Robert J. Delorenzo MD, PhD, MPH*
David A. Sun PhD**
* George B. Bliley III Professor of Neurology, Department of Neurology, Virginia Commonwealth University, Richmond, Virginia
** Research Scientist, Department of Neurology, Virginia Commonwealth University, Medical College of Virginia, Richmond, Virginia
Phenytoin [diphenylhydantoin, Dilantin (Parke-Davis, Morris Plains, NJ); Figure 57.1], first synthesized in 1908 by Biltz (6), was introduced for the treatment of epilepsy in 1938 by Merritt and Putnam (87). The success of phenytoin as an anticonvulsant was one of the major pharmacologic advances in treating neurologic diseases and favorably altered the lives of many people with epilepsy worldwide. Phenytoin produces effective anticonvulsant action without troublesome sedation, and it is one of the most effective compounds for treating generalized tonic-clonic seizures (33,39) and status epilepticus (34,35).
As the first anticonvulsant drug without hypnotic effects (96,97), phenytoin had a major impact on patient care (33, 34, 35, 36, 37). Overdoses of phenytoin experimentally and clinically had excitatory rather than depressant effects (89). Because of its clinical importance, phenytoin has been extensively studied in both clinical and laboratory investigations (19,38,39,44). Phenytoin has been useful in studies elucidating the pathophysiology of epilepsy. Woodbury and coworkers (129, 130, 131, 132, 133, 134, 135) have contributed extensively to the understanding of the effects of phenytoin on nervous tissue and have written several reviews concerning the history, clinical uses, and mechanisms of action of this compound (2,32,57,80,82,104). More recently, Macdonald's group (78, 79, 80, 81, 82) and other laboratories (118,125) have provided evidence of a direct effect of phenytoin on the sodium channel.
Although limitations of space do not permit a complete review of the mechanisms of action of phenytoin, this chapter summarizes its major properties and effects on nervous tissue.
MULTIPLE ACTIONS OF PHENYTOIN
Early Phenytoin Research
Early research on phenytoin was directed at characterizing its metabolism, toxic effects, and clinical efficacy. This research is relevant to the basic mechanism of action of phenytoin because numerous aspects of its effect on the body indicated that it probably had multiple sites of action on physiologic function.
After oral or intravenous administration, phenytoin is widely distributed in the body with high plasma protein binding, giving maximum plasma solubility at 37°C of approximately 75 µg/mL (64). Phenytoin is not given by intramuscular injection. If the pH of the carrying vehicle is adjusted well below 7.8, phenytoin precipitates out of solution. Thus, with intravenous injection the high local concentration of the drug in a neutral pH environment around the injection site causes the drug to precipitate out of solution, with consequent slow absorption. For many years, it was thought that adequate phenytoin levels could be obtained only after several days of phenytoin loading to saturate all of the fat and other body stores for phenytoin. Loading with appropriate intravenous doses of phenytoin can produce adequate serum concentrations within 10 to 20 minutes. This allows phenytoin to be administered for the treatment of status epilepticus and other acute seizure problems. It also enables laboratory investigators to administer phenytoin in high concentrations quickly to various animals. The wide distribution of phenytoin in the brain
and other body tissues has been studied extensively (91,135). Phenytoin clearly can have adverse effects on the skin, gastrointestinal system, and other major organ systems. The classic phenytoin facies is well documented and demonstrates that phenytoin affects other organ systems besides the central nervous system.
FIGURE 57.1. Structure of phenytoin.
The toxic effects of phenytoin are numerous and have been extensively characterized. Fortunately, phenytoin has a very high therapeutic index, and therapeutic levels can be maintained without serious side effects. Consequently, clinical overdoses of phenytoin are easily recognized and usually not severe. Phenytoin can produce nystagmus, ataxia, and instability of gait at lower toxic levels. At higher toxic levels, dysarthria, incoordination, and unsteadiness often are seen. At very toxic levels, exceeding 30 µg/mL, phenytoin can produce drowsiness, lethargy, and coma. In addition, phenytoin at very high toxic levels can directly damage the cerebellar vermis, producing midline ataxia and a widebased gait. Phenytoin at high toxic levels also can produce diplopia, hypotension, cardiac suppression, and even death. An important early finding indicated that phenytoin also can produce hyperexcitability, as well as irritability, hallucinations, and even psychotic reactions. These early clinical observations contributed to our understanding of the basic actions of phenytoin, indicating that it worked on multiple regulatory systems in the nervous system. Phenytoin, under certain conditions, clearly can act as an anticonvulsant and neuronal stabilizing compound. In other circumstances, however, it can be excitatory, and even cause psychotic phenomena.
Another hallmark of the multiple effects of phenytoin on the body is the gingival hypertrophy that can occur in patients younger than 21 years of age. This side effect has limited the use of phenytoin in children, although careful oral hygiene can minimize the problem, and research is being directed at inhibiting the effects of phenytoin on gingival cell growth. In adults, phenytoin does not produce much gingival hypertrophy. The developmental relationship of this side effect to the action of phenytoin and the specific cellular mechanisms by which phenytoin produces this effect are not clearly understood and could have important ramifications for developmental neurobiology. Phenytoin may cause mild hirsutism in some patients. It can have an adverse effect on the hematopoietic system and produce, in rare instances, megaloblastic anemia. Phenytoin rarely may produce lupus erythematosus, and is well known to exacerbate this condition.
These early clinical investigations demonstrated that phenytoin affected multiple organ systems and several physiologic and biochemical processes in the body. Much of the research directed at understanding the mechanisms of action of phenytoin was initiated to explain the many effects of this compound in clinical and laboratory settings.
