![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOXICOLOGY
Agonist
Pfizer Global Research and Development, Drug Safety Evaluation (M.D.A., G.R.L., D.K.B., G.L.C., S.W.S., W.M.K.) and Drug Metabolism Department (W.M.S.), Groton, Connecticut
Received for publication
December 16, 2002
Accepted
March 6, 2003.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
agonists.
)
and entered clinical development around the 1990s. Like currently marketed
thiazolidinediones such as pioglitazone
(Hoffman et al., 1991) and
rosiglitazone (Kramer et al.,
2001), this class of compounds sensitizes skeletal muscle and
adipose tissue to the effects of insulin
(Hulin et al., 1996;
Saltiel and Olefsky, 1996).
Unlike sulfonylureas or biguanides, which stimulate insulin secretion directly
from functioning pancreatic 
-islet cells, this class of compounds
increases the translocation of glucose transporters from intracellular storage
pools to the plasma membrane of insulin-sensitive tissues
(Hulin et al., 1996;
Saltiel and Olefsky, 1996;
Petrie et al., 1997). Although
thiazolidinediones were initially developed without specific knowledge of a
molecular mechanism of action, later work by researchers such as Lehmann et
al. (1995) subsequently
defined the mode of action and receptor target as agonist activity at the
peroxisome proliferator-activated receptor- (PPAR-). This
nuclear hormone receptor is expressed to a greater extent in adipose tissue
relative to skeletal muscle and liver tissue
(Braissant et al., 1996;
Vidal-Puig et al., 1996).
Heterodimerization of this receptor with the retinoid X receptor regulates
genes that control expression and translocation of glucose transporters. This
effect lowers peripheral insulin resistance and restores glycemic control by
enhancing glucose disposal in peripheral tissues.
Despite numerous articles publicizing the clinical efficacy and molecular
mechanism of action of thiazolidinediones, information in the peer-reviewed
literature regarding the safety profile of this class of compounds in multiple
animal species is incomplete. For example, ciglitazone reportedly caused
cataracts in animals (Kraegen et al.,
1989). Tanabe-174 (approximately 10x more orally efficacious
than ciglitazone) caused a nonregenerative anemia and cardiac hypertrophy in
rodents and dogs at drug levels near projected therapeutic levels in humans
(Williams et al., 1993),
whereas DRF-2189 caused hemodilution and cardiac hypertrophy in rats
(Lohray and Bhushan, 1999).
Troglitazone, the first approved compound of this class for humans, was
subsequently removed from the market-place after it was associated with a low
incidence of liver injury in diabetics
(Gale, 2001). The preclinical
evaluation and development of newer more potent drug candidates of this
pharmacological and/or structural class would be facilitated if more were
known regarding correlations between the safety profile of these agents in
animals and humans. Understanding the specificity and degree of concordance
would enhance risk assessment and management strategies before initiation of
clinical trials since this type of information can be used to set margins of
safety based on experience and develop biomarkers used in the conduct of
clinical trials. Still, complete disclosure of the preclinical safety
assessment profile of earlier clinical candidates along with more potent and
structurally diverse compounds compared with any of their adverse effects
observed in human clinical trials are warranted in the peer-reviewed
literature.
The present report describes our experience with the safety profile of
darglitazone in rats and cynomolgus monkeys. Darglitazone has an oral potency
at least 10 to 200x greater than troglitazone and ciglitazone,
respectively (Hulin et al.,
1996). At the receptor level, darglitazone is approximately 20 to
150x more potent than rosiglitazone and pioglitazone, respectively, at
displacing [3H]darglitazone from PPAR- binding (based on
unpublished IC50 values from Donald E. Wilder; Pfizer Global
R&D). Compared with in-house experience with an earlier less potent
thiazolidinedione (englitazone), darglitazone caused dramatic effects on white
and brown adipose tissue and unexpected morbidity/mortality in rats and
monkeys due to hydrothorax during the conduct of subacute safety studies at
maximum tolerated daily doses. Because of the similarity and severity of this
finding in two preclinical toxicology species, additional studies were
conducted to generate a perspective regarding the potential relevance of these
preclinical findings to humans, at that time in the absence of any relevant
clinical safety data regarding darglitazone or this class of compounds. The
relationship between deaths due to hydrothorax to the profound
hypertrophic/hyperplastic changes observed in BAT was investigated in rats to
determine a probable mechanism of action. The effects of darglitazone observed
in treated rats and monkeys (peripheral and pleural fluid retention) were
consistent with some of the clinical effects observed with this compound
during clinical trials. The similarities of these reversible findings in
humans treated with darglitazone to what is now known about other structurally
similar and diverse PPAR- agonists suggest that alterations in adipose
tissue, hemoglobin and hematocrit levels (hemodilution), and fluid retention
(edema) may represent a pharmacologic rather than a structural class
effect.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Materials. Streptozotocin (STZN), corticosterone (CORT), and dehydroepiandrosterone (DHEA) were obtained from Sigma-Aldrich (St. Louis, MO). STZN was dissolved in 0.01 M sodium citrate buffer (pH 7.4), and CORT was dissolved in peanut oil immediately before use. DHEA was administered in the diet to animals as a ground feed admixture at a level of 0.6%. Darglitazone was administered orally as the sodium salt, and all daily doses were expressed as the free acid equivalent based on an activity factor of 94.5%. All dose levels stated in the text are based on the weight of the free acid. Darglitazone, chemically known as (±)-5-(4-(3-(5-methyl-2-phenyl-4-oxazolyl)-propionyl)phenyl-methyl)-thiazolidine-2,4-dione·sodium salt, was synthesized according to good manufacturing processes and contained <0.5% impurities based on various analytical techniques designed to determine the presence of process related impurities. The compound was dissolved daily before use in distilled water and administered at a dosing volume of 10 ml/kg. Rats were dosed daily approximately 1 h after the beginning of the light cycle. Monkeys were dosed daily approximately 3 h before receiving their daily food ration at 11:00 AM.
