Cholecystokinin is one of the most abundant neuropeptides in the human CNS. It coexists with dopamine in ventral tegmental and substantia nigra neurons in rodents and primates, but the coexistence is less obvious in normal humans.1 It modulates central motor effects of dopamine through nigral or striatal cholecystokinin-A (excitatory) and cholecystokinin-B (inhibitory) receptors. The effect of the neuropeptide differs, however, depending on the animal species, the dose used, cotreatments, and site of injection.

Cholecystokinin is selectively decreased in the substantia nigra of patients with Parkinson's disease, and cholecystokinin-A antagonist binding is reduced in hemiparkinsonian monkeys. Cholecystokinin inhibits levodopa induced dyskinesias in parkinsonian monkeys,1 but proglumide, a cholecystokinin antagonist, did not improve motor signs in dyskinesia free patients with Parkinson's disease.2 Proglumide is, however, a weak non-selective cholecystokinin antagonist. Oral SR 27897B (SR; Sanofi Recherche), a highly selective and potent cholecystokinin-A receptor antagonist, penetrates the CNS and blocks cholecystokinin potentiation of dopaminergic neurotransmission.3 We evaluated the potential antidyskinetic effects of oral SR in parkinsonian patients using a placebo controlled double blind study design and a single challenge of apomorphine, a test used to determine the antidyskinetic properties of associated treatments.4 As cholecystokinin-A antagonism may modify gastrointestinal motility,3 and consequently the kinetics of oral levodopa absorption, parenteral apomorphine was preferred to oral levodopa.

Nineteen patients with Parkinson's disease, who had motor fluctuations and levodopa induced dyskinesias for 6 months, were included in the study, which was approved by the local ethics committee. All patients gave written informed consent. Eighteen patients completed the study. Although patients in the placebo group tended to have a longer Parkinson's disease course, no significant difference was found between the two treatment groups for age, sex, or medical history (table 1).

Patients were randomly allocated to the SR group or a placebo group using a ratio of 2:1, and SR or placebo was given once a day for 14 days. The initial SR dose of 1 mg/day was increased to 2 mg/day after 7 days, if tolerated by the patient. The minimal dose of apomorphine inducing dyskinesias was determined for each patient before starting the study. Apomorphine was administered in the morning in a fasting condition, after a 3 day administration of domperidone (60 mg), and a 12 hour withdrawal of antiparkinsonian drugs. Motor disability (table 1) and dyskinesias were then assessed in each patient in two identical apomorphine tests, before (day 1) and after the 14 day treatment (day 14). During the 14 days, patients kept a diary of abnormal events with special attention to dyskinesias.

The primary end point was the severity of dyskinesias/minute as evaluated by a videotaped standard procedure.4 The predominant type of dyskinesia (dystonic, ballistic, or choreic) and its severity from 0 (no abnormal movements) to 4 (abnormal movements resulting in severe disability) were scored once a minute for 90 minutes in the four limbs, trunk, neck, and face by one scorer (maximum score=28). Dyskinesia time profiles in each patient were also analyzed qualitatively by all investigators blind to treatment. The investigator and the patients globally assessed (from 0 to 5) changes in dyskinesia noted during the study.

Treatment with SR was tolerated well without marked adverse effects. One patient discontinued the study after 5 days of treatment due to severe dyskinesias and repeated falls. These problems were present before the study. Three patients (SR group two; placebo group one) stayed at the 1 mg dose level. Although three patients on SR and none on placebo reported an occasional increase in dyskinesias, daily levodopa induced dyskinesias were considered globally by both patients and investigators not to have been modified. The mean doses of apomorphine used for video testing were 5.0 mg (placebo group) and 4.6 mg (SR group). There was no significant difference in delay before turning on, the duration of the on state, the percentage of motor improvement, or in apomorphine induced dyskinesias between the two groups (table 1). Qualitative analysis failed to detect any differences either in the type of dyskinesias, their topography, or their timing (onset and end of dose dyskinesias, peak dose dyskinesias) before and after SR treatment.

In this study, no significant changes in drug induced dyskinesias and in motor disability were found when patients with Parkinson's disease were treated with the cholecystokinin-A antagonist SR 27897B. The fact that this selective antagonist was ineffective in our study may have been for several reasons. The dose may have been below or even above the response threshold. Indeed, in rats, striatal perfusion with high concentrations of cholecystokinin induces hypolocomotion, whereas perfusion with low concentrations induces dopaminergic-like contralateral rotation. The apomorphine test model may not be sensitive enough. This seems unlikely, however, as it has been used to show the antidyskinetic properties of fluoxetine, clozapine, and propranolol in small groups of patients.4 Furthermore, it permits the study of a wide range of dyskinesias, from dystonia to ballic and choreic dyskinesia, which result from differential activation of dopamine receptors. Thus the absence of an effect of SR on any type of dyskinesia suggests that cholecystokinin may not modulate dopamine release at the level of striatal cholecystokinin-A receptors. However, as cholecystokinin-A antagonist binding has been found reduced in a model of hemiparkinsonism in the monkey, it cannot be totally excluded that the absence of effect of SR results from a reduction in the density of striatum cholecystokinin-A receptors in patients with Parkinson's disease. Moreover, the dopamine agonist apomorphine acts postsynaptically, whereas cholecystokinin might act presynaptically—for example, by modulating dopamine release. An effect of SR on dopamine release would not be detectable in the apomorphine test. Finally, cholecystokinin-A antagonist may not have been effective if the modulatory effects of cholecystokinin on dopamine tone is mediated only through cholecystokinin-B receptors.5 A study of cholecystokinin-B antagonists and agonists should be considered in patients with Parkinson's disease.

In conclusion, this is the first administration in patients of a selective cholecystokinin-A antagonist. Our results show that, at least under our experimental conditions, defective dopamine systems in parkinsonian patients are not modified by the inhibition of cholecystokinin-A receptors.