| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Vol. 54, Issue 4, 589-598, December 2002
Center for Clinical Pharmacology, Department of Medicine, Harbor-University of California Los Angeles Medical Center, Torrance, California
Abstract
I. Introduction
II. Overview of Carnitine Homeostasis
A. Functions of Carnitine
B. Carnitine Balance in Humans
C. Tissue Carnitine Content
III. Impact of Pivalate-Generating Prodrugs on Carnitine Homeostasis
A. Metabolism of Pivalate and Cellular Impact of Pivaloylcarnitine Production
B. Impact of Net Pivaloylcarnitine Excretion
1. Theoretical Considerations.
2. Animal Models.
3. Human Studies.
4. Considerations in Disease Populations.
IV. Regulatory Labeling of Pivalate Prodrugs
V. Conclusions and Clinical Implications
Acknowledgments
References
 |
Abstract |
|---|
 Top NextReferences
|
|---|
Prodrugs that liberate pivalate (trimethylacetic acid) after hydrolysis have been developed to improve the bioavailability of therapeutic candidates. Catabolism of pivalate released by activation of a prodrug is limited in mammalian tissues. Pivalate can be activated to a coenzyme A thioester in cells. In humans, formation and urinary excretion of pivaloylcarnitine generated from pivaloyl-CoA is the major route of pivalate elimination. Because the total body carnitine pool is limited and can only slowly be replenished through normal diet or biosynthesis, treatment with large doses of pivalate prodrugs may deplete tissue carnitine content. Animal models and long-term treatment of patients with pivalate prodrugs have resulted in toxicity consistent with carnitine depletion. However, low plasma carnitine concentrations after pivalate prodrug exposure may not reflect tissue carnitine content and, thus, cannot be used as a surrogate for potential toxicity. The extent of tissue carnitine depletion will be dependent on the dose of pivalate, because carnitine losses may approximate the pivalate exposure on a stoichiometric basis. These concepts, combined with estimates of carnitine dietary intake and biosynthetic rates, can be used to estimate the impact of pivalate exposure on carnitine homeostasis. Thus, even in populations with altered carnitine homeostasis due to underlying conditions, the use of pivalate prodrugs for short periods of time is unlikely to result in clinically significant carnitine depletion. In contrast, long-term treatment with substantial doses of pivalate prodrugs may require administration of carnitine supplementation to avoid carnitine depletion.
| |
I. Introduction |
|---|
|
TopPreviousNextReferences
|
|---|
The use of prodrugs (drugs which undergo covalent modification after administration to yield biologically active molecules) is a well accepted strategy for improving the pharmaceutical, pharmacokinetic, or pharmacodynamic characteristics of a therapeutic agent. Prodrug activation may involve a structural modification to the single administered molecule, such as the conversion of the lactone ring of lovastatin to the active hydroxyacid. Alternatively, the prodrug may be broken down into two molecules, one of which is the active therapeutic moiety, and the other an inactive byproduct. The latter scenario occurs for prodrugs that are esters of the active drug with a carboxylic acid. These derivatives are intended to enhance oral absorption and systemic delivery of the active chemical entity. In vivo hydrolysis of the ester then yields the active drug and the carboxylic acid. The carboxylic acid is then further broken down through endogenous intermediary metabolism.
Pivalate (trimethylacetic acid, Fig. 1) has been used to generate prodrugs to increase oral bioavailability (for example cefetamet pivoxil and pivampicillin). After hydrolysis of the prodrug, pivalate like most carboxylic acids can be activated inside cells for further metabolism by acyl-CoA synthases to form pivaloyl-CoA. However, unlike the coenzyme A thioesters of acetate or isobutyrate, pivaloyl-CoA cannot be oxidized to carbon dioxide in mammalian cells. As a result, pivaloyl-CoA accumulates in cells and is a substrate for a variety of acyl-CoA transferases. These reactions generate pivaloyl conjugates that can be excreted, usually by the kidney. In humans, formation and urinary excretion of pivaloylcarnitine is the dominant route of pivalate elimination. The formation of pivaloylcarnitine, and its excretion in the urine, has the potential to perturb normal cellular function due to the important roles of carnitine in cellular homeostasis.
|
The current manuscript critically examines the theoretical, experimental, and clinical bases of the pivalate-carnitine interaction in humans. The review indicates that although long duration therapy with high doses of pivalate-containing prodrugs may result in clinically important changes in carnitine homeostasis, normal carnitine metabolism protects humans from pivalate-associated toxicity except under extreme conditions. Additionally, plasma or urinary carnitine concentrations are a poor predictor of potential pivalate-associated toxicity.
| |
II. Overview of Carnitine Homeostasis |
|---|
|
TopPreviousNextReferences
|
|---|
A. Functions of Carnitine
Carnitine (Fig. 1) is an endogenous molecule with important
functions in normal intermediary metabolism and cellular physiology (Bremer, 1983
). Biochemical reactions involving carnitine all involve
the reversible transfer of a carboxylic acid moiety (acyl group) from a
coenzyme A thioester (acyl-CoA) to form the corresponding carnitine
ester (acylcarnitine). This reaction is catalyzed by a family of
enzymes, the carnitine acyltransferases (Bieber, 1988). These enzymes
differ based on their subcellular localization and acyl group
structural specificity.
