Eisai Canada is the Canadian arm of the Japanese pharmaceutical company that, with its partner Biogen, developed Leqembi (lecanemab) — the first anti-amyloid therapy proven to slow the progression of early Alzheimer's disease rather than merely manage its symptoms. For a condition long treated with fatalism, that is a genuine turning point. But Leqembi is not a pill. It is an intravenous infusion that requires serial MRI scans — at baseline and before the 5th, 7th, and 14th infusions, then at one year — to monitor for a potentially serious side effect called ARIA (amyloid-related imaging abnormalities). Without timely MRI access, eligible patients cannot start treatment, and those already on it cannot safely continue.
The problem Eisai faced was structural, not scientific. Canada's MRI system was already running years behind demand, and no single actor controlled the fix. Radiologists, MRI technologists, neurologists, health-system administrators, patient advocates, and provincial policy actors each held one piece of the access problem — and had never been asked to solve it together.
The workshop convened around a single, deliberately constrained question:
The “we all” was the point. It signalled that no one organization — least of all a mid-sized pharmaceutical company — could solve this alone, and that the answer would have to be built by the people who collectively control the scan.
Canada's MRI system was over capacity before Leqembi arrived. Pre-pandemic wait times for MRI averaged roughly 89 days — nearly three times the recommended target — and COVID-19 pushed median waits past 130 days in some provinces. Under the Canadian Association of Radiologists' triage system, an early-Alzheimer's patient needing monitoring scans would be classified P4, “non-urgent” — the very back of a queue where lower-priority patients routinely wait five to six months.
Leqembi would add multiple MRIs per patient per year to that queue, with no national protocol to manage them and no funded place in the schedule. The clinical window is narrow: patients with mild cognitive impairment have a limited period in which the therapy helps. Delay the MRI, miss the window. A hypothetical patient — “Mary,” 68, newly eligible — would need four MRIs in her first seven months; at current wait times she could not realistically get any of them on schedule, and her neurologist could not safely keep her on therapy.
Eisai could do none of the things that would fix this. It could not build MRI capacity, hire radiologists or technologists, write clinical guidelines, or change provincial funding policy. Like Biogen before it, the company needed a coalition of system actors to solve a problem none of them — including Eisai — could solve alone. The system was opaque enough that even the team closest to it could not see it whole.
MRI access was not one obstacle but several interlocking ones, each owned by a different part of the system. None could be solved by Eisai, or by any single actor, alone.
| Dimension | Why it could not be solved alone |
|---|---|
| MRI workforce capacity | There were too few MRI technologists to run the machines that already exist, with high burnout, few training seats, costly certification, and barriers for internationally trained technologists. Expanding the workforce requires colleges, regulators, immigration authorities, CAMRT, and hospital employers to move together — no one of them controls the pipeline. |
| Prioritization and clinical guidelines | Early-Alzheimer's monitoring scans fall to P4 “non-urgent” under existing triage, and there is no Canadian radiology guideline for disease-modifying-therapy monitoring. Rewriting prioritization and protocols requires radiologists, the Canadian Association of Radiologists, neurologists, and provincial systems to agree on a standard none can mandate. |
| Funding and accountability | MRI is funded in fragments — radiologist, hospital, and technologist are each paid, but no one is accountable for overall access or outcomes, and diagnostic imaging is not an insured service nationwide. Changing the funding model and building accountability sits with provincial payers and governments, not the manufacturer. |
| Technology and process throughput | Scanners sit idle off-hours, scheduling is inefficient, no-shows waste slots, and shorter follow-up protocols and AI tools are unevenly adopted. Raising throughput without new machines requires hospitals, technologists, and radiologists to redesign workflows together. |
| Patient navigation and stigma | Patients struggle to navigate a fragmented system, self-advocate, and even understand their results, while stigma and fatalism suppress early help-seeking. Fixing the lived experience requires the Alzheimer Society, clinicians, and patients themselves to co-design the journey. |
| Policy, advocacy, and the “burning platform” | Sustained change depends on persuading provincial governments that timely imaging for dementia is worth funding against competing demands such as cancer. Making that case credibly requires clinicians, advocates, and former system leaders to advocate together — a manufacturer advocating alone has limited standing. |
The Mind Meeting convened 28 participants over three days in October 2023 — seven from Eisai Canada (about one-third of its head-office staff, including the General Manager) and 21 external stakeholders spanning radiology, neurology, MRI technology, health policy, patient advocacy, and a person living with dementia. Each held a piece of the problem the others could not solve without them.