General Experimental Design. Animals were assessed daily for clinical signs of toxicity. Body weight and food consumption were determined weekly. For toxicokinetic determinations, animals were periodically bled in the nonfasted state during the conduct of the study to quantitate serum drug levels. Potential changes in hematological parameters and serum chemistries were determined periodically during the course of study by collecting blood from rats (retroorbital sinus) or monkey (venopuncture) in the fasted state before receiving the daily dose of darglitazone. Urinalysis was performed on samples obtained from fasted animals after an overnight collection period. Samples were analyzed manually or using automated analyzers [Technicon H·1E (Technicon, White Plains, NY), Hitachi 704/717 (Boehringer Mannheim Corp., Indianapolis, IN), COBAS-BIO (Roche Diagnostics, Nutley, NJ), or CLINITEK 200 (Bayer Diagnostics, Tarrytown, NY)]. After treatment was completed, the animals were euthanized and necropsied for histological assessment of potential pathologic changes in all major organs.
Experiment I: Standard 3-Month Subacute Toxicity Assessment in Rats. Male and female rats were treated daily for 3 months by oral gavage at doses of 0, 1, 5, and 50 mg/kg/day (15 animals/sex/group).
Experiment II: Chronology and Reversibility Study in Rats. In a separate study, female rats were treated daily for 5 to 36 days by oral gavage at doses of 0 and 50 mg/kg/day (6 –20 animals/group). Nine treatment groups, along with corresponding controls, were treated as follows (days of treatment/days of reversal): 5/0, 10/0, 20/0, 20/20, 30/0, 30/30, 36/0, 36/19, and 36/40 to assess the onset and reversibility of the observed toxicity.
Experiment III: Role of Insulin or Insulin-Dependent/Regulated Processes. In another series of studies, four groups of female rats (12–18 animals/group) were treated by daily oral gavage at doses of 0 and 50 mg/kg/day for two months. Animals were assigned to either a control group (normal or STZN-pretreated) or a darglitazone treated group (normal or STZN-pretreated). Rats were rendered insulinopenic through the destruction of pancreatic islet cells by pretreating the animals with a single dose of STZN (75 mg/kg, i.v.) 8 days before initiation of darglitazone treatment. Fasting serum glucose levels before the start of treatment were: 103 ± 8 mg/dl in normal animals, 360 ± 154 mg/dl in STZN-pretreated animals administered vehicle, 99 ± 6 mg/dl in normal animals subsequently administered darglitazone for 2 months, and 345 ± 122 mg/dl (mean ± S.D.) in STZN-pretreated animals subsequently administered darglitazone for two months.
Experiment IV: Role of Endogenous Glucocorticoids. In this study, female rats were treated daily for 2 weeks by oral gavage at doses of 0, 5, and 50 mg/kg/day (eight animals per group). Surgically altered rats were obtained from Charles River Breeding Laboratories (Releigh, NC). Rats were either sham-operated (SHAM) or adrenalectomized (ADX) 2 weeks before initiation of treatment. ADX animals had their drinking water replaced with normal saline to maintain proper salt balance. CORT was administered in peanut oil at a dose of 20 mg/kg/day s.c. (1 ml/kg) to half of the SHAM and ADX animals placed on treatment. In a separate series of studies, normal animals were supplemented with 0.6% DHEA in ground diet. Darglitazone-treated animals were pair-fed to their ADX or DHEA respective controls to account of hyperphagic episodes observed in previous studies.
Experiment V: Subchronic Toxicity/Reversibility Assessment in Cynomolgus Monkeys. Male (3.4 –5.4 kg) and female (2.7–3.6 kg) cynomolgus monkeys (Charles River Primate Co.) were treated daily by oral gavage at doses of 0, 50, 75, and 100 mg/kg/day (2/sex/dose) for 2.5 to 3.5 months. Because morbidity/mortality was observed during the in-life portion of the study, animals given darglitazone at 50 mg/kg/day were treated for 106 days, with an interim necropsy on days 71 to 72, followed by a reversibility period of 50 days with the remaining animals. Animals given 0, 75, and 100 mg/kg/day were treated for 71 to 72 days, at which time an interim necropsy was performed. The remaining animals were followed for 85 days to assess reversibility of the drug effects.
Analysis of Serum Darglitazone Levels. Whole blood was collected periodically during the rat (retro-orbital sinus) and monkey studies (venopuncture). Serum samples were deproteinized with 1 N HCl before extraction with ethyl acetate. The organic phase was then alkalinized with 0.01 M NaOH, dehydrated with sodium sulfate, and evaporated to dryness before being reconstituted in mobile phase (45:55 v/v acetonitrile/10 mM phosphate buffer, pH 3.0) for analysis. The extracted serum samples were analyzed under isocratic conditions by reverse phase high-performance liquid chromatography using a Nova-Pak C18 column (150 x 3.9 mm, 4-µm packing) from Millipore Corp. (Milford, MA). Detection was at 280 nm, and the flow rate was 1.2 ml/min.
Fatty Acid Analysis. Tissues were suspended in methanolic sodium hydroxide and heated in an airtight tube at 100°C for 30 min. Upon cooling to room temperature, acidified boron trichloride was added to the tissue. The acidified suspension was heated at 85°C for 10 min. Upon cooling to room temperature, fatty acid methyl esters were extracted by the addition of a 50:50 mixture of hexane/ether. Fatty acid methyl esters were analyzed by injection of the organic phase into an HP 5970/5970A series gas chromatograph (Hewlett Packard, Palo Alto, CA) equipped with an HP-1 methyl silicone gum capillary column. Ramp conditions were raised at a rate of 20°C/min from 50 -250°C. Peaks were assigned based on comparisons with retention times with authentic standards.