The formation of acylcarnitines is a critical step in the mitochondrial
oxidation of long-chain fatty acids (Bremer, 1983). Long-chain
acyl-CoAs cannot cross the inner mitochondrial membrane to reach the
site of 
-oxidation. In contrast, the acylcarnitine is the substrate
for a transmembrane carrier, which moves the molecule into the
mitochondrial matrix where a second acylcarnitine transferase
regenerates the acyl-CoA for further oxidation. Thus, carnitine is an
obligate for normal mitochondrial fatty acid oxidation, and loss of
tissue carnitine may compromise cellular bioenergetics.
The coenzyme A pool of the cell is important in a large number of
catabolic and anabolic biochemical reactions. As the cellular total
coenzyme A pool is small and cannot be increased quickly, the cell is
dependent on continued turnover of acyl-CoAs to make coenzyme A
available for other reactions. Thus, if a specific acyl-CoA accumulates
in a cell, multiple biochemical pathways may be impaired. The formation
of acylcarnitines under these conditions serves to buffer the coenzyme
A pool from transient acyl-CoA accumulation and make coenzyme A
available for other reactions (Bieber, 1988). In pathologic conditions
of acyl-CoA accretion, this detoxification function of carnitine may be
extremely important and require large amounts of carnitine (Chalmers et
al., 1983). However, this interchange between the coenzyme A and
carnitine pools is ongoing and dynamic. Under a variety of physiologic
conditions, including fasting (Hoppel and Genuth, 1980) or exercise
(Hiatt et al., 1989), the carnitine pool is redistributed between
carnitine and acylcarnitines to reflect a change in the status of the
coenzyme A pool and metabolic state. Thus, such redistribution cannot
be viewed as intrinsically detrimental and must be viewed in the
context of the overall metabolic status of the cell or tissue.
Furthermore, definition of carnitine status requires knowledge of the
total carnitine availability (the sum of carnitine and all
acylcarnitines) and the distribution of the carnitine pool between
carnitine and acylcarnitines.
B. Carnitine Balance in Humans
All tissues that use fatty acids as a fuel source, or require
coenzyme A for cellular reactions, require availability of carnitine for normal function as discussed above. In humans, carnitine is derived
from dietary sources and endogenous biosynthesis, and biosynthesis is
adequate to meet total body carnitine needs (Rebouche and Seim, 1998).
Meat and dairy products are important dietary sources of
carnitine, and variations in dietary carnitine intake can impact plasma
and urinary carnitine contents without any physiologic or functional
impact on the individual (Lennon et al., 1986; Cederblad, 1987). Muscle carnitine content is also independent of dietary carnitine intake (Cederblad, 1987), again illustrating the dissociation between tissue stores and urine or plasma carnitine concentrations.
Carnitine biosynthesis involves a complex series of reactions involving
several tissues (Hoppel and Davis, 1986). The carbon backbone for
carnitine is derived from lysine. Lysine in protein peptide linkages
undergoes methylation of the 
-amino group to yield trimethyllysine,
which is released upon protein degradation. Muscle is the major source
of trimethyllysine. The released trimethyllysine is further oxidized to
butyrobetaine and ultimately hydroxylated to form carnitine. The last
reactions of carnitine biosynthesis occur primarily in the liver and
the kidney.
There are no catabolic reactions involving carnitine in mammalian
cells, and the only route of elimination is through urinary elimination. Both carnitine and acylcarnitines appear in the urine. In
normal kidneys, greater then 90% of the filtered carnitine is
reabsorbed (Engel et al., 1981; Rebouche et al., 1993). This reabsorption is saturable, with a transport maximum of 60 to 100 µM.
Urinary acylcarnitines derive from both filtered plasma acylcarnitines and products of renal metabolism. Filtered acylcarnitines are reabsorbed to varying degrees based on the specific acyl moiety present.
C. Tissue Carnitine Content
The total body carnitine pool is extremely dynamic, with carnitine
and acylcarnitines moving between the gastrointestinal tract (after
absorption), the liver (after biosynthesis), the kidney (for
elimination), and tissues such as heart or skeletal muscle that require
carnitine for function. Within a tissue, the carnitine pool is
redistributed between carnitine and acylcarnitines as metabolic shifts
occur. Neither carnitine nor acylcarnitines can efficiently diffuse
across plasma membranes, and tissue-specific transport systems exist to
move carnitine into and out of tissues (Brass, 1992; Kerner and Hoppel,
1998). As a result, tissues vary enormously in their total carnitine
content and the kinetics of carnitine homeostasis. Furthermore,
dramatic changes in carnitine homeostasis may occur in one biologic
compartment, which are not reflected in other compartments. The
specific gene and corresponding protein primarily responsible for renal
and muscle carnitine transport have been characterized (Wang et al.,
1999; Lahjouji et al., 2001).