Major Sites of Actions
Phenytoin is probably the most widely studied anticonvulsant. The numerous effects of phenytoin on electrophysiologic and biochemical systems in brain and other tissues have been extensively reviewed by Woodbury and coworkers (129, 130, 131, 132, 133, 134, 135) and others (32,39,78, 79, 80, 81, 82,104,107). These studies have shown that phenytoin acts on ion conductances, sodium-potassium adenosine triphosphatase (ATPase) activity, various enzyme systems, synaptic transmission, posttetanic potentiation (PTP), neurotransmitter release, and cyclic nucleotide metabolism. These findings suggest that phenytoin has many sites of action in the central nervous system and that it most likely interacts with numerous biochemical processes that regulate neuronal function.
A different point of view from this original theory is that phenytoin may interact with a few important major regulatory systems in the nervous system that could then regulate numerous other cellular processes controlled by these regulators (26, 27, 28, 29, 30, 31). Phenytoin has been shown to regulate sodium-potassium ATPase and sodium ion channels. This could have widespread regulatory effects on numerous excitatory and inhibitory systems. Phenytoin also has been shown to regulate calcium-calmodulin-dependent enzyme systems, which also may provide insight into the multiple effects of phenytoin. The effects of phenytoin on cyclic nucleotide metabolism levels also could provide a focal point for multiple effects on cell function. Thus, the effects of phenytoin on several second messenger systems, such as the cyclic nucleotides and calcium systems, might explain the widespread action of this compound on numerous cells and physiologic functions. However, no single action of phenytoin is likely to explain all of its diverse effects on the nervous system.
Whether phenytoin has few or numerous sites of action is still a matter of debate. Although the precise anticonvulsant effect or effects of this compound on neuronal tissue need to be elucidated, the effect of phenytoin on the sodium channel is becoming widely accepted as a major mechanism of action. The following topics represent those areas of research that scientists generally agree have some importance in the action of phenytoin.
EFFECTS OF PHENYTOIN ON NEURONAL EXCITABILITY
Phenytoin limits the development of maximal seizure activity and reduces the spread of seizure discharge from a seizure focus. Both of these experimental observations are pertinent to the clinical effects of phenytoin on generalized tonic-clonic seizures and focal epilepsy. A major anticonvulsant effect of phenytoin is believed to be its ability to
block the epileptogenic focus from recruiting surrounding neurons, preventing the spread of seizure discharge.
In contrast to phenobarbital, phenytoin does not significantly elevate the threshold for seizures induced by electrical stimulation with 60-Hz alternating current or by pentylenetetrazol, strychnine, or picrotoxin. In fact, phenytoin actually can potentiate the convulsant effect of pentylenetetrazol and picrotoxin. Thus, studies currently indicate that, although there are some similarities between anticonvulsant drugs, there are important differences in their mechanism of action and the way they produce control of seizure activity.
Despite its inability to elevate the seizure threshold of electrical stimulation of the brain with 60-Hz alternating current, phenytoin does slightly elevate the threshold for seizures induced by 6-Hz stimulation of the brain. Phenytoin's effect is not as dramatic as the effect of phenobarbital in this system, but is clearly distinct from its effect on high-frequency stimulation. Recent evidence, described later, on the effects of phenytoin on use-dependent inhibition of sodium channel function might be related to this interesting physiologic phenomenon.
Phenytoin blocks the tonic phase of tonic-clonic seizures induced by supramaximal electroshock (5,55,123,124). This effect of phenytoin in animals also has been documented in humans undergoing electroconvulsive therapy (123). Phenytoin also blocks the tonic phase of seizures induced by picrotoxin, pentylenetetrazol, and fluorothyl. In spinal preparations, phenytoin abolishes the tonic phase of seizures elicited by supramaximal electroshock (53, 54, 55). However, this effect requires higher doses than those that are able to block the tonic phase of seizures in the cerebral cortex.
Phenytoin reduces the prolonged increase in excitability and independent repetitive firing that occurs in the peripheral nerve after supramaximal rapid stimulation (118,121). The hyperexcitability of the peripheral nerve induced by low calcium or a combination of low calcium and low magnesium in the bathing medium also is reduced by phenytoin (73,104,105). These effects on the peripheral nerve suggest that phenytoin has an overall stabilizing effect on the neuronal membrane that may be related in some way to the effects of calcium or sodium on neuronal excitability.
Phenytoin prevents the spread of seizure activity in most areas of the central nervous system. However, its effects on seizure threshold are directed somewhat more toward the cerebral cortex. In several species, phenytoin has been shown to elevate seizure threshold in the hippocampus, amygdala, and anterior dorsal nucleus of the thalamus (1,21), but does not significantly affect the threshold in the reticular activating system and does not influence the sensory relay path to the pyramidal tract (7). Some of these results indicate that phenytoin is most effective in reducing seizure threshold in anatomic regions that contain numerous synaptic connections. Gangloff and Monnier (62) showed that phenytoin elevated the seizure threshold of the diencephalon, but Morell et al. (90) could not confirm this result. This effect was in contradiction to the effects of phenobarbital and trimethadione on this region of the nervous system. Morell et al. (90) felt that the inability of phenytoin to affect the diencephalon is consistent with some of its clinical effects. Phenytoin can block generalized tonic-clonic seizures but may not block tonic-clonic seizures of cortical origin. Phenytoin does not completely block the sensory or other prodromal signs associated with some partial complex seizures. These authors argued that some of these other effects might result from the inability of phenytoin to alter the seizure threshold in the diencephalon. However, at high concentrations, phenytoin may have marked effects on this structure, and the discrepancy in the findings may relate to the concentrations of the drug used in each experimental system.