Statistical Analysis. Comparisons between two treatment groups were made using an unpaired Students's t test. Comparisons among multiple treatments were made using a one-way analysis of variance followed by Fisher's protected least significant difference test. Statements of significance were based on a p value <0.05. All results are expressed as mean ± S.D.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
Darglitazone administration to rats was associated with a time- and dose-dependent increase in morbidity/mortality that became evident after the first month of consecutive daily exposure. The total incidence of morbidity/mortality, which occurred during the 3-month study, can be found in Table 1. The 2-fold higher incidence of mortality in the high-dose female rats corresponded with a 2-fold higher serum darglitazone level compared with high-dose males (22.4 ± 2.9 versus 10.5 ± 1.7 µg of darglitazone/ml of serum at Cmax, respectively, p < 0.05). Death was preceded by clinical signs of impaired respiration (cyanosis and dyspnea) and was ultimately attributed to the accumulation of a clear acellular nonclotting fluid in the thoracic cavity (hydrothorax) of all involved rats during post-mortem examination.
Progressive changes in the gross appearance, size, and palpable firmness of IBAT were observed to varying degrees in all treated male and female rats that survived the 3-month study and in all animals that died before the end of study. At the time of necropsy, there was a pronounced discoloration and increased size and firmness of IBAT in treated animals. IBAT was transformed in treated animals from a small soft, dark-rust brown colored fat mass to a large firm, off-white colored IBAT mass in treated animals (Fig. 1, A and B). In addition, the profound adipogenic activity of darglitazone was clearly evident in the TBAT areas that are found along the descending aorta and major lymphatic trunks and within the thoracic cavity and the perirenal area (Fig. 1, C and D). Examination of BAT by light microscopy revealed morphological transformation of adipocytes from small multiloculated to large uniloculated lipid droplet containing cells (photomicrographs not shown). There was also evidence of an increase in white adipose tissue mass. Although there were no pathological changes found in major organ systems (heart, lung, kidney, brain, bone marrow, and testes), there were significant changes in ovarian morphology of treated female rats. These changes consisted of a decrease in the number of corpura lutea as well as an increase number of atretic follicles (Table 2). The incidence and severity of these ovarian changes were dose-related. Subsequent analyses showed a treatment-related decrease (2-fold) in serum progesterone levels at daily doses of 5 and 50 mg/kg (data not shown).
|
|
Experiment II: Chronology and Reversibility Study in Rats. Because of the profound effects observed in the first 3-month rat study, female rats were used in subsequent studies to investigate 1) the chronology and reversibility of events associated with prolonged treatment with darglitazone (based on animal survival, hyperplastic/hypertrophic changes in BAT, red blood cell parameters, and ovarian morphology), 2) the probable relationship between thoracic fluid volume and the proliferation/firmness of BAT, and 3) hypothesized changes in the fatty acid composition of BAT. At 50 mg/kg/day (where serum darglitazone levels were 21.2 ± 8.9 µg/ml at Cmax), palpable increases in IBAT size and firmness were noted within 5 days of drug administration (Fig. 2). IBAT size and firmness (based on palpation) progressed as a function of time during treatment and was not fully reversed even after drug withdrawal (Fig. 2). Animal survival was not adversely affected until IBAT size and firmness were at a maximum. Within 21, 28, and 36 days of treatment 1/46 (2%), 3/46 (7%), and 8/46 (18%) of all treated animals expired or were euthanized in moribund condition (Fig. 2). Mortality increased with increasing length of treatment in the presence of hydrothorax. The incidence of morbidity/mortality paralleled the incidence of hydrothorax found at necropsy. Two additional animals died of hydrothorax during the 1-month recovery period.
|
The presence of hydrothorax was difficult to detect in rats until
potentially life-threatening and obvious signs of pulmonary dysfunction were
present. Therefore, we also determined whether there was a relationship
between the volume of fluid found in the pleural cavity and the firmness of
dorsal TBAT (determined at necropsy) with the size and firmness of IBAT (a
palpable clinical assessment). Trying to develop a clinically assessable and
relevant marker of this effect was necessary since the amount of fluid in the
thoracic cavity and the mass and firmness of dorsal TBAT and aortic BAT cannot
be followed during the in-life portion of the study. In general, thoracic
fluid volume increased with increasing TBAT tonicity such that the greatest
accumulation of thoracic fluid occurred in rats with TBAT tonicity 
10 mm
Hg (Fig. 3). The severity of
dyspnea and cyanosis observed before death also correlated well with the
presence and volume of pleural fluid found in the thoracic cavity at necropsy
(data not shown). By clinical assessment, there was a good correlation between
palpable IBAT size and firmness in animals with the greatest accumulation of
fluid. This suggested that in the rat IBAT size and firmness could be used as
surrogate markers of TBAT firmness and thoracic fluid accumulation when
present.
|
Fluid that was found in the pleural cavity at necropsy was further characterized to determine its etiology. Based on its appearance (clear), color (pale yellow), specific gravity (range 1.020 –1.027), cellularity (mostly acellular), and glucose (range 120 –182 mg/dl), protein (2.8 – 4.1 g/dl) and lactate dehydrogenase (30 –200 U/l) content, the characteristics of the fluid were consistent with a modified transudate derived from plasma. These results suggest that the pleural fluid was lymphatic in origin. From this, we deduced that the accumulation of fluid in the thoracic cavity may be due to impaired lymphatic drainage caused by the increased firmness and size of TBAT in the rat.
