Levodopa therapy

Normal motor function is dependent on the highly regulated synthesis and release of the transmitter dopamine by neurons projecting from the substantia nigra to the corpus striatum. Parkinson's disease leads to progressive degeneration of these neurons. Its core symptoms (tremor, rigidity, and bradykinesia) are the direct consequence of an insufficiency of intrasynaptic dopamine and are manifest when striatal levels of the transmitter amine are depleted by 70% or more.¹

Limitations of current levodopa therapy. Almost 30 years after its clinical introduction,² levodopa remains the standard of care for the treatment of PD. It is usually administered in combination with a peripherally acting decarboxylase inhibitor, such as carbidopa, that blocks the conversion of levodopa to dopamine outside the CNS. By this means, peripherally mediated adverse effects are reduced and centrally mediated antiparkinsonian activity is augmented. After oral ingestion, levodopa crosses the blood-brain barrier and is taken up by residual dopaminergic nerve terminals, in which it is decarboxylated to dopamine. The transmitter amine is stored in vesicles and protected against enzymatic degradation until it is released intrasynaptically in response to nerve terminal depolarization. The release of dopamine synthesized from exogenous levodopa thus acts to normalize dopamine concentrations at its postsynaptic receptors. As dopamine-mediated synaptic transmission is restored, parkinsonian symptoms are alleviated.

Although levodopa is a highly effective palliative that tends to normalize life expectancy for parkinsonian patients,³ there are significant differences between the normally functioning dopamine system and the restoration of function provided by standard levodopa treatment. These differences presumably are magnified as the population of nigrostriatal dopaminergic neurons declines. Ordinarily, the nigrostriatal pathway is essentially a tonically operating system: Therefore, intrasynaptic dopamine concentrations remain fairly constant over time.^4,5 In contrast, standard levodopa therapy in patients with advanced PD results, at best, in only relatively brief, episodic restoration of physiologic dopamine levels.⁶ For most of the dosing cycle, postsynaptic dopamine receptors are exposed to subthreshold transmitter concentrations, interrupted soon after each dose by a brief interval during which transmitter levels rise quickly through the physiologic range and probably far beyond.⁷ Recent evidence suggests that this nonphysiologic stimulus pattern may be central to the pathogenesis of motor complications.

In early PD intrasynaptic dopamine formed from exogenous levodopa is believed to be largely synthesized and stored in striatal dopaminergic terminals. It is from these terminals that the transmitter presumably is released at a relatively steady rate under normal neuronal control. Dopamine storage in nerve terminal vesicles explains why the effects of levodopa initially last for several hours despite a much shorter half-life (roughly 90 minutes) in plasma and in the CNS. Continuous, appropriate rate release of dopamine from nerve endings ensures the smooth, complication-free response to levodopa that patients with early stage parkinsonism typically enjoy.

With the progressive destruction of substantia nigra neurons, less exogenous levodopa enters the striatal terminals. Instead, increasing amounts are taken up and converted to dopamine in other decarboxylase-containing cells, especially nonaminergic neurons, glia, and endothelial cells.⁸ In the absence of mechanisms for storing or regulating the intrasynaptic release of dopamine, the newly synthesized amine quickly leaks from these cells into the interstitial space, from which it can diffuse and come into contact with nearby dopamine receptors. Under these circumstances, intrasynaptic dopamine concentrations begin to reflect the broad swings in plasma and cerebral levodopa levels that attend standard dosing regimens. Therefore, as a direct consequence of the declining ability of the dwindling population of dopaminergic terminals to store dopamine, the antiparkinsonian action of levodopa progressively shortens. This reduced duration of action accounts for the initial appearance of one of the earliest response complications, motor fluctuations of the "wearing-off" type.^9,10 As discussed below, increasing evidence indicates that this chronic intermittent stimulation of postsynaptic dopamine receptors also contributes to the subsequent appearance of the other major motor complications.

Clinical phenomenology of motor complications. Administration of levodopa to most parkinsonian patients eventually results in the appearance of one or more motor response alterations that ultimately compromise the drug's clinical usefulness.^3,6 These complications include motor fluctuations as well as choreiform and dystonic dyskinesias that appear at various times during the dose cycle. Of these, fluctuations of the "wearing-off" and "on-off" types and peak dose dyskinesias occur most commonly.

