Current issues of ACP Journal Club are published in Annals of Internal Medicine


Levodopa, with or without selegiline, and bromocriptine similarly reduced disability in patients with Parkinson disease

ACP J Club. 1994 Mar-April;120:36. doi:10.7326/ACPJC-1994-120-2-036

Source Citation

Parkinson's Disease Research Group in the United Kingdom. Comparisons of therapeutic effects of levodopa, levodopa and selegiline, and bromocriptine in patients with early, mild Parkinson's disease: three year interim report. BMJ. 1993 Aug 21;307:469-72.



To determine the optimal therapy for idiopathic, early Parkinson disease.


Randomized controlled trial with a mean follow-up of 3 years.


93 hospitals in the United Kingdom.


782 patients (mean age 63 y, 55% men) with idiopathic, early Parkinson disease. Exclusion criteria were previous failure to respond to dopaminergic drugs or incapacitating cognitive impairment.


Patients were allocated to 1 of 3 treatment groups: treatment 1, levodopa and benserazide, 62.5 mg 3 times daily (with a dose increase to 125 mg 3 times daily for ≥ 3 months before further increases were considered; n = 249); treatment 2, selegiline, 5 mg for a week, then 5 mg twice daily for 3 weeks with the addition of levodopa and benserazide at week 3 (same doses as in treatment 1; n = 271); or treatment 3, bromocriptine, 2.5 mg, with increases to a maximum dose of 40 mg 3 times daily administered in 3 divided doses (n = 262).

Main outcome measures

Mortality, and assessment of disability determined by the Hoehn and Yahr, modified Webster, and North Western University disability scales.

Main results

All treatment regimens led to an improvement in baseline disability scores after 12 months of treatment. Treatments 1 and 2 did not differ for changes in disability score (P = 0.35), but the improvement in Webster disability score during year 1 of follow-up was greater for treatments 1 and 2 compared with treatment 3 (P ≤ 0.006). The adjusted difference in Webster rating for treatment 3 compared with treatment 1 was 0.93 (95% CI 0.27 to 1.50), and for treatment 3 compared with treatment 2, 1.25 (CI 0.61 to 1.89). After a mean follow-up of 3 years, the incidence of dyskinesias and disabling on-off oscillations was higher with treatments 1 or 2 than with treatment 3 (Table). More patients withdrew from treatment 3 than from treatments 1 and 2 because of adverse effects or no response (P < 0.01).


Levodopa alone or in combination with selegiline and bromocriptine alone resulted in similar changes in disability scores in patients with idiopathic, early Parkinson disease after 3 years of follow-up.

Sources of funding: Parkinson's Disease Society of the United Kingdom; Roche Products; Britannia Pharmaceuticals; Sandoz Products.

For article reprint: Dr. A.J. Lees, Honorary Secretary, PDRG-UK, Department of Neurology, Middlesex Hospital, London, United Kingdom W1N 8AA. FAX 44-71-380-9816.

Table. Treatment groups 1 or 2 vs treatment group 3 in patients with idiopathic early Parkinson disease at a mean follow-up of 3 years*

Outcomes Comparison group Comparison group rate Treatment 3 rate RRI (95% CI NNT (CI)
Dyskinesia Treatment 1 27% 2% 1315% (501 to 3274) 4 (3 to 5)
Treatment 2 34% 2% 1685% (664 to 4126) 3 (3 to 4)
Oscillations Treatment 1 33% 5% 566% (286 to 1062) 4 (3 to 5)
Treatment 2 35% 5% 609% (313 to 1132) 3 (3 to 4)

*Treatment 1 consisted of levodopa and bensarazide, treatment 2 consisted of selegiline with levodopa and bensarazide, and treatment 3 consisted of bromocriptine. Abbreviations defined in Glossary; RRI, NNT, and CI calculated from data in article.


The study by the Parkinson's Disease Research Group refines the role of dopaminergic drugs in mild, early Parkinson disease and complements similar previous studies (1) by comparing levodopa with the dopamine agonist bromocriptine. It also compares levodopa with levodopa combined with the monoamine oxidase-type B inhibitor selegiline (L-deprenyl) to address the contentious issue of the neuroprotective effect of selegiline against Parkinson disease suggested by the DATATOP study (2).

Although this study was not blinded, the large number of patients, long study duration, drug regimens reflecting widespread clinical practice, and use of several standardized scales of disability are the strong design points.

The results confirm previous experience that none of these drug regimens is optimal or clearly superior to any other in the study. Bromocriptine improved function for all practical purposes to the same degree as levodopa and as the levodopa-selegiline combination but with many fewer late disabling dyskinesias and on-off fluctuations. Of the 262 patients taking bromocriptine, however, 41 failed to respond or deteriorated, and 69 had adverse effects severe enough to require withdrawal from the study. The levodopa-selegiline combination was not better than levodopa alone and did not retard the natural worsening of the disease.

This study reinforces the notion that the drugs should be tailored to each patient with early Parkinson disease. The results encourage clinicians to initiate treatment with bromocriptine more often with the likelihood that, if the patients do tolerate the drug, the long-term course may be better. Because the DATATOP studies have shown that selegiline significantly postponed the need for levodopa therapy, it is prudent to await replication of the negative results of the current study before discarding selegiline.

Nagagopal Venna, MD
Boston City HospitalBoston, Massachusetts, USA


1. Olanow CW. Single-blind double observer-controlled study of levodopa/carbidopa vs bromocriptine in untreated Parkinson patients. Arch Neurol. 1988;45:206.

2. Parkinson Study Group. Effects of tocopherol and deprenyl on the progression of disability in early Parkinson's disease. The Parkinson Study Group. N Engl J Med. 1993;328:176-83.