Plasma serves only to carry carnitine and acylcarnitines between tissues, and as such its concentrations are relatively low (Table 1). As carnitine serves no metabolic function within plasma, changes in plasma carnitine concentrations can only be interpreted in the context of other metabolic or tissue-specific information. Kidney, liver, heart, and skeletal muscle all contain carnitine at concentrations in excess of those found in plasma (Table 1). Due to the large amount of skeletal muscle, most of the total body carnitine is present in the skeletal muscle compartment, with very little in the plasma or extracellular compartments (Table 1).
|
Tissue carnitine turnover rates also vary widely (Brooks and McIntosh,
1975; Rebouche and Engel, 1984). The liver appears to rapidly
equilibrate with the plasma compartment, both in respect to total
carnitine content (albeit at concentrations higher than plasma) and
carnitine-acylcarnitine distribution (Brass and Hoppel, 1980). In
contrast, the large skeletal muscle carnitine pool interacts only
sluggishly with plasma. Thus, very large changes may occur acutely in
the plasma compartment through acute carnitine administration (Brass et
al., 1994) or acute carnitine depletion during hemodialysis (Guarnieri
et al., 1987; Wanner et al., 1987) without any significant perturbation
in skeletal muscle carnitine content or function.
Processes affecting net carnitine balance (biosynthesis, elimination,
absorption) will most immediately affect the plasma compartment.
However, the plasma carnitine pool interacts with the discrete
carnitine pools of specific organs, as discussed above. Thus, changes
in any aspect of carnitine homeostasis eventually will be reflected in
all tissues in the body. The relative impact at steady state will be
dependent on the size of the perturbation and the total body carnitine
pool. Before steady-state redistribution of the carnitine pool,
individual compartments may show large changes, as is the case with the
plasma carnitine depletion with hemodialysis. In the case of
hemodialysis, the plasma carnitine pool is restored from tissue stores
within 1 to 2 days after a hemodialysis session (Bartel et al., 1981;
Guarnieri et al., 1987; Wanner et al., 1987). Due to the large amount
of carnitine in tissue compared with extracellular fluid, the
extracellular compartment may be replenished with only a small
physiologically irrelevant decrement of the tissue stores (Table
2).
|
| |
III. Impact of Pivalate-Generating Prodrugs on Carnitine Homeostasis |
|---|
|
TopPreviousNextReferences
|
|---|
A. Metabolism of Pivalate and Cellular Impact of Pivaloylcarnitine Production
Pivalate generated from prodrug administration can enter cells for
further metabolism. As for other organic acids, formation of a
thioester with coenzyme A is a critical step in cellular metabolism of
pivalate. The ability to generate pivaloyl-CoA has been demonstrated
directly in rat hepatocytes (Ruff and Brass, 1991) and indirectly in
rat heart cells (Diep et al., 1995b). This reaction most likely occurs
in most mammalian cells. However, in contrast to most organic acids,
pivaloyl-CoA cannot be further metabolized in any mammalian cells. As
pivaloyl-CoA accumulates, the pivaloyl moiety can be transferred from
coenzyme A to carnitine to form pivaloylcarnitine. The cellular
coenzyme A and carnitine pools appear to be in equilibrium. Thus, the
degree of pivaloyl-CoA accretion will be reflected in the carnitine
pool as pivaloylcarnitine. In animal models, administration of pivalate
is associated with pivaloylcarnitine accumulation and a shift in the
carnitine pool toward acylcarnitines (Bianchi and Davis, 1991; Diep et
al., 1995a).
Despite the accumulation of pivaloyl-CoA and pivaloylcarnitine in
cells, cellular metabolism is well preserved in in vitro model systems.
A build-up of acyl-CoAs may be toxic if the compound inhibits other
cellular pathways directly or through depletion of coenzyme A (Brass,
1994). Similarly, the resultant shift of the carnitine pool toward
acylcarnitines may decrease carnitine availability. However, even at
very high concentrations (greater than 1 mM), pivalate does not have
major effects on glucose or lipid metabolism (Ruff and Brass, 1991; Ji
and Tremblay, 1993). This most likely reflects the lack of direct
toxicity of pivaloyl-CoA, and the preservation of adequate free
coenzyme A and carnitine under the kinetics of the synthase and
transferase reactions in the cell. Thus, direct or acute cellular
toxicity of pivalate is not a major concern with pivalate-containing prodrugs.
Cellular accumulation of acylcarnitines leads to export of the
compounds to the plasma, from which the acylcarnitine can be transported to other tissues for metabolism or transported to the
kidney for elimination in the urine. Specific transport mechanisms may
facilitate the cellular export of acylcarnitines across plasma membranes, with carnitine cotransported from the plasma into the cell
through an exchange mechanism (Sartorelli et al., 1985). Net
elimination occurs through urinary excretion of pivaloylcarnitine because no mammalian tissue can further catabolize pivaloyl-CoA.