The studies described in the foregoing paragraphs pioneered the research efforts into the mechanisms of action of phenytoin on the central nervous system. Phenytoin clearly produces dramatic and clinically useful suppression of the spread of seizure activity in the cortex and other regions of the brain. This effect is not universal, however, and is somewhat selective for specific types of seizures initiated by maximal electroshock but not by several chemical convulsants. In addition, phenytoin appears to be somewhat selective in seizure phase in that it inhibits the tonic phase more potently than the clonic phase of tonic-clonic seizures.
PHYSIOLOGIC EFFECTS OF PHENYTOIN
The well documented effects of phenytoin on seizure discharge and the spread of neuronal excitability set the stage for studying the physiologic mechanisms underlying these effects. Rapid advances in molecular neurobiology, with sophisticated intracellular and extracellular recording techniques, have greatly facilitated this research since the early 1980s. Phenytoin has been shown to modify several important physiologic processes, including PTP and sustained repetitive firing (2,80).
Effects of Phenytoin on Posttetanic Potentiation
PTP is a physiologic phenomenon that has been implicated in the development of hyperexcitable areas in the brain during seizure activity (122). PTP also is thought to be an important mechanism leading to high-frequency trains of impulses in excitatory brain circuits and to the spread of this activity to adjoining neurons, as well as to their propagation to distant neuronal aggregates, resulting in uncontrolled spread of excitation to the whole brain in the maximal tonic seizure discharge (101,107). Thus, PTP is
considered an important physiologic process inherent in normal neuronal circuitry that could regulate the spread of neuronal excitability.
PTP specifically refers to the augmentation of the postsynaptic compound action potential elicited by presynaptic stimulation after a repetitive stimulus (100,101,106). Thus, stimulating a neuronal circuit, after a number of intense repetitive stimulations (tetanus) of the same circuit, results in a more dramatic response than before tetanus. This tetanus, or intense stimulation, somehow alters the normal resting levels of excitability of the system and produces a hyperexcitable state. This phenomenon suggests that repetitive use of a neuronal pathway sensitizes it for a given time to enhanced discharge. These types of phenomena could develop a reverberating or “building” hyperexcitability in the neuronal circuitry and implicate normal neuronal mechanisms in the development and spread of the epileptogenic focus. Although this model is appealing, the question of whether PTP is a major mechanism for producing spread of the seizure focus in humans remains unanswered.
Phenytoin effectively blocks PTP (100,101), and this effect may represent one of its major sites of action in preventing the spread of seizures. Phenytoin inhibits PTP in spinal cord preparations as well as in preparations of stellate ganglion in the cat (50,101). In addition, phenytoin blocks PTP at intramedullary terminals and the neuromuscular junction (100). Not all anticonvulsants are effective in blocking PTP. Phenobarbital, trimethadione, and valproic acid have little or no effect on PTP. However, carbamazepine and the anticonvulsant benzodiazepines are effective in blocking PTP.
The mechanism by which phenytoin regulates PTP has not been clearly established. Both the accumulation of calcium in the nerve terminal during the tetanus and the ability of phenytoin to block sodium channels in a use-dependent fashion may contribute as mechanisms for phenytoin's action in blocking PTP (30,31,106).
Effects of Phenytoin on Sustained Repetitive Firing
A growing body of evidence (81, 82, 83, 84, 85, 86) indicates that sustained high-frequency repetitive firing (SRF) is an important property of vertebrate and invertebrate neurons and plays a role in regulating the excitability of the cell. SRF is manifested in several types of central nervous system neurons and may be involved in anticonvulsant drug action and epileptogenesis. Although no direct evidence has demonstrated the link between SRF and epilepsy, information obtained from in vitro studies on isolated neurons concerning SRF may have some bearing on altered neuronal excitability and anticonvulsant drug action.
Phenytoin is effective in regulating SRF (78, 79, 80, 81, 82, 83, 84, 85, 86) (Chapter 4). The correlation of specific anticonvulsant drug activity with actions on SRF has indicated that the SRF model is useful for studying drug effects against generalized tonic-clonic and maximal electroshock-induced seizures (78, 79, 80). Anticonvulsants effective against generalized absence seizures, such as ethosuximide and trimethadione, are not effective against SRF. Therapeutic cerebrospinal fluid levels of phenytoin in humans are within the concentration range that inhibits SRF in isolated cultured neurons (80).
The ability of phenytoin to limit SRF has been shown in a wide variety of neurons maintained in culture and several invertebrate preparations (78, 79, 80, 81, 82, 83, 84, 85, 86). Phenytoin's ability to limit SRF is an attractive hypothesis for some of the neuronal stabilizing effects of this drug. Evidence is now accumulating that the effects of phenytoin on SRF are mediated by the use-dependent blockage of sodium channels produced by phenytoin (Chapter 4).
EFFECTS OF PHENYTOIN ON SODIUM-POTASSIUM TRANSPORT AND SODIUM ION MOVEMENTS
In 1955, Woodbury (129) provided evidence that phenytoin played a major role in altering sodium ion movements across nerve cell membranes. This work set the stage for much of the following research over the next 30 years with regard to the effects of phenytoin on ion conductances in neuronal membranes. On the basis of calculated intracellular sodium concentrations, Woodbury suggested that phenytoin might regulate sodium transport in the brain by affecting sodium-potassium ATPase (22,130, 131, 132, 133, 134). The more recent electrophysiologic studies of Macdonald's group (78, 79, 80) have clearly established that concentration ranges of phenytoin that are consistent with anticonvulsant levels in humans produce a use-dependent blockage of the voltage-gated sodium channel.