Although the composition of dietary fat intake remained constant during the study, there were distinct biochemical changes in the fatty acid composition of TBAT from treated animals. Compared with controls, TBAT from treated animals (50 mg/kg/day for 36 days) had a 274% increase in the unsaturated fatty acid of palmitoleic acid (C16:1) combined with a 36% reduction in saturated fatty acid content of stearic acid (C18:0) (Table 3). There was a 16% reduction in the ratio of unsaturated to saturated fatty acids in TBAT. There were no apparent alterations in TBAT palmitic (C16: 0), linoleic (C18:2) or oleic (C18:1) acid content. Biochemical changes in TBAT fatty acid composition were noted as early as 5 days after the start of treatment (data not shown) and were not readily reversed one month after withdrawal of treatment (Table 3). Similar effects on fatty acid composition and a decreased ratio of saturated/unsaturated fats were also observed in the serum and white adipose tissue obtained from the same animals within 5 days after initiation of treatment (data not shown).
|
Other previously noted changes in red blood cell parameters and ovarian histology in treated animals occurred within 5 days and were either fully (hematology changes) or partially (ovarian changes) reversed within 1 month after discontinuation of drug treatment (data not shown). Treated animals as a group also periodically displayed evidence of hyperphagic episodes, consuming as much as 5 to 24% more feed on a given day than control animals.
Experiment III: Role of Insulin or Insulin-Dependent/Regulated Processes. To determine whether the adipogenic activity of darglitazone on BAT mass and the consequent development of hydrothorax-induced death were related to enhanced sensitivity to circulating insulin, we investigated whether animals rendered insulinopenic by pretreatment with STZN were susceptible to the mortality caused by hydrothorax and BAT hypertrophy/hyperplasia. Animals were treated daily for 60 days with 50 mg/kg/day darglitazone. Control (normal) animals and STZN-pretreated animals that received only vehicle survived the entire study. There was a 78% (14/18 animals) mortality rate in normal animals treated with darglitazone. In contrast, an 11% (2/18 animals) mortality rate was observed in STZN-pretreated animals treated with darglitazone. It should be noted that the two STZN-pretreated animals that died of hydrothorax were marginally diabetic (fasting serum glucose levels of <160 mg/dl during the first 3 weeks of study) and did not have histological evidence of pancreatic islet cell atrophy at necropsy. Only animals that died had any clinical signs of hydrothorax (dyspnea and cyanosis) and had pleural fluid volumes ranging between 1 to 15.5 ml found in the thoracic cavity at the time of necropsy. Therefore, only rats that were marginally insulinopenic and had no histologic evidence of sustained pancreatic damage from the STZN pretreatment succumbed to hydrothorax.
All normoglycemic rats treated with darglitazone developed the characteristic profound increase in prominence and firmness of IBAT mass, dorsal thoracic and perirenal BAT hypertrophy/hyperplasia, and follicular atresia. In contrast, half of the STZN-pretreated animals that received darglitazone (9/18 animals) had only small to moderate increases in BAT mass (dorsal thoracic, interscapular, and perirenal). Follicular atresia was also dependent upon endogenous insulin since STZN pretreatment reduced the incidence and severity of this effect compared with normal animals treated with darglitazone (data not shown).
Experiment IV: Role of Endogenous Glucocorticoids. Darglitazone-treated animals share similar features to genetically obese or VMH-lesioned animals (e.g., BAT enlargement, discoloration, and morphological changes in the relative content of certain BAT fatty acids, hyperphagia, and changes in ovarian function). Since these genetically obese or VMH-lesioned animals are overly sensitive to endogenous glucocorticoids, we investigated whether the adverse effects of darglitazone on BAT proliferation were related to enhanced sensitivity to endogenous glucocorticoids. Animals were either ADX or SHAM before study. Half of all ADX and SHAM animals received a replacement dose of corticosterone daily, as described under Materials and Methods.
After 2 weeks of study, untreated ADX and SHAM rats had an equivalent IBAT mass of 0.26 ± 0.06 and 0.34 ± 0.07 g (p > 0.05), respectively (Table 4). SHAM animals treated with darglitazone developed an IBAT mass approximately twice as large but were exposed to twice the serum levels of darglitazone (Cmax) compared with ADX animals treated at the same dose. Daily replacement of corticosterone resulted in the development of similarly sized IBAT mass and serum darglitazone levels (Cmax) in both treated SHAM and ADX animals. This effect was observed at both doses of darglitazone (5 and 50 mg/kg/day). These results were identical whether the effect on IBAT size was expressed as Cmax or as parent drug exposure over the course of 8 h (AUC0–8 h). The 50% reduction in IBAT size in darglitazone-treated ADX animals corresponded to the 50% reduction in serum drug levels rather than the nearly 100% loss of serum corticosterone.
|
DHEA supplementation in the diet of darglitazone-treated animals was ineffective in preventing BAT proliferation. There were no differences in BAT weight, serum darglitazone exposure (Cmax or AUC0–8 h), and baseline corticosterone levels between darglitazone-treated animals given a normal or DHEA-supplemented diet (data not shown). Although dietary supplementation with 0.6% DHEA effectively antagonizes the enhanced sensitivity to endogenous levels of corticosterone in genetically obese animals, the lack of an effect further illustrates that darglitazone does not enhance sensitivity to endogenous glucocorticoids.
Experiment V: Subchronic Toxicity/Reversibility Assessment in Cynomolgus Monkeys. During the subacute toxicity study in monkeys, there was clinical evidence of enhanced adipogenic activity consisting of clinically palpable increases in body fat around the eyelids, scrotum (males), breasts, and abdomen. One intermediate-dose (75 mg/kg/day) female was euthanized on day 16 because of its moribund condition (due to an unknown cause), while one low-dose (50 mg/kg/day) female died on day 106 as a result of complications from hydrothorax. This female was dyspneic before death and had 150 ml of fluid in the pleural cavity at necropsy. Of the six treated monkeys that were necropsied during the interim sacrifice on day 71 (1/sex/group), three animals (one from each treatment group) displayed varying degrees of hydrothorax (2–200 ml) and displayed clinical signs of dyspnea before being euthanized. The one female monkey from the intermediate dose that died on day 16 had 10 ml of a straw colored fluid in the abdominal cavity, whereas another monkey from the high-dose group had pericardial effusion.