B. Impact of Net Pivaloylcarnitine Excretion
1. Theoretical Considerations.
As discussed above, once
generated, pivaloylcarnitine will move from tissues to plasma, and
ultimately be excreted in the urine. At high pivalate doses, the loss
of pivaloylcarnitine may exceed normal physiological carnitine
excretion. The net loss will reflect the stoichiometry of the pivalate
dose and the percentage of the pivalate eliminated as
pivaloylcarnitine. Pivalate may be excreted as the glucuronic acid
conjugate in animals (Vickers et al., 1985), whereas pivaloylcarnitine
is the dominant route of pivalate elimination in humans. Thus, for each
mole of pivalate or pivalate-generating prodrug administered, there is
the potential to lose one mole of carnitine as pivaloylcarnitine in the
urine. This may be offset by decreased excretion of carnitine or other acylcarnitines, by increased synthesis or dietary intake, or by non-pivaloylcarnitine elimination of pivalate. However, this construct allows worst-case estimates to be made and clearly emphasizes that any
impact of pivalate prodrug administration on carnitine homeostasis will
be dependent on the dose of prodrug given and the duration of therapy.). Daily carnitine biosynthesis
is approximately 0.07 mmol, and dietary intake may be estimated as 0.28 mmol per day (Lombard et al., 1989). Thus, unless therapy was
maintained for an extended period of time, it would be predicted that
due to the large carnitine stores, pivalate prodrug administration
would not adversely affect metabolism. This theoretical conclusion is
supported by experiments in animals and clinical experience, as
detailed below.
2. Animal Models. The effect of pivalate administration on carnitine homeostasis and metabolism has been studied in animals. Animal models will not extrapolate directly to humans due to species differences in carnitine and pivalate metabolism, but nonetheless provide insight into the in vivo pivalate-carnitine interaction.
Bianchi and Davis (1991) administered sodium pivalate to rats for 8 weeks via incorporation into their drinking water. Within 4 days,
urinary total carnitine elimination was markedly elevated, and plasma
total carnitine concentration was depressed to less then 40% of
control values. In contrast, total carnitine content of skeletal muscle
was unaffected after 4 days. This illustrates the large pool size and
slow turnover of the skeletal muscle carnitine pool, and that plasma
concentration is not indicative of tissue stores. With 8 weeks of
treatment, plasma total carnitine concentrations fell to 25% of
control values, and total carnitine content of skeletal muscle was
approximately 50% of control values. Thus, cumulative dose is critical
in defining the impact of pivalate administration on carnitine stores.
Liver and heart carnitine content fell to variable degrees during the 8 weeks of treatment. Despite these apparently dramatic changes in
carnitine levels, animals on pivalate ate and gained weight at the same
rate as controls, and their metabolic response to starvation was
unaffected, except for an exaggerated ketosis. As ketone bodies are the
products of fatty acid oxidation, this accelerated ketosis is
surprising since decreased carnitine content might have been predicted
to decrease fatty acid oxidation and ketosis. In contrast, liver fatty
acid oxidation was apparently preserved despite the 8 weeks of pivalate
exposure and carnitine losses.
Diep et al. (1992) administered large doses of pivampicillin (630 mg/kg) orally to rats for 24 days. Plasma total carnitine concentrations fell by 60%, liver content by 70%, but heart total carnitine content was only 35% less than controls and skeletal muscle
content was 30% less than controls. The hepatic response to fasting as
assessed by plasma -hydroxybutyrate concentrations was mildly
accentuated, as seen by Bianchi and Davis (1991). Nakajima et al.
(1996) also demonstrated relatively preserved hepatic metabolic function despite pivalate-induced carnitine depletion. Adaptive compensatory mechanisms, including an increase in carnitine
biosynthesis, may further limit the net impact of pivaloylcarnitine
losses on tissue carnitine stores (Nakajima et al., 1999).
Broderick et al. (1995) administered pivalate (approximately 192 mg of
pivalate/kg of body weight) to rats for up to 28 weeks. This protracted
exposure resulted in a 60% decrease in heart total carnitine content.
In vitro assessment of these hearts revealed impairment in fatty acid
oxidation that limited cardiac work. In contrast, treatment of rats
with pivalate for two weeks was associated with a 24% reduction in
heart carnitine content but no change in cardiac function under control
conditions (Broderick et al., 2001). Two weeks of pivalate exposure was
associated with impaired recovery of cardiac function after a no-flow
ischemic insult in this rat model (Broderick et al., 2001).
Rat pups born with mothers who were treated with pivalate throughout
pregnancy demonstrated decreased plasma and tissue carnitine contents
(Ricciolini et al., 2001). This is consistent with the observation in
humans that neonatal plasma carnitine concentrations are influenced by
maternal carnitine concentrations (Novak et al., 1981). It is not clear
if maternal pivalate concentrations are high enough to result in fetal
pivalate exposure. If rats born from pivalate treated mothers are
continued on pivalate treatment for the first four months of life, the
animals develop hyperglycemia, insulin resistance and increased body
mass (Ricciolini et al., 2001). The relevance of this extreme pivalate
exposure to other models or potential therapeutic use of pivalate
prodrugs is unclear.
