Iatrogenic Cerebral Amyloid Angiopathy-Related Inflammation

By Samuel S. Bruce, MD

Synopsis: In this multicenter, retrospective case series, the authors sought to describe the co-occurrence of two rare variants of cerebral amyloid angiopathy: iatrogenic cerebral amyloid angiopathy, thought to be caused by prion-like spread of amyloid beta after seeding from neurosurgical procedures decades prior to symptom onset, and cerebral amyloid angiopathy-related inflammation, characterized by robust perivascular inflammatory response in amyloid-laden vessels. This series demonstrates that inflammation can complicate iatrogenic cerebral amyloid angiopathy, supporting a spectrum model of disease and urging multicenter study.

Source: Storti B, Negro G, Orsani G, et al. Iatrogenic cerebral amyloid angiopathy-related inflammation: A multicenter case series. Neurol Neuroimmunol Neuroinflamm. 2025;12(6):e200470.

Cerebral amyloid angiopathy (CAA) is a vasculopathy caused by deposition of amyloid beta (Aβ) in the cortical and leptomeningeal small vessels of the brain. There is increasing awareness of the clinical diversity within CAA, and the need for new knowledge regarding variants with distinct clinical, radiographic, and epidemiological features. Two rare variants have attracted recent attention: iatrogenic CAA (iCAA), thought to be caused by prion-like seeding and spread of Aβ after neurosurgical exposure to cadaveric dura decades prior to symptom onset, and CAA-related inflammation (CAA-ri), caused by an overactive immune response to vascular Aβ. There are limited data on the co-occurrence of these two variants, both of which can present at a younger age than the more common sporadic form of CAA that affects older individuals.

In this multicenter, retrospective case series, the authors screened two existing Italian databases of CAA patients (a total of 433 cases) to identify patients with iCAA who developed CAA-ri. Previously published diagnostic criteria for iCAA and CAA-ri that allow for in vivo classification of “possible” and “probable” diagnoses without histopathological confirmation were used to ascertain both conditions. Among 38 iCAA cases, six (15.8%) subsequently developed CAA-ri. Of these six, three met diagnostic criteria for probable iCAA, and three met diagnostic criteria for possible iCAA; five met diagnostic criteria for probable CAA-ri, and one met criteria for possible CAA-ri. Neurosurgery occurred at a mean age of 17.2 years (range, < 1 to 43 years), and CAA-ri presented at a mean age of 61.8 years (range, 48 to 79 years), for an average latency of 44.7 years. In three of the six patients, CAA-ri was the initial presentation of CAA.

All patients had hemorrhagic signatures of CAA, including superficial siderosis and lobar microbleeds, as well as radiographic evidence of inflammation. One patient had only subtle sulcal enhancement, while all the others had classic findings of confluent asymmetric white matter edema. Two patients additionally experienced intracerebral hemorrhage. The clinical manifestations of CAA-ri were variable, with four patients experiencing seizures, three experiencing fluctuations of consciousness, and all experiencing focal motor or sensory deficits of some sort. All patients received corticosteroids, with three patients having no appreciable clinical response and ultimately dying during follow-up. Two of the three patients with good response to corticosteroids were asymptomatic with no deficits at last follow-up, while the third had mild residual deficits.

Commentary

Rigorous study of both iCAA and CAA-ri thus far has proven difficult because of their rarity. Thus, the compilation of multiple cases with both subtypes is remarkable. Although our ability to draw firm conclusions from these cases is limited because of the small sample and retrospective nature of the data, this series still significantly increases our knowledge about these conditions. The nontrivial prevalence of inflammation in iCAA may indicate that CAA-ri is not itself a discrete phenotype of CAA, but a pathophysiological process that can be present in all forms of CAA, and may be more widespread than previously thought.

The disease courses described in this series offer some suggestion that inflammation in the setting of iCAA may be especially challenging. Whereas CAA-ri typically is responsive to corticosteroids and immunosuppressive treatments, three of the six patients included in this series had no clear improvement with corticosteroids and died during follow-up and two died within months of symptom onset. Symptomatic seizures also provided a frequent challenge in these patients, with four patients experiencing seizures as a symptom of CAA-ri and one additional patient experiencing seizures as the initial clinical manifestation of iCAA.

In summary, this case series provides early evidence that iCAA and CAA-ri may not necessarily be distinctive phenotypes of CAA but instead represent two separate pathophysiological processes that can occur simultaneously within a clinical spectrum. This case series represents an important first step in elucidating the interplay of these two fascinating entities and argues strongly for multicenter collaboration to prospectively study iCAA and CAA-ri and optimize our approaches to their diagnosis and management.

Samuel S. Bruce, MD, is Assistant Professor of Neurology, Weill Cornell Medicine.