JLE

Epileptic Disorders

MENU

Piracetam and levetiracetam: close structural similarities but different pharmacological and clinical profiles Volume 2, issue 2, Juin 2000

Figures

See all figures

Slight and subtle differences in molecular configuration can cause marked changes in pharmacological activity as, for example, the structural similarity of male and female sex hormones. Such is also the case with the structurally related molecules PIR and LEV which have different and distinct pharmacological profiles. Although many pyrrolidone derivatives exist, only a few are have been developed: PIR, aniracetam and pramiracetam (figure 1). Of these, the best known is PIR, which has been available for more than 25 years. LEV was initially evaluated in models of cognitive impairment with the primary objective of finding a drug more effective than PIR. Although the two compounds have not been directly compared, much is known of their respective activities. This paper reviews experimental and clinical data for both drugs which demonstrate their differing clinical profiles.

Chemistry

PIR (2-oxo-l-pyrrolidine-acetamide, C6H10N2O2) is a white crystalline powder with almost no odor and a slightly bitter taste. It is readily water- and alcohol-soluble, and is almost insoluble in acetone. The molecular weight of PIR is 142. LEV (formerly UCB L059) ((S)-alpha-ethyl-2-oxo-1-pyrrolidine acetamide, C8H14N2O2) is a white or off-white powder, soluble in water, with a molecular weight of 170.2. It exists as a racemically pure S-enantiomer.

Mode of action

Based on extensive in vitro experiments, including ligand-binding assays, the modes of action of PIR and LEV do not appear to derive from any interaction with known mechanisms involved in inhibitory or excitatory neurotransmission [1, 2]. Neither PIR nor LEV appears to influence membrane excitability [2-4]. In vitro and in vivo recordings of epileptiform activity from the hippocampus have shown that LEV inhibits burst firing without affecting normal neuronal excitability [3, 4]. In animal models, a brain-specific stereoselective binding site has been identified for LEV in synaptic plasma membranes of the central nervous system; PIR has a low affinity for this binding site [1]. Affinity for the binding site is correlated with seizure protection in the audiogenic seizure mouse model. In contrast, PIR is inserted into the neuronal membrane at the level of phospholipid heads [5] and probably affects the fluidity of animal and human brain membranes; this may explain its effects on memory [6] (table I).

Preclinical efficacy

Cognitive function

Piracetam. In preclinical studies with PIR, learning and memory were improved in animal models of a variety of behavioral procedures. The benefit of treatment with PIR was most pronounced in animals subjected to hypoxia [7, 8], drug intoxication (e.g. barbiturates) [9], electric shock [10], and aging [11]. In a classic study by Wolthuis [12], adult male rats were injected intraperitoneally with PIR at a dose of 100 mg/kg, or saline daily and submitted half an hour later to a training session. Rats placed at the entry to a Y-maze with light bulbs at the end of each branch were trained to avoid painful electric shock to the paw by selecting the illuminated branch. Investigators alternated lights at the end of the maze in a random fashion. During the first 3 days, no difference was found between the treatment groups, but after 10 days, PIR-treated rats displayed significantly more correct choices than the placebo group.

Improvement in learning and memory has also been demonstrated in goldfish. In a study by Bryant and colleagues [13], goldfish treated by immersion with PIR (400 mg in 4 l of water) were subjected to dark-avoidance training in which each goldfish was placed in the darkened compartment of a water tank and subjected to electric shock. A light stimulus in an unoccupied compartment remained illuminated and for a correct response the goldfish avoided shock by swimming from the dark into the lit compartment. Mean levels of correct active dark-avoidance response were higher in PIR-treated fish than in controls with a difference that was significant after 3 to 5 days of training.

Levetiracetam. LEV has been found to improve learning and memory in animals with underlying cerebral ischemia [14], drug-induced amnesia [15] and spatial learning disorder [16] (table II). In a model of scopolamine-induced amnesia, mice were randomized to receive scopolamine either alone or with LEV or a nootropic agent (PIR, oxiracetam, rolziracetam) [15]. In an acquisition test, each mouse was placed in a light compartment and allowed to enter a dark compartment where an electric shock was delivered through the feet. Twenty-four hours later the retention trial was initiated; the mouse was again placed in the light compartment with the time taken to enter the dark compartment (retention latency) recorded. Unlike oxiracetam and rolziracetam, both PIR and LEV significantly increased retention latencies and were therefore active in antagonizing the amnesic effect of scopolamine. PIR was most effective at 10 mg/kg. LEV was effective only at higher doses (17-54 mg/kg) and its effect was less (table II).