Other major clinical/pathological findings that were observed during the conduct of the study or interim necropsy (including one dead and one moribund monkey) were: 1) evidence of peripheral edema (all treated animals), 2) a mild but persistent reduction (10 –20%) in erythrocyte parameters (red blood cells, hemoglobin, and hematocrit) or hemodilution occurring in all treated animals within 2 weeks (See Table 5), 3) amenorrhea and lost of postovulatory progesterone surge occurring in all treated females irrespective of dose, 4) varying degrees of attenuated (low-voltage) QRS complexes were noted in all treated animals based on visual inspection of electrocardiograms, and 5) hypertrophy of subcutaneous adipose tissue in the abdominal wall and surrounding the eyes with or without periorbital edema (all treated animals), 6) hypertrophy of periaortic and perirenal adipose tissue (all treated animals) and scrotum (males only), mild erythroid hyperplasia of sternal marrow (all treated animals), and ectasia of the mammary ducts (one intermediate dosed male and female). These effects were observed at initial serum drug levels and exposures (Cmax:AUC0–24 h) of 60 ± 23 µg/ml:228 ± 54 µg · h/ml, 82 ± 21 µg/ml:400 ± 147 µg · h/ml, and 68 ± 33 µg/ml:410 ± 225 µg · h/ml in the low-, intermediate-, and high-dose group (day 1), respectively. Serum drug levels and exposure declined during the first 17 days of treatment, the results of which are also presented in Table 5.
|
During the 50-to 85-day post-treatment recovery period, monkeys in the intermediate- and high-dose groups showed complete reversibility of depressions in erythrocyte parameters within 2 weeks after discontinuing treatment. Menstrual cycle and postovulatory progesterone surges returned in all previously treated females within 4 weeks post-treatment. There was no evidence of pericardial effusion or depressed QRS amplitudes during the reversal period. Hypertrophy of subcutaneous adipocytes surrounding the eye and/or edema (low- and high-dose animals), erythroid hyperplasia of sternal marrow (low- and high-dose animals), and ectasia of mammary ducts (one intermediate- and all high-dose animals) were present but less pronounced at final necropsy after completion of the reversal period detailed above (data not shown).
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
The major adverse event identified in darglitazone-treated rats and monkeys
was the development of a life-threatening hydrothorax. The high incidence of
morbidity/mortality occurred at doses 5 mg/kg/day (serum drug exposure
2 µg/ml at Cmax) in the rat and 50 mg/kg/day
(serum drug exposure 37 µg/ml at Cmax) in the
monkey. Increased IBAT and TBAT size, changes in fatty acid composition, and
morphological changes in the appearance of BAT cells were observed in rats
within 5 days of treatment. As treatment progressed, IBAT and TBAT became
larger and firmer (based on touch and tonicity measurements) and were followed
by the accumulation of fluid (
15.5 ml in rats and 200 ml in monkeys)
within the thoracic cavity with clinical signs of peripheral edema, dyspnea,
and hemodilution. Severely affected animals eventually succumbed to
hydrothorax.
Mechanistically these effects appear to represent multiple aspects of an
exaggerated pharmacologic response to insulin sensitization coupled to
receptor target expression in select tissues. For example, fluid extravasation
into tissue, in the form of peripheral and pleural edema, can be caused by
several mechanisms. Troglitazone and rosiglitazone increase endothelial
permeability (see review by Wang et al.,
2002), whereas troglitazone, but not rosiglitazone, has
vasodilatory properties by itself (Walker
et al., 1998). Insulin has both vasodilatory
(McNally et al., 1995) and
hemodynamic (Baron, 1994)
properties. Hemodilution has been observed with other thiazolidinediones due
to plasma volume expansion, an effect thought to be due to reduced renal
sodium excretion (see review by Niemeyer
and Janney, 2002). Finally, hypertrophic/hyperplastic changes in
BAT, the development of hydrothorax and mortality in rats were
insulin-dependent since animals that were rendered totally insulinopenic with
STZN were devoid of these effects. Since STZN also profoundly reduces
expression of PPAR- in adipose tissue
(Vidal-Puig et al., 1996), it
appears that the protective effect of STZN is directly linked to significant
reductions in both endogenous insulin levels and PPAR- expression in
sensitive tissues.
Although many of the effects listed above appear to be common to other
thiazolidinediones, to our knowledge no one has reported life-threatening
hydrothorax in animals treated with thiazolidinediones. While BAT
proliferation (ciglitazone and troglitazone) and firmness of dorsal thorax in
rats (troglitazone) occurs with less potent thiazolidinediones
(Chang and Wyse, 1988;
Herman et al., 1997), only
nonlife threatening signs of labored breathing/rales were reported in
troglitazone-treated rats (de la Iglesia
et al., 1998). Differences in the physical characteristics of BAT
may have some bearing on this manifestation. Darglitazone has opposing effects
on the content of certain saturated/unsaturated fatty acids in BAT compared
with cultured adipocytes treated with troglitazone or pioglitazone
(Kurebayashi et al., 1997).
Moreover, manifestation of a life-threatening hydrothorax may also have a
mechanical component as well. Obstructive compression or narrowing of
lymphatic vessels due to the presence and increased firmness (tonicity) of
extensive BAT hypertrophy/hyperplasia around the descending aorta and major
thoracic lymphatic trunks (Fig.
1) may be implicated in the pathogenesis of hydrothorax. Ohtani
and Ohtani (1997) have shown
that obstruction or narrowing of the lymphatic trunks draining the
diaphragmatic lymph alone is sufficient to cause hydrothorax in rats. Since
the majority of pleural fluid (87%) is drained through the lymphatic system in
the dog (Nakamura et al.,
1988), it is conceivable that abnormal fat accumulation around the
major lymphatics may contribute to the development of hydrothorax in rats and
monkeys. Based on these findings and limited in-house experience with another
less potent thiazolidinedione (englitazone), the severity of these
PPAR-/insulin-dependent responses (BAT proliferation and firmness,
edema, and hydrothorax) would appear to be potency driven.
