[Skip to Navigation]
March 1994

Separating Parkinson's Disease From Normality: Discriminant Function Analysis of Fluorodopa F 18 Positron Emission Tomography Data

Author Affiliations

From the MRC Cyclotron Unit (Drs Sawle, Playford, Burn, and Brooks), Clinical Sciences Section, and the MRC Cyclotron Unit Biology Section (Dr Cunningham), Hammersmith Hospital, London, England, and the University Department of Clinical Neurology (Drs Sawle and Brooks), Institute of Neurology, National Hospitals for Neurology and Neurosurgery, London, England.

Arch Neurol. 1994;51(3):237-243. doi:10.1001/archneur.1994.00540150027011

Objective:  To explore the relationship between normal and parkinsonian fluorodopa F 18 (18F-6-l-fluorodopa [18F-dopal]) uptake data to identify clinically normal subjects who may have preclinical Parkinson's disease.

Design:  A statistical comparison of striatal fluorodopa F 18 positron emission tomography scan data from patients with Parkinson's disease and normal controls.

Setting:  Positron emission tomography unit within a postgraduate teaching hospital.

Main Outcome Measures:  Discriminant function analysis used to compare the pattern of striatal (left and right caudate and putamen) fluorodopa F 18 uptake in normal subjects and patients with Parkinson's disease. Results: The discriminant score that best separates patients with Parkinson's disease from normal controls is a function of the lowest putamen influx constant minus a function of the caudate influx constant values. Borderline low normal subjects have slightly low fluorodopa F 18 uptake throughout the striatum, whereas patients with early Parkinson's disease have low fluorodopa F 18 uptake in one putamen with preserved uptake in the caudate (for normal subjects, subtracting the caudate influx constants from a function of the lowest putamen value lowers the discriminant score, although it remains positive; for patients, subtracting a larger caudate value from a function of the putamen uptake value leads to a negative score). One control subject had a borderline low discriminant score, compatible with focal nigral pathological changes as expected in preclinical Parkinson's disease. A repeated scan taken 3 years later showed a marked reduction in fluorodopa F 18 uptake, suggesting progressive nigral dysfunction.

Conclusion:  Normal and parkinsonian fluorodopa F 18 uptake data differ both in the overall level of tracer uptake and in its spatial distribution. Subjects whose overall striatal fluorodopa F 18 uptake falls at the borderline of normal and parkinsonian values are likely to be normal if they have uniformly low uptake, but may have early or preclinical Parkinson's disease if uptake into putamen is very much lower than uptake into caudate.

Add or change institution