When a 68-year-old retired teacher in Leeds is told her husband's memory lapses are early-stage Alzheimer's, the cruel modern twist is that a treatment finally exists and the lecanemab NHS will not pay for it. For the first time in the history of the disease, two licensed drugs, lecanemab (sold as Leqembi) and donanemab (sold as Kisunla), can measurably slow the decline of early Alzheimer's by clearing the amyloid plaques that clog the brain. Both carry a UK marketing authorisation from the Medicines and Healthcare products Regulatory Agency (MHRA), meaning regulators have judged them safe and effective enough to prescribe. Yet the National Institute for Health and Care Excellence (NICE) has confirmed, in final guidance upheld through 2025 and into 2026, that neither drug offers value for money to the NHS so this family, like the roughly 982,000 people living with dementia in the UK, must either go without or pay privately. The same molecules approved in Brussels for European patients are, in effect, denied at home, and that contradiction has cracked open a dementia treatment divide that now stretches across the UK and the EU.
To understand why the Alzheimer's drug NICE rejection feels so bitter, you have to understand what these medicines actually do and, just as importantly, what they do not. Lecanemab and donanemab are monoclonal antibodies that bind to beta-amyloid, the sticky protein that accumulates into plaques in the Alzheimer's brain, and prompt the immune system to strip it away. In the pivotal CLARITY-AD trial, lecanemab slowed cognitive and functional decline by roughly 27% over 18 months compared with placebo; donanemab's TRAILBLAZER-ALZ 2 trial showed a comparable 22–35% slowing depending on how advanced the patient's tau pathology was. These are the first genuinely disease-modifying amyloid plaque drugs to reach the clinic, and that is a real scientific milestone. But the benefit is a slowing of decline, not a reversal or a cure: a person still progresses, just more slowly, buying perhaps four to seven extra months of preserved function over a year and a half. Patients and carers need to weigh that honestly against the burden of treatment.
The eligibility window is also narrow, which many families discover only after an anxious wait. These drugs work and are only licensed for people with mild cognitive impairment or mild dementia due to confirmed Alzheimer's, in the early stages where amyloid is present but damage is limited. Confirming that requires biomarker proof: either a specialist PET scan or a lumbar puncture (spinal tap) to measure amyloid in the cerebrospinal fluid, the kind of early dementia treatment Europe and Britain alike are still scrambling to deliver at scale. Crucially, genetics matter for safety. People who carry two copies of the APOE4 gene face a substantially higher risk of the drugs' signature side effect amyloid-related imaging abnormalities, or ARIA, which means brain swelling and microbleeds visible on MRI. Most ARIA is mild and symptomless, but a minority of cases are serious, and a handful of trial deaths were linked to it. That is why treatment demands repeated MRI monitoring, and why the EMA initially balked before carving out a lower-risk group.
This is where the regulatory divide becomes the heart of the story, and where the UK and EU pathways genuinely diverge. In Britain, drug approval is split: the MHRA decides whether a medicine is safe and effective enough to be licensed, while NICE decides whether the NHS should fund it given the price. The MHRA licensed lecanemab in August 2024 and donanemab in October 2024. NICE, however, looked at list prices and the enormous wraparound costs the PET scans, the fortnightly or four-weekly infusions, the serial MRIs, the specialist staff and concluded the benefit was too modest for the money. Its final guidance rejected both for routine NHS use, a Leqembi cost UK verdict and a donanemab UK verdict that essentially priced the drugs out of public provision. The manufacturers Eisai and Eli Lilly disputed the maths, but the appeals did not move the core decision. The result: licensed but unfunded, available in principle, unreachable in practice for nearly everyone relying on the NHS.
Across the Channel, the European Medicines Agency travelled a more tortuous road to a different destination. In July 2024 the EMA's human medicines committee initially refused lecanemab, judging that the modest benefit did not outweigh the risk of serious ARIA. Then, after the company requested a re-examination, the agency reversed itself in November 2024 and recommended approval but only for a restricted, lower-risk population: patients with just one copy or no copies of the APOE4 gene, explicitly excluding the double-carriers most likely to suffer dangerous brain bleeds. The European Commission granted EU-wide marketing authorisation in 2025, and the EMA has since been reviewing donanemab on similar terms. So the EMA Alzheimer's approval exists, but it is narrower than the UK licence and bolted to a genetic safety gate. That nuance Brussels approving a carefully fenced-off version of the same drug Britain licensed broadly but won't fund is the precise shape of the divide.
