed to the morbidity of the animals. In light of the death of these two dogs, 300 mg/kg dose groups of both sexes with concurrent control groups were added to the study.

In remaining animals, clinical signs including reduced food consumption, changes in fecal consistency (diarrhea, soft stool), emesis, and salivation were dose-related in incidence, primarily affecting the 300- and 1000 mg/kg dose groups. Biologically significant group mean body-weight loss (up to 1.5 kg) was noted in both sexes at 1000 mg/kg, but was reversed in males after their food ration was increased (Table 5Go).


View this table:
[in this window]
[in a new window]
 
TABLE 5 Body Weight Changes by Quarter and by Dose Group (Mean ± SD)
 
Heart murmurs were detected in one 100 mg/kg and one 1000 mg/kg male. The heart murmur was transient in the 1000 mg/kg animal, detected only during Weeks 44 to 46, but persisted in the 100 mg/kg animal from Week 44 until termination. The murmur in the 100 mg/kg dog was attributed to spontaneous endocardiosis (Jubb et al., 1985Go) after echocardiography and assessment of mitral valve morphology at termination. Neither animal had clinical signs (coughing, exercise intolerance) associated with cardiac insufficiency. Neither murmur was considered treatment-related.

Changes in clinical biochemistry parameters consistent with the pharmacology of the compound included dose-related decreases in serum cholesterol (decreased up to 86%) and lipoproteins (decreased up to 100%) relative to pretest (Fig. 3Go). Changes in clinical biochemistry parameters indicative of an effect on the liver (Fig. 4Go) were primarily restricted to doses >=300 mg/kg, noted in both sexes, and included moderate to marked increases (up to 24-fold) in serum AP and ALT with lesser increases (less than 5-fold) in serum AST and gamma glutamyltranspeptidase (GGT). Serum protein was decreased up to 22% primarily due to decreased albumin (up to 30%). Changes in clinical biochemistry parameters were generally maximal by Weeks 4 to 8 and either remained the same or declined by Week 52. Hepatic effects, as indicated by elevated ALT, were associated with concurrent hypocholesterolemia (Fig. 5Go). Abnormal red blood cell (RBC) morphology, characterized by the presence of burr cells, was noted at >=300 mg/kg, was dose-related in incidence and severity, and was found almost exclusively in animals with concurrent hypocholesterolemia (Fig. 6Go).



View larger version (38K):
[in this window]
[in a new window]
 
FIG. 3. Effect of avasimibe on mean serum cholesterol, HDL-C, LDL-C, and VLDL-C (n = 4/sex). LLN, lower limit of normal range; *p < 0.01 relative to control. Normal ranges have not been established for HDL-C, LDL-C and VLDL-C in dogs.

 


View larger version (36K):
[in this window]
[in a new window]
 
FIG. 4. Effect of avasimibe on mean serum AP, ALT, and total protein (n = 4/sex). ULN, upper limit of normal range; LLN = lower limit of normal range; *p < 0.01 relative to control.

 


View larger version (28K):
[in this window]
[in a new window]
 
FIG. 5. Correlation between individual animal serum ALT and serum cholesterol in dogs treated with avasimibe for 52 weeks. Data collapsed across both sexes and across all time points.

 


View larger version (35K):
[in this window]
[in a new window]
 
FIG. 6. Association of RBC morphology changes with serum cholesterol, obtained concurrently, in dogs treated with avasimibe for 52 weeks (combined sexes). Each circle represents data from an individual animal. LLN, lower limit of normal range.

 
Relative to control, mean total adrenal cholesterol was decreased 32% to 68% at 300 and 1000 mg/kg. The changes were characterized by marked decreases (up to 78%) in esterified cholesterol. Mean liver esterified cholesterol was significantly reduced (93%) in surviving 1000 mg/kg females but was not significantly affected in other treatment groups (Fig. 7Go).



View larger version (52K):
[in this window]
[in a new window]
 
FIG. 7. Effect of avasimibe on mean adrenal (top) and liver (bottom) total, free, and esterified cholesterol (n = 4/sex). *in bar, p < 0.01 for respective cholesterol fraction relative to concurrent control; *on top of bar, p < 0.01 for total tissue cholesterol relative to concurrent control. Note: {dagger}, added 300 mg/kg group and concurrent control.