"Wearing-off" phenomena are typically the first motor complication to become clinically evident. Also known as "end-of-dose" deterioration, these response fluctuations occur in temporal relation to dosing. They arise as a direct result of the shortening of levodopa's duration of action which, in turn, initially reflects an insufficient capacity of the striatum to store newly synthesized dopamine.¹¹ Later, continuing treatment and advancing dopamine neuron loss (and thus more prolonged exposure of downstream dopaminoceptive elements to intermittent stimulation) contribute to the increasing severity of "wearing-off" fluctuations and to the subsequent appearance of the response alterations that underlie motor fluctuations of the "on-off" type and peak dose dyskinesias.

"On-off" phenomena are characterized by sudden, random switches between the relatively treated ("on," or hyperkinetic) and untreated ("off," or hypokinetic) states. They are presumed to arise as a consequence of the increasing steepness in the levodopa dose-antiparkinsonian response relation that accompanies disease progression.¹¹ Under such circumstances, relatively small changes in circulating levodopa, and thus in striatal dopamine, become increasingly able to induce large shifts in dopaminergic transmission and thus in motor function. The choreiform dyskinesias that occur when striatal dopamine levels peak reflect the progressive reduction in the threshold for inducing these abnormal movements that also attends advancing disease.¹² Because there is no associated change in the threshold dose for ameliorating parkinsonian signs, the therapeutic index for levodopa essentially vanishes in late-stage PD. At this point, the dose of levodopa necessary to alleviate symptoms becomes sufficient to induce dyskinesias.

All motor complications tend to become more common and more severe with advancing disease. It has been estimated that about half of levodopa-treated patients will develop one or more response complications within 5 years of symptom onset.¹³ Long-term administration of levodopa is therefore associated with an erosion in clinical benefit that reflects a rise in adverse effects more than a decline in efficacy against core symptoms.

Pathogenesis of motor complications. Dietary, metabolic, and pharmacokinetic mechanisms were among the early theories offered to explain the appearance of motor complications in parkinsonian patients receiving long-term levodopa therapy. Although such factors can influence certain aspects of the response to levodopa, they fail to explain the totality of the motor response-complication syndrome. More recent preclinical and clinical data have implicated changes in basal ganglionic structures downstream from the degenerating nigrostriatal projections. Therefore, although presynaptic dopaminergic neuron degeneration is the initial event and accounts for early stage "wearing-off" fluctuations, postsynaptic alterations soon play the central role in the pathogenesis of all major motor response complications.^12,14,15

Current research has begun to elucidate the basis for these secondary changes in structures downstream from the nigrostriatal dopamine system. These changes now appear to be the direct result of the chronic intermittent stimulation of striatal postsynaptic dopaminergic receptors. Normally, the nigrostriatal system operates to maintain essentially stable intrasynaptic dopamine concentrations.⁹ In early PD, this situation presumably continues, as a result of the ability of dopaminergic terminals to store dopamine and thus to "buffer" the variations in dopamine synthesis associated with periodic levodopa precursor dosing.^9,10,16 With advancing disease, however, dopaminergic neurons continue to degenerate. In an attempt to compensate for this loss, residual dopaminergic neurons accelerate dopamine formation and rapidly release the newly synthesized amine rather than retaining it in storage vesicles.^17,18 In addition, as already noted, nondopaminergic neurons and other cells that possess significant decarboxylase activity become an increasingly important source of intrasynaptic dopamine.⁶ Just as in the few surviving dopaminergic neurons, dopamine, once synthesized in these cells, is immediately released. As a result, intrasynaptic dopamine concentrations begin to reflect the marked swings in precursor availability characteristically associated with the periodic ingestion of levodopa.¹⁰ Striatal dopaminergic synapses thus gradually convert from an essentially tonic to a pseudophasic mode of operation.¹¹

Lessons from animal models. Animal models of PD have been studied extensively to obtain a more detailed understanding of the CNS changes associated with chronic levodopa therapy. Recently, these models have been applied to the evaluation of basal ganglionic modifications that accompany levodopa-induced motor response alterations mimicking those manifested by parkinsonian patients. The results now suggest that a heightened response by striatal medium-sized GABAergic neurons to glutamatergic inputs from the cerebral cortex may be a factor in the pathogenesis of motor complications.

