Cerebral amyloid-β-related angiitis and iatrogenic cerebral amyloid angiopathy-related vasculitis: Implications for amyloid-related imaging abnormalities
Year: 2024
Type: article
Abstract: To the Editor: Cerebral amyloid angiopathy (CAA) occurs with age and is particularly common in patients with Alzheimer’s disease (AD). Its prevalence in AD was estimated at 48% according to a recent meta-analysis,1 and was as high as 100% in 1 autopsy study.2 CAA is over-represented in patients with APOE ε4 and ε2 alleles,3 although because of the high frequency of ε3 in the population, homozygous ε3 is the most common APOE genotype in patients with CAA4 and is present in about 30% of patients with moderate or severe CAA.4 CAA cannot be diagnosed definitively during life without tissue examination, but may be suggested by the presence of lobar hemorrhages or microhemorrhages in older individuals.5 A rare complication of CAA is CAA-related inflammation (CAA-RI).6 Its estimated prevalence was 1 in 769 000 according to 1 study.7 The neurological presentation of suspected cases of CAA-RI, broadly defined, includes cognitive decline, focal neurological deficits, encephalopathy, seizures, and headache.6 On magnetic resonance imaging (MRI), T2/FLAIR-hyperintense white matter lesions with lobar cortical microbleeds are characteristic, sometimes accompanied by cortical superficial siderosis and/or macrohemorrhage.6 The inflammatory pathology associated with CAA-RI is variable. Moreover, the pathology termed “CAA-RI” is defined differently in different studies. In one of the larger pathology studies to date, inflammatory pathology was dichotomized into granulomatous, angiodestructive amyloid-β(Aβ)-related angiitis (ABRA) and CAA with perivascular inflammation, the latter abbreviated “CAA-RI” to distinguish from ABRA.8 The authors concluded that diffuse cortical microbleeds on MRI in elderly patients with infiltrative white matter changes or prominent leptomeningeal enhancement was “highly suggestive of vascular inflammatory CAA.” For purposes of clinical and radiographic correlation, however, ABRA and “CAA-RI” were considered together. Clinical features specific to ABRA were therefore not validated. In another validation study, Auriel et al. defined “CAA-RI” as, “evidence of 1 or more CAA-positive vessels with lymphocytes in the vessel wall or perivascular histiocytes (not associated with hemosiderin),” further noting, “macrophages or bland lymphocytic perivascular cuffs associated with CAA-affected vessels were not considered diagnostic for CAA-RI because they are sometimes found in sporadic CAA.”9 Using this definition, probable and possible radiographic criteria for “CAA-RI” were proposed with reasonable sensitivity (82%) and specificity (68%–97%). However, ABRA as defined by Salvarani et al.8 was again not clinically validated. While it seems plausible that ABRA, as defined by Salvarani et al.,8 might behave more aggressively than CAA-RI of either definition, more research is needed to determine specific clinical and radiographic presentations associated with ABRA, that is angiodestructive, granulomatous inflammation associated with CAA. Importantly, and relevant to this correspondence, Scolding et al. demonstrated evidence of Aβ plaque phagocytosis in their series of 10 cases of ABRA, suggesting an immune response to Aβ in general in such cases, both parenchymal and vascular.10 We recently examined the brain biopsy of a 75-year-old woman who presented with new right-sided weakness and was diagnosed clinically with a left occipital hemorrhagic stroke. Initial MRI showed complicated FLAIR hyperintensity in the occipital lobes along with leptomeningeal enhancement. A left parietal craniotomy with open biopsy was performed. Histopathology demonstrated CAA with widespread granulomatous vasculitis, with transmural involvement of the blood vessel walls, although without fibrinoid necrosis (Figure 1). In addition to the granulomatous vasculitis, there were focally robust perivascular lymphoplasmacytic infiltrates, including B and T lymphocytes. Focal phagocytic involvement of amyloid plaque cores was also noted. Overall, the findings were characteristic of ABRA with additional CAA-RI. H&E-stained sections comparing ABRA with CAA-RI (A, C, E) and changes associated with lecanemab treatment (B, D, F). Both cases showed histiocytic infiltration of blood vessels walls with phagocytosis of amyloid (A–D), and multinucleated histiocytes (C, D). CAA-RI showed more pronounced lymphoplasmacytic inflammation in perivascular areas and in the leptomeninges (E). The case with lecanemab treatment showed widespread fibrinoid necrosis of blood vessel walls (F). Scale bars: A—100 µm, B—100 µm, C—50 µm, D—50 µm, E—200 µm, F—50 µm. Panels B, D, F are reprinted from Castellani et al. with permission from IOS Press. We were struck by the histological similarities between this case and our previously reported case of cerebral amyloid vasculitis associated with lecanemab infusion (Figure 1).11,12 Both showed a remarkably similar angiodestructive, granulomatous vasculitis involving blood vessels with CAA as well as focal amyloid plaque core phagocytosis. Subtle differences included lymphoplasmacytic inflammation in the untreated case and more pronounced fibrinoid necrosis of blood vessel walls in the lecanemab-treated case. More research is needed to determine whether the monomorphic inflammation and the widespread fibrinoid vascular necrosis are distinguishing features of exogenous anti-Aβ antibody therapy. Our research group and others have commented on the pathogenic similarities between CAA-RI (broadly defined) and the side effects of anti-Aβ antibody infusion (so-called amyloid-related imaging abnormalities or ARIA).11,13–16 It appears that both endogenous anti-Aβ antibodies in ABRA17 and exogenous anti-Aβ antibodies (eg, lecanemab11,18) target cerebral blood vessels with CAA. It further appears, based on the anecdotal reports, that lecanemab toxicity more closely resembles ABRA as described in Scolding et al.,10 among the inflammatory pathologies described to date. In both instances, there was granulomatous CAA vasculitis with fibrinoid necrosis and amyloid plaque core phagocytosis. Important questions that remain in light of these observations are the extent to which ARIA represents iatrogenic, granulomatous cerebral amyloid vasculitis, and the long-term brain health of individuals exposed to anti-Aβ therapy. None declared. None declared.
DOI: 10.1093/jnen/nlae068
Authors: Rudy J Castellani, Pouya Jamshidi
Institution: Northwestern University
Language: English
Subfield: Clinical Neurology
Field: Medicine
Domain: Health Sciences
Open Access status: bronze
OpenAlex ID: W4400387200