Cabergoline in the treatment of early parkinson's disease

The best drug treatment for newly diagnosed cases of Parkinson's disease (PD) is still debated. Levodopa (L-dopa) undoubtedly is efficacious, well tolerated, easy to handle, and a low-cost treatment; mean improvement at the beginning of treatment is greater than 50% and appears to remain stable for 2 to 3 years, but uniformly declines after this time. ^[1] Longterm side effects become increasingly frequent as treatment continues. ^[2,3] After 3 years of L-dopa medication, 25 to 30% of patients develop motor complications, the frequency reaching 50 to 80% after 5 to 10 years of treatment. ^[4] Although the reasons for the development of motor fluctuations remain obscure, some authors consider that L-dopa-related end-of-dose failure ("wearing off") may be attributable to dopamine storage failure ^[5] or to postsynaptic alterations possibly occurring in response to nonphysiologic fluctuations in synaptic dopamine concentration, ^[6,7] or both. Direct and long-lasting dopaminergic stimulation may delay or prevent oscillations in motor response. ^[7]

A substantial number of clinical observations reported in recent years indicates that patients started on pharmacologic treatment for their PD symptoms with dopamine agonists alone exhibit a much lower frequency of motor complications than patients started on L-dopa. ^[8,9] However, only a small proportion of patients derive a stable benefit and can be maintained under long-term treatment with dopamine agonists alone. Moreover, side effects are more frequent with dopamine agonists and the cost of treatment is higher. Alternative therapeutic strategies suggest commencing treatment with low doses of L-dopa and adding a dopamine agonist instead of increasing the L-dopa dosage, ^[10] starting treatment with a combination of low doses of L-dopa and a dopamine agonist, ^[9] or starting with a dopamine agonist and adding L-dopa when needed. ^[11] Each of these approaches may have a potential advantage over the use of L-dopa alone by delaying and reducing the occurrence of long-term motor complications. However, the preferred approach is still unclear as there have been few prospective comparative studies. In most of these trials, bromocriptine has been used, ^[12-16] but there is little information on pergolide ^[17,18] and lisuride. ^[19]

Cabergoline (CBG) is a new ergolinic derivative with an in vitro affinity for rat striatal dopamine receptors about two times lower than that of bromocriptine and pergolide; its elimination half-life has been estimated as about 65 hours, ^[20] and therefore CBG, besides being suitable for once-daily administration, may provide a fairly steady dopaminergic stimulation. PET studies in the monkey have indeed shown a 40% reduction of¹¹ C-raclopride striatal receptor occupancy 72 hours after CBG administration. ^[21]

The present double-blind study was aimed at assessing whether initial treatment with CBG in newly diagnosed PD patients may postpone the onset of motor fluctuations compared with starting treatment with L-dopa alone. In this paper, we report a planned interim analysis of the therapeutic efficacy of CBG compared with L-dopa during the first year of treatment. In addition, we examined the ability of CBG to postpone the initiation of L-dopa treatment and to possess L-dopa-sparing efficacy in the first year of treatment.

Patients and Methods.

The study is a multinational, multicenter, double-blind, parallel-group study in which newly and recently diagnosed PD patients (Hoehn and Yahr stage 1 to 3) with functional disability severe enough to require pharmacologic treatment were recruited. After obtaining informed consent, a total of 413 patients entered the baseline period of the study between October 1989 and July 1992 at 32 centers in Europe and South America. Prior treatment with L-dopa, selegiline, or dopamine agonists was a major exclusion criteria; previous treatment with amantadine or anticholinergic drugs was allowed, but these treatments had to be discontinued at least 4 weeks prior to the study. All the patients met the major inclusion/exclusion criteria with the exception of two patients (one in each group) who were subsequently diagnosed as suffering from multiple system atrophy; for the purpose of this analysis, these patients have not been excluded. Minor protocol violations were reported in a few patients and were considered of no relevance for the analysis: two patients (one in each group) had already received a dopamine agonist prior to study inclusion, one patient in the L-dopa group had already received selegiline, four patients (one in the CBG and three in the L-dopa group) discontinued their anticholinergic treatment less than 4 weeks prior to study start, one patient (78 years) was outside age range (75 years maximum) defined by the inclusion criteria.

