Blood-based biomarkers of Alzheimer's disease: potential utility in clinical practice

Abstract

Purpose of review

Blood-based biomarkers (BBMs) for Alzheimer's disease are beginning to enter clinical practice. As this integration advances, it is essential to critically examine their strengths, limitations, and readiness for broader clinical application.

Recent findings

Evidence increasingly supports the utility of BBMs for clinical management of Alzheimer's disease, with phosphorylated tau species, Aβ42/40 ratio, GFAP, and NfL among the most studied. Plasma p-tau forms have emerged as the most promising markers, showing strong correlations with amyloid plaque deposition and predictive value for disease progression. The WHO and the Global CEO Initiative have outlined minimum performance criteria for clinical use. While no BBM meets these benchmarks with a single cutpoint, adopting a two-cutpoint approach by introducing an intermediate category has enabled some assays to achieve the required accuracy. Several assays are now commercially available, and two have recently received FDA clearance to assist in confirming or ruling out amyloid-beta pathology.

Summary

BBMs could transform Alzheimer's disease diagnostics by enabling scalable, minimally invasive approaches for early detection and monitoring. As implementation advances, assay harmonization, assessment of demographic and physiological influences, and real-world validation across diverse populations remain essential to ensure reliability and equitable access.

INTRODUCTION

Blood-based biomarkers (BBMs) for Alzheimer's disease have advanced significantly with the development of ultrasensitive assay platforms and the discovery of novel targets [1,2]. Several BBMs with strong correlations with brain amyloid beta (Aβ) plaques (A), hyperphosphorylated tau neurofibrillary tangles (T), and neurodegeneration (N) – the three core pathological features of Alzheimer's disease – have recently emerged, positioning them as next-generation tools for diagnosis and disease monitoring [3,4^▪▪,5,6^▪▪,7]. These tests are transforming neuroscience research and neurology practice. Previously, antemortem assessment of in-vivo Alzheimer's disease pathology relied on cerebrospinal fluid (CSF) obtained through invasive lumbar punctures or costly neuroimaging techniques such as MRI and PET [8,9]. These modalities are generally limited to specialty care settings or urban academic centers, restricting access for the wider population [10,11]. The lack of time-effective and cost-effective diagnostic tools, together with the scarcity of dementia health professionals, has contributed to high rates of undiagnosed or delayed diagnosis of cognitive impairment [12,13]. These limitations are also expected to hinder the adoption of recent FDA-approved Aβ-targeting therapies, which require confirmation of Aβ pathology before initiation [14,15^▪].

To address the growing global burden of dementia, which affects approximately 55 million people worldwide, out of which 60–70% pertains to Alzheimer's disease, the WHO has proposed a Global Dementia Action Plan [16^▪▪]. One key priority is the development of highly sensitive and specific diagnostic biomarkers that are also cost-effective. Compared with CSF biomarkers and neuroimaging, Alzheimer's disease BBMs offer advantages such as being minimally invasive, lower cost, and suitable for widespread use. These biomarker tools are increasingly used for detection, prognosis, and tracking progression. Many therapeutic trials now incorporate BBMs as proxies for brain pathology to support participant selection and monitor treatment response, a strategy that is expected to streamline drug development for future disease-modifying therapies (DMTs) [17,18]. Additionally, healthcare systems have begun integrating BBMs into diagnostic workflows, with guidelines for appropriate use recently published [19^▪▪]. Notably, two BBM assays – Lumipulse p-tau217/Aβ42 (Fujirebio) and Elecsys pTau181 (Roche) – have received FDA approval for clinical use in detecting or ruling out brain Aβ plaques in adult patients, aged 55 years and older [20,21].

However, are Alzheimer's disease BBMs ready for prime time? Could they be used independently of neuroimaging in clinical practice? As BBMs move closer to implementation, it is critical to understand their strengths, limitations, and readiness for routine use. This review provides an overview of recent advances and highlights the opportunities and challenges that lie ahead. 

