Proposed Section Structure
This topic is best organized around: clinical background and unmet need; study design and methods; imaging, CSF, and genetic results; mechanistic interpretation; strengths and limitations; and implications for diagnosis, counseling, and future research. That structure matches the translational nature of the study, which sits at the interface of cerebrovascular neurology, molecular imaging, and biomarker science.
Highlights
Patients with iatrogenic cerebral amyloid angiopathy, or iCAA, had substantially higher amyloid PET burden than those with sporadic CAA despite being younger overall.
In sporadic CAA, higher Centiloid values tracked with lower CSF Aβ42 and lower Aβ42/Aβ40 ratio, but this expected biomarker coupling was not seen in iCAA.
APOE ε4, a well-established genetic risk factor in amyloid-related brain disease, was markedly less frequent in iCAA than in sporadic CAA.
Taken together, the data support the view that iCAA may reflect a biologically distinct, possibly exogenous, route of cerebral Aβ propagation rather than simply an earlier-onset form of sporadic disease.
Background
Cerebral amyloid angiopathy, or CAA, is characterized by deposition of β-amyloid within small- and medium-sized cortical and leptomeningeal vessel walls. Clinically, it is a major cause of lobar intracerebral hemorrhage, convexity subarachnoid hemorrhage, transient focal neurologic episodes, and progressive cognitive impairment in older adults. In routine practice, probable sporadic CAA is generally diagnosed using clinicoradiologic criteria, increasingly refined by MRI markers such as lobar microbleeds, cortical superficial siderosis, and white matter injury patterns.
Iatrogenic cerebral amyloid angiopathy is much rarer. It has been proposed to arise through transmission of Aβ seeds during neurosurgical procedures, particularly after exposure to cadaveric dura mater grafts or other contaminated materials used historically. This concept is biologically plausible and parallels broader ideas about templated protein misfolding, although the extent to which iCAA overlaps with or diverges from sporadic CAA at the biomarker level has remained uncertain.
That gap matters for several reasons. First, iCAA tends to present at a younger age than sporadic CAA, so diagnostic delay can be substantial if clinicians anchor too strongly on age-based expectations. Second, if iCAA follows a distinct pathogenic route, commonly used biomarkers may perform differently, with implications for diagnosis and mechanistic inference. Third, clarifying whether APOE-associated susceptibility is shared or not shared across iCAA and sporadic CAA may improve understanding of how host genetics interact with exogenous amyloid seeding.
The study by Storti and colleagues addresses these questions using three complementary readouts: amyloid PET quantified on the Centiloid scale, CSF measures of amyloid and tau, and APOE genotype.
Study Design and Methods
This was a single-center, cross-sectional study including adults who fulfilled diagnostic criteria for probable sporadic CAA or iCAA and who underwent both amyloid PET and CSF assessment between 2021 and 2024. The cohort comprised 95 patients in total: 24 with iCAA and 71 with sporadic CAA.
The age contrast between groups was notable and clinically intuitive. Median age was 56.5 years in iCAA versus 70 years in sporadic CAA. Because age and sex can influence biomarker expression and disease phenotype, the investigators used multivariable regression models adjusting for both variables when examining associations with amyloid PET burden.
Amyloid imaging used either 18F-flutemetamol or 18F-florbetaben PET, harmonized through Centiloid scaling. That choice is important because the Centiloid framework improves comparability across tracers and studies by converting PET signal to a standardized quantitative metric. CSF biomarkers were measured with Lumipulse assays and included Aβ42, Aβ40, the Aβ42/Aβ40 ratio, phosphorylated tau at threonine 181, and total tau.
The primary analytic question was whether amyloid PET burden differed between iCAA and sporadic CAA. Secondary questions included how PET burden related to CSF biomarkers within each group and whether APOE genotype distributions differed.
Key Findings
Amyloid PET burden was higher in iatrogenic CAA
The headline result was a significantly higher amyloid PET burden in iCAA than in sporadic CAA. Median Centiloid values were 57.3 in iCAA versus 30.5 in sporadic CAA, with a p value of 0.0026. This is not a subtle difference. In absolute terms, it suggests substantially greater fibrillar amyloid signal among patients with presumed iatrogenic disease.
The multivariable model reinforced the robustness of the finding. After adjustment for age and sex, iCAA diagnosis remained independently associated with an increase of 34.99 Centiloid units, with a 95% confidence interval of 15.25 to 54.74 and a p value of 0.001. In other words, the higher amyloid burden in iCAA was not simply explained by demographic imbalances between groups.
From a clinical standpoint, this is perhaps the most provocative observation in the paper. One might have expected younger patients with iCAA to have less cumulative amyloid deposition than older patients with sporadic CAA. Instead, the reverse was observed, favoring the interpretation that exogenous seeding may trigger efficient or accelerated amyloid propagation in susceptible cerebral vasculature.
CSF-PET relationships differed by disease subtype
In sporadic CAA, Centiloid values correlated inversely with CSF Aβ42 and with the Aβ42/Aβ40 ratio. That pattern is broadly consistent with established amyloid biology: as amyloid accumulates in brain tissue, soluble CSF Aβ42 falls, and the ratio to Aβ40 also declines. This PET-CSF coupling lends internal validity to the sporadic CAA biomarker signal.