Patients were titrated on either CBG or L-dopa (with carbidopa as a dopa decarboxylase inhibitor) over a maximum period of 24 weeks, from dose level 1 (0.25 mg/d CBG once daily or 100 mg/d L-dopa in three daily doses) to dose level 7 (4 mg/d CBG once daily or 600 mg/d L-dopa in four daily doses). Dose titration was carried out until achievement of the optimal daily dose (defined as a well-tolerated dose at which no further improvement as compared to previous dose was obtained) or until the maximum dose level was reached. Thereafter, long-term treatment with the identified optimal daily dose was initiated; in case of diminishing clinical control, L-dopa/carbidopa was added on an open basis; experimental treatment doses were decreased if adverse events occurred. Domperidone, if needed, could be administered.

PD disability was evaluated by means of the complete Unified Parkinson's Disease Rating Scale (UPDRS) ^[22] at baseline, at the end of the dose-titration period, and yearly thereafter. In addition, at monthly visits during the dosetitration phase and at 3-month visits in the stable-dose period, Factor III of the UPDRS was evaluated. Decreases of >30% in parkinsonian disability (UPDRS Factor III) versus baseline were considered indicative of satisfactory clinical improvement. Finally, at each visit a global judgment on the level of improvement compared with baseline conditions was expressed by means of the Clinical Global Impression (CGI) scale. Statistical analysis was performed by means of chi-square test ^[23] on the proportions of patients experiencing a 30% decrease in parkinsonian disability and on the proportions of patients requiring the addition of L-dopa/carbidopa, and by means of log-rank test ^[24] when the time to L-dopa/carbidopa addition had been analyzed. Baseline disability between subpopulations requiring or not requiring L-dopa/carbidopa addition was compared by means of Student's t test. ^[23] Safety was evaluated through reporting of all adverse events and intercurrent illnesses, irrespective of their possible relationship with the treatments. In addition, hematologic count and blood chemistry parameters, as well as ECG and chest x-ray were performed at baseline and every 6 months. Finally, blood pressure (BP) values were recorded at each visit and the proportion of patients with clinically relevant BP reductions, as well as the proportion of patients with orthostatic hypotension, were evaluated. The former was defined as a decrease in diastolic BP >or=to 15 mm Hg versus baseline associated with an absolute diastolic BP value <50 mmHg, or a decrease in systolic BP >or=to20 mmHg versus baseline associated with an absolute systolic BP value <90 mmHg, either in lying or in standing position; the latter was defined as a decrease of systolic BP in standing position >or=to20 mmHg compared with lying position.

Results.

Dosage.

(Table 2) provides a summary of CBG and L-dopa doses at the end of the dose-titration period and at the 1-year visit. Dose titration lasted, on average, from 4 to 5 months in both groups. In the CBG group, 187 patients reached the end of dose titration: mean optimal daily dose (SD) was 3 (0.9) mg/d; the corresponding dose at the end of dose titration for the 188 patients in the L-dopa group was 492 (122) mg/d. The 1-year visit was performed, on average, after 356 to 361 treatment days. At that time point, mean CBG dose was 2.8 (1) mg/d in the 175 patients still on study, while mean L-dopa dose was 468 (140) mg/d in the 176 patients still under treatment.

View this table:

• View inline
• View popup
• Download powerpoint

Table 2. Duration of treatment and treatment doses at the indicated time points

In those requiring addition of L-dopa/carbidopa, the time at which this was needed is shown in Figure 1 and detailed in Table 3. The proportion of patients requiring addition of L-dopa/carbidopa increased with each time point. The difference between the two groups in the proportion of patients requiring the addition of L-dopa/carbidopa was statistically significant (18% versus 10% at 6 months, 38% versus 18% at 1 year). However, it is worth noting that 62% of patients were still on monotherapy in the CBG group after 1 year. Also worth noting is the fact that mean doses of added L-dopa/carbidopa (see Table 3) were similar in the two experimental treatment arms (305 +/- 135 mg/d in the CBG-treated and 317 +/- 161 mg/d in the L-dopa-treated patients). Thus, there was a relevant sparing of levodopa in the CBG group compared with the L-dopa group, where the total mean dose of L-dopa (including the amount administered as experimental treatment) was 785 mg/d.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1. Survival analysis of time to L-dopa/carbidopa addition (p < 0.01).