Box 1.

no caption available

MOLECULAR NATURE OF ALZHEIMER'S DISEASE BLOOD-BASED BIOMARKERS

Alzheimer's disease BBMs represent molecular signatures of the pathological processes underlying the disease. Key Alzheimer's disease BBMs include Aβ peptides, various tau, and phosphorylated tau (p-tau) proteoforms, glial fibrillary acidic protein (GFAP), neurofilament light chain (NfL) [4^▪▪,22^▪▪]. Each biomarker is thought to originate from distinct cellular and molecular events in the brain. Aβ peptides, cleavage products of the amyloid precursor protein (APP), are the main constituents of Aβ plaques, with Aβ40 and Aβ42 being the dominant forms. Aβ42 aggregates more readily than Aβ40, leading to its sequestration in plaques and a decreased CSF Aβ42/Aβ40 ratio – a phenomenon that underpins its diagnostic utility. Tau pathology is another defining feature of Alzheimer's disease, and phosphorylated tau species have become central to Alzheimer's disease BBM research and clinical care. Tau hyperphosphorylation promotes neurofibrillary tangle formation, and several p-tau species – such as p-tau181, p-tau205, p-tau212, p-tau217, and p-tau231 – have been shown to correlate with Alzheimer's disease pathology. Beyond p-tau species, assays targeting specific tau forms or fragments have also demonstrated strong associations with disease state, offering additional molecular granularity. Examples include brain-derived tau (BD-tau), which selectively detects CNS-origin tau, as well as truncated tau fragments and regions within the microtubule-binding domain (MTBR) [23^▪,24].

Other biomarkers do not directly reflect Aβ or tau pathology but instead capture biological processes, such as neuroinflammation and axonal injury, that play critical roles in Alzheimer's disease pathogenesis. GFAP, an intermediate filament protein expressed by astrocytes, signals reactive astrogliosis – a process that can be protective or harmful depending on astrocytic response [25]. Elevated plasma GFAP is generally linked to astrocyte injury, neuroinflammation, or blood–brain barrier dysfunction, all of which are common in Alzheimer's disease [26]. NfL, another intermediate filament protein with essential role in maintaining the neuronal cytoskeleton and structural integrity, is released following axonal damage and serves as a marker of neuronal structural integrity loss [27]. NfL is elevated in several neurological conditions including Alzheimer's disease, frontotemporal lobar degeneration (FTD), multiple sclerosis, and traumatic brain injury (TBI). Although not specific to Alzheimer's disease, its elevation indicates neurodegeneration and complements other BBMs in staging disease severity [28–30].

Historically, the measurement of these biomarkers was hindered by their low concentrations in blood. Recent technological breakthroughs have overcome these limitations through the development of ultrasensitive platforms, including Single Molecule Array (Simoa), Meso Scale Discovery (MSD), Lumipulse, NUcleic acid-Linked Immuno-Sandwich Assay (NULISA), Beckman DxI 9000, Roche Elecsys, Fujirebio Lumipulse, and Ella, as well as mass spectrometry-based methods [4^▪▪,22^▪▪]. These advances have enabled the integration of BBMs into the AT(N) and the recently revised frameworks (AT₁T₂NISV) proposed by the Alzheimer's Association, establishing a biological basis for Alzheimer's disease diagnosis beyond clinical symptoms [5,6^▪▪].

DIAGNOSTIC UTILITY OF BLOOD-BASED BIOMARKERS IN ALZHEIMER'S DISEASE

BBMs have emerged as powerful tools to support the clinical management of Alzheimer's disease. Their potential has been evaluated in numerous studies, most of which focus on diagnostic accuracy of brain Aβ pathology. Among these biomarkers, p-tau species stand out as top performers, with p-tau181 and p-tau217 being the most extensively investigated [4^▪▪,31^▪,32^▪▪]. Head-to-head comparisons consistently show that p-tau217 outperforms p-tau181 for detecting Aβ pathology [33^▪,34–36^▪]. Meta-analyses report pooled sensitivity and specificity of 88.1 and 88.7% for p-tau217, compared to 80.5 and 76.4% for p-tau181 [32^▪▪]. Both biomarkers can be measured across multiple analytical platforms. In a recent meta-analysis, Pahlke et al. (2025) reported pooled sensitivities ranging from 49.3 to 91.4% and specificities from 75 to 96.7% across 11 analytical platforms for plasma p-tau217, with the Precivity-%p-tau217 IP-MS assay demonstrating the best overall performance. In contrast, plasma p-tau181 exhibited pooled sensitivities of 66.9–86.1% and specificities of 67.5–89.1% across eight immunoassays [4^▪▪]. Notably, while blood levels of p-tau biomarkers generally increase with the severity of Alzheimer's disease pathology, their trajectories are predictive of disease stage, highlighting the potential utility of different p-tau species for staging Alzheimer's disease. Specifically, p-tau231 and p-tau217 show early elevations as cognitively normal individuals transition to Aβ positivity, with p-tau231 rising first, followed shortly by p-tau217. P-tau181 increases later, as amyloid pathology becomes more widespread, whereas p-tau205 shows a marked rise much later, around the onset of clinical symptoms [37^▪,38^▪,39^▪▪].