Thus, studies of pivalate administration to animals confirm the loss of
carnitine as pivaloylcarnitine. Net losses occur preferentially in the
plasma and liver compartments. The degree of depletion is a function of
cumulative dose exposure. The large tissue carnitine stores allow
preservation of tissue function except under extreme degrees of
pivalate exposure and associated carnitine depletion.
3. Human Studies.
The formation of pivaloylcarnitine
following pivalate-containing prodrug administration has been confirmed
in human studies, and these reports demonstrate that urinary
pivaloylcarnitine excretion is the dominant route of pivalate
elimination (Vickers et al., 1985; Melegh et al., 1987, 1990; Konishi
and Hashimoto, 1992). Melegh et al. (1987), studying children receiving
a 7-day course of pivampicillin (2 g/day), demonstrated that the
pivaloylcarnitine excretion was associated with a 39% decrease in
plasma total carnitine, and a 73% decrease in plasma carnitine
concentration. This study also demonstrated that urinary
pivaloylcarnitine clearance approximated creatinine clearance, implying
lack of pivaloylcarnitine reabsorption in the kidney and explaining the
rapid fall in plasma carnitine concentrations.
confirmed the acute effects of pivampicillin and
pivmecillinam to increase urinary pivaloylcarnitine excretion and
decrease plasma carnitine concentrations. They studied a group of
children on long-term treatment with a combination of pivampicillin and
pivmecillinam for urinary tract infections. These patients had been on
the antibiotics for as long as 30 months, and demonstrated dramatic
decreases in their plasma carnitine concentrations. Furthermore, in two
patients who underwent muscle biopsy, muscle carnitine content was
decreased more then 75% compared with reference values. Although no
specific signs or symptoms were associated with these low carnitine
contents, the authors note that they had previously seen two patients
with hypoketotic hypoglycemia with pivaloylcarnitine in their urine.
Furthermore, patients on the pivalate prodrugs frequently complained of
being tired, and two subjects had discontinued antibiotic therapy due
to nonspecific symptoms. Thus, long-term treatment with pivalate
prodrugs may cause substantial depletion of muscle carnitine stores.
However, the degree of depletion required to produce clinical symptoms
or signs could not be identified. Of note, in patients with inherited
muscle carnitine deficiency due to a mutation in the plasma membrane
transport system, carnitine therapy is effective in relieving symptoms
despite yielding only very low muscle carnitine contents (Schulte et
al., 1990; Kerner and Hoppel, 1998).
Diep et al. (1993) studied an additional 6 patients who had been on
pivaloyl-containing antibiotics for as long as 24 months. As expected,
these patients all showed low serum carnitine concentrations, and with
discontinuation of therapy, serum total carnitine concentrations did
not return to the reference range until periods ranging from 5 to 12 months. Only one of the six patients demonstrated symptoms or signs
potentially related to carnitine deficiency. This subject, an
82-year-old woman who had received pivmecillinam for 8 months, reported
muscle weakness, anorexia, and cognitive impairment. This patient had
blunted ketosis with starvation and dicarboxylic aciduria. The
metabolic abnormalities and symptoms resolved ten months after
pivmecillinam therapy was discontinued. The patient had no signs or
symptoms of cardiac dysfunction.
Holme and colleagues (1992) studied 17 children who had been treated
with a combination of pivmecillinam and pivampicillin for at least 1 year. All had low serum carnitine concentrations (average 2.0 µM).
Several patients were described as having nonspecific symptoms such as
tiredness, behavioral problems or weakness, which resolved after
discontinuation of the antibiotic or with carnitine supplementation.
Four patients had muscle biopsies for carnitine quantitation and
demonstrated carnitine contents that were less then 20% of control
values. Six of the patients underwent a fasting test to assess the
metabolic status of the liver. Three of these patients developed
hypoglycemia during the fast and were characterized by the authors as
having a blunted ketogenic response. Blunted ketosis and hypoglycemia
would be consistent with impaired liver fatty acid oxidation. The
authors emphasized the relationship between cumulative dose and
carnitine depletion, and estimated that half the body stores would be
lost after 5 months of therapy.
The effects of 7 to 8 weeks of treatment with pivmecillinam (1200 mg/day) were studied in seven adults (Abrahamsson et al., 1995b).
Muscle total carnitine content fell by 55% during the treatment
period. Serial echocardiography demonstrated a 10% decrease in left
ventricular mass, but a number of functional assessments, including
left ventricular ejection fraction and fractional shortening, were
unaffected by the treatment.
