The Missing Synapse

A Mandate Without a Mechanism

The National Plan to End Parkinson’s Act, passed with bipartisan support in 2024, created the National Parkinson’s Project and placed MJFF at the center of its implementation. It also imposed the first statutory congressional reporting deadline in MJFF’s history — and that deadline, January 2, 2026, passed without a seated Advisory Council. Despite nominations closing in March 2025 and sustained advocacy from the foundation and its partners, HHS did not seat the Advisory Council until April 10, 2026 — over a year late. Legislative approval produced no coordinated delivery mechanism, and a compressed execution window has now replaced a missed one.

That delay is a symptom of the underlying structural condition. MJFF is the connective tissue across a system it cannot command — partnering with NIH and NINDS on research priorities, lobbying the EPA on environmental risk factors like paraquat, coordinating disease advocacy organizations, and now accountable to a federal reporting timeline it did not control. The mandate covers not just Parkinson’s disease but atypical parkinsonisms including PSP, CBD, and MSA. No single organization, including MJFF, holds authority across that system. And the system still has no standing mechanism for producing recommendations across it that the actors who must act on them have each co-owned.

At the same moment, MJFF’s core scientific mandate has undergone a hard pivot. The May 2026 rebranding of the Parkinson’s Progression Markers Initiative to the Parkinson’s Precision Medicine Initiative is not a marketing update. It reflects a fundamental realization that Parkinson’s disease is heterogeneous — multiple subtypes, driven by distinct underlying biologies, requiring biologically defined patient cohorts rather than generic recruitment. The alpha-synuclein seed amplification assay, validated on PPMI samples, detects the disease in 93.3% of sporadic Parkinson’s patients and 95.9% of GBA mutation carriers — but in only 67.5% of LRRK2 mutation carriers overall, and in just 34.7% of LRRK2 carriers who do not exhibit olfactory deficits. That divergence means the assay that unlocked precision medicine simultaneously proved that no single biomarker can qualify patients across the full disease spectrum. Clinical trial design has become exponentially more complex at the exact moment MJFF is scaling its research network to 32 new international teams.

The Infrastructure That Doesn’t Yet Exist

The April 2026 expansion of the Collaborative Research Network, announced jointly with Aligning Science Across Parkinson’s, commits $261 million to fund 32 new teams across 187 institutions in 24 countries. The stated ambition is a standardized, open-access global toolkit of research resources. The operational reality is that nearly 400 highly competitive investigators, governed by different university legal frameworks, operating under different data privacy regimes, and incentivized by publication priority rather than interoperability, must now agree on shared rules for how data, tools, intellectual property, and biological materials move across institutions. Those rules are a condition of receiving the funding — but enforcing consistent compliance across competing university legal teams, GDPR privacy regimes, and researchers incentivized by publication priority rather than interoperability is a different problem entirely. Without a standing mechanism for resolving these tensions as they arise, the network risks producing parallel silos rather than cumulative science.

The same fragmentation risk runs through MJFF’s precision medicine data infrastructure. The LITE program — governed by an MJFF steering committee, implemented by the University of Dundee, feeding into the Global Parkinson’s Genetics Program and the Accelerating Medicine Partnership — requires seamless handoffs between clinical sites, corporate partners like Metabolon, academic implementers, and open-access federal platforms. Each handoff is a point where incompatible data standards, privacy constraints, or IP ambiguity can silently degrade the pipeline. If those handoffs aren’t forced into alignment by a deliberate process, the multi-million dollar data architecture degrades into isolated lakes that cannot communicate.

Finally, the late-May 2026 failure of the Phase 2b LUMA trial for the LRRK2 inhibitor BIIB122 — co-developed by Biogen and Denali Therapeutics — surfaces the clinical pipeline risk that runs beneath all of this. LUMA enrolled 648 patients and ran for up to 144 weeks. The drug achieved over 90% peripheral LRRK2 kinase inhibition and a 30% reduction in a downstream biomarker of LRRK2 activity. It failed to slow disease progression on a single endpoint. The program has now been discontinued for idiopathic Parkinson’s disease; only the Phase 2a BEACON study, restricted to confirmed LRRK2 mutation carriers, continues. As Chief Mission Officer Todd Sherer noted publicly, each failure advances the field’s understanding of Parkinson’s biology — and that is precisely the point. When commercial partners exhaust their tolerance for biological complexity, the burden of de-risking the targets returns to MJFF. No standing mechanism currently exists to reconvene the right stakeholders, reset shared assumptions, and produce a co-owned path forward that each actor is prepared to execute on their side of the ecosystem.