 
Microscopic changes were restricted to the liver and adrenal gland at doses >=300 mg/kg. Other than changes noted in the 2 moribund 1000 mg/kg females, hepatic changes were characterized by dose-related moderate to marked accumulation of hepatocellular lipofuscin pigment in 5 of 8 dogs at 1000 mg/kg and 3 of 8 dogs at 300 mg/kg (Fig. 8aGo). The lipofuscin accumulation correlated with tan/brown liver discoloration noted grossly. Multifocal to disseminated macrophage and/or mixed cell infiltrates, which were often associated with increased hemosiderin, were noted in 5 of 8 dogs at 1000 mg/kg and 3 of 8 at 300 mg/kg. Minimal multifocal hepatocyte single-cell necrosis was noted in the 2 surviving females and 1 of 4 males at 1000 mg/kg (Fig. 8bGo). Adrenal cortical changes were characterized by minimal to mild cortical cytoplasmic vacuolation in 2 of 8 dogs at 300 mg/kg and 4 of 8 at 1000 mg/kg. Minimal to mild fibrosis of the adrenal cortex was noted in 1 dog at 300 mg/kg and 4 dogs at 1000 mg/kg (Fig. 9Go). Adrenal weights were unaffected.



View larger version (113K):
[in this window]
[in a new window]
 
FIG. 8. Liver sections from male dog treated with 1000 mg/kg for 52 weeks. (a) Section stained for lipofuscin revealed moderate to marked accumulations in centrilobular regions. (b) Minimal single cell necrosis (arrow) was noted in the 2 surviving females and 1 of 4 males (hematoxylin and eosin stain).

 


View larger version (91K):
[in this window]
[in a new window]
 
FIG. 9. Adrenal gland from (a) control dog, (b) dog treated with 1000 mg/kg avasimibe for 52 weeks, and (c) dog treated with 100 mg/kg PD 138142–15 for 5 weeks; g, zona glomerulosa, f, zona fasciculata; r, zona reticularis (100x original magnification for all micrographs, stained with hematoxylin and eosin). The adrenal gland from the avasimibe-treated dog demonstrates minimal vacuolation and fibrosis of the zona reticularis/fasciculata. The adrenal gland from the PD 138142–15–treated dogs shows moderate vacuolation and cortical atrophy of the zona fasciculata and reticularis, which was associated with decreased adrenal weight.

 
Toxicokinetic parameters are summarized in Table 6Go. Although intra-animal variability was apparent, particularly at the higher doses, less than dose-proportional increases in Cmax and AUC values were evident at doses above 100 mg/kg. The 300 mg/kg dose groups of both sexes, which were added on later than the remaining groups, had Cmax and AUC values less than the 100 mg/kg dose groups. Cmax and AUC values varied inconsistently between Weeks 12 and 49, and no conclusions about differences in blood concentrations between the 2 time points could be made. No sex differences in toxicokinetic parameters were noted. Relative to control, hepatic CYP was increased in all drug-treated groups with a corresponding increase in END activity (Table 6Go).


View this table:
[in this window]
[in a new window]
 
TABLE 6 Fifty-Two-Week Study, Toxicokinetic Parameters
 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The results of these studies demonstrated a clear difference in severity of toxicity when compared to previous ACAT inhibitors evaluated in our laboratories. One of these inhibitors, PD 132301, was lethal to dogs within 2 weeks at doses as low as 50 mg/kg (Dominick et al., 1993bGo). Another ACAT inhibitor, PD 138142–15, was lethal to dogs within 5 weeks at a dose of 100 mg/kg (Wolfgang et al., 1995Go). The Day 1 Cmax and AUC values for dogs treated with 100 mg/kg PD 138142–15 were approximately 1.5 to 2-fold higher than Day 1 Cmax and AUC values noted at 1000 mg/kg in the 13-week study with avasimibe. However, due to autoinduction, Week 12 Cmax and AUC values from the highest surviving dose-group of PD 138142–15 treated dogs (30 mg/kg), in which both hepatic and adrenal lesions were noted, were actually comparable to values found at 10 mg/kg after 13-weeks of treatment with avasimibe. This suggests that avasimibe is intrinsically less toxic to dogs than previous ACAT inhibitors, and differences in toxicity are not due to systemic exposure.

Hepatic effects appeared to be associated with concurrent hypocholesterolemia in the 52-week study, which was demonstrated by a striking association between ALT elevations and low serum cholesterol (Fig. 5Go). This result was more than a simple dose response, since dogs at 300 mg/kg, with normal cholesterol, also had normal ALT values, while those dogs with serum cholesterol below the normal range had elevated ALT. Additionally, three 300 mg/kg (2M, 1F) dogs, with no hepatic pathology at termination, had Week 52 serum cholesterol levels within the normal range while 4 of 5 animals with hepatic pathology had Week 52 serum cholesterol levels at or below the lower limit of the normal range (98 and 99 mg/dl for males and females, respectively). The single animal with normal Week 52 cholesterol and hepatic pathology had only minimal hemosiderin pigment associated with macrophages. In addition to these findings, it was clear that the data did not support any speculation about accumulation of hepatic cholesterol due to inhibition of tissue ACAT.