In contrast to the dramatic effects of long-term pivalate-prodrug
administration, Abrahamsson et al. (1994) demonstrated less dramatic
effects of short-term treatment. Six patients were given 1200 mg of
pivmecillinam for 12 days. Carnitine concentrations fell from an
average 42.8 µM to 11.6 µM. In contrast, muscle total carnitine
content did not change, emphasizing the compartmentalization of
carnitine homeostasis. Two subjects had a blunted ketosis during a 36-h
fast post-treatment as compared with pretreatment, which was
interpreted by the authors to reflect clinically significant hepatic
carnitine deficiency. However, two subjects showed equivalent ketotic
responses, while two subjects did not demonstrate ketosis during the
pre- or post-treatment fast, making generalizing the results difficult.
In a follow-up study, the same authors (Abrahamsson et al., 1995a)
demonstrated decreases in serum carnitine concentrations after 7 days
of pivmecillinam therapy.
Ito et al. (1995) also observed a decrease in blood carnitine
concentrations with short-term (3-14 days) treatment with a pivalate
prodrug, and suggested that the hypocarnitinemia was associated with
increased ammonia concentrations. However, the blood ammonia
concentrations were within the normal range, no patient developed any
symptoms, and the ammonia concentrations pretreatment were not
different from those during drug therapy. Statistically significant
changes in ammonia concentration were only demonstrated when comparing
post-treatment to concentrations during therapy, raising the
possibility that the acute infection may have contributed to the
differences noted. Melegh et al. (1997) suggested that short-term
pivampicillin therapy for the treatment of bacterial pharyngitis was
associated with a decrease in the use of fat as an energy source and an
increase in the respiratory exchange quotient.
Therapy with carnitine during a course of treatment with pivalate
prodrugs can correct the hypocarnitinemia associated with pivaloylcarnitine excretion (Holme et al., 1992; Melegh et al., 1997;
Pap et al., 1999). However, the clinical significance of this
correction is unknown. In potentially symptomatic patients with severe
carnitine depletion secondary to long-term pivalate exposure, carnitine
supplementation may accelerate symptom relief (Holme et al., 1992).
Thus, the clinical experience with pivalate-containing prodrugs is
congruent with both the theoretical considerations and the animal
models. Short-term exposure (less then 2 weeks) and molar pivalate
exposures equivalent to less then 10 to 20% of total body carnitine
stores are associated with hypocarnitinemia, but no significant
decrease in muscle carnitine stores and no clinical signs or symptoms.
Treatment for periods of 1 to 2 months are associated with decreases in
muscle carnitine content, but subjects remain asymptomatic due to the
extent of body carnitine stores. In contrast, long-term treatment
(greater then 6 months) with pivalate prodrugs may result in dramatic
depletion of carnitine stores and signs or symptoms consistent with
carnitine deficiency.
4. Considerations in Disease Populations.
The effect of
pivalate-prodrug administration in humans may be accentuated by
underlying disease or metabolic conditions. As carnitine depletion is
stoichiometric with the pivalate dose, the impact may be increased if
body carnitine stores are significantly below normal. A rare inherited
mutation in a critical carnitine transport will result in systemic
carnitine deficiency and inadequate carnitine stores (Stanley, 1995;
Pierpont et al., 2000; Lahjouji et al., 2001). These patients,
estimated as fewer then 100 worldwide, present with severe metabolic
disturbances in childhood. Similarly, patients with inherited organic
acidurias characterized by acyl-CoA accretion may develop carnitine
insufficiency, or secondary carnitine deficiency, due to increased
acylcarnitine excretion (Chalmers et al., 1983). These patients will be
under the care of a physician skilled in metabolic disorders and will
often be on carnitine supplementation as part of their therapeutic regimen.
; Hiatt et al., 1992). Muscle carnitine
stores (per gram of muscle) have been shown to be preserved in diabetes
(Cederblad et al., 1977), alcohol abuse (de Sousa et al., 1988),
patients with protein-calorie malnutrition (Wennberg et al., 1992),
liver disease (Wennberg et al., 1992), and in woman compared with men
(Lennon et al., 1986; Opalka et al., 2001). Muscle skeletal content has
been reported to be unaffected by aging (Cederblad et al., 1976), or to
be unaffected in women with men showing a small decline with aging
(Opalka et al., 2001). Thus, these diverse populations would not
represent substantially increased theoretical risk during short-term
exposure to pivalate prodrugs.
Carnitine stores will also be relatively reduced if muscle mass is
decreased. This decrease may result from small stature, with normal
muscle per body weight, or with a decrease in muscle mass per body
weight (for example, in paraplegics). Thus, children will have smaller
carnitine stores then adults. However, if pivalate exposure is
normalized per kilogram of body weight, no increase in risk will result
due to the stoichiometric nature of the pivalate-carnitine interaction.
Premature infants may represent an exception to this due to a lower
carnitine biosynthetic capacity and incompletely developed body
carnitine stores (Shenai and Borum, 1984). Muscle carnitine stores
appear to be near adult levels in full term infants (Shenai and Borum,
1984). A reduction in muscle mass per body weight will not affect total
carnitine stores dramatically except in conditions such as morbid
obesity, where pivalate-containing prodrugs might be best dosed on a
lean body mass basis to yield equivalent pivalate exposures.