Why Conventional Approaches Won’t Close the Gap

Why Process Is the Variable That Matters

Each of these challenges shares not just a structural root cause, but a structural solution requirement. The science does not need more analysis — the research teams, trials, and data already exist. What the ecosystem is missing is not more meetings run sequentially, but the connective function that converts distributed expertise into shared commitment. That function is closed by a process that produces recommendations the actors who must execute them have genuinely co-owned, stress-tested against each other’s real constraints, and are each prepared to carry back into their own organizations.

A landmark McKinsey study of 1,048 major organizational decisions established that the quality of the decision-making process predicts strategic outcomes six times more powerfully than the depth or quantity of the analysis feeding it — and that top-quartile process organizations earned a 6.9 percentage-point performance premium over bottom-quartile ones.¹ For an organization navigating the NPP’s compressed execution window, the CRN’s open-science governance challenge, and the post-LUMA pipeline reset simultaneously, that finding has a specific implication: the analytical case for what needs to happen across these files is not the constraint. Every actor in the ecosystem has access to the same PPMI datasets, the same CRN publications, the same regulatory precedents. The constraint is whether the right actors can be brought into a process structured to produce recommendations they each have skin in implementing.

The research on multi-stakeholder deliberation makes the structural risk precise. When expert participants from different organizations are presented with the same evidence, their judgments about what should happen next vary by a median of 55% — because those judgments are shaped as much by institutional position, which voice spoke first, and how the question was framed as by the data itself.² Applied to MJFF’s operating environment, this is not an abstract finding. Scientific advisory boards, federal steering committees, and sequential working groups across HHS, NIH, NINDS, and advocacy organizations each produce recommendations shaped by those dynamics. The recommendations that emerge may reflect which agency representative spoke with the most authority in a given meeting more than the underlying strategic logic. Structured process is the mechanism that narrows that variance — not by suppressing expert judgment, but by creating conditions where competing constraints are visible to everyone in the room before any recommendation is formed.

In multi-stakeholder ecosystems spanning federal agencies, academic consortiums, and competing advocacy groups, the cognitive dynamics that work against coherent recommendations intensify precisely when the stakes are highest. Anchoring on familiar coordination mechanisms crowds out structural options. Groupthink within institutional silos produces consensus around what each organization can already do, rather than what the system needs all of them to do simultaneously. Overconfidence in bilateral agreements — the assumption that NIH and MJFF aligning is the same as the full inter-agency system aligning — generates plans that hold in bilateral conversations and collapse when the broader ecosystem encounters them. The same dynamic played out in complex biomedical consortiums before structured convening was introduced: both the Alzheimer’s Disease Neuroimaging Initiative and the Accelerating Medicines Partnership required deliberately designed multi-stakeholder processes — not additional analysis — to produce data-sharing agreements and enrollment commitments across competing academic sites.⁴ A process designed to surface and manage these dynamics is not a procedural addition to the work. It is what makes the work produce recommendations that hold.

Six Challenges. One Structural Root Cause.

Each of the following challenges was identified through MMG’s independent research as a file where the primary execution barrier is structural — not scientific, not financial, and not a function of MJFF’s internal capability. In each case, the actors who must move hold conflicting incentives, and no standing forum exists to produce recommendations they have each co-owned and are prepared to implement.

From Fragmented Mandate to Co-Owned Path: What Each Stage Produces

The gap between what the NPP mandate requires and what MJFF can currently execute is not closed by more inter-agency meetings run sequentially. It is closed by a different kind of process: one designed from the outset to produce recommendations that the actors who must execute them have genuinely co-created, tested against competing institutional constraints, and are prepared to champion back through their own reporting chains. The three stages of that process — Analyze, Diverge, Converge — each do something distinct, and their discipline lies in keeping those stages separate. What follows applies the structure to two of the challenges above. The mechanics are identical; the content differs.

Worked Example A — National Parkinson’s Project and the inter-agency delivery chain

Analyze: Putting the real constraints in front of the real owners

The NPP Advisory Council was seated on April 10, 2026 — more than a year after the January 2026 congressional deadline. The compressed execution window that followed is the starting condition, not a manageable complication. What the Analyze stage does with that fact is not recount it, but use it to surface what each agency in the delivery chain actually controls — and what they cannot commit to without internal authorization they have not yet sought.

An HHS implementation lead knows which NPP deliverables require Secretary-level sign-off and which can move at the program level. An NIH representative knows which research priorities are locked by existing grant structures and which are genuinely negotiable in the next funding cycle. An EPA liaison knows what “engagement on environmental risk factors” can realistically produce within their mandate and what falls outside it entirely. A CurePSP advocate knows which atypical parkinsonism priorities have political traction and which will be deprioritized the moment the room stops watching. That information exists at every inter-agency meeting. In unstructured settings it is rarely stated explicitly, because the format does not create conditions for it and no one is responsible for drawing it out before the group moves to agenda items.