| Stakeholder | Role in the problem | Why their absence would have stalled the solution |
|---|---|---|
| Radiologists and neuroradiologists | Read and interpret MRIs, set imaging protocols, and shape how scans are prioritized. | Any guideline or prioritization change designed without the people who read the scans would miss the real interpretive bottlenecks and lack the credibility needed for radiologists elsewhere to adopt it. |
| MRI technologists (MRTs) | Operate the scanners; their numbers and retention set the system's true throughput. | Capacity recommendations built without the workforce that actually runs the machines would optimize the wrong constraint — equipment, not the staff shortage that idles it. |
| Neurologists and geriatricians | Diagnose Alzheimer's, prescribe the therapy, and order the monitoring MRIs. | Without the prescribers who live the treatment window, the plan would not connect MRI access to the clinical reality of starting and safely continuing therapy. |
| Health-system and policy leaders (former ADMs, health-coalition leads) | Understand provincial funding levers and how to build a case governments will act on. | Advocacy recommendations made without insiders who know how funding decisions are really made would be naive, and the “burning platform” case would not land with payers. |
| Patient advocates (Alzheimer Society, CanAge, Dementia Advocacy Canada) | Hold public-awareness channels, navigation support, and the patient voice in policy. | Awareness and navigation recommendations would be unworkable without the organizations that actually deliver them — and the patient-centred framing would be hollow. |
| A person living with dementia | Carries the lived reality of diagnosis, stigma, missed reminders, and navigating the system. | Every recommendation about patient experience risked optimizing for the system's convenience; the patient's presence kept the work honest. |
| Eisai Canada cross-functional team (7 roles) | Holds the therapy, the evidence, and the resources to convene and help execute. | The convener and eventual coordinating partner had to hear the constraints first-hand to build a plan it could actually support — not one handed to it second-hand. |
The conventional response to a system-readiness problem of this kind is to commission research, engage key opinion leaders one at a time, and brief stakeholders through individual meetings or small advisory panels. Eisai had pursued these channels and understood the problem technically. The MRI wait-time challenge had, in fact, been studied for years — by the Canadian Association of Radiologists, by CAMRT, and in a 2023 national review of wait-time strategies. The diagnosis-by-experts existed.
But one-on-one KOL engagement and advisory boards hit the same ceiling. Like advisory boards, they put the company at the centre and ask external stakeholders to react to its framing, separately, one conversation at a time. The radiologist's view of the access problem differs from the neurologist's, which differs from the administrator's — and none of them sees the whole system. That format produces a map of individual opinions, not a shared diagnosis or a coalition. Tellingly, the same national review concluded that reducing wait times demands a coordinated, multi-disciplinary approach — precisely what separate conversations cannot create.
Eisai needed Canada's MRI system to absorb a significant new volume of monitoring scans for early-Alzheimer's patients — but it could not build capacity, hire radiologists or technologists, write clinical guidelines, or change provincial funding policy on its own. The radiologists, technologists, neurologists, administrators, and policy experts who together controlled the bottlenecks had never been asked to solve the access problem as a single, integrated system. The workshop narrowed its lens to Ontario, Alberta, and British Columbia, and to patients with mild cognitive impairment and early Alzheimer's, so that the recommendations would be concrete enough to pilot.
