Neurology · Medical Literature
A systematic review and meta-analysis of metalloproteinases in patients with Alzheimer's disease and mild cognitive impairment.
Abstract
BackgroundVarious studies have reported altered expression of metalloproteinases in Alzheimer's disease (AD); however, expression profiles of each metalloproteinase during cognitive decline have not yet been fully characterized.ObjectiveThe purpose of this systematic review was to generate a comprehensive overview of metalloproteinases and their cognate inhibitors expression in AD and mild cognitive impairment (MCI), across sample matrices, to determine whether metalloproteinases are dysregulated in AD and may have predictive power in individuals with cognitive decline.MethodsAn electronic literature search was conducted in PubMed, EMBASE, Scopus and MEDLINE from inception to December 2024. Sixty-one publications reporting metalloproteinase and inhibitor levels in 8576 patients with AD or MCI, and 7333 controls were included in the systematic review, twenty-one of which were extracted for meta-analysis. Standardized mean difference (SMD) was used to illustrate comparisons, and the Newcastle-Ottawa scale to assess bias.ResultsHigher levels of cerebrospinal fluid (CSF) tissue inhibitor of metalloproteinase-2 (TIMP-2; p = 0.0003) were observed in the AD group and in patients with MCI (p = 0.0009) compared to cognitively healthy controls. Following sensitivity analysis, significantly higher levels of CSF MMP-10 (p = 0.0005) and lower plasma TIMP-2 (p = 0.004) were also noted in patients with AD. TIMP-3, across all sample matrices, was decreased in patients with MCI versus controls (p = 0.01).ConclusionsSignificantly altered levels of metalloproteinases and their inhibitors were verified between patients with AD and MCI, representing potential biomarkers and prospective therapeutic targets for cognitive decline. This study was registered with PROSPERO, CRD42024628202.
Clinical relevance
Are metalloproteinases and their cognate inhibitors (TIMPs) differentially expressed in patients with Alzheimer's disease or mild cognitive impairment compared to cognitively healthy controls, and do they hold promise as biomarkers or therapeutic targets for cognitive decline? __SNAPSHOT__ {"studyDesign":"Systematic review with meta-analysis (PROSPERO-registered: CRD42024628202)","setting":"Not reported (multi-national aggregate; literature search across PubMed, EMBASE, Scopus, and MEDLINE from inception to December 2024)","sampleSize":"Systematic review: 61 publications; n=15,909 total (AD/MCI patients: n=8,576; controls: n=7,333). Meta-analysis subset: 21 publications","population":"Adults with clinically or research-diagnosed Alzheimer's disease (AD) or mild cognitive impairment (MCI), compared to cognitively healthy controls; precise inclusion/exclusion criteria not reported in abstract","intervention":"Measurement of metalloproteinase (MMP) and tissue inhibitor of metalloproteinase (TIMP) levels across biological sample matrices (cerebrospinal fluid, plasma, and others)","comparator":"Cognitively healthy controls","primaryOutcome":"Standardized mean difference (SMD) in metalloproteinase and TIMP expression levels between AD/MCI patients and controls across sample matrices","followUp":"Not reported (cross-sectional study-level comparisons; no longitudinal follow-up specified)","fundingSource":"Not reported"}
Key findings
The meta-analysis identified statistically significant dysregulation of specific metalloproteinase pathway analytes across both AD and MCI populations. CSF TIMP-2 was significantly elevated in both the AD group (p=0.0003) and in patients with MCI (p=0.0009) relative to cognitively healthy controls. Following sensitivity analysis — indicating these findings were not fully robust in the primary analysis — significantly higher CSF MMP-10 (p=0.0005) and lower plasma TIMP-2 (p=0.004) were observed in patients with AD. TIMP-3 was significantly decreased across all sample matrices in patients with MCI versus controls (p=0.01). Absolute effect sizes (SMDs with 95% confidence intervals), NNT/NNH, and heterogeneity statistics (I²) are not reported in the available abstract. The divergence in TIMP-2 directionality between CSF (elevated) and plasma (decreased) in AD represents a notable matrix-specific finding that warrants independent replication. Adverse events are not applicable to this study design.
Practice implications
This systematic review and meta-analysis, the largest of its kind aggregating data from 15,909 individuals, provides the most comprehensive characterization to date of metalloproteinase and TIMP expression profiles across the AD-MCI continuum and across biological matrices. The findings suggest that TIMP-2 (CSF elevation) may be a consistent signal in both AD and MCI, while CSF MMP-10 elevation and plasma TIMP-2 reduction appear more specific to AD, though both emerged only after sensitivity analysis and thus should be interpreted cautiously. TIMP-3 reduction in MCI across matrices adds a potentially early signal of interest. However, because only an abstract is available for this review, critical meta-analytic parameters — including effect sizes with confidence intervals, measures of heterogeneity (I²), publication bias assessment (e.g., funnel plot/Egger test), and subgroup definitions — cannot be evaluated. The clinical utility of these analytes as diagnostic or prognostic biomarkers is not established by this study; the review was not designed to assess sensitivity, specificity, or predictive accuracy thresholds. Patients included span across multiple studies with likely heterogeneous diagnostic criteria, demographics, and disease stages. Findings are not actionable for individual patient care at this time. __LIMITATIONS__ Only the abstract was available for appraisal; full meta-analytic statistics (effect sizes, 95% CIs, I², publication bias) cannot be independently assessed.,High expected between-study heterogeneity due to pooling across diverse diagnostic criteria, disease stages, sample matrices, assay platforms, and study populations.,Several key findings (CSF MMP-10, plasma TIMP-2) emerged only after sensitivity analysis, suggesting instability in the primary pooled estimates and raising concerns about multiplicity without correction.,Cross-sectional study-level design precludes any inference about longitudinal biomarker trajectories, conversion risk, or causal relationships.,The Newcastle-Ottawa Scale, used for bias assessment, has known limitations in grading observational studies and does not fully capture assay variability or pre-analytical confounders.,No data on sensitivity, specificity, area under the ROC curve, or predictive accuracy; biomarker diagnostic utility cannot be determined.,Matrix-specific contradictions (CSF TIMP-2 elevated vs. plasma TIMP-2 decreased in AD) could reflect true biology or could represent analytical, pre-analytical, or population heterogeneity artifacts.,Funding source not reported, limiting assessment of potential conflicts of interest.,Generalizability is constrained by the demographic profiles of the constituent studies, which are not characterizable from the abstract alone. __BOTTOM_LINE_CATEGORY__ promising_but_preliminary __BOTTOM_LINE__ This large systematic review and meta-analysis of 15,909 individuals identifies statistically significant dysregulation of CSF TIMP-2 (elevated in both AD and MCI), CSF MMP-10 (elevated in AD), plasma TIMP-2 (decreased in AD), and TIMP-3 (decreased across matrices in MCI), supporting biological plausibility for metalloproteinase involvement in cognitive decline. These findings position specific metalloproteinases and TIMPs as candidate biomarkers and therapeutic targets, but the absence of diagnostic accuracy metrics, unstable sensitivity-analysis-dependent results, and uncharacterizable between-study heterogeneity mean that no change in clinical biomarker practice is currently warranted. Prospective longitudinal studies with standardized assay platforms and predefined diagnostic accuracy thresholds are required before these analytes can be considered for clinical translation.