Sodium-potassium ATPase and its regulation by phenytoin have been extensively studied and reviewed (22,23). The early research studies showed that phenytoin, under some conditions, increased the activity of the sodium-potassium ATPase. In brain synaptosomes, phenytoin increased sodium-potassium ATPase activity after systemic administration to animals (63,75) and after administration in vitro (60,128). Phenytoin also increased the activity of sodium-potassium ATPase in the adrenal medulla (65). Some conflict developed in this field when it was found that under some experimental conditions phenytoin did not affect sodium-potassium ATPase in in vitro brain synaptosome experiments. These results have been carefully examined by numerous investigators (22), and it now appears that the difference in experimental results relates to the ratio of sodium to potassium in the experimental systems.
Deupree (48,49) concluded that phenytoin, under certain conditions, does not affect sodium-potassium ATPase and that earlier studies may be explained by the contamination of phenytoin with potassium, which would increase the potassium ratio. However, Delgado-Escueta and Horan (22) reviewed these data and found that the effects of phenytoin on active transport of potassium in synaptosomes occur under conditions in which the potassium content of the cell is lowered and the sodium concentration is increased. This latter condition is thought to be more closely analogous to the environment of the epileptogenic focus.
These results may explain why phenytoin lacks toxic effects on normal neuronal function, but can have a marked neuronal stabilizing effect during an excitable discharge. Under normal conditions, phenytoin may not play a role in regulating sodium-potassium ATPase activity. However, in an epileptogenic focus, where the ratio of sodium to potassium across the membrane may be altered, phenytoin may regulate the activity of this important membrane enzyme system. These studies on sodium and sodium-potassium ATPase represented a major advance in understanding the mechanism of action of anticonvulsant drugs. This research represents one of the first neurochemical insights into the mechanism of action of phenytoin and serves as a model for its testing in numerous other biochemical systems.
Phenytoin Effects on Sodium Conductances
After the initial observations by Woodbury that phenytoin may regulate neuronal excitability by affecting sodium permeability across the cell membrane, several relevant investigations over the next 30 years used new techniques for studying the action of phenytoin on neuronal tissue. The membrane “stabilizer” effect of phenytoin and its ability to prevent repetitive electrical activity (13,59,71,73,121) indicated that phenytoin's action on sodium conductance might be an important area for further research. Lipicky et al. (76) observed that phenytoin decreased the early sodium current in the voltage-clamped squid axon, suggesting that the drug decreased the number of open channels in the early phase of the action potential. Johnson and Ayala (70) also observed in Aplysia that phenytoin decreased sodium influx. Further evidence for an effect of phenytoin on sodium influx was provided by Swanson and Crane (115) using guinea pig cerebral cortical slices, and by Schwarz and Vogel (109) in voltage-clamp experiments on single myelinated neurofibers. These studies also suggested that phenytoin decreased the action potential amplitude and increased the threshold to fire. These results led to the hypothesis that phenytoin might reduce the conduction velocity by affecting sodium currents.
More recent observations by DeWeer (50) and Perry et al. (94) on the isolated squid axon provided additional evidence that phenytoin affects sodium influx. These investigators postulated that phenytoin behaved like tetrodotoxin in blocking sodium channels. Their studies also confirmed the observations of Schwarz and Vogel (109) that phenytoin induced membrane hyperpolarization. Thus, numerous studies have demonstrated that phenytoin has a significant effect on sodium influx in neuronal membranes (20,105). The results are consistent with the original observations of Woodbury and may explain some of the neuronal stabilizing effects of this anticonvulsant.
Use-Dependent Inhibition of Sodium Channels
A major contribution of Macdonald's laboratory (78, 79, 80) has been to demonstrate, using electrophysiologic techniques and kinetic analysis, that phenytoin interacts with sodium channels at concentrations found in the plasma of patients treated for epilepsy. Phenytoin directly reduces the frequency of SRF of action potentials in isolated neurons maintained in culture (78, 79, 80, 81, 82, 83, 84, 85, 86). An important aspect of this effect of phenytoin on the action potential was that it did not reduce the amplitude or duration of a single action potential but reduced only the ability of the neuron to fire trains of action potentials at high frequency. Outside-out patch recordings in hippocampal neuronal cultures demonstrated that phenytoin more effectively inhibited late sodium channel openings, believed to underlie ictal epileptiform activity, than the transient sodium channel openings that comprise the peak sodium current (110). This ability of phenytoin to limit high-frequency repetitive firing was voltage dependent, increased after depolarization, and reduced by hyperpolarization. The limitation of firing was prolonged enough to last for several hundred milliseconds.
It was postulated from these studies that one of phenytoin's anticonvulsant actions may be to shift the sodium channel to an inactive state similar to the normally occurring inactive state of the channel, but from which recovery was delayed (80). The ability of phenytoin to limit SRF implies an action that occurs only under abnormal conditions, where neurons are firing repetitively at high frequencies. Thus, phenytoin can decrease the high-frequency spread of seizure discharge during the development of seizure activity without inhibiting the normal, less frequent firing of the neuron under background conditions. This experimental finding, along with the finding that phenytoin preferentially inhibits late sodium channel openings (110), is attractive because it seems to explain why phenytoin has so few sedative or cognitive effects relative to its potent anticonvulsant action. Other studies have supported the findings of Macdonald's group and further substantiated the effect of phenytoin on sodium channels using electrophysiologic techniques. Schwarz and Vogel (109) showed that phenytoin produced a voltage-dependent block of sodium channels that could be removed by hyperpolarization
in mammalian myelinated nerve fibers. Phenytoin caused a shift of the steady-state sodium channel inactivation curve to the more negative voltages in these experiments. Also, phenytoin reduced the rate of recovery of sodium channels from inactivation. These studies showed that under normal conditions, sodium channels recovered from the inactivation state in a few milliseconds after a 500-millisecond depolarization to 25 mV. In the presence of phenytoin, however, recovery from the inactive state took greater than 90 milliseconds. In addition, phenytoin was shown in this preparation to produce a frequency-dependent block of action potentials. The advantage of this experimental model was that it could be used to apply more sophisticated kinetic analysis. These studies suggested that the effect of phenytoin on sodium channel inhibition assumed first-order kinetics, indicating that the anticonvulsant was binding at one site near or at the sodium channel.
