Keywords
Alzheimer Disease - Immunotherapy - Amyloid - Cost-Benefit Analysis
INTRODUCTION
Alzheimer's disease (AD), the commonest form of dementia, is a devastating, life-limiting
illness which has profound implications for individuals, families, and society. Estimates
suggest that the world-wide prevalence is now 55 million rising to 139 million by
2050 as the population ages, and those costs exceed $1.3 trillion.[1]
At the core of AD is beta-amyloid (β-amyloid) pathology. Numerous lines of evidence
suggest that accumulation of toxic β-amyloid species culminating in the deposition
of extracellular plaques is an early, upstream, and central pathological process.[2]
[3]
While β-amyloid by itself may cause subtle cognitive problems,[4] a significant proportion of elderly individuals harbor ostensibly asymptomatic β-amyloid
pathology for many years.[5] By mechanisms not fully elucidated, but likely to involve inflammation, susceptible
individuals then undergo a cascade of pathological processes including the accumulation
of tau pathology, synaptic loss, and neuronal cell death leading to the emergence
of the progressive cognitive symptoms, increasing dependence, and premature death.[6]
[7]
The centrality of β-amyloid in the pathogenesis of AD has made it a major target for
drug development over the last two decades. Initial attempts to induce β-amyloid clearance
using a range of different immunotherapies were unsuccessful, variably due to combinations
of side-effects, suboptimal therapeutic efficacy, insufficient dosing, and/or inappropriate
patient selection, that is, not requiring evidence for β-amyloid pathology at recruitment.[6] However, the situation has changed radically in recent years with the publication
of 2 pivotal, positive, phase-3 studies of the monoclonal antibodies lecanemab[8] and donanemab[9] in patients with mild cognitive impairment (MCI) or mild dementia due to AD.
PROVEN AND POSSIBLE BENEFITS AND POTENTIAL OF AMYLOID IMMUNOTHERAPY
PROVEN AND POSSIBLE BENEFITS AND POTENTIAL OF AMYLOID IMMUNOTHERAPY
Despite differences in the specific β-amyloid moieties targeted by the two drugs—protofibrils
in the case of lecanemab, and plaques for donanemab—the results of the trials are
broadly similar. Given intravenously every 2 weeks (lecanemab) or monthly (donanemab)
over an 18-month period, both resulted in robust and extensive clearance of β-amyloid;
both led to significant slowing of cognitive and functional decline by 35% (donanemab)
and 27% (lecanemab) compared with placebo; and both had effects on relevant non-cognitive
outcomes including quality of life.[8]
[9] While the duration of the trials was too short to definitively determine disease-modification
effects, these are suggested by changes in other, presumed downstream, biomarkers
of AD such as ptau217 (reflecting abnormal phosphorylated tau due to β-amyloid), glial
fibrillary acidic protein (GFAP; reflecting astrocytic pathology), and microtubule-binding
region of tau containing the residue 243 (MTBR-tau243; reflecting tau tangle pathology).[10]
[11]
REAL AND POTENTIAL SIDE-EFFECTS OF AMYLOID IMMUNOTHERAPY
REAL AND POTENTIAL SIDE-EFFECTS OF AMYLOID IMMUNOTHERAPY
Both drugs are associated with significant side-effects. Besides frequent and relatively
minor infusion-related reactions, there are potentially more serious side-effects,
including amyloid-related imaging abnormalities (ARIAs) comprising brain swelling
(ARIA-E), or the emergence of new microhemorrhages (ARIA -H).[12] In the phase-3 clinical trials, ARIA was seen in 21% of patients treated with lecanemab[8] and 39% with donanemab.[8]
[9] Among patients who developed ARIA, ∼ 80% of cases in clinical trials were asymptomatic
and detected on surveillance magnetic resonance imaging (MRI) scans alone;[13] most cases occurred in the first 3 months of dosing; and > 80% resolved spontaneously
within 2 to 4 months,[8]
[9] following which dosing could be safely resumed in the majority. A small but clearly
important number of patients did develop symptoms including blurred vision, headaches,
unsteadiness, or dysphasias, and rare deaths have now been reported[9]—often but not exclusively in patients with preexisting risk factors. There is now
consensus that risk for ARIA is increased in patients with preexisting brain swelling
or intracerebral bleeding, and in individuals harboring one, and particularly two,
copies of the ApoE-E4 risk allele.[12] These data have informed inclusion/exclusion criteria, requirements for MRI safety
monitoring, and genetic testing reflected in appropriate use criteria,[14] and, ultimately, in licensing decisions made by individual countries (see below).
The observation that individuals treated with amyloid immunotherapies as a group have
excess brain shrinkage as measured using serial MRI has led to concerns that these
drugs might, contrary to predictions, be associated with excess neurodegeneration. However, a recent comprehensive review concluded that this excess
volume loss is much more likely to be explained by β-amyloid removal.[15] While longer-term follow-up data are required, these authors proposed that this
phenomenon be termed amyloid-related pseudoatrophy (ARPA).
CONTROVERSIES
After more than two decades since the last drug was licensed for AD, one might expect
that the emergence of two new therapies with proven efficacy in large well-run phase-3
trials would be greeted with universal enthusiasm. However, the clinical and research
community has been far from united, with very robust and heated debate on a range
of topics including, but not limited to, clinically meaningfulness, risk-benefit ratio,
and cost effectiveness.