Plasma Aβ42/Aβ40 also becomes abnormal very early during Alzheimer's disease pathogenesis [38^▪] and can be measured on several technical platforms, including IP-MS assays and automated immunoassays. Head-to-head comparisons generally favor IP-MS assays, with the Washington University IP-MS (IP-MS-WashU) method showing the best overall performance and achieving the highest weighted average area under the curve (AUC) (0.846) across 21 studies reviewed by Brand et al. (2022) [40–42]. The meta-analysis by Pahlke et al. in 2025 reported pooled sensitivities of 59.3–90.1% and specificities of 61.5–83.3% across 11 different Aβ42/Aβ40 assays [4^▪▪]. The HISCL immunoassay by Sysmex, not included in the review by Brand et al., was identified as the top performer; however, this conclusion was based on a single study involving two cohorts with a combined total of 397 participants [43]. Despite some plasma Aβ42/Aβ40 assays showing strong discrimination for brain Aβ pathology, the relatively small effect size (~15% difference between Alzheimer's disease and controls) [44], and the instability of Aβ peptides [45], make results highly susceptible to preanalytical and analytical variability, limiting routine clinical use. Furthermore, Aβ42/Aβ40 plateaus quickly as amyloid burden increases, reducing its value for tracking disease progression.

Plasma GFAP correlates positively with Aβ burden, both in early stages and across the disease continuum. In a study of 871 participants across three cohorts, GFAP distinguished Aβ-PET positive from negative individuals with AUC values of 0.69–0.86 and was also elevated in symptomatic stages [46^▪▪]. A meta-analysis of 31 studies found significantly higher GFAP levels in Aβ-positive compared to Aβ-negative individuals, with a pooled standardized mean difference (SMD) of 0.895 and a pooled SMD of 1.149 comparing Alzheimer's disease versus controls [47]. Despite its strong discriminatory potential, the clinical implementation of plasma GFAP faces several challenges. Plasma GFAP levels exhibit substantial intra-individual (test–retest) and inter-individual variability, necessitating wider diagnostic cut-offs to ensure clinically meaningful differences beyond natural variation [48]. GFAP concentrations are also influenced by demographic and physiological factors, including age (higher in older individuals), sex (higher in female individuals), circadian rhythm, and postprandial food intake, complicating interpretation. Moreover, GFAP can be elevated in non-Alzheimer's disease neurological conditions such as Parkinson's disease, Huntington's disease, multiple sclerosis, stroke, and TBI [49,50].

Both plasma NfL and BD-tau reflect neurodegeneration, but they differ in disease specificity. NfL is a nonspecific marker of axonal injury, elevated in Alzheimer's disease as well as other neurodegenerative conditions [51^▪]. Plasma NfL is consistently higher in Alzheimer's disease compared to controls, but with smaller effect sizes for MCI, indicating that NfL becomes abnormal later in the disease continuum. For example, Zhao et al. in 2019 reported pooled SMDs of 1.15 comparing Alzheimer's disease versus controls and 0.40 for MCI versus controls in their meta-analysis involving 42 studies. Similar patterns were found by Sahrai et al.[52] in 2023 in their meta-analysis involving 15 studies, with corresponding SMDs of 0.808 and 0.361, respectively [29]. Due to its nonspecific nature and limited ability to detect early Alzheimer's disease, NfL is not ideal as a diagnostic biomarker. In addition, similar to GFAP, plasma NfL levels are also strongly influenced by demographic and physiological factors such as age, BMI, and comorbidities [53–55]. In contrast, BD-tau, measured with assays targeting a CNS-specific tau epitope (exon 4-5 junction) absent in peripheral tissues, is specific to Alzheimer's disease-related neurodegeneration. Plasma BD-tau correlates strongly with CSF tau and autopsy-confirmed Alzheimer's disease neuropathology [56^▪▪,57^▪] and, unlike NfL, demonstrates robust discrimination between Alzheimer's disease and non-Alzheimer's disease dementias. However, due to its recent introduction, validation in larger and diverse cohorts is still needed to confirm its clinical utility.

