Editorial: Real-World Outcomes of Disease-Modifying Therapies Highlight the Need for Diagnostic Biomarkers in Early Alzheimer’s Disease

Abstract

The promise of targeted humanized monoclonal antibody therapies to amyloid β and tau protein in Alzheimer’s disease from clinical trial data has not been realized when put to the test in real-world studies and practice. There have been regulatory approvals for diagnostic blood tests in China, Japan (HISCL Aβ42/40), and the United Kingdom (UK) (PrecivityAD2). On May 16, 2025, the US FDA approved the Lumipulse G blood test, which utilizes the plasma pTau217/Aβ1–42 ratio, for the diagnosis of cerebral amyloid plaques in symptomatic patients of 55 years or more. Biological biomarkers and targets are currently being evaluated in phase 1 to phase 3 clinical trials, to rapidly implement less invasive blood-based assays to detect tau species and neurofilament light (NfL). However, clinical validation of the diagnostic value of identifying blood biomarkers still requires support from amyloid positron emission tomography (PET) brain scans and/or the results of cerebrospinal fluid (CSF) analysis. This editorial aims to identify how real-world outcomes of new disease-modifying therapies highlight the need for diagnostic biomarkers for early Alzheimer’s disease and the current status of biomarkers identified by blood plasma, CSF, and imaging.

Alzheimer’s disease is the most common type of dementia and represents an increasing global public health challenge as populations are living longer [1]. Until 2021, the only treatments for patients with Alzheimer’s disease were symptomatic and aimed to improve cognitive and behavioral symptoms, rather than disease-modifying therapies to delay or slow disease progression [2]. A recently published systematic review and meta-analysis study by Kusoro and colleagues evaluated 13 cohort studies published up to December 2024 that included more than 30,000 patients with a diagnosis of dementia [3]. In this analysis, the age of onset of dementia ranged from 54 to 93 years, and the average time to diagnosis was 3.5 years, with younger age at onset and frontotemporal dementia having longer diagnostic delay [3]. Therefore, timely diagnosis of Alzheimer’s disease is a major global challenge and a significant factor in outcomes from any disease-modifying therapies that may explain some recent real-world findings [4].

The promise of targeted humanized monoclonal antibody therapies to amyloid β and tau protein from clinical trial data has not been realized when put to the test in real-world studies and practice [4,5]. For example, in June 2021, the disease-modifying therapy aducanumab, the first humanized recombinant monoclonal antibody targeting amyloid β, received accelerated approval by the US Food and Drug Administration (FDA) to treat Alzheimer’s disease and mild cognitive impairment [4, 6]. Clinical trial data showed that one year of monthly intravenous infusion of aducanumab in 165 patients with mild Alzheimer’s disease reduced brain amyloid β levels in a dose-dependent and time-dependent manner, and positron emission tomography (PET) imaging showed that almost 50% of the patients no longer had cerebral amyloid [7]. However, 40% of patients experienced amyloid-related imaging abnormalities (ARIA), which included edema (ARIA-E) or micro-hemorrhage (ARIA-H) [7,8]. The limitations of screening costs, treatment costs, and side effects resulted in a loss of support from healthcare insurers in the US [8]. In February 2024, production of aducanumab ceased [8].

It is important to recognize that disease-modifying therapies are not a cure for Alzheimer’s disease. Currently, the two primary disease-modifying monoclonal antibodies in Alzheimer’s disease are lecanemab and donanemab [9,10]. Clinical trial data have shown that when treated in the early stages of the disease, patients may experience delayed disease progression and an increased period of independence [10,11]. However, in real-world situations, any benefits of treatment will depend on the diagnosis of Alzheimer’s disease in the early stages and balance any benefits with increasingly recognized treatment toxicities and recently recognized challenges and inequalities in patient selection [12].