Cholesterol represents a significant component of the RBC membrane and has pronounced effects on membrane fluidity (Telen, 1993Go). Hypercholesterolemia causes significant alterations in RBC morphology (Telen, 1993Go), and the presence of echinocytes (burr cells) noted in burn patients has been attributed to hypocholesterolemia (Harris et al., 1981Go). Therefore, since abnormal RBC morphology was observed almost exclusively in animals with concurrent hypocholesterolemia, it seems plausible that RBC effects noted in this study were due, at least indirectly, to an exaggerated pharmacological effect.

If toxicity was the result of exaggerated pharmacology, the greater effect on serum cholesterol noted in the 52-week study may partially explain the increased severity of effects noted in the 52-week study relative to the 13-week study, at the same doses and within the same timeframe. The differences in severity of toxicity cannot be explained by differences in plasma drug concentrations since the Week 12 Cmax and AUC values at 1000 mg/kg in both studies were similar.

Our previous ACAT inhibitors, PD 132301 and PD 138142–15, also induced hepatic toxicity characterized by elevated AP and transaminases. RBC changes were not noted, possibly since the studies were only 2 and 13 weeks in duration, respectively (Dominick et al., 1993bGo; Wolfgang et al., 1995Go). In the case of PD 138142–15, moribundity of 2 dogs at 100 mg/kg was associated with extreme hypocholesterolemia (<=27 mg/dl). While it is tempting to speculate about a causal association between hypocholesterolemia and hepatic toxicity and/or RBC morphologic changes, it seems unlikely that they are directly related. Chronic treatment of dogs with statin drugs, which reduced serum cholesterol as much as 88%, produced a plethora of effects including cataracts, CNS toxicity, and gall bladder, as well as hepatic, toxicity (Berry, 1988; Gerson, 1989; Hartman, 1996), quite a different profile from the toxicity in dogs treated with avasimibe. Marked serum cholesterol depletion alone cannot explain the divergent effects noted in all these studies. Differences may be due to tissue specificity of cholesterol depletion or other pharmacological effects such as marked depletion of LDL-C or VLDL-C.

The minimal, but statistically significant, changes noted in serum AP activity at doses of 10 and 100 mg/kg (Fig. 3Go) were similar between the 2 studies. These findings were not associated with hepatic pathology, and AP increases are a consequence of hepatic microsomal induction in both dogs (Robertson et al., 1993Go) and humans (Perucca, 1987Go). Therefore, these findings were not considered toxicologically relevant.

Adrenal changes noted in these studies were different from adrenal lesions reported for other ACAT inhibitors (Fig. 9Go). The ACAT inhibitors PD 132301–2, PD 138142–15, and FR145237 produced significant decreases in adrenal weight, corresponding to marked zonal atrophy/ablation of the zona fasciculata and reticularis at pharmacologic doses in as little as 2 weeks when orally administered to dogs (Dominick et al., 1993bGo; Matsuo et al., 1996Go; Wolfgang et al., 1995Go). Avasimibe did not produce adrenal pathology in the 13-week study and adrenal effects in the 52-week study were restricted to minimal to mild cortical cytoplasmic vacuolization and fibrosis at doses >=300 mg/kg. Adrenal weights were unaffected, lesions were microscopic and of limited severity, and as such, were not considered physiologically significant. Although adrenal function was not assessed, our previous experience with ACAT-induced adrenal toxicity in dogs demonstrated that functional effects, as measured by adrenal corticotrophic responsiveness (ACTH challenge), were not evident until frank morphologic lesions were present (Wolfgang et al., 1995Go). It has been hypothesized that ACAT inhibitors induce accumulation of free cholesterol, which is responsible for adrenal toxicity (Warner et al., 1995Go). This was clearly not the case with avasimibe. The absence of adrenal findings in the 13-week studies and the minimal findings in the 52-week study were evident in spite of marked depletion of adrenal total and esterified cholesterol (decreased up to 68% and 78%, respectively, at Week 52) and little change in free cholesterol. Based on these findings, we conclude that frank adrenal toxicity should not be considered a necessary class effect of ACAT inhibitors.

