Premotor Parkinson's disease: concepts and definitions

Abstract

Parkinson's disease (PD) has a prodromal phase during which nonmotor clinical features as well as physiological abnormalities may be present. These premotor markers could be used to screen for PD before motor abnormalities are present. The technology to identify PD before it reaches symptomatic Braak Stage 3 (substantia nigra compacta [SNc] involvement) already exists. The current challenge is to define the appropriate scope of use of predictive testing for PD. Imaging technologies such as dopamine transporter imaging currently offer the highest degree of accuracy for identifying premotor PD, but they are expensive as screening tools, and abnormalities on these studies would only be evident at Braak Stage 3 or higher. Efficiency is greatly enhanced by combining imaging with a prescreening test such as olfactory testing. This 2-step process has the potential to greatly reduce costs while retaining diagnostic accuracy. Alternatively, or in concert with this approach, evaluating high-risk populations (eg, patients with rapid eye movement behavior disorder or LRRK2 mutations) would enrich the sample for cases with underlying PD. Ultimately, the role of preclinical detection of PD will be determined by the ability of emerging therapies to influence clinical outcomes. As such, implementation of large-scale screening strategies awaits the arrival of clearly safe and effective therapies that address the underlying pathogenesis of PD. Future research will establish more definitive biomarkers capable of revealing the presence of disease in advance of SNc involvement with the promise of the potential for introducing disease-modifying therapy even before the development of evidence of dopamine deficiency.

Figure 1

The PARS pyramid: In this conceptual model there are 4 stages that precede clinically manifest PD: pre-physiological, pre-clinical, pre-motor and pre-diagnostic. There are a relatively large number of individuals that have a pre-disposition to develop PD, but only a fraction progress to each succeeding level. Ultimately, the number of individuals who develop clinically manifest PD is a small fraction of the at-risk pool.

Figure 2

Two-by-two table illustrating the four scenarios that could apply to diagnostic tests (true (A) and false (B) positives and true (D) and false (C) negative results). Sensitivity is calculated as A/A+C; Specificity is D/B+D. Positive predictive value (PPV) is calculated as A/A+ B and NPV is calculated as D/C+D. Because the denominator for PPV and NPV include values from both the affected (disease is present) and unaffected (disease is absent) groups, their value depends on the relative proportion of people from each group in a given population or study cohort. As the proportion of unaffected people increases, PPV also decreases and NPF increases.

Figure 3

Shows the effect of increasing prevalence on the ratio of false-positive to true-positive tests for less common disorders. Even very specific tests have high false-positive rates when a disease s not common. The figure shows that a test with 99% specificity still gives the same number of false positive tests as true positive tests in a disorder with population prevalence of 1/1000 (similar to PD). Assuming 95% sensitivity and 99% specificity, as prevalence increases, the proportion of positive tests that are true positives increases from about 50% to over 90%. (adapted from Tanner, Annals of Epidemiology 1996, 6:438–441).