Lamberty and Klitgaard [16] investigated the effect of LEV and other antiepileptic drugs on the spatial learning performance of rats in a maze. At doses known to suppress motor seizures (17-54 mg/kg), LEV did not alter cognitive performance in fully amygdala-kindled rats; higher doses of LEV (170 mg/kg) were also without this effect. In contrast, valproate, clonazepam, phenobarbital and carbamazepine all induced cognitive impairment (table II).

Both PIR and LEV have thus been found effective in many models of cognitive performance but LEV was less efficient than PIR and less consistently so.

Antiepileptic effects

The pharmacology of PIR and LEV has been studied in many animal models of seizures and epilepsy. These include classical screening models of chemically- and electrically-induced seizures, as well as models of epilepsy in kindled animals, pilocarpine- and kainic acid-induced seizures and in genetically susceptible animals.

Classical screening tests
Supramaximal electroshock (MES) and subcutaneous pentylenetetrazole (PTZ) tests

The results of these tests were negative in rats after large doses of PIR (3,000 mg/kg) (MES) and 1,420 mg/kg (PTZ). With LEV, both the MES test (after intraperitoneal doses up to 500 mg/kg) [17] and the PTZ test [18] were negative (table III). LEV did however show significant activity against seizures induced by submaximal PTZ stimulation [18].

Models of partial seizures.
Amygdala-kindled rats [19]

In one study [17], rats were subjected to daily electrical stimulation for 21 days to induce a progressive increase in seizure activity and after-discharge duration. When they were fully kindled, LEV was injected intraperitoneally at doses of 13 to 108 mg/kg. Significant reductions in the severity and duration of both focal and secondary generalized seizures occurred even at the lowest dose tested and a dose-related increase in potency was observed with higher doses. LEV also significantly increased the focal seizure threshold after a dose of 54 mg/kg and decreased the duration of after-discharges. Compared with PIR, LEV was much more potent as an antiepileptic in kindled rats. Doses of at least 216 mg/kg PIR were needed to obtain significant antiseizure activity whereas comparable effects with LEV were seen with 27 mg/kg. In addition, at the doses tested, LEV did not cause motor impairment or other adverse effects while PIR induced moderate ataxia at a dose of 432 mg/kg. These results reflect the relative activities of LEV and PIR in a model of partial seizures and demonstrate the potent antiepileptic effects of LEV. In addition, LEV has been shown to retard the development of amygdala-kindling in rats suggesting that it may also posses antiepileptogenic properties [20].

Seizures induced by pilocarpine and kainic acid

Seizures mimicking human complex partial seizures can be induced by systemic administration of both high-dose pilocarpine and kainic acid. Seizure protection in these models is considered to predict efficacy against partial seizures in humans. LEV has been found to protect against secondarily generalized seizures caused by systemic administration of maximal convulsive doses (CD97) of pilocarpine (375 mg/kg intraperitoneally) and kainic acid (13 mg/kg subcutaneously) in rats; the minimal active intraperitoneal doses were 17 and 170 mg/kg [21]. PIR was not tested in this model.

Genetic models

Genetic models using audiogenic-seizure sensitive rodents are frequently used as a primary screening tool for antiepileptic activity. In one study in audiogenic mice, Matagne and Gower [22] compared the antiepileptic effects of LEV with those of PIR and other nootropic drugs. Protection against both clonic and tonic convulsions elicited by a 90-decibel sound in audiogenic mice was provided by LEV intraperitoneally at an effective dose (ED50) of 26 mg/kg (clonic) and 9 mg/kg (tonic). In contrast, PIR exhibited only weak antiepileptic activity with no inhibition of clonic convulsions; tonic convulsions were reduced but only at a high dose of 589 mg/kg. In audiogenic-seizure sensitive rats, a high intensity sound causes wild running followed by tonic-clonic convulsions [23]. LEV in single doses between 5.4 and 96 mg/kg provided dose-dependent protection against wild running and tonic-clonic convulsions, effects that were more pronounced 90 minutes after administration than after 45 minutes. At 90 minutes, the ED50 values were 16.0 (95 percent CI intervals: 3.1-34.6) mg/kg and 9.0 (95 percent CI intervals: 1.3-17.3) mg/kg for wild running and tonic-clonic convulsions respectively [24].

