Table 1

Summary of potential new structural and functional imaging markers for CAA

Imaging marker	Evidence of potential as a biomarker in CAA	Limitations
MRI visible perivascular spaces in the centrum semiovale (CSO-PVS)	Severe or high-grade CSO-PVS commonly observed in CAA19–22 Pilot data show that, in those with CAA, CSO-PVS severity is associated with Aβ burden (as measured by PiB)23	Non-specific (age-related); present in a number of other conditions24
Cortical atrophy	Thinner cortices in those with sporadic CAA compared with healthy controls; occipital, temporal, posterior parietal and medial frontal areas affected14 (figure 1)	Difficult to differentiate between atrophy secondary to parenchymal Aβ and that due to vascular Aβ in sporadic CAA
Visual functional MRI	Patients with CAA have abnormal BOLD responses to a visual stimulus (alternating checkerboard), with reduced response amplitude and prolonged time both to peak and to baseline15 25 Those with CAA show a decline in BOLD amplitude that is detectable at 1 year; longitudinal difference in BOLD amplitudes was significantly lower in CAA compared with controls26 Potentially of interest as a surrogate marker of vascular health in clinical trials	Clinical implications of this remain unclear; due to technical factors at this stage, this is predominantly a research tool limited to academic medical centres
Network measures	Lower global efficiency of brain network in those with CAA; occipital, parietal and posterior temporal lobes most affected103 Reduced efficiency correlated with Aβ burden (as measured by PiB) and impaired executive function and processing speed103 Global efficiency shows a longitudinal decline with time (1.3 years) in those with CAA, and is associated with deteriorating executive function104	Difficult to differentiate between network effects of parenchymal Aβ versus those due to vascular Aβ in sporadic CAA Mainly a research tool limited to academic medical centres
Amyloid PET imaging using [¹¹C] PiB-PET and [¹⁸F] compounds	In those with CAA, regions with high PiB retention area associated with subsequent haemorrhage27 Although PiB-PET may not reliably distinguish between patients and age-matched controls,28 early phase (1–6 min) uptake can do this29 The occipital/posterior cingulate ratio of PiB uptake is different for those with CAA versus those with AD29 PiB-PET and [¹⁸F] florbetapir binding is able to distinguish between CAA-associated ICH and hypertension-associated ICH30	Amyloid PET unable to differentiate between vascular and parenchymal Aβ Diagnostic accuracy for CAA seems limited Few data on change over time in CAA

Aβ, amyloid-beta; AD, Alzheimer’s disease; BOLD, blood-oxygen level-dependent; CAA, cerebral amyloid angiopathy; ICH, intracerebral haemorrhage; PET, positron emission tomography; PiB, Pittsburgh B compound.