The possibility that these adipogenic effects were linked to enhanced
sensitivity to endogenous glucocorticoids was quickly disproved. Despite some
similarities in the clinical profile of darglitazone-treated rats to
genetically obese or VMH-lesioned animals (i.e., adipogenic activity in the
form of BAT proliferation, discoloration, morphological changes, and increases
in palmitoleic and oleic acid content, hyperphagic episodes, and changes in
ovarian function and histology) (Haessler
and Crawford, 1965; Saito et
al., 1985), we proved darglitazone did not enhance sensitivity to
endogenous glucocorticoids. Even though ADX or DHEA dietary supplementation
effectively corrects these abnormalities in obese and VMH-lesioned animals,
these same surgical and dietary manipulations were without substantial effect
in darglitazone-treated animals. The partial protective effect observed in ADX
animals was due to differences in drug exposure rather than changes in the
glucocorticoid status or sensitivity of the animal
(Table 4). Therefore, the
profound adipogenic activity of darglitazone observed here in vivo and by
others in vitro (Rabelo et al.,
1997) can be explained by its high affinity binding to the
PPAR- (approximately 50 nM; unpublished information by D. E. Wilder,
Pfizer Global R&D), a receptor implicated in the differentiation of white
or brown adipocyte precursors to mature adipocytes
(Brun et al., 1996). Binding to
this receptor by darglitazone (Camirand et
al., 1998) and other thiazolidinediones
(Foellmi-Adams et al., 1996;
Sears et al., 1996) also
induces terminal differentiation (i.e., expression of uncoupling protein) of
brown adipocytes in culture. Therefore, the adipogenic activity of
thiazolidinediones is strongly related to PPAR- binding and potency
rather than enhanced glucocorticoid sensitivity.
The clinical significance of these effects on adipose tissue proliferation,
hemodilution, abnormal fluid retention/distribution, and hydrothorax could not
be discounted as a rodent specific-phenomenon. These effects were also
observed in darglitazone-treated monkeys. In humans, troglitazone promotes
subcutaneous fat proliferation (Mori et
al., 1999). Peripheral edema has also been reported with this
class of compounds (Niemeyer and Janney,
2002). Pleural edema has rarely been associated with troglitazone
or rosiglitazone treatment but was reversible upon discontinuation of drug
treatment (Koshida et al.,
1998; Inoue and Sano,
2000; Wang et al.,
2002). Peripheral edema and pleural effusion (one case) was also
observed during early clinical trials with darglitazone (information on file,
Pfizer Global R&D) and were also both reversible. This generalized
phenomenon is not linked to the chemical class since recent clinical trials
with a highly selective and more potent PPAR- agonist, farglitazar (a
L-tyrosine-based nonthiazolidinedione), showed a dose-dependent
weight gain, peripheral edema and decreases in hemoglobin
(Fiedorek et al., 2000;
Raz et al., 2000;
Wilson et al., 2000). These
clinical findings suggest some degree of concordance of this phenomenon
(adipose hypertrophy/hyperplasia, peripheral and pulmonary edema, and
hemodilution) in rats and monkeys with humans. Given the clinical effects of
GI262570, the phenomenon itself is most probably linked to the pharmacologic
target itself as a function of potency and administered dose rather than the
thiazolidinedione structure.
Other effects observed in rats and monkeys range from reversible changes in
erythrocyte parameters to follicular atresia. Reversible depressions in
erythrocyte parameters have also been reported in troglitazone-treated rodents
(Herman et al., 1997). In both
cases, bone marrow depression was not evident, suggesting that the effect was
not the result of hematotoxicity. Unlike Tanabe-174 where a nonregenerative
anemia occurred in rats and dogs (Williams
et al., 1993), the present findings were consistent with the
hemodilution effect of this pharmacologic class of compounds.
Alterations in female ovulatory cycle and fertility may be caused by
multiple mechanisms. Blackwell et al.
(1998) proposed that
darglitazone enhanced insulin action in reproductive tissues within the body
based on work by Poretsky and Kalin
(1987). Additional work by
Barbieri et al. (1983)
implicates insulin in causing an androgen excess that would in turn cause
follicular atresia with the loss of estrogen producing cells. Recently, Gasic
et al. (2001) have shown that
troglitazone inhibits 3-hydroxysteroid dehydrogenase, an effect that
would directly reduce ovarian progesterone production. Therefore, we postulate
that the hormonal alterations observed with darglitazone might be the result
of enhanced sensitivity to insulin or inhibition of this enzyme. Further work
would be needed to confirm either proposed mechanism. The clinical relevance
of this effect was not determined during clinical trials.
In summary, darglitazone, an insulin-sensitizing agent for the treatment of
NIDDM, caused BAT proliferation, hemodilution, and abnormal estrous cycling in
rodents and nonhuman primates. Prolonged treatment at high doses was
associated with morbidity/mortality in rodents caused by the development of
hydrothorax. The adverse effects of thiazolidinedione treatment in rats and
monkeys manifest themselves clinically as subcutaneous fat accumulation,
peripheral edema, hemodilution, and anecdotal cases of pulmonary edema. Many
of these effects are fully reversible. The concordance between some of the
effects observed in preclinical safety assessment studies and clinical studies
should aid in setting appropriate margins of safety for future clinical trials
with a new generation of potent thiazolidinediones or PPAR- agonists
for the treatment of NIDDM.
|
Acknowledgements |
|---|
binding studies. |
Footnotes |
|---|
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
ABBREVIATIONS: NIDDM, noninsulin-dependent diabetes mellitus;
PPAR-, peroxisome proliferator-activated receptor-; BAT, brown
adipose tissue; IBAT, interscapular brown adipose tissue; TBAT, thoracic brown
adipose tissue; STZN, streptozotocin; CORT, corticosterone; DHEA,
dehydroepiandrosterone; SHAM, sham-operated; ADX, adrenalectomized; VMH,
ventromedial hypothalamus; AUC, area under the concentration curve;
Cmax, concentration of drug at peak serum levels.
