MJFF Asks: What should the next funding program on PD Subtypes look like?

In response to this RFI, it is worth mentioning that the MJFF has been investing in efforts to better define PD subtypes for the last few years. In 2007, we launched a ‘PD Subtypes’ program to help collect more data and build consensus around the existence of such subtypes. The program ended in 2008 and below is a summary of the program and its outcomes.

Why did MJFF launch a PD Subtypes program?

Variability in the clinical features and course of sporadic Parkinson’s disease suggest the existence of potential disease subtypes that may differ in etiology, pathogenesis, and response to treatments. Indeed, many clinicians report being able to ‘predict’ to some degree how a patient may progress by the types of presenting symptoms they see. But a consensus definition and agreement for such subtypes has generally been lacking. Detailed characterization and validation of PD subtypes could significantly alter our understanding of PD, improve our ability to treat patients with existing therapies, and enhance our ability to develop novel treatments.  Information about PD subtypes could also substantially improve the design of future PD clinical studies by improving patient selection, reducing variability, and focusing on relevant outcome measures.  To address these questions, MJFF funded seven grants totaling a little more than $700,000 in 2007 as part of the ‘PD Subtypes’ program.  Each project used existing PD clinical databases for retrospective data mining to analyze the variability in the progression of individual PD symptoms (e.g., tremor, posture and gait, cognitive dysfunction, etc) and/or to identify clusters of clinical features that predict different PD prognoses.

What were the general outcomes of the funded studies?

Despite the fact that several groups used distinct patient databases (including a Netherlands based cohort, the Kaiser Permanente database, and the DATATOP and CALM-PD studies) several researchers reached surprisingly similar conclusions in the identification of PD subtypes.  Three broad groups of patients were identified:

1.     Patients with a benign (slow) progression that were marked by tremor as a dominant feature at disease onset.

2.     Patients with a benign (slow) progression who experienced L-dopa-induced dyskinesias and other motor complications.

3.     A group of patients with worse prognosis – including posture and gait disturbances and cognitive dysfunction.

Despite the similarities, not all the funded projects were able to confirm the existence of these groups.  In fact, some of the funded researchers proposed an alternative hypothesis that PD subtypes do not definitely exist and that patients fluctuate between these different categories based on various factors including stage of disease and responses to medication.  Thus, a definitive consensus that such subtypes exist could not be reached.

What actions were taken after the original PD Subtypes program?

 While the heterogeneity of PD clinical presentation has been appreciated for some time, the consistent findings across the various databases analyzed provide some credence to the idea that distinct PD subgroups may exist.  Whether these subgroups represent distinct PD subtypes, with distinct underlying etiology and predictable progression and prognosis remains to be determined.  This type of natural history information is critical to improving our ability to treat PD as well as to develop and test novel interventions.  In 2009, the MJFF launched a funding program specifically focused on the Postural Impairment and Gait Disturbance (PIGD) subgroup previously described by Jankovic et al (1990) with the goal to increase our understanding of the pathological substrates and biological features that define or predict postural instability and gait disturbance in people with Parkinson’s disease. In parallel, the MJFF encourages researchers to follow-up on the initial findings reported above to provide more clarity on these PD subtypes. 

The MJFF has identified improving the understanding of PD subtypes as a priority research area for 2010.  Improving our understanding of PD subtypes can directly optimize the use of current PD meds (do some treatments work better in some types of patients?) as well as improve our design of clinical studies for testing novel therapeutics by improving patient selection and selection of outcome measures.  We are seeking input from the PD online research community to help the MJFF identify the critical questions that should be addressed in studying PD subtypes as well as the types of studies that can be designed to address these challenges.  It is our goal to design a funding initiative around this critical question but we need the input and recommendations from the experts in the field.  We ask you to help contribute to our decision-making process by posting your ideas and suggestions.

Among the benefits of enhancing our knowledge on PD subtyping are the better focusing of target populations for clinical trials. At stake is who may and may not benefit from participating in the many research avenues surfacing in the near future.

One of the windows into the PD subtypes is the phenotypic complexity of movement itself, which we simply generalize as bradykinesia, and believe essential for PD diagnosis. Some patients are described as akinetic, that is, they exhibit extreme poverty of movement, and others as hypokinetic, implying that their reduction in amplitude is more prominent than any reduction in speed. The UPDRS-III asks clinicians and researchers to consider "slowness, reduced movement amplitude, hesitations, and arrests in ongoing movement" to generate a single 0-4 severity score for each of the repetitive-alternating movement tasks, assuming that abnormalities in speed, amplitude, and rhythm are roughly equivalent. Not only are we missing critical information on PD subtyping in our every-day rating of patients but our clinical trials are growingly relying on an independent bradykinesia endpoint formed from the non-validated practice of adding "bradykinesia related" items from the UPDRS-III.  

Our teams in Toronto and Cincinnati have been using kinematical and clinical means to sort patterns of speed-and-amplitude impairment of movement in PD. We had earlier reported that graded impairments of speed (but not of amplitude) parallel motor impairment (UPDRS-III). Conversely, graded impairments of amplitude (but not of speed) parallel disability (Schwab & England). More recently (data shared at the last PSG symposium), we are finding that amplitude is worse than speed or rhythm across all patients and tasks. However, only speed significantly improves with dopaminergic medications. Therefore dopaminergic responsiveness appears to be significant in “bradykinetics” (predominantly speed-impaired patients) but not in “hypokinetics”  (predominantly amplitude-impaired patients). To bring the relevance to the PD subtyping discussion, clinical trials of dopaminergic drugs may be more favorable in a population of predominantly "bradykinetics" (one PD subtype) but potentially unfavorable or negative when predominantly "hypokinetics" (another PD subtype) are recruited.

Alberto Espay posted about how PD diagnostic categories can gloss over the phenotypic complexity of movement and therefore hide the possible presence of disease subtypes.  Is this also true about the phenotypic complexity of mood, anxiety, and cognition? Can a closer look at the details of rating scales and such data as responsiveness to specific drug treatments, time course, severity, variability, and age at onset of various psychiatric and cognitive symptoms of PD contribute to the discovery of non-motor components of PD subtypes?

In addition to the discussion going on about the MJFF PD Subtypes funding program, here are 10 other recent PD Online posts about PD Subtypes, by:

• Katie Hood: Subtypes -what would you do?
• Susan Bressman: LRRK2 G2019S mutation produces a clinical spectrum that lends itself to detailed analyses
• Joseph Jankovic: Clinical PD subtypes: tremor-dominant and postural-instability-gait-difficulty (PIGD)
• Thomas G Beach: PD clinical subtypes
• Haydeh Payami: PD subtypes: First thing you do - katie....
• Holli Kawadler: Is there a parkin clinical subtype?
• John Q Trojanowski: Heterogeneous phenotypes from autosomal dominant mutations are not uncommon
• Todd Sherer: Linking PD genetics to clinical subtypes
• Avi Orr-Urtreger:Genotype – phenotype correlations
• Todd Sherer: Interesting results from North Arab-Berber population comparing LRRK2, PINK1, and genetically undefined PD

In practicality most people are in between tremor dominant and PIGD, so maybe the better question is whether there are PD subtypes relevant for PD therapeutic development.

The delineation of PD into subtypes may have some commercial and/or grant funding advantages since the subtypes may be classified as "orphan diseases" which place them in  a different funding category with the FDA.  Classification as an orphan disease provides an opportunity for the investigator to apply for orphan drug status for their therapy which establishes seven years of exclusivity after an NDA has been approved by the FDA, a decided commercial advantage equal to, if not superior, to the length of an exclusive patent.