Objective: The objective of this study was to investigate the effect o
f selegiline first as monotherapy and then in combination with levodop
a in the early phase of PD. Methods: A total of 157 de novo PD patient
s were randomized to receive either selegiline or placebo in a double-
blind study until levodopa therapy became necessary. Thereafter, the d
rugs were withdrawn for an 8-week washout period to evaluate the possi
ble symptomatic effect of selegiline. Results: Analysis of Kaplan-Meie
r survival curves for each group showed that selegiline delayed signif
icantly the need for levodopa therapy (p = 0.028). The semiannual rate
of disability progression was slowed down significantly in the selegi
line group analyzed with the Unified Parkinson's Disease Rating Scale
(total and motor scores, p < 0.001). Selegiline had a ''wash-in'' effe
ct (i.e., an initial symptomatic amelioration of PD at 6 weeks and 3 m
onths). However, after the 8-week washout period, no significant diffe
rences in the deterioration of disability between the groups was revea
led in any of the scales, suggesting that besides having a slight symp
tomatic effect, selegiline mag also have neuroprotective effects. Simi
larly, the progression of symptoms from baseline to the end of the was
hout period was significantly slower (p = 0.033) in the selegiline gro
up when the progression was adjusted by the time to reach the end poin
t. Selegiline was well tolerated. Conclusions: Selegiline delayed sign
ificantly the need to start levodopa in early PD. After a 2-month wash
out period (before the start of levodopa therapy) no significant sympt
omatic effect of selegiline was seen in comparison with the placebo gr
oup, supporting the concept of neuroprotective properties of the drug.