PROGNOSTIC UTILITY OF ALZHEIMER'S DISEASE BLOOD-BASED BIOMARKERS

Evaluating prognostic performance requires cohorts with longitudinal follow-up, making the acquisition of suitable samples considerably more challenging than for diagnostic studies. Nevertheless, an increasing number of investigations have examined the prognostic utility of these BBMs, supporting their ability to predict future clinical trajectories across both clinical cohorts and population-based samples. Overall, these studies consistently highlight the value of baseline p-tau181, p-tau217, GFAP, and NfL, while Aβ42/Aβ40 generally shows limited prognostic relevance [58^▪,59–64]. Among preclinical Alzheimer's disease participants, p-tau217 shows the strongest predictive power for progression to Alzheimer's disease dementia [61,63,64]. For instance, in a community-based cohort of approximately 2148 dementia-free individuals, Grande et al.[61] in 2025 reported that higher plasma levels of p-tau181, p-tau217, NfL, and GFAP were associated with an increased 10-year risk of incident dementia, achieving AUC values between 70.9 and 76.8% for predicting Alzheimer's disease dementia, while Aβ42/Aβ40 and total tau showed no significant predictive value. P-tau217 slightly outperformed other BBMs for predicting incident Alzheimer's disease dementia, whereas NfL showed superior performance for all-cause dementia. Notably, predictive accuracy improved when p-tau217 was combined with either NfL or GFAP. Similarly, in the ADNI cohort of 233 nondemented participants, elevated baseline levels of p-tau181, p-tau217, GFAP, and NfL were linked to steeper cognitive decline on MMSE and composite measures of executive function and memory, while Aβ42 – but not Aβ40 – was associated with memory performance [59].

Plasma NfL, while not an ideal diagnostic biomarker due to limited sensitivity for early Alzheimer's disease detection, is a robust prognostic marker for disease progression, independent of amyloid status, particularly in clinical settings. A meta-analysis of 19 studies demonstrated that higher baseline plasma NfL levels were associated with faster subsequent cognitive decline, as measured by multiple instruments including MMSE, ADAS, CDR, ADNI-S, and ADCS-ADL [65]. It outperformed GFAP, p-tau181, total tau, and Aβ42/Aβ40 in predicting worsen cognitive scores assessed by ADAS-Cog11 and Clinical Dementia Rating sum of box (CDR-SB), greater loss of whole brain volume, and increasing ventricular volume in the T2 Protect Alzheimer's disease, a negative 48-week, phase-2, placebo-controlled trial of troriluzole in mild-to-moderate Alzheimer's disease [60].

CURRENT LANDSCAPE OF CLINICAL USE OF ALZHEIMER'S DISEASE BLOOD-BASED BIOMARKERS

The strong performance of Alzheimer's disease BBMs, supported by a growing body of evidence, has generated significant enthusiasm for their integration into clinical practice. BBMs provide a minimally invasive approach for risk stratification, diagnostic confirmation, prognostic assessment, treatment monitoring, and clinical trial participant screening across diverse healthcare settings [19^▪▪,62,66^▪▪]. To support the adoption of BBMs in clinical practice, both the WHO and the Global CEO Initiative have proposed minimum performance criteria for their implementation. For diagnostic use to confirm amyloid pathology, both organizations recommend a sensitivity and specificity of at least 90% in the intended-use population [16^▪▪,67^▪▪]. Additionally, the Global CEO Initiative has outlined an alternative implementation pathway for BBMs as a triage tool prior to confirmatory testing, such as amyloid PET or CSF analysis [67^▪▪]. Under this approach, the Global CEO Initiative recommends a sensitivity of at least 90% and a specificity of a least 85% in primary care settings, and a specificity between 75 and 85% in secondary care settings, depending on the availability of follow-up testing.

Given the high prevalence of dementia, its prolonged disease course, and the potential harm of an inaccurate diagnosis, such as emotional distress, unnecessary interventions, and missed treatment opportunities, BBM testing is currently recommended only for people with cognitive impairment and should be paired with a full clinical assessment to guide Alzheimer's disease management [68^▪]. Meta-analyses by Pahlke et al.[4^▪▪] in 2025, which evaluated the diagnostic accuracy of several classical BBMs (p-tau217, %p-tau217, p-tau181, p-tau231, and Aβ42/Aβ40), each across multiple platforms, show that none of the currently available assays meet the recommended performance threshold of at least 90% sensitivity and specificity for confirmatory testing. However, several p-tau217 assays come close, achieving sensitivity and specificity above 85%. To optimize diagnostic accuracy to meet the performance criteria, a two-threshold approach has been proposed, incorporating distinct high cut-off and low cut-off and creating three result categories: positive, intermediate, and negative [67^▪▪]. This strategy achieved 92% sensitivity and 96% specificity for the Lumipulse and ALZpath p-tau217 assays in 427 participants with MCI, while classifying 20 and 39% of results as indeterminate, respectively [69]. Similarly, in a large multicenter cross-sectional study by Ahn et al.[70] in 2024, p-tau217 levels measured with the ALZpath assay consistently met confirmatory testing standards, achieving sensitivity and specificity at least 90% across subgroups defined by age, sex, BMI, and apolipoprotein E (APOE) ε4 status within the cognitively impaired population.