Similar studies were performed on isolated mammalian neurons by Wakamori et al. (127). These studies investigated the effect of phenytoin on hippocampal pyramidal neurons isolated from CA1 regions from 1- to 2-week-old rats. Phenytoin produced a negative shift in the steady-state inactivation curve for sodium channels in these cells and also produced frequency-dependent block of sodium channels. Thus, phenytoin has been shown to inactivate sodium channels in a use-dependent manner in both mammalian myelinated nerve fibers as well as isolated neurons in culture.
Phenytoin also has been shown to affect human sodium channels (125). Electrophysiologic studies of the human brain sodium channels expressed in the oocytes were blocked by phenytoin in a voltage-, frequency-, and time-dependent fashion. The authors (125) concluded that the effects of phenytoin on human sodium channels were similar to those in cultured neurons, rat myelinated nerve, and rat hippocampal pyramidal neurons (80).
More recent studies have investigated the specific kinetics and binding actions of phenytoin to sodium channels (80,118,125). Phenytoin appears to stabilize the inactive form of the sodium channel in a voltage-dependent fashion. Kinetic data suggest that opening of the sodium channel allows the phenytoin molecule to diffuse through the channel and bind to a receptor site on the inside of the membrane surface. Thus, the effect occurs in a voltage-dependent manner. The affinity of phenytoin for the sodium channel is short-lived, and the channel returns to normal activity within milliseconds. Therefore, the sodium channel can quickly recover from this use dependent block.
The characteristics of this binding and the elucidation of the possible site of the phenytoin sodium channel have been actively pursued (14,69,99,118). Site-directed mutagenesis studies have shown that three amino acids in the S6 transmembrane α helix of domain IV of the pore-forming α subunit glycoprotein of the sodium channel are critical for phenytoin binding. Mutation of these three amino acids has been found to reduce phenytoin binding to the inactivated state of the sodium channel (14). The anticonvulsants carbamazepine and lamotrigine, and local anesthetics, share this common receptor site, as well as the mechanism of stabilizing the inactivated state of the channel (14). A two-phenyl-ring structure also is common to these drugs and has been implicated in forming the ligand for the binding interaction (69).
The effect of phenytoin on limiting sustained repetitive firing through use-dependent blockage of the sodium channel provides a biochemical mechanism for preventing rapid neuronal discharge from spreading from one neuron to another without interfering with normal action potential communication between neurons.
EFFECTS OF PHENYTOIN ON NEUROTRANSMISSION
Phenytoin has long been known to depress synaptic transmission. Since the early 1970s, new electrophysiologic techniques have allowed the inhibitory effects of phenytoin to be studied in more detail. It appears that phenytoin can inhibit depolarization-dependent synaptic transmission but can increase the frequency of miniature end-plate potentials (MEEPs) at rest in the synapse (136). These mechanisms provide insight into the ability of phenytoin to be both excitatory and inhibitory in the nervous system.
It has been shown in in vitro synaptosome preparations that phenytoin inhibits norepinephrine and acetylcholine release (27, 28, 29). Studies (27, 28, 29, 30, 31) have shown that phenytoin can inhibit neurotransmitter release from synaptosomes by blocking calcium entry during depolarization. In addition, under conditions where calcium enters the synaptosomes through an ionophore, phenytoin still could inhibit neurotransmitter release by direct inhibition of intracellular synaptosomal biochemical processes, such as protein phosphorylation and other calmodulin-regulated events. These studies provide the first evidence that at least two mechanisms exist for explaining the effects of phenytoin on synaptic transmitter release. Phenytoin most likely blocks transmitter release during an action potential by minimizing or limiting calcium entry and by having a specific effect on other molecular processes in the nerve terminal that modulate transmitter release. The effects of phenytoin on MEEPs are postulated to be the result of increased intracellular calcium concentrations induced by the drug. Phenytoin not only inhibits calcium uptake into synaptosomes but can block calcium uptake into mitochondria. As an important intracellular calcium buffering system, the mitochondria keep the calcium concentration in the nerve terminal at a low level. By inhibiting this process, phenytoin can slightly elevate intrasynaptosomal calcium, causing hyperexcitability manifested as a significant increase in MEEPs.
EFFECTS OF PHENYTOIN ON CALCIUM CHANNELS AND CALCIUM SYSTEMS
Phenytoin has been shown to inhibit calcium influx in numerous preparations. The mechanisms by which phenytoin inhibits calcium influx are not completely understood. However, several studies have provided convincing evidence that this drug regulates the calcium conductances in nerve preparations as well as in other tissues (92).