The US FDA approved two anti-amyloid therapeutic monoclonal antibodies, lecanemab in 2023 and donanemab in 2024, based on clinical trial data that showed significantly reduced amyloid plaques on brain magnetic resonance imaging (MRI) [10,11]. However, in a phase 3 clinical trial, lecanemab slowed cognitive decline by 25% at 18 months, which was only moderately less than placebo after 18 months [10]. Also, in a phase 3 clinical trial, donanemab slowed Alzheimer’s disease progression by 22.3% compared with placebo after 76 weeks of treatment [11]. Additionally, initial clinical trial findings suggested that lecanemab led to fewer amyloid-related imaging abnormalities (ARIA) compared to donanemab [10,13]. Notably, the US FDA requires that both disease-modifying therapies have boxed warnings about ARIA, as these imaging findings are usually asymptomatic and should be monitored with magnetic resonance imaging (MRI) scans to detect either bleeding or cerebral edema, and cautions that ARIA can be severe enough to be life-threatening [14]. Homozygosity for apolipoprotein E4 (APOE4) is a significant genetic risk factor for Alzheimer’s disease, but it is still unclear whether this is associated with an increased risk for ARIA [15].

Donanemab has now been approved in 13 countries, and its approval and launch in Europe are expected in 2025 [16]. On August 22, 2024 the UK Medicines and Healthcare products Regulatory Agency (MHRA) approved both lecanemab and donanemab for the early stages of Alzheimer’s disease in adults, but the UK National Institute for Health and Care Excellence (NICE) has not recommended them for routine use, as the benefits of both lecanemab and donanemab for patients without two copies of APOE4 are too small to outweigh the benefits and safety issues [17]. Investigator-led head-to-head clinical trials of lecanemab and donanemab are required, as well as long-term trials combined with real-world data [12,16]. Interpretations and assumptions from the results of controlled clinical trials can be significantly changed when tested in real-world clinical practice, when real-world data are used to support real-world evidence [5]. Recently approved disease-modifying therapies for Alzheimer’s disease, which showed such promise in clinical trials, have undergone a recent real-world ‘trial by fire’ in Alzheimer’s disease [12].

Phase 3 clinical trial study participants showed major differences from real-world patients as they were selected with minimal comorbidities, unlike patients who can also have stroke, heart failure, endocrine disease, or traumatic brain injury [12]. There are concerns regarding the duration and mode of treatment, as lecanemab requires an initial infusion every two weeks for 18 months, which is typically administered at infusion centers [18]. Because the symptoms of dementia can continue even after clearance of amyloid plaque, this supports the need for maintenance therapy [19]. However, the duration, dosing, and treatment frequency remain to be determined [19]. The availability and cost of treatment with lecanemab and donanemab result in a disparity in access to these therapies between countries and within countries [12,20].

Any new biomarker associated with Alzheimer’s disease or other types of dementia is not a disease predictor but is an association that supports early clinical diagnosis [21]. There is still a lack of public and clinical awareness of the presenting diagnostic features of Alzheimer’s disease [21]. The diagnosis of Alzheimer’s disease has relied on invasive methods in patients who present with symptoms, often at an advanced stage [1,22]. Invasive diagnostic procedures, including cerebral positron emission tomography (PET) imaging and analysis of cerebrospinal fluid (CSF), are costly and invasive methods that are not widely accessible [22]. There is an urgent requirement for early diagnosis and screening for Alzheimer’s disease, which can then drive a more effective approach to treatment, prevention, and possible screening. In the past year, the challenge for early diagnosis of Alzheimer’s disease has become realized with several blood biomarkers that may have a future role in diagnosis and screening [23]. There have been regulatory approvals for diagnostic blood tests in China, Japan (HISCL Aβ42/40), and the United Kingdom (UK) (PrecivityAD2) [23]. On May 16, 2025, the US FDA approved the Lumipulse G blood test (Fujirebio Diagnostics), which utilizes the plasma pTau217/Aβ1–42 ratio, for the diagnosis of cerebral amyloid plaques in symptomatic patients of 55 years or more [23]. However, a limitation of blood biomarker tests is the high rate of false positives and false negatives, which are expected in peripheral biomarker tests for conditions that affect the brain [23]. Therefore, in practice, the clinical validation of the diagnostic value of identifying blood biomarkers still requires support from amyloid PET brain scans and/or the results of CSF analysis [23].