Valproate (2-propylpentanoic acid) is a short-chain fatty acid used
therapeutically as an anticonvulsant, primarily in pediatric populations (Wallace, 1988). Like pivalate, valproate can be activated to the corresponding acyl-CoA (valproyl-CoA). Unlike pivalate, valproyl-CoA can be metabolized by mammalian tissues to a variety of
metabolites (Bjorge and Baillie, 1991; Li et al., 1991; Sugimoto et
al., 1996). Although valproylcarnitine can be generated from valproyl-CoA, it is a quantitatively minor pathway of valproate elimination (Millington et al., 1985; Melegh et al., 1990) and thus,
valproate-induced absolute secondary carnitine deficiency is not a
major consideration. Most studies (Ohtani et al., 1982; van
Wouwe, 1995; Castro-Cago et al., 1998) but not all (Hirose et al.,
1998) demonstrate that valproate therapy is associated with a mild
decrease in plasma carnitine concentrations (carnitine concentrations
approximately 20 µM versus a normal range of 30 to 50 µM). Data on
tissue carnitine contents during valproate therapy are not available.
Lower plasma carnitine concentrations may be more common in
valproate-treated individuals with underlying disease (Ohtani et al.,
1982; Beghi et al., 1990), but the differential effects of the disease
compared with valproate are difficult to define.
Valproate therapy is associated with a rare, severe hepatoxicity that
is similar to Reye's syndrome (Bryant and Dreifuss, 1996). This
hepatotoxicity has focused attention on the metabolic effects of
valproate. Although pivaloyl-CoA appears to be relatively metabolically
inert except for coenzyme A sequestration (Ruff and Brass, 1991),
valproyl-CoA and its metabolites have profound effects on hepatic
intermediary metabolism (Turnbull et al., 1983; Ito et al., 1990;
Ponchaut et al., 1992). Although carnitine supplementation can
"normalize" the plasma carnitine pool during valproate therapy (Kossack et al., 1993), there are no data demonstrating that carnitine supplementation changes the incidence or severity of the Reye's-like toxicity associated with valproate. Similarly, no data are available that demonstrate an association between the status of the carnitine pool and the development of this rare hepatic toxicity. Studies that
report carnitine abnormalities in patients with valproate-associated hepatotoxicity (Krahenbuhl et al., 1995) are difficult to interpret due
to the expected changes resulting from the metabolic derangements of
the acute hepatic failure. Thus, the data do not support a theoretical
concern that coadministration of pivalate-prodrugs with valproate
increases the risk of severe hepatotoxicity, and non-carnitine related
effects of valproate metabolites are likely to play an etiological role
in this syndrome. As the host factors responsible for the idiosyncratic
development of hepatotoxicity are not known, there is no basis to link
any aspect of pivalate metabolism with this syndrome.
| |
IV. Regulatory Labeling of Pivalate Prodrugs |
|---|
|
TopPreviousNextReferences
|
|---|
Cefditoren pivoxil was recently approved for marketing in the United States and is the only pivalate-containing prodrugs currently marketed in the United States. The U.S. cefditoren pivoxil label notes that the drug is contraindicated in patients with carnitine deficiency or metabolic defects associated with clinically significant carnitine deficiency. Cefditoren pivoxil is intended for short-term therapy, and thus consistent with the concepts discussed above, the label indicates that neither measurement of carnitine concentrations nor carnitine supplementation are recommended with its use.
The labeling of pivalate prodrugs marketed in other countries reflects varied responses to the potential toxicity by manufacturers and regulatory agencies. In Japan, the package inserts for cefteram pivoxil, cefetamet pivoxil, and cefcapen pivoxil all contain a statement that use may reduce serum carnitine. The label for cefetamet pivoxil in Germany contains a contraindication statement for patients with primary carnitine deficiency, as well as for patients on hemodialysis or those with diabetes mellitus. The label also warns against use with valproate. As discussed above, these contraindications without respect to duration of therapy in hemodialysis, diabetes and with valproate are theoretical at best and may not be supported by a careful review of carnitine homeostasis during pivalate exposure. Similar warnings appear on the pivampicillin label in Sweden, with specific mention of unstable diabetes, premature children, patients on valproate, and patients with very small muscle mass. In contrast, the pivampicillin labeling in Canada emphasizes the absence of clinical adverse events with short-term use despite the lowering of serum carnitine concentrations and concludes that neither the monitoring of serum carnitine concentrations or the administration of carnitine supplementation are indicated during pivampicillin use. The United Kingdom labeling for this product cautions against use in patients with carnitine deficiency or those using valproate.