Effects of Phenytoin on Calcium Channels and Sequestration Mechanisms
Studies by Ferrendelli and coworkers (57,58,111,112) elegantly demonstrated that phenytoin inhibits depolarization-dependent calcium influx in preparations of presynaptic nerve terminals in vitro. This work provided the initial evidence that phenytoin inhibits both sodium and calcium influx during depolarization and suggests that these conductances are affected independently of each other. Phenytoin has been shown moderately to inhibit L-type calcium channels in neuronal (68,108) and nonneuronal preparations (88,103). Phenytoin also inhibits T-type calcium currents (68,119). Recent studies have demonstrated that phenytoin inhibition of T-type currents may be subject to subunit variation of the T-type calcium channel (74,120). Phenytoin caused a moderate blockade of native dorsal root ganglia T-currents, but a complete, although lower-affinity, blockade of currents from transfected α1G subunits in HEK293 cells. Interestingly, a complete, lower-affinity block similar to that of α1G T-currents was seen in some α1H currents, whereas other α1H T-currents demonstrated a partial, higher-affinity blockade similar to the native DRG neurons (120). The investigators suggested that there could be an as yet unknown regulatory factor present in DRG neurons, and some HEK293 cells, that selectively regulates α1H subunit T-currents (120). Further investigation is required to elucidate the complete role of phenytoin on depolarization-dependent calcium influx and the potential for anticonvulsant action.
Phenytoin also blocks calcium sequestration in a number of different preparations. Calcium uptake is inhibited in the intact neuromuscular junction preparations as well as in the synaptosome preparation (102,135). Several other studies on the effects of phenytoin on calcium uptake and metabolism have been extensively reviewed (117,131). Cytosolic calcium levels in gingival fibroblasts are modulated by phenytoin (89), possibly providing mechanistic insight into the gingival hyperplasia associated with phenytoin use in children.
An overwhelming body of evidence suggests that phenytoin can inhibit calcium influx during depolarization and also inhibit the uptake and sequestration of calcium in the nerve terminal after its entry. Thus, phenytoin has both a depressive effect by blocking calcium uptake and a potentially excitatory effect by blocking the uptake and sequestration of calcium in the nerve terminal. This latter effect could result in prolonged elevated calcium concentrations in the nerve terminal after tetanic stimulation or the spread of repetitive firing. This molecular insight might have some bearing on the clinical observations that phenytoin can have both neuronal stabilizing and anticonvulsant properties, as well as cause hyperexcitability in the nervous system. Depending on the balance in the system, phenytoin could either suppress neuronal activity by decreasing calcium entry or cause hyperexcitability of the nervous system by delaying the recovery of resting intracellular calcium in the nerve terminal after repetitive discharge. These results have led several investigators (30,31) to suggest that the effect of phenytoin on calcium metabolism may explain some of its anticonvulsant activity and effects on hyperexcitable nervous tissue.
Effects of Phenytoin on Calmodulin Target Enzymes
Calmodulin is a major calcium-binding protein that mediates some of the second messenger effects of calcium on cell function (16,17,72). Calmodulin binds calcium, and this calcium-calmodulin complex can regulate several enzyme systems in the cell (17). Major enzymes regulated by calcium and calmodulin are the calcium- and calmodulin-dependent protein kinases such as calmodulin kinase II. Phenytoin inhibits calcium-calmodulin-regulated protein phosphorylation in neuronal preparations and in preparations of presynaptic nerve terminals (24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46). The ability of phenytoin to regulate this major calcium-dependent enzyme system suggests that phenytoin modulates many of the second messenger effects of calcium in the nervous system, which may provide a major pathway by which phenytoin can regulate many cellular processes. The effect of phenytoin on calmodulin-regulated systems may occur at higher-than-physiologic concentrations, and thus may account for some of the toxic effects of phenytoin on neuronal tissue. The precise role of phenytoin inhibition of calcium-calmodulin-regulated enzyme systems is an important area for further investigation.
Because calcium has been implicated in regulating many physiologic processes in the brain, the influence of phenytoin on some of these processes by blocking calcium entry, regulating intracellular calcium levels, or affecting one or more major calcium-calmodulin-regulated enzyme system could account for some of the broad anticonvulsant and toxic effects of phenytoin on the central nervous system.
EFFECTS OF PHENYTOIN ON CHLORIDE PERMEABILITY
Phenytoin increases chloride conductance in mammalian cortical neurons as well as in crayfish stretch receptor neurons
(3). This chloride conductance is associated with the γ-aminobutyric acid A (GABAA) receptor. At nanomolar concentrations, phenytoin postsynaptically modifies the gating mechanism, thereby decreasing the rate of closing of the chloride channel. Phenytoin apparently enhances the chloride conductance of the GABAA receptor, which underlies the inhibitory postsynaptic potential, leading to increased hyperpolarization of the neuronal membrane. Therefore, similar to benzodiazepines and barbiturates, phenytoin apparently enhances the effect of the GABAA receptor.
Although phenytoin-enhanced GABAA receptor chloride currents are an attractive anticonvulsant mechanism, phenytoin's inability effectively to inhibit pentylenetetrazol-induced seizures in animals somewhat discredits this hypothesis. Pentylenetetrazol-induced seizures are very sensitive to benzodiazepines, working through the GABAA-benzodiazepine complex. This anticonvulsant complex functions to increase the GABAA receptor chloride current. If phenytoin's anticonvulsant properties are similarly mediated through increased chloride conductances, then phenytoin should inhibit pentylenetetrazol-induced seizures. Because phenytoin is ineffective in this model, the significance of phenytoin's action on chloride currents requires further investigation.
EFFECTS OF PHENYTOIN ON BIOCHEMICAL SYSTEMS
Phenytoin, particularly at toxic concentrations, affects numerous biochemical systems and processes, as noted in the reviews by Woodbury (131, 132, 133). A complete discussion of all of these effects goes beyond the scope of this chapter. However, several effects of phenytoin on major second messenger and biochemical systems are worth discussing in some detail because they may have an important contribution to some of the anticonvulsant or toxic effects of this drug.
