Dopamine Agonists in Early Therapy for Parkinson Disease: Title and subTitle BreakPromise and Problems

The successful use of levodopa, the immediate precursor of dopamine, to reverse motor deficits in Parkinson disease (PD) revolutionized treatment.¹ But for most patients with PD, the benefits of levodopa therapy have been tempered by adverse effects such as involuntary movements or hallucinations. These adverse effects occur in as many as 90% of patients receiving long-term treatment, although whether they are actually caused by levodopa treatment, or simply reflect progression of the underlying disease, is unknown.²

Some evidence suggests that levodopa may injure substantia nigra neurons.^{3 - 4} Thus, the patient with PD is thrust into an almost untenable therapeutic double bind—does effective short-term symptom relief cause permanent nerve cell loss and ultimately a more malignant disease course? As a result, patients with PD commonly are not treated with levodopa until motor disability interferes significantly with independent functioning. The emotional and economic costs of this treatment approach include early retirement, lost recreational time, and changes in interpersonal relationships.

Initial therapy with a dopamine-receptor agonist may provide a solution to this dilemma. Initiating treatment of PD with a dopamine-receptor agonist is theoretically superior to levodopa, which must be metabolized to dopamine within the brain. Agonists bypass this step, acting at the intact postsynaptic receptor. However, initial treatment with bromocriptine, the first agonist available, was discouraging.⁵ Interest in initial agonist therapy was rekindled after the completion of the placebo-controlled trials of pramipexole or ropinirole. In these studies, motor benefit was sustained for 6 months or longer and motor adverse effects were rare.^{6 - 8} Whether initial use of either agent could confer any advantage over initial therapy with levodopa awaited controlled trials comparing these drugs.

In this issue of THE JOURNAL, the Parkinson Study Group reports the result of such a trial.⁹ The primary goal of this randomized, double-blind controlled clinical trial comparing initial therapy with pramipexole or levodopa/carbidopa combination (henceforth called levodopa) in patients with PD was to determine the risk of developing any of 3 dopaminergic motor adverse effects: dyskinesias, wearing off of motor benefit, or on-off motor fluctuations. Open-label supplementation with levodopa was permitted and dosages of the 3 agents were similar to those used in clinical practice. After 2 years, motor adverse effects, particularly dyskinesias (reduced risk, 67%), were less common in the group treated with pramipexole. On the other hand, although motor benefit was improved compared with baseline function in both groups, the magnitude of improvement was greater in the levodopa-treated group. The benefit of levodopa was observed by both physicians and patients. Hallucinations, somnolence, and edema (generalized and peripheral) were more common in the pramipexole-treated group, and caused drug withdrawal in only 4% of patients. Nausea caused drug withdrawal in 4% of patients taking levodopa and none taking pramipexole.

A 5-year comparison of ropinirole vs levodopa reported earlier this year by a group of European and Canadian investigators showed similar results.¹⁰ Initial therapy with ropinirole was associated with a reduction in dyskinesias, including disabling dyskinesias, and a longer time until the occurence of dyskinesias compared with levodopa. Unlike pramipexole, ropinirole did not reduce the risk of wearing off of motor benefit. As with pramipexole, motor benefit was reduced and hallucinations were more common (17% treated with ropinirole vs 6% levodopa) in those treated with ropinirole, even after open-label levodopa supplementation. A third report comparing initial therapy with the agonist cabergoline (not available in the United States) to initial therapy with levodopa also found reduction in dyskinesias in patients receiving monotherapy.¹¹ A fourth study comparing initial treatment with pergolide to initial levodopa treatment is in progress.¹²

These reports support the hypothesis that initial treatment of PD with a dopamine-receptor agonist reduces the risk of dyskinesias, and possibly of other motor adverse effects. The price paid is less motor benefit, and a greater risk of certain adverse effects. The critical unanswered question is whether dopamine agonists prevent disease progression. Determination of the best course in treating any patient with early PD, then, must be individualized, depending on patient preferences regarding the expected benefits and adverse effects of the drugs. The following questions may be useful in addressing the trade-offs.

Is pramipexole therapy neuroprotective? Is levodopa therapy neurotoxic? The ideal therapy in early PD would reverse symptoms, or at least stop disease progression. No such treatment is known. Neither a neuroprotective effect for pramipexole nor a neurotoxic effect for levodopa is supported by the dopamine transporter imaging results reported in the current study by the Parkinson Study Group.⁹ An ongoing trial also conducted by the Parkinson Study Group and sponsored by the National Institutes of Health to determine the effects of early vs late administration of levodopa in PD includes a similar evaluation using iodine 123 2-β-carboxymethoxy-3-β-(4-iodophenyl)tropane (β-CIT) and single photon emission computed tomography.¹³ The results of this work will provide further insight into whether levodopa treatment is harmful in PD.