View this table:

• View inline
• View popup
• Download powerpoint

Table 3. Frequency of patients requiring L-dopa/carbidopa addition and mean doses (SD) of added (A) and cumulative (C) L-dopa/carbidopa at the indicated intervals

Discussion.

We performed this interim analysis to assess the efficacy of CBG in newly diagnosed PD patients over the short- to middle-term treatment period, i.e., up to the end of 1 year, without interfering with the main, long-term study endpoint represented by the time to the onset of motor fluctuations.

The number of patients entered in this doubleblind study was quite large, and the sample size was therefore adequate to draw some relevant conclusions. L-dopa, as expected, proved to be a very efficacious treatment for early PD, with as many as 86.7% of patients showing an improvement greater than 30% of their baseline PD disability as evaluated by the total score of UPDRS Factor III. Nonetheless, patients treated with CBG, alone or with subsequent addition of L-dopa, showed a similar success rate, with as many as 80.9% of patients showing the same extent of improvement at 1 year. If observation is limited to patients remaining on monotherapy, 79.4% in the CBG group versus 86.5% in the L-dopa group were improved to the extent defined above. In both cases, the proportion of successes did not differ significantly between the two groups. This result is particularly relevant, especially in the light of the high proportion of patients (62%) who remained under cabergoline monotherapy.

These results compare favorably with the literature on the most widely used dopamine agonist, bromocriptine. In one of the largest controlled studies versus L-dopa so far performed, ^[13] where patients were classified by response rate, 29 of 66 patients randomized to bromocriptine (43.9%) remained on monotherapy at 1 year, with a clinically relevant improvement (20% or higher reduction of the Columbia University Rating Scale total score) in 31% of the cases, compared with 59% of the L-dopa-treated patients.

Consistent with some literature data, ^[25] the patients able to remain on dopamine agonist monotherapy in our study were slightly, but significantly, less disabled at baseline than those requiring L-dopa/carbidopa addition within the first year of treatment.

Although L-dopa/carbidopa addition was required earlier in the CBG group than in the L-dopa group, this should not be regarded as a limitation. The use of dopamine agonists in newly diagnosed patients is aimed at postponing L-dopa treatment as much as possible, ^[14] and at keeping L-dopa dosage as low as possible in patients not completely controlled by the dopamine agonist. ^[26] Looking at results from this point of view, in 62% of the CBG-treated patients, L-dopa use was completely avoided for 1 year, with an improvement in motor disability indistinguishable from that offered by L-dopa treatment. In the remaining 38% of patients who required L-dopa/carbidopa addition, those on CBG treatment required similar doses to those on L-dopa.

This study employed only a low dose of CBG (up to a maximum of 4 mg/d). CBG doses up to 6 mg/d have been safely used in PD patients with motor fluctuations, with doses as high as 18 mg/d being given to particularly severe cases. ^[27] Higher doses of CBG might give greater benefit in de novo patients.

The mean improvement from baseline of the UPDRS Factor III total score was higher in the L-dopa-treated group than in the CBG group. The between-treatment difference was statistically significant, but of doubtful clinical relevance at all time intervals: the maximum between-treatment difference in the improvement versus baseline, seen in the subgroup on monotherapy at 1 year was 4 points. Similar considerations apply to UPDRS Factor II total score at the 1-year endpoint, while all other UPDRS Factors did not differ significantly between the two groups. The largest randomized study in newly diagnosed PD patients ^[28] comparing bromocriptine (n = 263) with L-dopa (n = 249) or L-dopa plus selegiline (n = 271) reported a significantly lesser improvement in motor disability (Webster Rating Scale) following bromocriptine administration versus L-dopa. Although, in that study, ^[28] the magnitude of the difference was relatively small, the dropout rate in the bromocriptine group in the first year was as high as 40%, while the dropout rate in the same period on L-dopa was only 5%. Adverse events accounted for 38% of the total dropouts in the bromocriptine group compared with only 7.5% in the L-dopa group. ^[28] In the present study of CBG compared with L-dopa, the dropout rate, both total (18% versus 14%) or due to adverse events (10.6% versus 6.3%), was essentially the same in the two groups. As expected, ^[29,30] in the present study, the rate and frequency of adverse events were higher in the CBG group (75.8%) than in the L-dopa group (65.7%), though severe cases occurred with a similar frequency in either group (15.0% versus 13.7%). This difference relates mainly to a few clusters of events (gastric upset, hypotension and related symptoms, and peripheral edema) that were more common in those on CBG.