Effects of Phenytoin on Cyclic Nucleotide Metabolism
Phenytoin regulates the metabolism of adenosine 3′,5′-monophosphate (cyclic AMP) and guanosine 3′,5′-monophosphate (cyclic GMP). Both of these cyclic nucleotides have been implicated as major second messengers in cell function. The effects of phenytoin on cyclic nucleotide metabolism have been studied and reviewed by Ferrendelli (56,58). Phenytoin depresses the basal levels of cyclic GMP in the cerebellum in vivo. Phenytoin also prevents electroshock convulsion-induced elevation of brain cyclic AMP and cyclic GMP levels in the cerebral cortex. Phenytoin also inhibits the elevation of brain cyclic nucleotides caused by depolarizing agents that increase sodium influx in synaptosome fractions.
These studies suggest that phenytoin may act directly on nucleotide metabolism or that the effects may be secondary to some neuronal stabilizing effects of phenytoin that block depolarization-induced production or metabolism of cyclic nucleotides. The possible relationship of these effects of phenytoin to its anticonvulsant action or toxic effects is an important area for further research.
Effects of Phenytoin on Neurotransmitter Systems
As discussed earlier, phenytoin affects neurotransmitter release and metabolism in numerous preparations [see review by Woodbury (131)]. Phenytoin inhibits the release of norepinephrine and other neurotransmitters in vivo and from intact nerve terminal preparations in vitro. Phenytoin also inhibits neurotransmitter reuptake. These effects are dose dependent because different effects are attained at different concentrations of phenytoin. These results suggest that phenytoin may act at different sites in regulating transmitter release and uptake.
Phenytoin also decreases the concentration of glutamic acid in the brain and increases the concentration of glutamine and GABA. The effect of phenytoin on GABA systems has been observed in different preparations and in different species. Thus, phenytoin may play a role in regulating the level and metabolism of this major inhibitory neurotransmitter. Furthermore, phenytoin affects the metabolism and activity of acetylcholine. These examples indicate that phenytoin may play an important role in regulating neurotransmitter systems in the brain by affecting the metabolism, storage, release, or uptake of these compounds, possibly through its effects on calcium and cyclic nucleotide second messenger systems.
Effects of Phenytoin on Calmodulin Systems
Phenytoin inhibits calcium-calmodulin-dependent protein kinase in neuronal preparations (4,24,31,43). Phenytoin's ability to regulate protein phosphorylation may play an important role in regulating the anticonvulsant or side effects of this compound on neuronal and nonneuronal tissue. Because calcium-dependent protein phosphorylation may play a major role in regulating numerous cellular processes, the ability of phenytoin to inhibit this system could account for many of its effects on cellular metabolism and, ultimately, on numerous cells in the body. The role of phenytoin in regulating other calmodulin-controlled enzyme systems needs further investigation.
NEUROPROTECTIVE EFFECTS OF PHENYTOIN
Phenytoin's ability to block calcium entry suggests that it may be a potent neuroprotective agent in ischemia or
anoxia. Studies from Taft et al. (116) demonstrate that phenytoin protects against ischemia-induced neuronal cell death in a well characterized gerbil forebrain ischemia model. This model entails brief bilateral carotid occlusion that results in complete forebrain ischemia. After 5 minutes of carotid occlusion, almost complete neuronal destruction in the CA1 sector of the hippocampus is observed, with preservation of other hippocampal neuronal structures. This is a very controllable experimental model for studying effects of neuroprotective agents. Phenytoin treatment (200 mg/kg) blocked the ischemia-induced neuronal cell death. Phenytoin was also neuroprotective with longer durations of ischemia and was more effective at higher concentrations.
These studies indicate that phenytoin may be a useful neuroprotective agent for the treatment of brain injury or cerebrovascular disease. The precise mechanism by which phenytoin accomplishes its neuroprotective effects has not been clearly established (66). The ability of phenytoin to inhibit calcium entry or calcium-dependent enzymes may play an important role in this neuroprotective effect. However, the possibility of reducing depolarization through block of the sodium channel or subsequent seizure activity associated with neuronal injury also must be considered. Phenytoin also inhibits spreading depression in retinal preparations (15). Phenytoin therefore may have an important role in the future as a neuroprotective or adjunct neuroprotective agent.
PHENYTOIN RECEPTOR MOLECULES
Receptor neuropharmacology has played a major role in the development of many drugs since the early 1980s. Specific receptors for the benzodiazepines, steroids, catecholamines, and other neuroleptic compounds have provided important mechanisms of action for several of these neuropharmacologic agents. However, at present, except for the high-affinity benzodiazepine receptor (10), there is no evidence for a specific anticonvulsant binding site. Recent studies (8,9) have identified a novel class of benzodiazepine-binding proteins that bind benzodiazepines with a potency series and a therapeutic concentration range that are consistent with the effects of the benzodiazepines on SRF and effects on maximal electroshock-induced seizures in animals. Studies also have indicated that phenytoin effectively competes in therapeutic concentrations with the benzodiazepines for this low-affinity receptor. These results suggest that these novel benzodiazepine receptors also may bind phenytoin in therapeutic concentrations and provide evidence of a specific binding protein for phenytoin in the central nervous system. Further investigations must be conducted to determine the significance and relevance of this phenytoin-binding protein to its actions as a stabilizer of neuronal membranes. It is clear that research directed at identifying anticonvulsant receptors would have important implications in developing new therapeutic agents to regulate seizures and in understanding mechanisms of action of anticonvulsant drugs. This is an area that needs further research.
FIGURE 57.2. Enzymatic conversion of fosphenytoin to phenytoin, formate, and phosphate by phosphatases.