| |
V. Conclusions and Clinical Implications |
|---|
|
TopPreviousNextReferences
|
|---|
Animal and clinical studies support a model in which pivalate
prodrug administration results in pivaloyl-CoA production that does not
adversely affect cellular function. The pivaloyl-CoA accumulation in
turn will result in pivaloylcarnitine production, a shift in the
distribution of the carnitine pool toward acylcarnitines, and a net
increase in total carnitine urinary losses. The impact of these losses
on the extracellular carnitine pool may be dramatic, but without
clinical sequelae. Tissue total carnitine content will only be impacted
as the urinary losses are sustained for a period of time. It is likely
that the liver will be the first organ affected, but despite loss of
carnitine and a shift in distribution from carnitine toward
pivaloylcarnitine, hepatic function is not adversely affected by
short-term exposure to pivalate. Continued pivalate exposure for months
may more dramatically impact the liver carnitine pool and lead to
depletion of the large muscle carnitine stores. Even under these
conditions, residual carnitine stores appear sufficient to maintain
critical tissue functions, and symptoms appear only under the most
extreme conditions. Thus pivalate prodrugs intended for chronic use,
such as the antiretroviral adefovir dipivoxil, incorporate carnitine
supplementation as part of the dosing regimen (Kahn et al., 1999).
The above model is expected to be only minimally impacted by a variety of underlying disease conditions. Conditions of carnitine deficiency, either primary or secondary to a metabolic defect, might dramatically diminish carnitine stores and/or increase carnitine requirements. However, these patients are rare, should be under the care of a physician, and are often on carnitine supplementation that will mitigate the pivalate effects. Other chronic diseases appear to be associated with relative preservation of carnitine stores.
Clinical experience with pivalate prodrugs appears to support this clinical safety profile with short-term use. It is estimated that between 1996 and 2001, almost 1 billion treatment days experience was accumulated world wide with pivalate prodrugs (data from IMS Health, Fairfield, CT). Despite this broad exposure, published reports of clinical toxicity have been limited to patients with long duration treatment. It is likely that a broad spectrum of populations have been exposed to these drugs without any risk groups being identified. Thus, this experience is also consistent with the safety of short-term use of pivalate prodrugs.
The development of pivalate-containing prodrugs may facilitate optimization of key characteristics for novel therapeutic agents. These drugs will result in perturbations in carnitine homeostasis. In the absence of underlying carnitine deficiency, this change in carnitine metabolism is well tolerated unless the cumulative dose of pivalate (in moles) exceeds the scope of the body carnitine stores. Thus, in circumstances where chronic treatment is anticipated, pivalate-based modifications should be avoided during drug development. The changes in carnitine homeostasis should not be viewed as contraindications for the development of pivalate-prodrugs intended for short-term use with small cumulative doses (for example less than 0.5 mmol/kg of body weight) when pivalate provides properties superior to other potential carboxylic acids or other strategies.
| |
Acknowledgments |
|---|
|
TopPreviousNextReferences
|
|---|
The author thanks Dr. Charles L. Hoppel for his thoughtful comments on this manuscript. The author is a consultant to TAP Pharmaceutical Products, Inc. (Lake Forest, IL) and Sigma-Tau Pharmaceuticals, Inc. (Gaithersburg, MD). TAP Pharmaceutical Products provided support for the preparation of this review.
| |
Footnotes |
|---|
Address correspondence to: Dr. Eric P. Brass, Center for Clinical Pharmacology, Harbor-UCLA Medical Center, 1124 West Carson Street, Torrance, CA 90502. E-mail: ebrass{at}ucla.edu
| |
References |
|---|
|
TopPrevious |
|---|
a risk irrespective of age.
Biochem Mol Med
55:
77-79[CrossRef][Medline].-oxidation of valproic acid in rat liver mitochondrial preparations.
Drug Metab Dispos
19:
823-829[Abstract].metabolism and functions.
Physiol Rev
63:
1420-1480mass spectrometry.
Clin Chem Acta
145:
69-76[CrossRef][Medline].
0031-6997/02/5404-589-598$7.00
PHARMACOLOGICAL REVIEWS
Copyright © 2002 by The American Society for Pharmacology and Experimental Therapeutics
This article has been cited by other articles:
|
|
 |
|
  A. Koch, B. Konig, G. I. Stangl, and K. Eder PPAR{alpha} Mediates Transcriptional Upregulation of Novel Organic Cation Transporters-2 and -3 and Enzymes Involved in Hepatic Carnitine Synthesis Experimental Biology and Medicine, March 1, 2008; 233(3): 356 - 365. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. D. Cornwell and R. G. Ulrich Investigating the Mechanistic Basis for Hepatic Toxicity Induced by an Experimental Chemokine Receptor 5 (CCR5) Antagonist Using a Compendium of Gene Expression Profiles Toxicol Pathol, June 1, 2007; 35(4): 576 - 588. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. Rytting and K. L. Audus Novel Organic Cation Transporter 2-Mediated Carnitine Uptake in Placental Choriocarcinoma (BeWo) Cells J. Pharmacol. Exp. Ther., January 1, 2005; 312(1): 192 - 198. [Abstract] [Full Text] [PDF]
|
|
| ||||||