The Genetic Absence Epilepsy Rat from Strasbourg (GAERS) model is based on a strain of Wistar rats selected by inbreeding for the spontaneous occurrence of EEG spike and wave discharges and clinical absence-like manifestations [25]. Extensive research in this inbred strain of rats indicates that it is representative of human absence epilepsy. LEV markedly suppressed spontaneous spike wave discharges at doses of 5.4-170 mg/kg with no dose-effect relationship.

Preclinical safety

Preclinical toxicity studies with PIR and LEV have been conducted in mice, rats and dogs after single and repeated administration of both oral and intravenous dosage forms. All data relating to these toxicity studies are unpublished and on file with UCB. The general toxicity of both compounds was evaluated in acute, subacute and chronic toxicity studies in the mouse, the rat and the dog while carcinogenicity studies were conducted in the mouse and rat. Mutagenicity was evaluated in bacterial and mammalian cells in vitro and the mouse in vivo. Reproduction toxicology was conducted in the mouse, rat and rabbit.

Acute toxicity

Single-dose studies in mice, rats and dogs indicate a low potential for acute toxicity with both compounds. Oral and intravenous data in mice, rats, dogs and, for LEV the monkey, summarized in table IV, show very high values for the maximum non-lethal dose of oral and intravenous PIR. With LEV no deaths occurred after oral administration and the intravenous maximum non-lethal dose in mice and rats was 750 mg/kg in the mouse, 1,000 mg/kg in the rat and >= 1,000 mg/kg in the rat and the dog. No treatment-related macroscopic abnormalities were seen in any species either at necropsy or in those dying on study. Assuming a maximum anticonvulsant dose of 54 mg/kg LEV p.o. in animals, the compound displays a high safety margin after single dose oral administration in rodents and in dogs. Death or other irreversible toxicity occurred only at doses 15 to 20 times higher than this level.

Subacute and chronic toxicity

Repeated administration of both compounds was well tolerated. Rats receiving up to 2,000 mg/kg/d PIR for six months and beagle dogs receiving 500 mg/kg/d for one year were without abnormal findings. In the latter study, in a high-dose group receiving up to 10,000 mg/kg/d for the last 13 weeks, there were no deaths and symptoms were only slight.

With LEV, mortality was seen only after intravenous administration of >= 900 mg/kg in rats. In general, clinical signs across studies and species were minimal with the most consistent observations being neuromuscular effects, salivation and emesis in dogs. In the rodent only, treatment-related changes were seen in the liver and kidney after administration of LEV. In the liver, a reversible increase in liver weight and hypertrophy of centrilobular hepatocytes was observed at doses of 300 mg/kg/d in both sexes in rats and mice. In the kidney, treatment-related changes were seen in male rats; these were dose-related and considered to be significant at doses of LEV of 200-300 mg/kg/d or more in gavage studies and 50 mg/d in dietary studies. This pathology, consisting of hyaline droplet nephropathy accompanied by exacerbation of chronic progressive nephropathy, was considered to be male rat-specific and associated with alpha2-microglobulin accumulation in the proximal tubules that was shown in further studies not to be toxicologically relevant to man. There was no target organ identified in the dog. No deaths, organ failure or other irreversible toxicity was seen after long-term oral treatment up to doses of 1,800 mg/kg/d in the rat, 960 mg/kg/d in the mouse and 1,200 mg/kg/d in the dog.

Carcinogenicity and mutagenicity

Neither PIR nor LEV was carcinogenic in lifetime feeding studies in mice or rats. Neither drug was mutagenic. The major metabolite of LEV, UCB LO57, was found to have a low potential for toxicity in animals and was not mutagenic.

