Neurology · Medical Literature
Biomarkers of angiogenesis in Alzheimer's disease: Systematic review and meta-analysis.
Abstract
BackgroundCerebrovascular dysfunction can contribute to the pathophysiology of Alzheimer's disease (AD) and trigger angiogenesis. To date, no prior systematic review and meta-analysis (SRMA) study has attempted to qualitatively and quantitatively review existing literature examining angiogenic factors in AD.ObjectiveIn this review, we aimed to identify markers of angiogenesis in biofluids that can differentiate between individuals with AD and healthy older adults, and inform the role of angiogenesis in AD.MethodsUsing Medline (1946 to August 4, 2021), the literature was systematically searched for articles according to PRISMA guidelines. Angiogenesis and AD terms were searched as keywords and mapped to MeSH headings. A total of 18 studies (including 28 dependent effect sizes) were included in the meta-analysis. Hedges' g was selected as the effect size of interest.ResultsA random-effects model including all eligible biofluid studies revealed a small and nonsignificant overall effect size (combined Hedges' g = -0.26, 95% CI [-1.45, 0.92], p = 0.647), with significant heterogeneity (I2 = 98.6%, p < 0.0001). Moderator analysis revealed no significant difference based on which specific angiogenic biomarker was examined (i.e., vascular endothelial growth factor (VEGF) versus others).ConclusionsVEGF-related marker levels were not significantly different in AD and normal aging. Based on our findings, few studies have examined fibroblast growth factor and platelet-derived growth factor-related markers; however, we believe they warrant further investigation. Identifying novel angiogenic biomarkers could elucidate the role of vascular mechanisms in AD and reveal potential therapeutic targets.
Clinical relevance
Can circulating biofluid markers of angiogenesis reliably differentiate individuals with Alzheimer's disease from healthy older adults, and what role does angiogenesis play in AD pathophysiology? __SNAPSHOT__ {"studyDesign":"Systematic review with meta-analysis (PRISMA-compliant)","setting":"Multi-study synthesis; literature search of MEDLINE (1946 to August 4, 2021); institutional affiliation includes University of California, Irvine (based on author affiliations)","sampleSize":"18 studies contributing 28 dependent effect sizes included in the meta-analysis; total participant count not reported in the available abstract","population":"Adults with clinically diagnosed Alzheimer's disease compared to healthy older adult controls; specific age ranges, sex distribution, and diagnostic criteria vary across included primary studies","intervention":"Measurement of angiogenic biomarkers (including VEGF and related factors, fibroblast growth factor, platelet-derived growth factor, and other angiogenesis-related proteins) in biofluids (blood-based and/or CSF)","comparator":"Healthy older adults (cognitively normal controls)","primaryOutcome":"Standardized mean difference (Hedges' g) in biofluid angiogenic marker levels between AD patients and healthy controls","followUp":"Not applicable (cross-sectional comparisons within included studies); study-level follow-up not reported","fundingSource":"Not reported in the available abstract"}
Key findings
The primary pooled analysis across all eligible biofluid studies using a random-effects model yielded a small and statistically nonsignificant overall effect size (combined Hedges' g = −0.26, 95% CI [−1.45, 0.92], p = 0.647), indicating no detectable aggregate difference in angiogenic biomarker levels between AD patients and healthy older adults. Critically, the meta-analysis demonstrated extreme and statistically significant heterogeneity across included studies (I² = 98.6%, p < 0.0001), rendering the pooled estimate unreliable and of limited interpretive value on its own. Moderator analysis did not identify a significant difference based on the type of angiogenic biomarker examined (VEGF vs. other markers). Specifically, VEGF-related marker levels were not significantly different between AD and normal aging groups. The authors noted that fibroblast growth factor- and platelet-derived growth factor-related markers have been examined in very few studies. NNT/NNH are not applicable to this diagnostic biomarker meta-analysis. No adverse events are relevant to this study design.
Practice implications
This systematic review and meta-analysis establishes that, based on studies published through August 2021, peripheral biofluid levels of angiogenic factors — including VEGF — do not reliably distinguish individuals with AD from healthy older adults at the aggregate level. The near-total heterogeneity (I² = 98.6%) signals that the included studies differ fundamentally in population characteristics, biomarker assay methodology, biofluid type (plasma vs. serum vs. CSF), disease staging, and/or concomitant cerebrovascular comorbidities, making any single pooled estimate clinically uninterpretable without subgroup context. Moderator analyses failed to resolve this heterogeneity by biomarker class. The study does NOT establish that angiogenesis is irrelevant to AD pathophysiology; rather, it demonstrates that the existing biofluid literature is too heterogeneous and methodologically variable to draw firm conclusions. Patients with mixed dementia, vascular dementia, or concurrent cerebrovascular disease were likely variably included or excluded across primary studies, further limiting generalizability. Results do not support the clinical use of any angiogenic biomarker for AD diagnosis at this time. IMPORTANT LIMITATION OF THIS SUMMARY: Only the abstract and partial publisher-page data were available; moderator analyses beyond biomarker type (e.g., by biofluid compartment, AD severity, or assay platform) may have been reported in the full text and could not be incorporated here. __LIMITATIONS__ Extreme heterogeneity (I² = 98.6%) renders the pooled effect size clinically uninterpretable and undermines confidence in any aggregate conclusion.,Only 18 studies with 28 effect sizes were included — a small evidence base that is likely underpowered for robust moderator analyses.,Literature search was restricted to MEDLINE and was conducted through August 2021 only; studies in other databases (EMBASE, Cochrane, PsycINFO) and more recent publications were excluded.,The abstract is truncated; full moderator analyses (e.g., by biofluid type, AD diagnostic criteria, assay platform, or disease severity) could not be assessed from available text.,Included studies likely varied substantially in AD diagnostic criteria (clinical vs. biomarker-confirmed), biofluid compartment (plasma, serum, CSF), analyte assay methodology, and control group definitions — all major sources of heterogeneity that were only partially interrogated.,Cross-sectional design of primary studies precludes any inference about the temporal relationship between angiogenic changes and AD onset or progression.,Very few studies examined fibroblast growth factor or platelet-derived growth factor markers, limiting conclusions about the broader angiogenic pathway.,Publication bias assessment results were not described in the available abstract.,Total sample size across included studies was not reported, preventing assessment of overall statistical power.,Funding source not reported; potential conflicts of interest cannot be evaluated. __BOTTOM_LINE_CATEGORY__ low_certainty __BOTTOM_LINE__ This first systematic review and meta-analysis of biofluid angiogenic markers in Alzheimer's disease found no statistically significant aggregate difference in angiogenic factor levels — including VEGF — between AD patients and healthy older adults (Hedges' g = −0.26, p = 0.647), with near-total heterogeneity (I² = 98.6%) across the 18 included studies. No angiogenic biofluid marker can currently be recommended for clinical use in AD diagnosis. Future well-standardized studies examining fibroblast growth factor, platelet-derived growth factor, and other angiogenic regulators in clearly phenotyped, biomarker-confirmed AD cohorts are needed to determine whether the vascular-angiogenic axis holds diagnostic or mechanistic utility.