The workshop ran each of its six topics through the same three-stage discipline. In Analyze, small cross-functional teams surfaced the real constraints — 180 distinct issues across the six topics. In Diverge, they generated options across organizational boundaries — 125 candidate solutions in all — before committing to any. In Converge, each team built the recommendations it would own, and the full room voted to prioritize them — consolidating 125 options into 19 recommendations. What follows is how that played out, topic by topic. The mechanics were identical; the content differed.
Analyze. This team's job was to create urgency that would move the public and policymakers. The binding constraint was profile: dementia lacks the organized advocacy machinery cancer enjoys, MRI capacity is invisible to the public, and governments react to crises rather than preventing them. Idle off-hours scanner time, no-shows, and a coming “tsunami” of dementia cases went largely unrecognized.
Diverge. Options ranged widely: unleashing existing private-clinic capacity and extended hours; text-message reminders co-designed with patients (a person with dementia confirmed that without a reminder, the appointment is missed); a hope-and-innovation change campaign with a centrally managed “backbone” not-for-profit holding province-specific capacity; borrowing from organized cancer bodies such as Cancer Care Ontario and BC Cancer; and a data-driven business case quantifying the cost of treating patients late.
Converge. The team committed to three recommendations: make the case for greater system efficiency and use of private MRI clinics; create a change campaign linking disease-modifying therapies with hope and innovation; and make a concise, data-driven business case for the appropriate use of those therapies. The private-clinic-efficiency recommendation ranked second of all 19 in the room's vote.
Analyze. The team confronted the system's true bottleneck: too few MRI technologists to run the machines that already exist, with high burnout (worsened by Ontario's 24/7 operation), too few training seats — BC had a single school with twelve — costly certification, and immigration barriers for internationally trained MRTs whose pay is constrained by collective agreements.
Diverge. Options included accelerating certification for internationally trained technologists and addressing immigration and NOC-code barriers; using funds such as the Surgical Innovation Fund to cut licensing costs; expanding seats and accelerated programs (e.g., Mohawk College); and modelling the staffing ratios, workflows, and roles of an ideal MRI unit to reduce burnout and turnover.
Converge. The team committed to three recommendations: reduce barriers to the MRT career path in Canada (in partnership with CAMRT); improve retention of MRTs by addressing burnout and incentives; and create a model of the ideal, best-practice MRI unit. Reducing MRT career barriers tied for third in the overall prioritization vote.
Analyze. The team mapped a guidelines vacuum: there was no Canadian radiology standard for monitoring disease-modifying therapies, early-Alzheimer's scans defaulted to P4 “non-urgent,” and protocols for booking, acquisition, and interpretation varied by jurisdiction. Patients had no navigation support and struggled to self-advocate.
Diverge. Options included Canadian-context guidelines specifying sequences, timing windows, and ARIA-aware reading protocols; incorporating AI reading practices; partnering with the Canadian Association of Radiologists to disseminate and endorse standards; and co-designing patient guides with the Alzheimer Society, built with diversity in mind so no patient is disadvantaged.
Converge. The team committed to three recommendations: identify appropriate radiology standards and guidelines for disease-modifying therapies; collaborate with the Canadian Association of Radiologists to disseminate them, with education on ARIA; and help patients navigate the MRI system and advocate for themselves. The guidelines recommendation tied for third in the vote.
Analyze. The team named the structural flaw: MRI is funded in fragments — radiologist, hospital, and technologist each paid, but no one accountable for overall access or outcomes — diagnostic imaging is not an insured service nationwide, and dementia has no dedicated funding envelope, producing rationing and competition with cancer for resources.
Diverge. Options spanned making MRI a funded service under the Canada Health Act; establishing dedicated dementia funding streams; building demonstration sites that deliver bundled, end-to-end dementia care to prove the model and where MRI funding fits; and standing up an accountability working group with national, dementia-specific performance measures, since “you can't manage what you don't measure.”