Yet an EU licence is not the same as EU access, and this is the analytical point that gets lost in the headlines. A central EMA authorisation simply opens the door; each member state still negotiates its own price and reimbursement, builds its own diagnostic capacity, and decides through its national health-technology bodies whether to pay. Germany, with its dense network of memory clinics and university hospitals, is positioned to roll out treatment relatively quickly for privately and statutorily insured patients who qualify. France's HAS and other national assessors are weighing cost-effectiveness questions strikingly similar to NICE's. Smaller and central or eastern European systems may have the legal right to prescribe but lack the PET scanners, neurologists and MRI slots to do so. The upshot is a continental postcode lottery: a patient in Munich may start infusions while an identical patient in a rural region waits indefinitely. Europe's divide is therefore not UK-versus-EU as a clean binary, but a patchwork in which Britain's blanket NHS refusal sits at one extreme and a handful of well-resourced EU regions at the other.
Why did two respected regulators and their funding gatekeepers reach such different answers from the same trial data? Partly it is institutional design: NICE applies an explicit cost-per-quality-adjusted-life-year threshold that forces a stark yes-or-no on value, whereas the EMA's job stops at the benefit-risk question and leaves money to member states. Partly it is risk appetite the EMA's exclusion of APOE4 double-carriers shows a regulator willing to approve only a safer subset, while the MHRA licensed more broadly and pushed the value judgement downstream. And partly it is infrastructure realism. NICE openly doubted the NHS could deliver the diagnosis-and-monitoring pathway these drugs require without diverting resources from everything else dementia care needs. With UK dementia numbers projected to climb from around 982,000 today to 1.4 million by 2040, and Alzheimer's the most common cause, the system-level question can we even run this safely at scale? weighed as heavily as the price tag.
For families unwilling to wait for policy to catch up, the practical path now runs through private medicine, and it begins with diagnosis. The first step is an early, accurate diagnosis confirming that symptoms stem from early-stage Alzheimer's with amyloid present which means getting onto a dementia diagnosis NHS memory clinic pathway for assessment, or paying for private cognitive testing and biomarker confirmation via PET or lumbar puncture. This matters because the drugs are useless, and unlicensed, for moderate or advanced dementia or for non-Alzheimer's causes; many hopeful enquirers are turned away simply because their disease is too far along or never amyloid-driven in the first place. An APOE genetic test is also part of the workup, both to gauge ARIA risk and, in the EU, to determine eligibility at all. Anyone considering this should expect weeks of investigation before a single infusion is even on the table.
Then comes the true cost, which is far larger than the drug's sticker price. Private Alzheimer's treatment UK clinics, a small but growing number of which began offering lecanemab and donanemab through 2025, quote in the region of £60,000 to £80,000 a year once everything is counted: the medicine itself, the infusion appointments every two to four weeks, the baseline and repeat MRI scans to catch ARIA, the consultant oversight and the diagnostic workup. The Kisunla price and Leqembi cost UK figures quoted as headline drug prices often in the low tens of thousands understate the real bill, because monitoring is not optional; it is the safety scaffolding without which the treatment should not be given. Donanemab does offer one potential efficiency: trials suggest patients can stop once their amyloid is cleared, capping the duration, whereas lecanemab is conceived as ongoing. Families should ask any clinic exactly what is and isn't included, how ARIA would be managed, where emergency MRI access sits, and what happens if treatment must be stopped.
The questions worth pressing a private provider on are pointed: Is the clinic licensed and are its prescribers experienced specifically in amyloid therapy and ARIA management? What is the full annual cost including all scans and consultations, not just the drug? How will my APOE status affect my risk and is it being tested? What is the plan if I develop brain swelling or bleeding who monitors, how often, and is there rapid imaging? And honestly, given my stage of disease, how much benefit is realistic for me? A reputable clinic will welcome these and decline patients who fall outside the evidence; a less scrupulous one may oversell hope to people whose disease is too advanced to benefit. The amyloid plaque drugs are powerful but unforgiving of poor patient selection, and the burden of fortnightly infusions and repeated scans is itself significant for someone in their seventies.
Step back, and the deepest issue is not really the price of two antibodies it is that the UK and much of the EU lack the diagnostic infrastructure to use any disease-modifying Alzheimer's drug fairly, even a cheaper one arriving tomorrow. Without enough PET scanners, blood biomarker tests entering routine use, neurologists and MRI capacity, early diagnosis remains a privilege of the well-resourced and the well-located, which is precisely how a postcode and price lottery entrenches inequity. The arrival of lecanemab and donanemab has, if nothing else, exposed that gap: NICE's refusal is in part an admission that the pipework isn't ready. For the family in Leeds and the millions behind them, the rational response is to act on what is controllable now seek an early, specialist assessment so options stay open, get clarity on APOE status and stage, scrutinise any private quote with cold eyes, and press MPs, MEPs and charities such as Alzheimer's Society and Alzheimer Europe to fund the diagnosis capacity that makes any future treatment, approved in Brussels or licensed in London, actually reachable at home.
Comments
Post a Comment