1 Current address: Lilly Research Laboratories, P.O. Box 708, Greenfield, IN
46140. 
2 Current address: Pfizer Global R&D, La Jolla Labs, 11085 North Torrey
Pines Rd., La Jolla, CA 92037.
Address correspondence to: Dr. Michael D. Aleo, Pfizer Global R & D, Groton Laboratories, Drug Safety Evaluation, MS 8274-1229, Eastern Point Rd., Groton, CT 06340. E-mail: michael_d_aleo{at}groton.pfizer.com
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Barbieri RL, Makris A, and Ryan KJ (1983) Effects of
insulin on steroidogenesis in cultured porcine ovarian theca.
Fertil Steril 40:
237-241.[Medline]
Baron AD (1994) Hemodynamic actions of insulin.
Am J Physiol 267:
E187-E202.
Blackwell DK, Lundeen GR, Coleman GL, and Stadnicki SW
(1998) Effects of darglitazone on reproductive parameters in
female Long-Evans rats and cynomolgus monkeys (Abstract). Fund Appl
Toxicol 42 (Suppl):
101.
Braissant O, Foufelle F, Scotto C, Dauca M, and Wahli W
(1996) Differential expression of peroxisome
proliferator-activated receptors (PPARs): tissue distribution of PPAR-
,
- and - in the adult rat. Endocrinol
137:
354-366.[Abstract]
Brun RP, Tontonoz P, Forman BM, Ellis R, Chen J, Evans RM, and
Spiegelman BM (1996) Differential activation of adipogenesis by
multiple PPAR-isoforms. Gene Dev
10:
974-984.
Camirand A, Marie V, Rabelo R, and Silva JE (1998)
Thiazolidinediones stimulate uncoupling protein-2 expression in cell lines
representing white and brown adipose tissues and skeletal muscle.
Endocrinol 139:
428-431.
Chang AY and Wyse BM (1988) The hypoglycemic effect of
ciglitazone in obese, hyperglycemic animal models. Prog Clin Biol
Res 265:
177-192.[Medline]
Fiedorek FT, Wilson GC, Frith L, Patel J, Abou-Donia M, and Study
Group PPA20005 (2000) Monotherapy with GI262570, a tyrosine-based
non-thiazolidinedione PPAR agonist, improves metabolic control in type
2 diabetes mellitus patients (Abstract). Diabetes
49 (Suppl 1):
A38.
Foellmi-Adams LA, Wyse BM, Herron D, Nedergaard J, and Kletzien RF
(1996) Induction of uncoupling protein in brown adipose tissue.
Synergy between norepinephrine and pioglitazone, an insulin-sensitizing agent.
Biochem Pharmacol 52:
693-701.[CrossRef][Medline]
Gale EA (2001) Lessons from the glitazones: a story of
drug development. Lancet
357:
1870-1875.[CrossRef][Medline]
Gasic S, Nagamani M, Green A, and Urban RJ (2001)
Troglitazone is a competitive inhibitor of 3-hydroxysteroid
dehydrogenase enzyme in the ovary. Am J Obstet Gynecol
184:
575-579.[CrossRef][Medline]
Haessler HA and Crawford JD (1965) Alterations in the
fatty acid composition of depot fat associated with obesity. Ann NY
Acad Sci 131:
476-484.[Medline]
Herman JR, Metz AL, McGuire EJ, de la Iglesia FA, and Masuda H
(1997) Subchronic toxicity of the antidiabetic troglitazone in
Wistar rats (Abstract). Fund Appl Toxicol
36 (Suppl):
273.[CrossRef]
Hoffman C, Lorenz K, and Colca JR (1991) Glucose
transport deficiency in diabetic animals is corrected by treatment with the
oral antihyperglycemic agent pioglitazone. Endocrinol
129:
1915-1925.
Hulin B, McCarthy PA, and Gibbs EM (1996) The
glitazone family of antidiabetic agents. Curr Pharm
Des 2:
85-102.
de la Iglesia FA, Herman JR, McGuire EJ, Gough AW, and Masuda H
(1998) Chronic toxicity study of the antidiabetic troglitazone in
Wistar rats (Abstract). Fund Appl Toxicol
42 (Suppl):
50.
Inoue K-I and Sano H (2000) Troglitazone-induced
pulmonary edema (Letter). Arch Intern Med
160: 871.
Koshida H, Shibata K, and Kametani T (1998)
Pleuropulmonary disease in a man with diabetes who was treated with
troglitazone (Letter). N Engl J Med
339:
1400-1401.
Kraegen EW, James DE, Jenkins AB, Chisholm DJ, and Storlien LH
(1989) A potent in vivo effect of ciglitazone on muscle insulin
resistance induced by high fat feeding of rats. Metab
38:
1089-1093.[CrossRef][Medline]
Kramer D, Shapiro R, Adler A, Bush E, and Rondinone CM
(2001) Insulin-sensitizing effect of rosiglitazone (BRL-49653) by
regulation of glucose transporters in muscle and fat of Zucker rats.
Metab Clin Exp 50:
1294-1300.
Kurebayashi S, Hirose T, Miyashita Y, Kasayama S, and Kishimoto T
(1997) Thiazolidinediones down-regulate stearyl-CoA desaturase 1
gene expression in 3T3–L1 adipocytes. Diabetes
46:
2115-2118.[Abstract]
Lehmann JM, Moore LB, Smith-Oliver TA, Wilkison WO, Willson TM, and
Kliewer SA (1995) An antidiabetic thiazolidione is a high
affinity ligand for peroxisome proliferative activated receptor-
(PPAR). J Biol Chem
270:
12953-12956.
Lohray BB and Bhushan V (1999) Indole-containing
thiazolidine-2,4-diones as novel euglycemic and hypolipidemic agents:
DRF-2189. Drugs Future
24:
741-757.
