Objectives: To determine whether cerebrospinal fluid (CSF) homovanilli
c acid (HVA) concentration in subjects with early, mild Parkinson's di
sease (PD) treated with the monoamine oxidase type B inhibitor selegil
ine hydrochloride differs from that of control subjects receiving plac
ebo. Our hypothesis is that if selegiline offers neuroprotection in su
ch patients, the HVA levels should not decrease over time as much as i
n those receiving placebo. A second objective was to define the kineti
cs of recovery of HVA concentration after discontinuation of selegilin
e therapy. Design: During the controlled clinical trial DATATOP (depre
nyl [selegiline] and tocopherol antioxidative therapy of parkinsonism)
(which examined the effects of selegiline and tocopherol in 800 subje
cts with early, untreated PD), the CSF HVA concentration was measured
at baseline and again 4 weeks after the study end point (need for levo
dopa therapy) was reached and medications were withdrawn (n=265). Base
d on an interim analysis, the lumbar puncture protocol tvas modified,
such that subjects who reached the study end point were randomly assig
ned an interval of 0 days or 2, 6, or 8 weeks between discontinuation
of selegiline therapy and the lumbar puncture (n=215). Setting: In the
hospital, after overnight bed rest and fasting. Patients: The 800 sub
jects with early, mild PD who participated in the DATATOP controlled c
linical trial. Intervention: The four treatment arms were (1) selegili
ne-placebo and tocopherol-placebo, (2) selegiline-placebo and active t
ocopherol (2000 IU/d), (3) active selegiline hydrochloride (10 mg/d) a
nd tocopherol-placebo, and (4) active selegiline hydrochloride (10 mg/
d) and active tocopherol (2000 IU/d). Main Outcome Measure: Cerebrospi
nal fluid HVA concentrations. Results: The CSF HVA concentration at ba
seline did not correlate with disease duration or severity; the mean (
+/-SD) HVA concentration was 34.7 +/- 17.0 ng/mL. In the 265 subjects
who underwent analysis 4 weeks after the study end point was reached a
nd medications were withdrawn, the decline in HVA concentration was si
gnificantly greater in subjects assigned to receive selegiline (9.2 +/
- 12.7 ng/mL) than in subjects not receiving selegiline (3.2 +/- 14.4
ng/mL), indicating persistent monoamine oxidase (MAO) inhibition by se
legiline. Tocopherol had no effect. Results from the modified protocol
revealed that HVA concentration increased with time to approximately
the same levels as determined in controls by 60 days but showed no cle
ar final plateau level. At 0 days, HVA concentration was reduced from
baseline by less than one third, indicating only partial inhibition of
MAO activity by selegiline. Conclusions: Measurements of CSF HVA conc
entrations (1) indicate the long duration of MAO inhibition by selegil
ine, (2) have limited utility as a marker of severity or progression i
n PD, (3) indicate that selegiline does not provide sufficient MAO inh
ibition to test adequately the oxidative stress hypothesis of the caus
e of PD, and (4) lend no support for a protective role of selegiline i
n slowing the progression of PD.