Several Alzheimer's disease BBM assays are now available in clinical practice. In May 2025, the FDA approved the first blood-based confirmatory test for Aβ pathology – the Lumipulse p-tau217/Aβ42 ratio – for use in symptomatic patients aged 55 and older. Approval was based on a clinical study of 499 plasma samples from individuals with cognitive impairment, demonstrating 91.7% sensitivity and 97.3% specificity, with fewer than 20% of results falling into the indeterminate range [20]. In October 2025, the FDA approved a second blood-based test – Roche's Elecsys pTau181 – as a triage tool for adults aged 55 and older with cognitive symptoms in primary care, designed to rule out Aβ pathology. Approval was based on a multicenter, noninterventional study involving 312 participants, which demonstrated a negative-predictive value (NPV) of 97.9%. This positions Elecsys pTau181 as an initial screening tool to reduce unnecessary PET scans or CSF testing and guide patient referral for appropriate treatment [21]. In addition, several CLIA-certified assays – though not FDA-cleared – are available, including C2N Diagnostics’ PrecivityAD [71], Quest Diagnostics’ AD-Detect (based on Aβ42/Aβ40 ratio with optional p-tau217) [72], and Lucent Diagnostics’ LucentAD Complete (a multimarker panel combining p-tau217, Aβ42/Aβ40, GFAP, and NfL) [73]. These tests vary in purpose, performance, and interpretation but collectively represent a major step toward integrating BBMs into routine clinical care.

CHALLENGES FOR REAL-WORLD IMPLEMENTATION OF ALZHEIMER'S DISEASE BLOOD-BASED BIOMARKERS

Despite significant progress, several challenges hinder the widespread clinical adoption of Alzheimer's disease BBMs. First, it is important to note that their reported performance largely reflects studies conducted in highly resourced, tightly controlled research environments and among populations with limited diversity. Measurements were typically performed retrospectively on banked blood samples in a few batches, rather than through real-time testing [74]. Moreover, most Alzheimer's disease BBM studies have focused on analytical performance or diagnostic accuracy, while far fewer have assessed their impact on diagnostic decision-making, patient outcomes, or broader societal implications [75]. Real-world performance data derived from routine clinical applications in diverse populations, including those in low-income and middle-income countries (LMICs), remain unavailable, making it difficult to assess their true translatability.

One key challenge in bringing Alzheimer's disease BBMs into clinical practice is that health systems are not yet fully prepared for their accurate and appropriate use [76]. Because any screening test can potentially cause harm such as stigmatization, inappropriate treatment, emotional distress, or financial burden, both healthcare professionals and the general public need appropriate education to avoid misuse or misinterpreting. This is especially important, as these tests move beyond specialized memory clinics into primary care and broader screening settings. Frontline clinicians will need practical training and clear guidance on when to use Alzheimer's disease BBMs, how to interpret results in context, and how to communicate uncertainties. Importantly, currently FDA-approved Alzheimer's disease BBMs are intended to help confirm or rule out Alzheimer's pathology in individuals with cognitive impairment and not a replacement for standard clinical evaluation.

Another key challenge is the influence of demographic factors, comorbid conditions, and social determinants of health on Alzheimer's disease BBM levels, raising concerns about the need for subgroup-specific interpretation of test results [53,54,76,77]. For instance, both NfL and GFAP show strong associations with age [53,78]. Similarly, findings from the multigenerational Young Finns Study revealed significant associations between BBM levels and factors such as age, APOE ε4 carrier status, glucose metabolism, and dyslipidemia [79]. Chronic comorbid conditions, highly prevalent among older adults with dementia [e.g. 46% coronary artery disease, 46% chronic kidney disease (CKD), and 34% chronic heart failure] [80], also demonstrate significant associations with BBM levels, with differences between participants with and without CKD approaching the effect size observed between individuals with and without elevated brain amyloid [81^▪].