The Analyze stage creates those conditions through structured pre-work and facilitated constraint mapping before any recommendation is formed. Participants leave with a shared picture of where the delivery chain actually breaks — not the version in the legislative record, but the one each agency lead carries into the room and rarely states at a congressional advisory council session.

Diverge: Generating options a room of federal stakeholders won’t reach on its own

The most common failure mode in expert inter-agency processes is not disagreement — it is premature convergence on the least-contentious option. Senior representatives arrive with positions shaped by what their agency has already approved internally. The most authoritative voice — typically the largest funder or the most senior federal official — anchors the group early. Options that require an agency to change its workflow, accept accountability outside its traditional scope, or subordinate its reporting timeline to a shared one are raised once, met with qualified support, and quietly set aside.

The Diverge stage prevents that dynamic by separating option generation from option evaluation. Participants work in small, cognitively diverse teams on defined sub-problems before any cross-group synthesis occurs. Structurally challenging options — those that require real changes to how agencies fund, report, or share authority — are not filtered out before the people who would need to authorize them have examined them. The output is a genuine options space for inter-agency governance, reporting architecture, and stakeholder coordination that an advisory council plenary is not structured to produce.

Converge: Decisions built to survive the return trip

The Converge stage is where the structural difference between a solving sprint and an advisory council session becomes most visible. The HHS implementation lead is in the room when inter-agency ownership over specific NPP deliverables is worked through. The NIH representative is present when research priority sequencing that determines the congressional reporting framework is evaluated. The advocacy organization representatives are at the table when the allocation of responsibilities across Parkinson’s and atypical parkinsonism priorities is forced onto the agenda. The process does not strip these actors of their organizational authority — they return to champion and implement what emerged. What it ensures is that the recommendations they carry back were shaped by their own constraints, tested against every other actor’s constraints in the same room, and co-created with the people who must act on them simultaneously.

A recommendation built this way — with an HHS lead, an NIH program officer, an MJFF policy director, and a patient council representative all present when the trade-offs were forced — carries a different weight in congressional reporting, in inter-agency coordination, and in internal approval processes than one that arrived in a working group report. The NPP’s compressed execution window makes that difference concrete: the question is not whether alignment will eventually be reached, but whether it is reached before the reporting window closes.

Worked Example B — Board alignment and the cost of the expectations gap

Analyze: Naming the expectations gap and who actually holds each lever

The late-May 2026 failure of the Phase 2b LUMA trial is the clearest concrete trigger for the board alignment challenge. BIIB122 achieved over 90% peripheral LRRK2 kinase inhibition and a measurable reduction in a downstream biomarker of LRRK2 activity. It still failed to slow disease progression on a single endpoint. The program has been discontinued for idiopathic Parkinson’s disease. From a scientific standpoint — as Chief Mission Officer Todd Sherer noted publicly — each failure advances the field’s understanding of Parkinson’s biology. From a finance-oriented board member’s standpoint, the question is different: what does failure mean for pipeline prioritization, partner confidence, and the next funding cycle?

The Analyze stage starts from that gap rather than assuming it away. Finance-oriented board members appointed from CapitalSpring and Khosla Ventures bring a venture and private-equity mindset to a research environment that is iterative and nonlinear by nature. Neither orientation is wrong. The misalignment arises when each side is operating on a different implicit model of what progress looks like, what failure means, and how risk should be communicated — a difference that goes unmapped because the governance formats that connect the board to scientific leadership are not designed to surface it.

Diverge: Building a shared framework the board will actually use

The Diverge stage generates options that neither scientific leadership nor finance-oriented board members would propose on their own, because the most useful options sit across the boundary between their respective frames. A shared program-evaluation rubric that combines scientific-learning value with mission-ROI metrics is one such option — neither a standard grant-reporting format nor a venture portfolio scorecard, but a hybrid built by people who understand both and must work with the result. A staged-disclosure model that gives finance board members earlier visibility into negative signals — before they surface in a formal program review — is another. A structured onboarding process for new board members that translates the scientific operating model into investment-familiar language is a third.

As with the NPP example, the Diverge stage is bounded by what was established in Analyze. An option survives only if someone in the room can own it. That constraint prevents the exercise from producing aspirational frameworks that collapse when board governance cycles resume. The goal is not comprehensiveness; it is a set of actionable options each participant could commit to carrying back.

Converge: A governance model that holds when the next trial fails

The Converge stage forces the commitment that a board meeting, by design, is not built to extract. The scientific leadership commits to a specific reporting format that translates mechanism data into language the finance board members have agreed is legible to them. The finance board members commit to a specific program-risk threshold that MJFF’s grants team has agreed is scientifically defensible. The executive team commits to a communications cadence that gives the board early visibility into program status without requiring a formal review session for every signal. Each of these is a genuine trade-off — scientific staff accept more translation work; finance board members accept a longer definition of “progress” — and those trade-offs hold because the people making them were in the room when the other side’s constraints were surfaced.