In rats, intramesencephalic administration of the potent neurotoxin 6-hydroxydopamine causes rapid and profound degeneration of dopaminergic neurons. The number of dopamine neurons destroyed depends on the site and the amount of 6-hydroxydopamine infused. Standard techniques typically eliminate more than 95% of nigral neurons, a situation approximating that occurring in advanced PD.¹⁹ However, rats with this degree of dopamine system loss do not manifest precisely the same motor abnormalities as humans do. For example, although they exhibit unmistakable hypokinesia, rats do not display obvious rigidity or resting tremor.

When 6-hydroxydopamine is administered unilaterally, as is usually the case, rodents develop postural asymmetry and, when they are given levodopa, they characteristically rotate away from the side of the lesion. For several decades, this turning model has been used to select dopaminomimetics that will be effective in the relief of parkinsonian symptoms.²⁰ More recently, it has become apparent that this rat model may prove no less useful for studies of the pathogenesis and treatment of motor response complications. Rats unilaterally lesioned with 6-hydroxydopamine and then given levodopa by twice-daily injection in combination with a peripheral decarboxylase inhibitor develop progressive motor response alterations resembling those found in similarly treated parkinsonian patients.¹⁹ Daily measurements reveal a progressive shortening in response duration that becomes statistically significant within 3 weeks.

Therefore, parkinsonian rats, like parkinsonian patients, manifest"wearing-off" phenomena. This response change cannot be explained by an increasing loss of dopamine neurons because, in the rat, the lesion does not progress at the time these measurements are made. The twice-daily administration of levodopa to rats given a smaller amount of 6-hydroxydopamine, which destroys less than 95% of nigral neurons and thus mimics an earlier stage of PD, does not produce any measurable shortening in response time.¹⁹ Again, this result is comparable to what is observed in parkinsonian patients, i.e., clinically evident"wearing-off" fluctuations do not occur in the initial stages of disease. Taken together, these findings indicate that "wearing-off" phenomena appear only when a relatively high proportion of nigrostriatal dopamine neurons is lost.

Parkinsonian rats given several weeks of twice-daily levodopa treatment also develop other response alterations that may serve as models of human motor complications. For example, the frequency with which they manifest an"off" response to an otherwise effective dose of levodopa increases significantly.²¹ Similarly, although parkinsonian rats do not manifest choreiform dyskinesias in response to dopaminomimetic drugs, their peak turning velocity may serve as an index of the magnitude of their antiparkinsonian response and thus to the severity of associated dyskinesias. Chronic levodopa therapy leads to a progressive increase in maximal rotatory velocity.²² The appearance of motor response alterations in parkinsonian rats after several weeks of levodopa treatment occurs only under certain conditions. As already noted, an extensive loss of the nigrostriatal dopaminergic system must be present. In addition, levodopa must be administered intermittently (bid injections were ordinarily used). Animals receiving the same daily dose of levodopa by continuous round-the-clock infusion do not manifest these response changes.²³

The animal model data therefore indicate that both an intermittent mode of levodopa administration and a profound loss of dopaminergic neurons may be necessary for motor response alterations to become manifest. With a sufficient loss of dopamine neurons, the response to levodopa can be affected rather quickly-within a few weeks in 6-hydroxydopamine-lesioned rats¹⁹ or MPTP-lesioned primates,²⁴ and a few months in humans with MPTP-induced parkinsonism.²⁵

An extrapolation of these animal model observations to the treatment of human PD appears to cast doubt on the usefulness of deferring the introduction of dopaminomimetic therapy in the hope of limiting motor complications. Onset occurs when the number of dopaminergic terminals is no longer sufficient to buffer the swings in levodopa levels associated with intermittent precursor administration. On the other hand, these observations do appear to support the view that the early use of levodopa treatment strategies that afford a more continuous stimulation of dopaminoceptive elements might act to delay or even prevent onset of motor response complications in parkinsonian patients.

Recent studies have sought to identify brain structures that contribute to the levodopa-associated changes underlying motor response alterations.¹² In rodent models, the response modifications that attend chronic intermittent levodopa treatment were initially elicited by acute challenge with levodopa. Because the effects of levodopa on motor function depend on its decarboxylation to dopamine in presynaptic dopaminergic terminals or elsewhere, the associated motor response alterations in both rats and humans could reflect changes at either the presynaptic or the postsynaptic level. To determine whether the changes might be occurring only postsynaptic to the dopamine system, acute challenge studies were conducted with apomorphine. This drug acts directly to stimulate postsynaptic dopaminergic receptors. Its motor effects do not depend on the integrity of the presynaptic dopaminergic system. In both parkinsonian rodents and parkinsonian patients, apomorphine reproduced the response changes observed with levodopa.^16,20,26 These findings indicate that motor fluctuations and peak dose dyskinesias essentially reflect changes occurring in structures downstream from the nigrostriatal dopamine system. Although dopamine terminal degeneration probably accounts for the initial appearance of "wearing-off" phenomena in patients with PD, current evidence clearly indicates that secondary changes at the postsynaptic level eventually are responsible for most of the underlying shortening in motor response duration.