Converge. The team committed to: create a complete dementia program of care via scalable demonstration sites; advocate to make MRIs a funded service nation-wide; and establish an accountability working group on MRIs with dementia-specific KPIs. The complete dementia program of care was the single highest-ranked recommendation across all six topics.
Analyze. The team targeted throughput without new machines: scanners sitting idle from Friday night to Monday morning, inefficient scheduling, no-shows, and uneven adoption of shorter follow-up protocols and AI tools. Crucially, the team noted that efficiencies still require technologist time — so process gains cannot substitute for the workforce fix.
Diverge. Options included shorter protocols for follow-up scans, block booking, patient assistants, and customized requisitions; centralized scheduling with pre-filled forms, reminders, and online access; AI for booking, communication, and faster scan analysis; and low-field or mobile units to free higher-field machines and reach rural, remote, and Indigenous communities.
Converge. The team committed to: leverage technology and process improvements to raise scan-appointment efficiency; implement AI and technology solutions for booking and communication; and increase capacity by using low-field machines where feasible. The efficiency recommendation ranked among the highest priorities in the room's vote.
Analyze. The team's task was to convert insight into a coordinated advocacy plan. The constraint was fragmentation: thirteen health jurisdictions, scattered one-off government investments, no rigorous tracking, and a manufacturer with limited standing to advocate alone — all while the system remained oriented to supportive care rather than treatment.
Diverge. Options included defining a single, research-grounded advocacy “ask” with a backbone organization to drive it; short-term profile-raising projects such as a policy book, a gap analysis, and a summit; building a coalition with people who have lived experience so their priorities lead; and reframing the policy ask around the shift from supportive care to treatment.
Converge. The team committed to: define the policy and advocacy “ask” and a plan to support it; raise the issue's profile through short-term advocacy projects; build a coalition with people with lived experience; and optimize the shift from supportive care to treatment. The lived-experience coalition and the supportive-care-to-treatment shift both drew strong cross-stakeholder support.
Across the six topics, the room produced a single coherent strategy rather than six disconnected wish-lists. The 125 candidate solutions consolidated into 19 prioritized recommendations, which clustered into three themes: shift policy, funding, and mindset; fix guidelines and the care model; and rebuild workforce, technology, and processes. When the recommendations were ranked by vote, the system-level moves rose to the top — a complete dementia program of care, greater efficiency and private-clinic use, expanding the MRT workforce, and Canadian guidelines for therapy monitoring.
Eisai Canada left the workshop with three things a pharmaceutical company cannot buy or commission:
One-on-one meetings and KOL briefings give you a map of individual perspectives. They don't give you a shared diagnosis or a cross-sector action plan, because each conversation happens in isolation. The radiologist's view of the access problem differs from the neurologist's, which differs from the health administrator's — and none of them has visibility into the full system. What Eisai needed was a process that put all those views in the same room simultaneously, so the recommendations would fit together as a coherent system rather than a set of unconnected asks.
Recommendations that come from outside — a consultant's report or a pharma company's internal team — land as suggestions. Recommendations built by the neurologists, radiologists, technologists, and policy experts who manage the MRI system daily land as commitments those people made publicly, in front of their peers. That distinction is the difference between a well-received report and a strategy that actually moves.
Through structure. MMG designed the engagement across six topics — Burning Platform, Workforce, Prioritization & Guidelines, Funding & Accountability, Technology, Data & Processes, and Policy & Advocacy — so every dimension of the access failure was examined simultaneously. The design ensures the recommendations address the full pipeline rather than the slice any one discipline controls. The three days are intensive, but the output is a coherent strategy, not a list of individual suggestions.
The underlying pattern — a complex system problem no single actor controls, where the people who hold the pieces have never been asked to solve it together — appears across sectors. MMG has run structurally identical engagements for federal government agencies, not-for-profits, and commercial organizations facing the same dynamic. If your problem requires the willing participation of stakeholders you can't mandate, this is the model worth examining.