Is prevention of dyskinesias important? Mild-to-moderate dyskinesias may cause more social disability than motor impairment. Patients often are completely unaware of dyskinesias that are obvious to family members or physicians.¹⁴ In these clinical trials,^{9 - 10} efficacy related to motor function improvement was greater in the levodopa-treated groups, for whom dyskinesias were more common. No patients withdrew from either study because of dyskinesias. However, while dyskinesias are often not disabling in the early years of treatment, avoiding dyskinesias might prevent the development of more severe dyskinesias later in the course of disease. The pathogenesis of dyskinesias is unknown. One theory is that exposure to levodopa "primes" the brain for production of dyskinesias.¹⁵ If this is true, and if dyskinesias, once present, are progressive, then use of pramipexole or ropinirole to delay the development of dyskinesias would result in less disability late in the course of the illness. For example, disabling dyskinesias were more common in the levodopa group than in the ropinirole-treated group.¹⁰ The reduced frequency of dyskinesias with initial agonist therapy would be useful for patients for whom social embarrassment or impairment of fine motor tasks is a problem. If initiating agonist therapy reduces the development of disabling dyskinesias, this additional benefit would be especially important for persons with PD in whom longer survival is most likely (such as those of younger age and in good health).

When is prevention of wearing off of motor benefit important? The second motor adverse effect that occurred less frequently with pramipexole monotherapy was the wearing off of motor benefit. The pathogenesis of wearing off is not completely understood, although diminished neuronal reserve due to loss of nigral dopamine neurons is thought to be a key component.¹⁶ If loss of nigral neurons is central to the process, then using treatment strategies to avoid this loss are critical. Whatever the cause of wearing off, the inability to sustain motor function in PD is a source of significant disability. Dopamine agonists help to avoid this frustrating phenomenon and may be useful particularly in more active persons in whom daily responsibilities require a predictable level of motor performance.

Are hallucinations and somnolence disabling? Hallucinations and somnolence are recognized causes of serious disability in persons with PD.^{17 - 19} Hallucinations were nearly 3 times more frequent in persons treated with pramipexole (9% vs 3% levodopa) or ropinirole (17% vs 6%). Somnolence, including sleepiness while driving, was also more common in the pramipexole-treated group, and a larger proportion of ropinirole-treated patients were affected (32% pramipexole, 27% ropinirole, 17% or 19% levodopa). Both adverse effects may have many causes. Nonetheless, cautious use of initial agonist therapy is advised if other risk factors are present, such as older age, cognitive impairment, or sleep disorder. For other patients, the potential benefits of initial agonist treatment may mitigate the possibility for adverse effects. Education of these patients and their family members is critical. Vigilance should be especially high during the early weeks of therapy and when doses are increased. Driving should be avoided during these times, and driving long distances, particularly alone, discouraged. Throughout treatment, careful monitoring and prompt dosage reduction when early signs of adverse effects occur are important to reduce the risk of serious sequelae of these adverse effects.

Should edema be avoided? While edema prompted withdrawal of the medication for some patients, the pathogenesis of this effect is unknown. A similar but less robust association was suggested with ropinirole (14% ropinirole vs 5.6% levodopa). Further investigation of the cause of edema and its consequences and treatment are important topics for future study.

Are these results applicable to all patients with PD? Volunteers in clinical trials of PD tend to be relatively younger and healthier when compared with the entire PD population. The results of the study reported by the Parkinson Study Group, therefore, are likely the best expected outcome of initial agonist treatment. Few data exist for older patients or for those with comorbid dementia or other illnesses.²⁰ For these groups, the benefits and risks of initial agonist therapy remain to be determined in controlled trials.

Few clinical trials of antiparkinsonian therapy have extended beyond a year, yet persons with PD commonly live decades after therapy is initiated. As the proportion of the US population at risk for PD increases, a long-term approach to therapy assumes increasing importance. Understanding the relationship between the initial treatment of PD and outcomes such as morbidity, mortality, and responsiveness to therapies for advanced disease (for example, surgical interventions, stem cell therapy, gene therapy) will be needed. Other outcomes, such as the cost-effectiveness of a particular therapy, are virtually unexplored. Controlled trials lasting many years and long-term follow-up of former participants in pivotal clinical trials will provide some answers to these questions. Until then, the current findings of the Parkinson Study Group will help inform patient-physician decisions regarding early therapy for PD.