Striatal GABA neuron changes. The foregoing conclusions focused attention on possible changes in striatal GABAergic efferent neurons that receive dopaminergic input from the substantia nigra. Initial studies indicated that effects on the binding characteristics of striatal dopaminergic receptors in 6-hydroxydopamine-lesioned rats given intermittent levodopa, just as in parkinsonian patients receiving the same therapeutic regimen, were insufficient to account for the observed alterations in motor response.^27,28 On the other hand, an evaluation of peptide cotransmitters used by these GABAergic neurons suggested profound functional modifications.^29,30 Striatal GABAergic neurons that largely express the D₂ dopamine receptor subtype project primarily to the internal segment of the globus pallidus via the external globus pallidus and subthalamic nucleus; they utilize enkephalin and neurotensin as peptide co-transmitters.³¹ In contrast, striatal GABAergic neurons that largely express the D₁ dopamine receptor subtype mainly project directly to the internal globus pallidus and contain dynorphin and substance P. Lesioning the dopamine system with 6-hydroxydopamine significantly increases enkephalin and neurotensin levels. Subsequent intermittent levodopa treatment produced additional changes, most notably a large rise in both dynorphin and neurotensin.³⁰ Continuously administered levodopa did not have this effect. At the same time, mRNA levels for these neuropeptides were also significantly elevated, suggesting that the concentration changes reflected accelerated synthesis rates.³¹ Limited pharmacologic studies have suggested that these peptide alterations may have important implications for motor performance.³² Clearly, they can be presumed to reflect substantial functional alterations in these GABAergic projection neurons at a time when levodopa-associated response alterations have made their appearance.

Additional insight into the nature of the functional alterations that occur in dopaminergically deafferented striatal GABAergic neurons after chronic intermittent levodopa therapy has been gained through studies with selective dopamine receptor agonists. By this means it is possible to distinguish contributions made by the direct D₁ dopamine receptor-mediated pathway from those related to the indirect D₂-mediated projections. Levodopa-treated parkinsonian rats evidencing motor response alterations were found to have a profoundly reduced response to agonists that stimulate mainly D₁ dopamine receptors.²⁸ In contrast, motor responses to a selective D₂ agonist were markedly increased. Thus, dopaminergic responsivity of the direct striatonigral pathway is attenuated, whereas responsivity of the indirect striatopallidal pathway is augmented.

These results may indicate that an imbalance between striatal output pathways that are influenced by D₁ and D₂ dopamine receptor stimulation contributes to the pathogenesis of motor complications. Striatal dopaminoceptive GABAergic neurons are densely innervated by glutamatergic projections from the cerebral cortex.³³ Alterations in these glutamate inputs might therefore help to bring about the functional changes that occur in the GABAergic output systems. For example, it could be hypothesized that excessive stimulation of glutamatergic receptors of the N-methyl-D-aspartate (NMDA) subtype on striatal GABAergic neurons produces a hyperfunctional response. Under such circumstances, D₁ dopamine receptor-mediated stimulation might be limited by a ceiling effect, whereas D₂ receptor-mediated inhibition might be enhanced. Studies in both animal models and in parkinsonian patients lend support to this possibility. Systemic administration of the selective NMDA receptor antagonist MK-801 to levodopa-treated parkinsonian rats markedly reduces motor response alterations.^26,34 Moreover, co-administration of an NMDA receptor antagonist with intermittent levodopa treatment substantially blocks the onset of the characteristic response changes.²¹ Finally, intrastriatal administration of the NMDA antagonist produces far greater effects than when this drug is given systemically or directly into other basal ganglionic structures.³⁴

These observations indicate that the functional changes produced by NMDA receptor blockade occur mainly at receptors located on striatal medium-sized GABAergic neurons. In addition, these findings are consistent with the view that glutamatergic hyper-stimulation of striatal GABAergic neurons may be a factor in the appearance of motor complications.