Lack of standardization to minimize measurement variability is another major challenge that must be addressed for the clinical implementation of Alzheimer's disease BBMs. Measurement variability arises from three sources: preanalytical, analytical, and postanalytical, with preanalytical variation accounting for more than 60% of errors in clinical laboratories [82]. BBM levels can be influenced by several preanalytical factors, including blood tube type and additives, time of day for collection, delays before and after centrifugation, processing temperature, short-term and long-term storage conditions, centrifugation parameters, and freeze–thaw cycles, as previously reviewed [22^▪▪,83,84]. Currently, no certified reference measurement procedures exist to ensure assay comparability.

Existing clinically used assays rely on venipuncture-based blood specimens and require advanced laboratory equipment, such as ultra-low temperature freezers and centrifuges. These requirements limit the broad application of Alzheimer's disease BBMs in resource-limited settings, including home sampling and studies conducted in remote areas. Alternative sampling methods, such as dried plasma spots, specialized blood tubes to prolong biomarker stability, and point-of-care devices are critical to facilitate and expand the use of BBMs in these settings. Although several studies have demonstrated the potential utility of alternative sampling [85,86^▪▪,87^▪▪–88^▪▪], further research is needed to confirm their reliability and clinical applicability.

Alzheimer's disease is a complex condition involving multiple brain pathologies. Its multifaceted nature is reflected in the updated research framework for Alzheimer's disease diagnosis and staging, which expands beyond amyloid (A), tau (T), and neurodegeneration (N) to include additional processes such as inflammation (I), vascular changes (V), and synucleinopathy (S) [89]. This revision underscores the need for a broader biomarker panel to enable precise clinical management. However, most current BBMs are primarily associated with amyloid or neurodegeneration. Recent emerged plasma MTBR and N-terminal containing tau fragment assays have shown promise as a tau tangle biomarker, but larger validation studies are needed to confirm this association [23^▪,24]. Furthermore, BBMs targeting mixed pathologies remain lacking, highlighting the need for continued development and validation of novel assays for comprehensive disease management.

CONCLUSION

Alzheimer's disease BBMs are increasingly applied in research and clinical practice to rule out individuals without Alzheimer's disease pathology or confirm pathology in cognitively impaired patients. Multiple assays are now available in clinics, including FDA-approved and CLIA-certified platforms; however, validation in real-world settings is essential to ensure clinical translatability. Their use can reduce reliance on invasive CSF sampling and costly amyloid PET imaging while accelerating confirmatory testing, treatment decisions, and clinical trial enrollment. Future research should prioritize assay standardization to minimize analytical variability, development of additional biomarkers to capture the multifaceted pathology of Alzheimer's disease, and evaluation across diverse populations considering demographic factors, comorbidities, and social determinants to enable equitable and reliable implementation.

Acknowledgements

The authors thank members of the Karikari Laboratory and their network of collaborators.

Financial support and sponsorship

The Karikari Laboratory was supported by NIH/NIA (R01 AG083874, U24AG082930, P30 AG066468, RF1AG077474, R01 AG083156, R37 AG023651, R01 AG025516, R01 AG073267, R01 AG075336, R01 AG072641, P01 AG025204), NIH/NINDS (U01 NS131740, U01 NS141777), NIH/NIMH (R01 MH108509), Aging Mind Foundation (DAF2255207), DoD (HT94252320064), the Anbridge Charitable Fund, and a professorial endowment from the Department of Psychiatry, University of Pittsburgh. The content of this article is solely the responsibility of the authors and does not necessarily represent the official views of the funders.

Conflicts of interest

T.K.K. has served as an adhoc consultant and/or advisory board member for Quanterix Corporation, SpearBio Inc., Neurogen Biomarking LLC., Alzheon, Siemens Healthineers and Neurogen Biomarking LLC., outside the submitted work. T.K.K. and X.Z. are inventors on patents and provisional patents regarding biofluid biomarker methods, targets, and reagents/compositions that may generate income for the institution and/or self should they be licensed and/or transferred to another organization. T.K.K. has received royalties from Bioventix for the transfer of specific antibodies and blood biomarker assays to third party organizations. The remaining author declares no competing interests.

REFERENCES AND RECOMMENDED READING

Papers of particular interest, published within the annual period of review, have been highlighted as:

• ▪ of special interest

• ▪▪ of outstanding interest