The LUMA result makes the timing concrete. The post-trial period is the moment when a board’s implicit model of what MJFF is doing is most likely to diverge from what MJFF’s scientific leadership believes the foundation should do next. Converging those models while the LUMA finding is still fresh — before it has hardened into competing narratives — is what allows the foundation to re-enter the pipeline planning cycle with a board that is genuinely aligned, rather than one whose alignment will be tested again at the next program review.

The Process That Sits Naturally Beside What MJFF Already Does

MJFF convenes constantly, and it does it at scale. In the past year alone the calendar has included the NPP Advisory Council’s inaugural operating sessions, the CRN global consortium launch across 187 institutions, the PPMI rebranding and scientific steering process, federal inter-agency coordination across HHS, NIH, NINDS, and the EPA, and the ongoing cadence of scientific advisory board and board of directors meetings. These processes do real work: they maintain the relationships the foundation depends on, surface the scientific evidence, align stakeholders on a shared understanding of where the field is, and generate the institutional trust that later action draws on. The question is not whether they are valuable — they are — but what each format is built to produce, and what naturally falls to a different one.

A federal advisory council session is built for a particular kind of work. At its best it confirms inter-agency alignment on priorities, surfaces new evidence from the agencies present, and builds the shared understanding that the reporting mandate requires. What the form does not do — anywhere — is force trade-offs between competing delivery options or test those options against the people who must implement them within specific agency constraints. Those actors are rarely in the room in working mode; senior representatives attend in their place, and alignment reached at that altitude tends to stay general. Options are described rather than pressure-tested. No one is asked to name what they will personally deliver, by when, against what measurable outcome. None of this is a failing of the council’s organizers. It is the boundary of the format, and knowing where that boundary sits is what signals when a second kind of session is due.

There is a clear test for when that moment has arrived. Look across the NPP delivery calendar and ask one question: are the same inter-agency coordination challenges returning to successive agendas in much the same terms? Where complex challenges are being resolved and moved off the list, the existing mechanisms are doing their job. Where a few hard problems keep resurfacing — inter-agency ownership, reporting sequencing, atypical parkinsonism allocation — that recurrence is itself the signal: the challenge has moved past the point where more discussion adds much, into territory where progress depends on a different kind of process, one built to bring constraint owners into the room together, generate options the standing council would not reach, and force a committed decision before the congressional window closes.

The same test applies to the board alignment challenge. A standard board meeting is the right forum for a great deal of what MJFF’s governance requires. It is not the right forum for building the shared expectations framework that would allow a finance-oriented board and a scientific leadership team to navigate a trial failure together, in real time, without the interpretation gap that currently runs between them. That framework is built in a different kind of session — one specifically designed to put the constraints of both sides on the table simultaneously, generate options neither would propose on their own, and produce a governance model each party has co-owned. With the LUMA result fresh and the post-trial pipeline planning cycle open, the value of that session is at its highest precisely now.

The choice, then, is not between the processes MJFF already runs and something that replaces them. It is whether to set, alongside the convening and governance work the foundation does so well, the occasional session whose specific purpose is to convert alignment into committed action while the window is still open. The two are complements: a federal advisory council session can establish that inter-agency coordination on the NPP must improve and that each party intends to contribute; a structured, time-boxed deliberation is what turns that shared intention into a sequenced delivery plan with named owners and a congressional reporting date attached to it.

The Window Is Already Narrowing

The National Parkinson’s Project Advisory Council was seated over a year late. The compressed execution window that followed is not getting longer. The CRN’s 32 new teams are scaling into a governance structure whose shared rules have not yet been written. The LUMA failure has returned pipeline de-risking to MJFF at the exact moment its research network is most complex. Each of these challenges has a time dimension: the window to shape the inter-agency operating model, write the open-science compliance framework, and reconvene the right stakeholders around a revised translational path is open now, and it will not remain so.

More analysis will not produce inter-agency alignment on NPP deliverables. More sequential meetings will not build open-science compliance across 400 competitive investigators who have never been in the same structured process. What converts a mandate — federal, scientific, or organizational — into coordinated action is a process that brings the right actors together, surfaces what each of them actually controls, generates options the room would not reach on its own, and produces recommendations each participant has co-owned and is prepared to implement. MJFF is uniquely positioned to be the synapse this ecosystem is missing — the organization with the credibility, reach, and scientific standing to convene the actors who cannot convene themselves. Mind Meeting Group designs and facilitates the process that makes that role executable.