Further evidence in support of this possibility derives from studies in nonhuman primates. Rats do not develop choreiform dyskinesias, but monkeys that have been rendered parkinsonian by dopamine system lesioning with the neurotoxin MPTP and then given chronic intermittent levodopa therapy do develop such dyskinesias. Systemic administration of certain NMDA antagonists to these animals significantly reduces peak dose dyskinesias at doses that have no effect on the antiparkinsonian response to levodopa.³⁵ Because not all NMDA receptor antagonists have the same effect on motor complications, an interaction with specific subtypes of NMDA receptors may be important for an optimal response. Limited observations in levodopa-treated parkinsonian patients involving the NMDA antagonists dextrorphan, dextromethorphan, and amantadine also suggest that some drugs of this type can limit motor response complications and thus further implicate glutamatergic hyperfunction in the pathogenesis of human motor response complications (Verhagen Metman L, et al., unpublished observations).^36,37 Recent investigations have attempted to determine a basis for the postulated excessive glutamatergic influence on striatal GABAergic neurons. Preliminary evidence from one such study suggests that the binding affinity of striatal NMDA receptors increases rather than declines in parkinsonian rats that develop levodopa-induced response modifications (Papa SM, et al., unpublished observations). These results appear to be more compatible with the hypothesis that upregulation of NMDA receptor sensitivity, rather than hyperfunction of cortical glutamatergic afferents, explains the beneficial effects of NMDA antagonists on motor complications. Current studies of NMDA receptor subunit composition and phosphorylation state may yield additional insight into this matter.

Mechanisms by which chronic intermittent stimulation of dopaminergic receptors on striatal medium-sized GABAergic neurons might enhance the synaptic efficacy of their NMDA receptors have recently begun to be elucidated.²² NMDA receptors expressed on striatal GABAergic neurons largely are confined to the distal tips of their dendritic spines.³³ Dopamine receptors are located nearby on the same spines, although somewhat more proximally. Components of the signaling pathways linking the dopamine and glutamate receptors have recently become better understood.³⁸ Present evidence suggests that the cAMP-protein kinase A-mediated pathway contributes to D₁ dopamine receptor-associated activation of NMDA receptors, whereas a calcium-calmodulin-dependent kinase II pathway participates in the signaling cascade linked to D₂ receptors. It is tempting to speculate that activation of these signal transduction cascades results in the site-specific hyperphosphorylation of NMDA receptor subunits, leading to long-term potentiation of their synaptic efficacy. As a result, striatal GABAergic system function changes in ways that favor the appearance of motor response complications.

Significance of the data for treatment of Parkinson's disease. In rodent and primate models of PD, levodopa-associated motor response changes can be reliably induced by intermittent dopaminergic stimulation. Clinical experience suggests that the same situation pertains in parkinsonian patients, although no data from definitive evaluations of the prophylactic value of relatively continuous dopaminomimetic administration are yet available. However, studies have been performed to evaluate the palliative benefit derived from converting parkinsonian patients who develop motor complications while receiving a standard (usually 3-6 times daily) levodopa regimen to a continuous levodopa infusion. Ten days of round-the-clock, optimal dose IV administration progressively reduced all motor complications.³⁹ After a continuous subcutaneous infusion of a dopamine agonist lasting 3 months, the beneficial effects on preexisting motor complications were even more dramatic.⁴⁰

Taken together, these studies indicate that the motor fluctuations and peak dose dyskinesias complicating dopaminomimetic therapy of relatively advanced PD are attributable, at least in part, to the intermittent administration of levodopa and can be partially reversed by treatments that provide more continuous dopamine replacement. It is therefore not inconceivable that the continuous stimulation of striatal dopaminergic receptors from the outset of dopaminomimetic treatment will confer prophylactic benefit by limiting the changes that occur in striatal GABA neurons as a consequence of their chronic nonphysiologic stimulation.