The decreased plasma Cmax and AUC values at 300 mg/kg, relative to 100 mg/kg in the 52-week study, are unexplained. Although different bulk drug lots were used in the study, the drug lot used for the 300 mg/kg group was the same at each time point as those used for the remaining groups. Therefore, the discrepancy cannot be explained by chemical specification differences in bulk drug lots. Additionally, the toxicity findings were clearly dose-related, supporting the contention that the toxicity noted in the 52-week study was more clearly associated with dose and pharmacodynamics than with plasma drug concentrations. Induction of CYP 3A and associated decreases in plasma drug concentrations subsequent to Day 1 were noted in all repeated dose studies. This finding has been confirmed in an early clinical trial of avasimibe (Vora et al., 1997Go).

In conclusion, avasimibe is an ACAT inhibitor that has minimal adrenal toxicity in dogs, with dose limiting toxicity defined by readily monitored and reversible changes in hepatic function. The excellent safety profile of avasimibe has been born out in early clinical trials of the compound (Koren et al., 1998Go).


    NOTES
 
1 To whom correspondence should be addressed. Fax: (734) 622-3478. E-mail: donald.robertson{at}wl.com. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Berry, P. H., MacDonald, J. S., Alberts, A. W., Molon-Noblot, S., Chen, J. S., Lo, C. Y. L., Greenspan, M. D., Allen, H., Durand-Cavagna, G., Jensen, R., Bailly, Y., Delort, P., and Duprat, P. (1998). Brain and optic system pathology in hypocholesterolemic dogs treated with a competitive inhibitor of 3-hydroxy-3-methylglutaryl coenzyme A reductase. Am. J. Pathol. 132, 427–443.[Abstract]

Bocan, T. M., and Guyton, J. R. (1985). Human aortic fibrolipid lesions: Progenitor lesions for fibrous plaques exhibiting early formation of the cholesterol-rich core. Am. J. Pathol. 120, 193–206.[Abstract]

Bocan, T. M., Mueller, S. B., Uhlendorf, P. D., Newton, R. S., and Krause, B. R. (1991). Comparison of CI-976, an ACAT inhibitor, and selected lipid-lowering agents for antiatherosclerotic activity in iliac-femoral and thoracic aortic lesions: A biochemical, morphological, and morphometric evaluation. Arterioscl. Thromb. 11, 1830–1843.[Abstract]

Dominick, M. A., Bobrowski, W. A., MacDonald, J. R., and Gough, A. W. (1993a). Morphogenesis of a zone-specific adrenocortical cytotoxicity in guinea pigs administered PD 132301–2, an inhibitor of acyl-CoA: cholesterol acyltransferase. Toxicol. Pathol. 21, 54–62.[ISI][Medline]

Dominick, M. A., McGuire, E. J., Reindel, J. F., Bobrowski, W. F., Bocan, T. M. A., Gough, A. W. (1993b). Subacute toxicity of a novel inhibitor of acyl-CoA: cholesterol acyltransferase in beagle dogs. Fundam. Appl. Toxicol. 20, 217–224.[ISI][Medline]

Gerson, R. J., MacDonald, J. S., Alberts, A. W., Chen, J., Yudkovitiz, J. B., Greenspan, M. D., Rubin, L. F., and Bokelman, D. L. (1990). On the etiology of subscapular lenticular opacities produced in dogs receiving HMG-CoA reductase inhibitors. Exp. Eye Res. 50, 65–78.[ISI][Medline]

Harris, R. L., Cottam, J. L., Johnston, J. M., and Baxter, C. R. (1981). The pathogenesis of abnormal erythrocyte morphology in burns. J. Trauma 21, 13–21.[ISI][Medline]

Hartman, H. A., Myers, L. A., Evans, M., Robinson, R. L., Engstrom, R. G., and Tse, F. L. (1996). The safety evalution of fluvastatin, an HMG-CoA reductase inhibitor, in beagle dogs and rhesus monkeys. Fundam. Appl. Toxicol. 29, 48–62.[ISI][Medline]

Hayes, T. J., Roberts, G. K., and Halliwell, W. H. (1989). An idiopathic febrile arteritis syndrome in the dog: Beagle pain syndrome. Toxicol. Pathol. 17, 129–37.[ISI][Medline]

Jubb, K., Kennedy, P., and Palmer, N., (1985). Pathology of the Domestic Animals, 3rd ed. Academic Press, New York.