Alzheimer’s disease shares with other neurological diseases, such as Parkinson’s disease and multiple sclerosis, pathological and physiological changes that occur long before clinical symptoms are seen, which supports research on biomarkers for early disease and screening [24]. In 2016, the FDA-NIH Biomarker Working Group introduced a classification framework encompassing seven categories of biomarkers and related endpoints [25]. These include biomarkers for diagnosis, monitoring, prognosis, predictive biomarkers, pharmacodynamic response, safety, and susceptibility (risk) biomarkers [25]. In 2018 and 2024, the combined National Institute on Aging and the Alzheimer’s Association (NIA-AA) diagnostic criteria identified a list of diagnostic biomarkers for Alzheimer’s disease that could be identified by proteomics analysis in plasma and CSF, and brain imaging biomarkers (Table 1) [24,26]. In Alzheimer’s disease, amyloid accumulation is followed by tau aggregation, followed by structural changes in the brain that lead to cognitive decline [26]. The importance of disease biomarkers is highlighted by the fact that biological biomarkers and targets are currently either used or are being investigated in clinical development and in phase 1 to phase 3 clinical trials, with plasma and CSF markers in 26% of trials targeting inflammation and amyloid [27]. CSF biomarkers for Alzheimer’s disease, including the FDA-approved Aβ42/40, phosphorylated tau 181 (pTau181), and total tau tests, and tests for neurofilament light (NfL), are currently used in clinical and trial settings [23,27]. The aim is now to move quickly to implementing less invasive blood-based assays to detect tau species and NfL, as demonstrated in the FDA approval of a pTau217/Aβ42 AD blood test in May 2025 [23,27]. Alzheimer’s disease is now recognized as a heterogeneous condition that may require a more personalized and targeted approach to treatment, which will need refinement in biomarker analysis [28,29].

Table 1.

Biomarker analysis in Alzheimer’s disease using imaging, blood plasma, and cerebrospinal fluid (CSF).

	Imaging	Blood plasma	Cerebrospinal fluid (CSF)
Diagnostic biomarkers
A: (Aβ proteinopathy) T1: (phosphorylated and secreted AD tau)	Amyloid PET –	– p-tau217	– –
Hybrid ratios	–	%p-tau217	p-tau181/Aβ42, t-tau/Aβ42, Aβ42/40
Staging, prognosis, and treatment response biomarkers
A: (Aβ proteinopathy) T1: (phosphorylated and secreted AD tau)	Amyloid PET –	– –	– –
Hybrid ratios:	–	%p-tau217	p-tau 181/Aß42, t-tau/Aβ42, Aβ42/40
T2: (AD tau protein enteropathy)	Tau PET	MTBR-tau243, other p-tau forms (e.g., p-tau205)	MTBR-tau243, other p-tau forms, non-phosphorylated mid-region tau fragments
N: (neutropil injury, dysfunction, or degeneration)	Anatomic MRI, FDG PET	NfL	NfL
I: (inflammation)	–	GFAP	GFAP
Co-morbid pathology biomarkers
N: (dysfunction, or degeneration of neuropil)	Anatomic MRI, FDG PET	NfL	NfL
V: Vascular brain injury	Cerebral infarction on MRI or CT	–	–
S: a-synuclein	–	–	aSyn-SAA

AD – Alzheimer’s disease; CT – computed tomography; FDG – fluorodeoxyglucose; GFAP – glial fibrillary acidic protein; MRI – magnetic resonance imaging; MTBR – microtubule binding region of the tau protein; NfL – neurofilament light chain; PET – positron emission tomography; SAA – serum amyloid A. Adapted from: The 2024 revised criteria from the Alzheimer’s Association Workgroup [26].

Conclusions

In Alzheimer’s disease, early detection of pathological changes before clinical symptoms develop remains a significant unmet need, as are biomarkers that can predict and monitor disease progression, and targets for treatment that should be both disease-specific and patient-specific. Drug treatments alone may not provide the solution for prevention or treatment of Alzheimer’s disease, even if reliable blood biomarkers for early diagnosis do become available. Identifying and controlling risk factors for Alzheimer’s disease is the third arm in managing this significant global public health problem.