Newer therapies: how well do they meet the challenges of Parkinson's disease? Ideally, the symptomatic treatment of PD should ameliorate symptoms but avoid clinically significant adverse effects. Data from the animal models described above indicate that the intermittent stimulation of striatal dopaminergic receptors may play a role in the pathogenesis of the most common forms of motor response complications. In parkinsonian animals, the response alterations associated with intermittent levodopa administration not only are reversed but also are prevented by therapeutic regimens that provide continuous dopaminergic receptor stimulation. Moreover, continuous levodopa infusions have documented palliative value in patients with advanced disease.³⁹ Stable stimulation of dopaminergic receptors may also confer prophylactic benefit at earlier stages of disease.²¹ Unfortunately, levodopa is too acidic and too insoluble for convenient long-term parenteral administration. At present, no ideal alternatives are available clinically.

Improved formulations of levodopa. Controlled-release formulations are designed to provide more constant plasma levodopa concentrations and thus more physiologic stimulation of striatal dopamine receptors.^41,42 Two currently marketed examples are levodopa-carbidopa CR and levodopa-benserazide HBS. One of the relatively minor drawbacks of these formulations has been the reduction in the bioavailability of levodopa compared to standard levodopa preparations.^43,44 The adverse effects associated with these controlled-release formulations are essentially no greater than those associated with traditional formulations of levodopa.

The controlled-release formulations have the advantage of permitting a reduction in the number of daily doses, albeit at the risk of compromising the benefits that arise from their ability to provide more continuous dopaminergic stimulation. On the other hand, controlled-release preparations are slower to exert their antiparkinsonian effects. As a result, many patients feel the need for additional doses of standard levodopa, especially on awakening in the morning. In addition, controlled-release levodopa formulations sometimes produce little or no response, leading to a concern about their reliability. These limitations may explain the relatively limited acceptance of these theoretically improved formulations.⁴¹

Dopamine receptor agonists. Dopamine agonists have a number of theoretical advantages over levodopa for the treatment of PD, including the possibility of being just as effective while having a duration of action sufficient to provide relatively continuous dopaminergic replacement. On the basis of early observations in the experimental animal, dopamine agonists were selected initially for use in PD because of their ability to interact with D₂ dopamine receptors.^45,46 Although dopamine interacts with all dopaminergic receptors, it binds more avidly to the D₂ family of receptors, especially those of the D₃ subtype. Therefore, among the currently marketed dopamine agonists, all still largely target the D₂ family of receptors. However, some data from animal models have suggested that D₁ stimulation, alone or in combination with D₂ stimulation, may be preferable.^47-49 Clinical studies have yet to resolve this issue. Both bromocriptine (which stimulates both D₂ and D₃ receptor subtypes but antagonizes the D₁) and pergolide (a D₂- and D₃-preferring agonist) have been widely used clinically.⁵⁰ Ropinerole and pramipexole, also D₂- and especially D₃-preferring agonists, are about to be introduced in the United States market.^51,52

In theory, early monotherapy with a dopamine agonist should eliminate the need for levodopa and, by virtue of such agonists' longer duration of action, possibly even delay the onset of motor complications.⁵³ Unfortunately, clinical results have yet to fully live up to this scenario. Although the severity of the fluctuations often declines after the introduction of a dopamine agonist, no existing drug of this type matches the antiparkinsonian efficacy of levodopa, particularly in patients with advanced disease. Additional levodopa is usually required after 2 or 3 years of agonist monotherapy.^54,55

No effect of dopamine agonists on prolonging the latency to onset of motor complications has yet been convincingly documented, possibly because their duration of action, although far longer than that of levodopa, is still too short.⁵³ On the other hand, several new long-acting dopamine agonists are now under development. These include cabergoline, which has a prolonged elimination half-life and therapeutic effects lasting 1 or 2 days,^56,57 and a lipid-soluble agonist, N-0923, which is suitable for continuous transdermal administration.⁵⁸

Inhibitors of levodopa metabolism. Motor fluctuations and later choreiform dyskinesias appear when intrasynaptic dopamine concentrations begin to reflect the shifts in circulating levodopa levels that follow ingestion of each dose. In theory, any drug that prolongs the plasma half-life of levodopa or the striatal half-life of dopamine should lessen the risk for these complications.⁵⁹ Dopamine is catabolized by monoamine oxidase (MAO), which deaminates it intraneuronally, and by catechol O-methyltransferase (COMT), which methylates dopamine extraneuronally.⁶⁰ Inhibition of MAO therefore tends to increase intraneuronal dopamine concentrations, whereas inhibition of COMT elevates extraneuronal concentrations of dopamine.