There has been much debate about whether the statistically significant observed cognitive
benefits based on the absolute changes observed between the treated and placebo, are
“clinically meaningful.” Detractors have argued that the observed benefits are, for example, smaller than
those seen with standard AD treatments such as acetylcholinesterase inhibitors.[16] Others have argued that these changes are clinically meaningful either based on their interpretation of the absolute difference
or as demonstrable benefits on non-cognitive measures, including quality of life and
career burden. Another argument is that different standards might be appropriate for
disease-modifying therapies in which the benefits are hoped to be maintained or to
increase over time, noting that some post-hoc analyses of the trial data have suggested
the latter, showing divergence of treatment benefits over time.[17]
[18]
At the time of writing, one or both drugs are licensed for use in the United States
(US), Japan, China, South Korea, Hong Kong, Israel, Great Britain, and United Arab
Emirates, but not in Australia. A reflection of varying and changing opinions is that
the European Medicines Agency originally rejected a license for lecanemab, only to
reverse this decision on appeal within a few months.[19] Similarly, even in countries endorsing their use, there are differences regarding
who is eligible. Thus, in the United Kingdom (UK) and the European Union (EU) neither
drug is licensed for use in ApoE4 homozygotes based on a decision that, in this group,
the risk (such as ARIA) outweighs the benefit.[19]
[20]
[21]
[22]
A related but different question is: When a drug is licensed, who should pay for it?
This is a highly complex and very health-system specific calculus based on perceived
risk/benefit, system readiness, duration of therapy, and differing views/requirements
based on equity and models of payment. Not surprisingly, different healthcare systems
have arrived at different conclusions. Thus, in the US, up to 80% of the costs of
treatment may be reimbursed by Medicare.[23] Conversely, at the time of writing, in the UK, the National Institute for Health
and Care Excellence (NICE) has rejected both drugs for routine use within the free-at-the
point of care National Health Service (NHS), citing insufficient cost-effectiveness
when the totality of costs—the drug, means of determining inclusion (using positron-emission
tomography [PET] scans or cerebrospinal fluid [CSF] analysis), regular infusions over
a minimum of 18-months, safety MRI scans, and clinical monitoring—are combined.[24]
WHERE ARE WE NOW?
As a result of the issues outlined above, there remains significant disparity into
how these drugs have been used in clinical practice to date. Where these drugs are
being prescribed, for example in the US and Japan, experience is slowly growing about
their use in a real-life setting. While, thus far, approximately 15 thousand people
have been treated worldwide it seems that, at least in the early-adopter and often
specialist services offering treatment, it can be done safely and effectively. Experience
is informing practicalities of administration and protocol changes to minimize risk.
Ongoing use and registries will further inform the longer-term risk/benefits and cost-effectiveness
in different populations: it is notable, for instance, that rates of ARIA seem much
lower in some countries (Japan) than others (the US).[20]
OUR OPINION
In our view, there can be no doubt that these drugs are effective in slowing cognitive
decline, and that they impact on the core pathologies that underpin AD. While side-effects
can be significant, both the trial and real-world data suggest that these can be managed
safely. It is, of course, important to bear in mind that AD is an inevitably progressive
and ultimately fatal illness. There is a risk of excessive medical paternalism in
denying patients the opportunity to make an informed decision about risks and benefits.
On this basis, we believe strongly that these drugs should be prescribable, and so
concur with the positive licensing decisions made by most countries who have thus
far opined.
When it comes to who should/will receive these medications, applying licensed inclusion/exclusion
and appropriate use criteria to the UK population, it has been estimated that 14%
of all those seen in memory services could be eligible.[25] It is unfortunate but perhaps not surprising that despite pleas over many years
that preparation to deliver disease-modifying therapies was needed, the health care
system in the UK—and in most countries—is simply not ready to deliver these treatments
at the scale required.[26] The licensing of lecanemab and donanemab must serve as a wake-up call to urgently
upscale our abilities to diagnose and offer novel treatments to patients with dementia;
new developments—perhaps most notably the advent of blood tests for AD pathologies—will
hopefully smooth this transition.[27] In our view, the reconfiguration of services to deliver these treatments would be
likely to result in improvements in diagnosis and management that extend beyond those
eligible for β-amyloid-targeting therapies, analogous to the broad and sustained improvements
in stroke care that resulted from system reconfiguration to deliver thrombolysis.[28] Moreover, there are around 127 drugs in the development pipeline for AD, and reconfigured
services for lecanemab and donanemab will result in much improved system readiness
to deliver other emerging treatments in due course.[29]
The costs of administering intravenous treatments on a fortnightly or monthly basis
and monitoring for side-effects using MRI are clearly a major barrier. The development
of next generation immunotherapies that can be administered subcutaneously[30] or via shuttle technologies[31] that require fewer doses and are hopefully safer will hopefully make this more feasible.
If longer term data were to show the continued accrual of benefits and resulting reduction
in health and social care resource use, then this could also improve the calculated
cost effectiveness. However, there are important questions about how cost-effectiveness
should be determined when much of the care costs are borne by families and carers
and may not, therefore, be adequately factored into health economic models. In the
meantime, it is essential that we continue to gain real life experience of how to
use the two licensed drugs we have in a range of health care settings and populations.
Accordingly, in our view, cautious administration to selected patients in settings
able to deliver treatments with appropriate monitoring is the right way forward.
Bibliographical Record
Jonathan M. Schott, Charles R. Marshall. Amyloid immunotherapy for Alzheimer's disease:
the case for cautious adoption. Arq Neuropsiquiatr 2025; 83: s00451807718.
DOI: 10.1055/s-0045-1807718
