**Background:** Alzheimer's disease (AD) is the most common form of dementia, contributing to 60%–70% of cases. With approximately 50 million people worldwide affected by dementia and projections of 82 million by 2030 and 152 million by 2050, the global societal cost was estimated at US$818 billion in 2015 (1.1% of global GDP). AD has a long presymptomatic phase, likely lasting decades, between the beginning of neuronal degeneration and first clinical symptoms. Pharmacological and behavioral interventions at the preclinical stage can prolong the time until clinical symptoms appear, creating a need for reliable early diagnostic tools. This narrative review summarizes early diagnostic markers of AD based on both micro and macro levels of pathology, discusses their shortcomings, and proposes a novel biomarker—the hippocampal-to-ventricle ratio (HVR).
**Methods:** This review was based on presenting comprehensive background of early diagnostic markers of AD. The authors compiled markers into micro-level and macro-level categories, discussed advantages and disadvantages of each, and ultimately proposed the ratio of gray matter volume to ventricle volume as a potentially superior marker.
**Key Results:** Micro-level biomarkers include cerebrospinal fluid (CSF) measures of total tau (T-tau), phosphorylated tau (P-tau), and amyloid-beta (Aβ) isoforms. CSF T-tau is associated with acute neuronal injuries reflecting neurodegeneration from days to weeks, while P-tau reflects chronic phosphorylation specific to AD. Decreased CSF Aβ42 strongly indicates imminent AD, and the CSF Aβ42/Aβ40 ratio is more clearly linked to AD than Aβ42 alone. Positron emission tomography (PET) using Pittsburgh compound-B ([11C]-PIB) for amyloid and tau agents, as well as fluorodeoxyglucose (FDG) PET for detecting decreased brain metabolism, are also used. However, CSF biomarkers require lumbar puncture and PET testing is expensive, hindering routine clinical implementation.
Macro-level biomarkers focus on structural MRI, particularly hippocampal volume (HV). The medial temporal lobe (MTL), including the hippocampus, is where AD neuropathology emerges and develops dramatically according to Braak stages. HV is recommended as a supplementary biomarker by the European Federation of the Neurological Societies, the European Medicines Agency, the National Institute on Aging and the Alzheimer's Association, and the International Working Group. Studies show smaller HV is positively correlated with decreased Aβ42 and increased amyloid and tau. The conversion rate from mild cognitive impairment (MCI) to AD ranges from 35% to 50.5% within 3 years, with some suggesting conversion within 1 year lies above 10% (though a meta-analysis suggested less than 10%). A meta-analysis including 27 studies supports HV as a good predictor of MCI-to-AD conversion. However, HV shows large variation among populations due to genetic, developmental, and environmental factors, undermining its validity. A meta-analysis of 33 studies found little support for the "bigger-is-better" hypothesis, with the relationship between HV and memory being positive but weak in older people.
Hippocampal atrophy through longitudinal assessment provides dynamic information but requires repeated assessments sufficiently spaced in time. Ventricular expansion has been shown to correctly indicate presence of AD and MCI, with some studies reporting better performance in measuring AD progress compared with hippocampal atrophy.
The authors propose HVR as a comprehensive biomarker that embraces information from HV, hippocampal atrophy, and ventricular expansion simultaneously while avoiding repeated measurements. HVR takes into account information about HV and relates it to surrounding ventricular enlargement, presenting a dynamic feature from cross-sectional data. Preliminary evidence from Bartos et al. (2019) showed that the ratio of hippocampus and the inferior part of the lateral ventricle allowed better discrimination of AD and control groups than absolute HV. Validation in a preclinical sample showed both age and memory had stronger negative correlations with HVR than with HV alone. A potential downside is the high time cost of manual segmentation (1.5–2 hours per brain), though automated algorithms like MAGeT can reduce segmentation time to approximately 30 minutes.
**Clinical Implications:** The authors argue that HVR could overcome limitations of established structural markers by controlling for individual variations in hippocampal size unrelated to neurodegeneration. If validated in clinical studies with MCI and AD patients, HVR could improve early diagnostic accuracy and potentially shift the sensitivity of structural MRI earlier into the preclinical phase. The ratio approach could be extended to other brain regions affected by AD even earlier than the hippocampus, such as the transentorhinal region. HVR assessment comes at low structural and economic costs with low patient burden, making it highly feasible for routine clinical practice. The authors conclude that ratio measures between atrophied structures and directly adjacent regions that have occupied the atrophied region can be more informative than absolute structural volumes alone, potentially helping with early preclinical diagnosis and intervention to lower individual and societal costs associated with dementia.