Years ago, when decarboxylase inhibitors such as carbidopa were first combined with levodopa, the objective was to reduce the side effects associated with the peripheral formation of dopamine, and the hope was that decarboxylase inhibitors would also prolong the plasma half-life of levodopa. The objective of reducing adverse events was achieved, but the plasma half-life of levodopa was little affected, because another enzyme (COMT) took over and degraded levodopa into 3-O-methyldopa.

Levodopa is degraded by COMT in several organs, notably the gut, the liver, and the brain. In the gut, COMT causes a considerable loss of levodopa by converting it to 3-O-methyldopa; once absorbed, levodopa is further degraded by hepatic COMT. Therefore, blocking COMT in the gut and liver should improve the bioavailability of levodopa and extend its duration of action. In the brain, levodopa and its metabolite, dopamine, are both degraded by COMT. This effect of COMT contributes to lower dopamine concentrations and shortens the duration of action of each levodopa dose. Therefore, blocking COMT in both the brain and the periphery might result in further benefit.

Two well-tolerated COMT inhibitors have recently been developed. One, entacapone, blocks COMT in the periphery but does not cross the blood-brain barrier.⁶⁰ When given with levodopa, entacapone prolongs its antiparkinsonian action by about 50% without increasing the severity of dyskinesias.^60,61 The duration of dyskinesias, however, is increased, reflecting the extension of the plasma half-life of levodopa.

Unlike entacapone the other compound, tolcapone, does cross the blood-brain barrier. Pharmacokinetic studies indicate that, as expected, it significantly increases the area under the dose-concentration curve and the half-life of levodopa,⁶² thus prolonging the duration of"on" time and the duration of dyskinesias, but not their peak severity. It also reduces the total daily requirement for levodopa.^59,63 These useful clinical effects appear to be related mainly to the ability of this drug to block COMT in the gut, thus increasing levodopa uptake by limiting its catabolism.⁵⁹

Results with MAO inhibitors have been somewhat less encouraging. Since the late 1960s, selegiline has been extensively tested in patients with PD. These studies were based primarily on the possibility that by inhibiting dopamine degradation in the brain selegiline would prolong the duration of action of levodopa and thus smooth out its effects in advanced disease. Clinical trials, conducted mainly in Europe, indicated that selegiline, when added to levodopa, could indeed confer some palliative benefit.^64,65 Moreover, selegiline was found to be generally well tolerated. However, it does tend to exacerbate levodopa-associated adverse events, such as peak dose dyskinesias, by increasing peak levodopa concentrations, an effect not usually seen after COMT inhibition.⁶⁶ At present, selegiline is administered primarily in the hope that it will slow the progression of PD.^67,68 Its symptomatic effects have been accorded relatively little attention except for their potential to confound the interpretation of neuroprotective trial results.

Future therapeutic prospects. Improved palliative treatments for PD will be a continuing requirement for patient welfare and thus an enduring goal for pharmaceutical discovery in the foreseeable future. Especially interesting in this respect is the possibility of finding drugs that interact not with cell surface receptors but rather with intraneuronal signal-transduction components. Clearly, however, the major unmet medical need for patients who suffer from PD is a treatment(s) that can delay the progressive destruction of dopaminergic neurons. Fortunately, current neuroscientific research is providing an ever-expanding array of promising therapeutic leads. At present, perhaps the most compelling research questions involve the way in which genetic and environmental factors can trigger a cascade of events that culminate in apoptotic death of dopamine neurons.

Answers may be more rapidly forthcoming with the recent announcement that a single missense mutation near the hydrophobic center of theα-synuclein molecule can cause an autosomal dominant form of PD.⁶⁹ This finding raises many interesting possibilities. For example, replacement of an alanine with a threonine residue in synuclein may favor adoption of a β-pleated rather than an α-helical secondary structure.⁶⁹ β-Sheet formation can lead to a self-aggregating protein, not unlike the amyloid deposits in Alzheimer's disease, that has toxic potential by virtue of its chemical properties or its mere physical presence. However, only a small proportion of familial or sporadic PD can be explained by the synuclein alteration. Abnormalities in other genes must also contribute to the pathogenesis of this disorder and must continue to be sought. Now, however, neurobiological research can be much more focused on how dopaminergic neurons die. Specifically, we have a far better opportunity to begin identifying pathogenetic links in the chain of events that connect initial etiologic factors with the ultimate neuronal destruction. The discovery of only a single critical component in this lethal cascade might provide an extraordinary target for definitive therapeutic intervention.