Reproduction toxicity

Reproduction toxicity studies were performed in the mouse, rat and rabbit, after oral administration of both PIR and LEV, to evaluate their effects on fertility, general reproductive performance, fetal and peri- and post-natal development. No effect was observed with either drug on fertility and reproductive performance. In the pregnant rabbit, oral PIR in doses up to 2,700 mg/kg/d had no adverse effect on pregnancy or the development of offspring. No fetal malformations were found in teratology studies in rats with PIR. When LEV was administered to female rats throughout pregnancy and lactation (0, 70, 350, 1,800 mg/kg/d), there was no maternal toxicity. Doses >= 350 mg/kg/d were associated with minor skeletal abnormalities and retarded offspring growth while higher dosing with 1,800 mg/kg/d led to an increase in mortality and behavioural changes in pups.

During the period of organogenesis, LEV was administered to pregnant mice (0, 1,000, 3,000 mg/kg/d), rats (0, 400, 1,200, 3,600 mg/kg/d) and rabbits (0, 200, 600, 1,800 mg/kg/d). LEV was not embryotoxic or teratogenic in mice at these doses. In rats, decreased fetal weight and minor skeletal abnormalities were seen at a dose of 3,600 mg/kg/d and in rabbits an increase in embryofetal mortality, an increase in minor fetal skeletal abnormalities and a decrease in fetal weight were seen at doses of 600-1,800 mg/kg/d.

No abnormal developmental or maternal effects were seen at doses of LEV up to 1,800 mg/kg/d in rats treated during the last third of gestation and throughout lactation. The studies overall indicate that PIR and LEV are unlikely to cause serious adverse effects on human fertility, the developing human embryo and in the pre- and post-natal period at the proposed therapeutic dose levels.

Clinical uses

PIR is indicated for the treatment of cognitive impairment including memory disturbances in age-related cognitive function or decline [26], the early stages of dementia [27, 28] and post-stroke aphasia [29-33]. PIR appeared to have only a weak effect against seizures in preclinical studies. However, Terwinghe and colleagues [34] first described dramatic improvement in a patient with post-anoxic myoclonus treated with PIR (12 g/d). Since then many case reports and open studies [35] and two double-blind studies [36, 37] have confirmed the efficacy of PIR in cortical myoclonus. Guerrini et al. [38] reported efficacy in cortical myoclonus of the Angelman syndrome. Remy and Genton [39] described the use of high doses up to 45 g/d and Genton et al. [40] the long term effects of PIR in myoclonus for up to 10 years.

Levetiracetam. The results of placebo-controlled studies in a total of 904 patients (LEV 592, placebo 312) showed that LEV, when added to existing treatment, is effective in controlling partial seizures in a dose range of 1,000-3,000 mg/d [41-43]. When these results were pooled, responder rates (percentage of patients achieving seizure reduction of >= 50% were 28% of patients on 1,000 mg/d, 32% on 2,000 mg/d and 41% on 3,000 mg/d. In addition, findings from case reports and data in juvenile myoclonic epilepsy [44] and a pharmacodynamic study evaluating photoparoxysmal EEG response [45] and a study with the primary aim of examining drug safety [46], show that LEV has the potential for broad spectrum antiepileptic activity and may thus be effective in the control of generalized seizures.

The results of these trials confirm the high therapeutic index shown in preclinical studies. The drug was generally well tolerated and was without severe adverse reactions. Effects most frequently reported included somnolence, asthenia and dizziness, usually transient and of no more than moderate severity. No changes were shown in a study designed to assess cognitive performance [47].

CONCLUSION

Despite similar chemical structures, PIR and LEV have distinct pharmacological profiles and clinical effects. The unique identity of each drug may be the result of different modes of action. In preclinical and clinical studies, PIR shows efficacy in the treatment of impaired cognition. It also exhibits protection against myoclonus but at very high dosages. In contrast, LEV exhibits fewer nootropic effects but potent protection against seizures in all models of epilepsy. LEV will be further investigated as an antiepileptic drug. The two compounds might share a common activity against myoclonus. Although it has been tried mostly in patients with focal epilepsy, LEV shows promising activity against generalized seizures and should be evaluated in myoclonic seizures and other forms of generalized epilepsy.

Received March 10, 2000 / Accepted May 11, 2000