The complement system in Alzheimer's disease: evaluating biomarker potential in a complex neuroimmune context.

Abstract

Neuroinflammatory processes are increasingly recognised as important modulators of Alzheimer's disease (AD) progression, driving interest in immune-related biomarkers beyond classical pathological measures. Among these, the complement system has attracted attention because of its interactions with amyloid-β (Aβ) and tau pathology, genetic associations with AD risk, and evidence of activation within affected brain regions. However, these biological observations do not directly translate into straightforward biomarker signals. Complement activity is highly dynamic, spans multiple activation and regulatory states, and may reflect both central and peripheral immune processes. This complexity limits interpretation when complement markers are assessed in isolation, as age, systemic inflammation, vascular comorbidity, and blood-brain barrier integrity can influence measured levels. Current evidence does not support complement-derived biomarkers as stand-alone diagnostic classifiers comparable to established amyloid, tau, and neurodegeneration (AT(N)) measures. Their independent or additive value within multimodal biomarker frameworks remains unclear, partly because of cohort heterogeneity, incomplete assay harmonisation, uncertain tissue-source attribution, and limited longitudinal validation. This review critically evaluates complement-derived measures as biologically informative markers of neuroimmune activity in AD, distinguishing biological plausibility from analytical and clinical utility. We argue that their most defensible current role is within multimodal biomarker frameworks, where they may provide context-specific information on inflammatory state rather than function as independent diagnostic, staging, or treatment-monitoring tools. Progress toward clinical application will require rigorous standardisation, mechanistic clarification, and validation across large, longitudinal, and diverse cohorts.