Kieft, K. A., Bocan, T. M. A., and Krause, B. R. (1991). Rapid on-line determination of cholesterol distribution among plasma lipoproteins after high-performance gel filtration chromatography. J. Lipid. Res. 32, 859–866.[Abstract]

Koren, M., Insul, W., Schrott, H., Davila, M., Heinonen, T., McClain, W., and Black, D. (1998). ACAT inhibitor avasimibe lowers VLDL and TG in patients with hypertriglyceridemia. Circulation 98, I240.

Krause, B. R., Anderson, M., Bisgaier, C. L., Bocan, T., Bousley, R., DeHart, P., Essenburg, A., Hamelehle K., Homan, R., Kieft, K. A., McNally, W., Stanfield, R. L., and Newton, R. S. (1993). In vivo evidence that the lipid-regulating activity of the ACAT inhibitor CI-976 in rats is due to inhibition of both intestinal and liver ACAT. J. Lipid Res. 34, 279–294.[Abstract]

Lee, H. T., Sliskovic, D. R., Picard, J. A., Roth, B. D., Wierenga, W., Hicks, J. L., Bousley, R. F., Hamelehle, K. L., Homan, R., Speyer, C., Stanfield, R. L., and Krause, B. R. (1996). Inhibitors of acyl-CoA:cholesterol acyltransferase (ACAT) as hypocholesterolemic agents. CI-1011: An acyl sulfamate with unique cholesterol-lowering activity in animals fed non-cholesterol-supplemented diets. J. Med. Chem. 39, 5031–5034.[ISI][Medline]

Matsuo, M., Hashimoto, H., Suzuki, J., Iwanami, K., Tomoi, M., and Shimomura, K. (1996). Difference between normal and WHHL rabbits in susceptibility to the adrenal toxicity of an acyl-CoA: cholesterol acyltransferase inhibitor, FR145237. Toxicol. Appl. Pharmacol. 140, 387–392.[ISI][Medline]

Perucca, E. (1987). Clinical implications of hepatic microsomal enzyme induction by antiepileptic drugs. Pharmacol. Ther. 33, 139–144.[ISI][Medline]

Robertson, D. G., Krause, B. R., Welty, D. F., Wolfgang, G. H. I., Graziano, M. J., Pilcher, G. D., and Urda, E. (1995). Hepatic microsomal induction profile of carbamic acid [[2,6-bis(1-methylethyl)phenoxy]sulfonyl]2,6-bis(1-methylethyl)phenyl ester, monosodium salt (PD 138142–15), a novel lipid-regulating agent. Biochem. Pharmacol. 49, 799–808.[ISI][Medline]

Robertson, D. G., Loewen, G., Walsh, K. M., Dethloff, L. A.. Sigler, R. S., Dominick, M. A. and Urda, E. R. (1993). Subacute and subchronic toxicology of CI-986, a novel anti-inflammatory compound. Fundam. Appl. Toxicol. 20, 446–455.[ISI][Medline]

Roth, B. D. (1998). ACAT inhibitors: Evolution from cholesterol-absorption inhibitors to antiatherosclerotic agents. Drug Disc. Today 3, 19–25.[ISI]

Sliskovic D. R., and Trivedi, B. K. (1994). ACAT Inhibitors: Potential anti-atherosclerotic agents. Curr. Med. Chem. 1, 204–225.

Suckling, K. E., and Stange, E. F. (1985). Role of acyl-CoA-cholesterol acyltransferase in cellular cholesterol metabolism. J. Lipid Res. 26, 647–671.[ISI][Medline]

Telen, M. J. (1993). The mature erythrocyte. In Wintrobe's Clinical Hematology, 9th ed. (G. R. Lee, T. C. Bithel, J. Foerster, J. W. Athens, and J. N. Lukens, Eds.), pp. 101–103. Febiger & Lea, Philadelphia.

Vora J., Stern R., and Lathia C. (1997). Clinical pharmacokinetics of CI-1011, an ACAT inhibitor. Pharm. Res. 14(Suppl.), S505.

Warner, G. J., Stoudt, G., Bamberger, M., Johnson, W. J., and Rothblat, G. H. (1995). Cell toxicity induced by inhibition of acyl coenzyme A: Cholesterol acyltransferase and accumulation of unesterified cholesterol. J. Biol. Chem. 270, 5772–5778.[Abstract/Free Full Text]

Wolfgang, G. H. I., Robertson, D. G., Welty, D. F., and Metz, A. L. (1995). Hepatic and adrenal toxicity of a novel lipid regulator in beagle dogs. Fundam. Appl. Toxicol. 26, 272–281.[ISI][Medline]