Fosphenytoin [ACC-9653; CI 982; Cerebyx (Parke-Davis)], first synthesized in 1973 by Stella and Higuchi (114), is a disodium phosphate ester of phenytoin. As a highly water-soluble prodrug of phenytoin, fosphenytoin offers an alternative composition of this anticonvulsant that lacks many of the adverse effects of phenytoin administration. As discussed earlier, intramuscular phenytoin administration is painful and ineffective owing to drug precipitation induced by the lower pH effect of phenytoin injection in tissue (93). Intravenous administration of phenytoin also is painful and can cause vascular tissue damage (113) and phlebitis (95). Rapid intravenous administration of phenytoin can cause cardiovascular collapse, presumably because of the 40% propylene glycol vehicle (77). Fosphenytoin, because of its high water solubility [over 4,000 times greater than phenytoin (126)], is an excellent alternative to phenytoin for parenteral administration. Fosphenytoin does not produce the vascular and cardiac toxicities of phenytoin preparations and therefore is safer and can be administered more rapidly.
Fosphenytoin has no known intrinsic activity. Fosphenytoin's anticonvulsant effects only become apparent after its rapid and complete metabolism to phenytoin by phosphatases (Figure 57.2) in the heart, lungs, liver, spleen, kidneys,
and small intestine, (98). At this time, no drugs are known to alter the conversion of fosphenytoin to phenytoin (18). The bioavailability of phenytoin derived from intravenous fosphenytoin administration is rapid and virtually identical to that of intravenous phenytoin (12) (Figure 57.3). Like phenytoin, protein binding of fosphenytoin to plasma proteins like albumin is high (95% to 99%) (67). However, in contrast to phenytoin, this binding is nonlinear (67). This extensive binding displaces phenytoin from albumin and increases the plasma free fraction of phenytoin (51,52). Direct renal excretion of fosphenytoin is minimal and clinically insignificant (12).
FIGURE 57.3. Plasma total phenytoin concentrations derived from fosphenytoin (circles) and from phenytoin (squares) after simultaneous intravenous infusion of 150 mg phenytoin equivalents are virtually identical over time. (From Browne TR, Szabo GK, McEntegart C, et al. Bioavailability studies of drugs with nonlinear pharmacokinetics: II. absolute bioavailability of intravenous phenytoin prodrug at therapeutic phenytoin serum concentrations determined by double-stable isotope technique. J Clin Pharmacol1993;33:89-94, with permission.)
Fosphenytoin has fewer local adverse effects at the administration site than phenytoin. Adverse effects of fosphenytoin in the central nervous system (nystagmus, headache, ataxia, and somnolence) are attributed to the phenytoin derived from fosphenytoin metabolism, rather than the prodrug itself. Electrocardiographic changes and hypotension sometimes seen with rapid phenytoin administration are not seen with fosphenytoin (52). Transient paresthesias that resolve without further consequence have been reported with intravenous administration of fosphenytoin (11,61). These paresthesias have not been associated with phenytoin use, but are not associated with permanent symptoms.
Fosphenytoin is a better preparation for rapid administration and for parenteral indications of phenytoin (47). The U.S. Food and Drug Administration has approved fosphenytoin for intramuscular loading and maintenance dosing in adults and children older than 5 years of age and for the intravenous treatment of status epilepticus. Fosphenytoin shares the same contraindications as phenytoin. Although fosphenytoin is more expensive than phenytoin, economic savings can be realized with this drug over phenytoin because of the savings associated with decreased adverse effects at the site of administration, decreased time and supplies associated with restarted and new intravenous lines, and decreased risk of cardiovascular complications.
UNIFYING MECHANISMS OF ACTION—SUMMARY
There is no single action of phenytoin that can completely account for its numerous effects on neuronal and nonneuronal tissue. The preponderance of evidence suggests that phenytoin may produce its numerous effects by regulating several important aspects of cellular function. Phenytoin's ability to regulate sodium transport across neuronal membrane is a major mechanism of action that almost certainly underlies some of its clinical effects on neuronal tissue. The use-dependent inhibition of sodium channels characteristic of phenytoin provides an important potential mechanism allowing phenytoin to regulate excitability ictally, but not interictally, during normal neuronal activity. Further research on the effect of phenytoin on sodium channels potentially will elucidate how its molecular effect might underlie its specific clinical effects. The ability of phenytoin to modulate sustained repetitive firing may underlie its ability to inhibit the tonic phase of generalized tonic-clonic seizures. The ability of phenytoin to regulate calmodulin and cyclic nucleotide second messenger systems could account for some of its widespread effects on cellular processes. It is difficult to find a biochemical or physiologic process that is not in some way regulated by cyclic nucleotides or calcium. Thus, the effects of phenytoin on
these second messenger systems would be dramatically amplified in terms of the diverse clinical and toxic effects that might result from its use. These effects may account for the wide diversity of phenytoin's actions. The ability of phenytoin to regulate and inhibit voltage-dependent neurotransmitter release at the synapse also may play an important role in its anticonvulsant action. Although the precise mechanisms of this effect are not known, it is clear that inhibition of calcium channels and calcium sequestration by phenytoin in the nerve terminal plays an important role in the excitatory and inhibitory actions of this anticonvulsant. The effect of phenytoin on PTP also may underlie some of its important anticonvulsant properties. Some of the effects of phenytoin on PTP may be mediated at the molecular level by its effect on calcium and sodium systems. Advances in molecular neurobiology and neuroscience have increased our understanding of the mechanisms of action of phenytoin. Major advances described in this chapter shed light on how phenytoin may mediate neuronal stabilizing and excitatory phenomena. More recent advances have lead to the synthesis and characterization of fosphenytoin, which shares the anticonvulsant mechanisms of action of phenytoin while reducing some of the adverse effects associated with parenteral administration.