[Skip to Navigation]
Comment & Response
June 29, 2020

Dopamine Metabolite Biomarkers and Testing for Disease Modification in Parkinson Disease—Reply

Author Affiliations
  • 1Department of Neurology, Georgetown University Medical Center, Washington, DC
  • 2Laboratory for Dementia and Parkinsonism, Georgetown University Medical Center, Washington, DC
  • 3Translational Neurotherapeutics Program, Parkinson's Foundation Center of Excellence, Georgetown University Medical Center, Washington, DC
  • 4Lewy Body Dementia Research Center of Excellence, Department of Neurology, Georgetown University Medical Center, Washington, DC
JAMA Neurol. 2020;77(8):1039-1040. doi:10.1001/jamaneurol.2020.1931

In Reply LeWitt correctly describes homovanillic acid (HVA) as an unreliable biomarker of Parkinson disease (PD) progression,1 but our study2 distinctively evaluated dopamine metabolism in response to nilotinib rather than a biomarker of PD progression, and participants were not taking monoamine oxidase B inhibitors. The hypothesis of this phase 2 study that nilotinib is a disease-modifying drug is based on previous animal data3 and explored a mix of biomarkers that included dopamine metabolism, α synuclein, and tau proteins. Methodologically, physiologically based population pharmacokinetics/pharmacodynamics studies showed nilotinib effects in a single center, thus reducing variability in sample collection, storage, shipping, and processing. First, participants received a random single dose of nilotinib vs placebo at baseline and to control for levodopa, the data were normalized to levodopa, 200 mg, which is the lowest dose that participants received as standard of care. Second, cerebrospinal fluid (CSF) HVA peaks in patients with de novo PD around 1.5 to 2 hours after administration of levodopa, 200 mg, and it remains constant up to 4 hours4; therefore, CSF was collected within 2 hours after the last levodopa dose and lumbar punctures were staggered at 1, 2, 3, and 4 hours after nilotinib administration (to obtain the area under the curve). Indeed, additional data in levodopa-naive patients with Alzheimer disease5 show that nilotinib significantly alters central nervous system (CNS) HVA and 3,4-dihydroxyphenylacetic acid (DOPAC) compared with placebo. Third, mass spectrometry was optimized to measure HVA and DOPAC but not monoamines or other metabolites. Both HVA and DOPAC are primary metabolites (DOPAC is a primary, not the primary metabolite) of dopamine. Furthermore, plasma HVA and DOPAC do not indicate CNS pathogenesis, but they provide excellent controls for changes of CNS dopamine metabolism. The brain contributes a small percentage (15%) of circulating plasma HVA,6 which is mostly because of intestinal dopamine metabolism. Plasma HVA did not change between groups, indicating that the observed changes in CSF HVA are because of CNS metabolism in the nilotinib vs placebo groups. Conversely, DOPAC is almost exclusively derived from the CNS metabolism of several monoamines, including dopamine, and it is secreted into the circulation and discarded via urination. There was a dose-dependent increase in plasma and CSF DOPAC in nilotinib vs placebo, indicating that plasma DOPAC was secreted from the CNS. Taken together, although nilotinib may have symptomatic effects via increased endogenous dopamine, it is totally unreasonable to evaluate the current data from a dopamine-centered perspective, as this phase 2 study investigated a biomarker system that is modified by nilotinib. Disease-modifying therapies should not only affect dopamine transmission, but they must also modify other pathologies, including the misfolding of α-synuclein and tau, which are integral to the brain’s motor and nonmotor systems. In addition to HVA and DOPAC, nilotinib stabilized normal CSF synuclein and tau levels, but reduced α-synuclein oligomers (150 mg only) and hyperphosphorylated tau in a dose-dependent manner.2 Measurement of these molecules together as a kit is the underlying hypothesis to evaluate nilotinib as a disease-modifying drug. Adequately powered phase 3 studies will determine a definitive correlation between this biomarker mix and motor and nonmotor symptoms in PD.

Add or change institution