The results of this interim analysis indicate that CBG, administered once daily at a maximum dose of 4 mg/d either in monotherapy or in association with L-dopa/carbidopa, has therapeutic efficacy in the first year of treatment approaching that of L-dopa administered q.i.d. at maximum doses of 600 mg/d in newly diagnosed PD patients. The majority of the patients was maintained on monotherapy up to 1 year with improvement in motor and global disability closely similar to that obtained with L-dopa therapy. Whether initial treatment with CBG in de novo cases of PD will delay and/or reduce the frequency of emergence of motor fluctuations and dyskinesias, compared with initial therapy with L-dopa, remains undetermined.

Other members of the PKDS009 Collaborative Study Group.

B. Bergamasco, Dept. of Neurology, Ospedale Le Molinette, Torino, Italy; A. Mamoli, Dept. of Neurology, Ospedali Riuniti, Bergamo, Italy; M. Scarpa, M. Scarano, Dept. of Neurology, Policlinico, Modena, Italy; N. Canal, M. Franceschi, Dept. of Neurology, Ospedale San Raffaele, Milano, Italy; A. Albanese, A.R. Bentivoglio, Neurology Clinic, Universita Cattolica, Roma, Italy; P. Barone, G. Cicarelli, Dept. of Neurology, II Universita di Medicina, Napoli, Italy; R. Malesani, Dept. of Neurology, University of Padova, Padova, Italy; F. Micheli, Dept. of Neurology, Hospital General Jose de San Martin, Buenos Aires, Argentina; S. Muchnik, L. Pliego, Dept. of Neurology, Instituto de Investigaciones Medicas, Buenos Aires, Argentina; M. Somoza, H.E. Gori, Dept. of Neurology, Hospital Alvarez, Buenos Aires, Argentina; R. Dominguez, A. Famulari, Dept. of Neurology, Sirio Libanes Hospital, Buenos Aires, Argentina; R. Aljanati, Dept. of Neurology, University of Montevideo, Montevideo, Uruguay; B. Mikkelsen, Neuromedicinsk Afdelning, Hjorring Sygehus, Hjorring, Denmark; J.M. Olive, Faculty of Medicine, Hospital Municipal de San Juan, Reus, Spain; J. Burguera, Dept. of Neurology, La Fe Hospital, Valencia, Spain; F. Miquel, Dept. of Neurology, Hospital Valle d'Ebron, Barcelona, Spain; S. Gimenez-Roldan, D. Mateo, Dept. of Neurology, Hospital "Gregorio Maranon," Madrid, Spain; H. Allain, M. Verin, Dept. of Clinical Pharmacology, Faculty of Medicine, University of Rennes, France; J.D. Degos, Dept. of Neurology, Henri Mondor Hospital, Creteil, France; A. Destee, Neurologic Clinic, B-Chu Hospital, Lille, France; J.M. Warter, Dept. of Neurology, Neurologic Clinic, Strasbourg, France; M. Baldy Moulinier, Dept. of Neurology, Neurologic Clinic, Montpellier, France; F. Bernardi, C. Chillotti, Dept. of Neurosciences "B.B. Brodie," University of Cagliari, Cagliari, Italy; J. Peres Serra, M. Calopa Garriga, Dept. of Neurology, Hospital Princeps de Espana, Barcelona, Spain; R. Benecke, Dept. of Neurology, University of Dusseldorf, Dusseldorf, Germany; R.M.N. Pirotta and M. Grossoni, Pharmacia SpA, Biometrics and Data Management Dept., Milano, Italy.