McNally PG, Lawrence IG, Watt PA, Hillier C, Burden AC, and
Thurston H (1995) The effect of insulin on the vascular
reactivity of isolated resistance arteries taken from healthy volunteers.
Diabetologia 38:
467-473.[Medline]
Mori Y, Murakawa Y, Okada K, Horikoshi H, Yokoyama J, Tajima N, and
Ikeda Y (1999) Effect of troglitazone on body fat distribution in
type 2 diabetic patients. Diabetes Care
22:
908-912.[Abstract]
Nakamura T, Tanaka Y, Fukabori T, Iwasaki Y, Nakagawa M, and Kira S
(1988) The role of lymphatics in removing pleural liquid in
discrete hydrothorax. Eur Respir J
1: 826-831.[Abstract]
Niemeyer NV and Janney LM (2002)
Thiazolidinedione-induced edema. Pharmaco-therapy
22:
924-929.[CrossRef][Medline]
Ohtani Y and Ohtani O (1997) Obliteration of the
lymphatic trunks draining diaphragmatic lymph causes peritoneal fluid to enter
the pleural cavity. Arch Histol Cytol
60:
503-510.[CrossRef][Medline]
Petrie J, Small M, and Connell J (1997)
"Glitazones", a prospect for non-insulin-dependent diabetes.
Lancet 349:
70-71.[CrossRef][Medline]
Poretsky L and Kalin MF (1987) The gonadotropic
function of insulin. Endocr Rev
8: 132-141.
Rabelo R, Camirand A, and Enrique Silva J (1997)
3',5'-Cyclic adenosine monophosphate-response sequences of the
uncoupling protein gene are sequentially recruited during darglitazone-induced
brown adipocyte differentiation. Endocrinol
138:
5325-5332.
Raz I, Kler L, Edwards RC, McNeil SJ, and Study Group PPAB2002
(2000) GI262570, a non-thiazolidinedione PPAR agonist,
improves metabolic control in patients with type 2 diabetes mellitus when
combined with glibenclamide treatment (Abstract).
Diabetes 49 (Suppl 1):
A129.
Saito M, Minokoshi Y, and Shimazu T (1985) Brown
adipose tissue after ventromedial hypothalamic lesions in rats. Am
J Physiol 248:
E20-E25.
Saltiel AR and Olefsky JM (1996) Thiazolidinediones in
the treatment of insulin resistance and type II diabetes.
Diabetes 45:
1661-1669.[Abstract]
Sears IB, MacGinnitie MA, Kovacs LG, and Graves RA
(1996) Differentiation-dependent expression of the brown
adipocyte uncoupling protein gene-regulation by peroxisome
proliferator-activated receptor . Mole Cell
Biol 16:
3410-3419.[Abstract]
Vidal-Puig A, Jimenez-Linan M, Lowell BB, Hamann A, Hu E,
Spiegelman B, Flier JS, and Moller DE (1996) Regulation of
PPAR gene expression by nutrition and obesity in rodents. J
Clin Invest 97:
2553-2561.[Medline]
Walker AB, Naderali EK, Chattington PD, Buckingham RE, and Williams
G (1998) Differential vasoactive effects of the insulin
sensitizers rosiglitazone (BRL 49653) and troglitazone on human small arteries
in vitro. Diabetes 47:
810-814.[Abstract]
Wang F, Aleksunes LM, Reagen LA, and Vergara CM (2002)
Management of rosiglitazone-induced edema: Two case reports and a review of
the literature. Diabetes Tech Ther
4: 505-514.
Williams GD, Deldar A, Jordan WH, Gries C, Long GG, and Dimarchi RD
(1993) Subchronic toxicity of the thiazolidinedione, Tanabe-174
(LY282449) in the rat and dog (Abstract). Diabetes
42 (Suppl 1):
59A.
Wilson GC, Abou-Donia M, Frith L, Patel J, Fiedorek FT, and Study
Group PPA20005 (2000) Monotherapy with GI262570, a tyrosine-based
non-thiazolidinedione PPAR agonist, significantly reduces triglyceride
and increases HDL-C concentrations in patients with type 2 diabetes mellitus
(Abstract). Diabetes 49 (Suppl 1): A39.[CrossRef]
This article has been cited by other articles:
|
|
 |
|
  G. J. Hausman, S. P. Poulos, T. D. Pringle, and M. J. Azain The influence of thiazolidinediones on adipogenesis in vitro and in vivo: Potential modifiers of intramuscular adipose tissue deposition in meat animals J Anim Sci, April 1, 2008; 86(14_suppl): E236 - E243. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. D. Aleo, C. M. Doshna, and K. A. Navetta Ciglitazone-Induced Lenticular Opacities in Rats: In Vivo and Whole Lens Explant Culture Evaluation J. Pharmacol. Exp. Ther., March 1, 2005; 312(3): 1027 - 1033. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Rival, A. Stennevin, L. Puech, A. Rouquette, C. Cathala, F. Lestienne, E. Dupont-Passelaigue, J.-F. Patoiseau, T. Wurch, and D. Junquero Human Adipocyte Fatty Acid-Binding Protein (aP2) Gene Promoter-Driven Reporter Assay Discriminates Nonlipogenic Peroxisome Proliferator-Activated Receptor {gamma} Ligands J. Pharmacol. Exp. Ther., November 1, 2004; 311(2): 467 - 475. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. P. Combs, U. B. Pajvani, A. H. Berg, Y. Lin, L. A. Jelicks, M. Laplante, A. R. Nawrocki, M. W. Rajala, Albert. F. Parlow, L. Cheeseboro, et al. A Transgenic Mouse with a Deletion in the Collagenous Domain of Adiponectin Displays Elevated Circulating Adiponectin and Improved Insulin Sensitivity Endocrinology, January 1, 2004; 145(1): 367 - 383. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