By contrast, no significant correlations between Centiloid values and CSF amyloid markers were observed in iCAA. This dissociation is one of the most interesting translational findings in the study. It implies that the usual relationship between deposited fibrillar amyloid and soluble CSF amyloid may not hold, or may hold less reliably, in iCAA.
Several interpretations are possible. The simplest is that iCAA represents a different spatiotemporal pattern of Aβ propagation, perhaps more heavily vascular or regionally heterogeneous, with CSF biomarkers that are less tightly linked to global cortical PET signal. Another possibility is that once exogenous seeding initiates vascular amyloid deposition, downstream CSF changes plateau or become uncoupled from PET-detectable fibrillar accumulation. A third explanation is methodological: with only 24 iCAA cases, power to detect correlations was limited. Still, the contrast with sporadic CAA is directionally important.
APOE ε4 was much less common in iatrogenic CAA
APOE ε4 was markedly less frequent in iCAA than in sporadic CAA: 4.5% versus 34.2%. This divergence strongly supports the idea that iCAA is not merely sporadic CAA occurring earlier in life. In sporadic amyloid disorders, APOE ε4 is associated with amyloid aggregation and deposition, including in Alzheimer disease and, to varying extents, CAA-related phenotypes. The low ε4 frequency in iCAA suggests that strong host genetic predisposition may be less necessary when the initiating event is exogenous Aβ exposure.
This does not mean genetics are irrelevant in iCAA. Rather, it shifts the framing from inherited susceptibility as a primary driver to host environment and exposure history as potentially dominant determinants. For clinicians, it also means that absence of APOE ε4 should not be falsely reassuring when iCAA is in the differential diagnosis.
How should tau measures be interpreted?
The abstract specifies that CSF p-tau181 and total tau were measured, but the principal signal emphasized in the report centers on amyloid markers and APOE genotype. Without stronger result details in the abstract, tau findings should be interpreted cautiously. In CAA research more broadly, tau can reflect coexisting neurodegeneration, Alzheimer copathology, or nonspecific neuronal injury, rather than vascular amyloid alone. The fact that the paper foregrounds Aβ markers and PET burden suggests these were the most informative discriminators between iCAA and sporadic CAA.
Mechanistic Interpretation
The study supports a divergent pathobiology between iCAA and sporadic CAA. The combination of higher PET amyloid burden, weaker CSF-PET correlation, and lower APOE ε4 enrichment is difficult to reconcile with a simple age-shifted version of the same disease process.
A plausible model is that iCAA begins with exogenous inoculation of Aβ seeds during neurosurgical exposure. Those seeds may preferentially propagate along vascular or perivascular pathways and subsequently recruit endogenous Aβ into deposited aggregates. If true, PET may capture a high fibrillar burden while CSF soluble Aβ levels reflect a more complex and less linear biology than in spontaneous disease. This model aligns conceptually with experimental work showing seeded Aβ aggregation and with prior case series linking iCAA to historical neurosurgical materials.
At the same time, caution is warranted. PET tracers detect fibrillar amyloid but do not directly distinguish vascular from parenchymal plaque deposition. Many patients with CAA, especially older ones, also harbor concomitant Alzheimer pathology. Thus, increased Centiloid burden in iCAA could reflect vascular amyloid, parenchymal amyloid, or both. Even so, the between-group difference remains clinically meaningful because the overall amyloid phenotype appears more intense in iCAA despite less APOE ε4 enrichment.
Clinical Relevance
For vascular neurologists and memory specialists, the findings have several practical implications.
First, younger age should not reduce suspicion for CAA when hemorrhagic or MRI features are suggestive and there is a compatible remote neurosurgical history. In fact, pronounced amyloid PET positivity in a relatively young patient may strengthen consideration of iCAA rather than argue against it.
Second, CSF amyloid biomarkers may need more nuanced interpretation in iCAA. In sporadic CAA, lower CSF Aβ42 and lower Aβ42/Aβ40 ratio alongside higher amyloid PET fit the expected model. In iCAA, absence of this relationship means that a nonclassic CSF profile should not automatically exclude substantial cerebral amyloid deposition.
Third, APOE genotyping, while not a diagnostic test on its own, may contribute to mechanistic framing. A low prevalence of ε4 in iCAA suggests that conventional genetic risk architecture differs from sporadic disease. For counseling and case adjudication, this can be informative when exposure history is uncertain or incomplete.
Fourth, the study reinforces the importance of harmonized biomarker programs in rare cerebrovascular disorders. Standardized PET quantification using Centiloids and automated CSF platforms such as Lumipulse make cross-cohort comparisons increasingly feasible. That is essential in diseases like iCAA, where single-center numbers remain modest.
Strengths and Limitations
Strengths
The work integrates imaging, fluid biomarkers, and genetics in the same cohort, which is a major strength. It also uses standardized quantitative PET methodology and adjusts key analyses for age and sex. Importantly, it addresses a clinically important but understudied form of CAA for which mechanistic data are sparse.
Limitations
The study is cross-sectional, so it cannot determine temporal sequence or rate of amyloid accumulation. The sample size, particularly for iCAA, is small, which limits precision and may obscure within-group biomarker relationships. Single-center recruitment raises the possibility of referral bias, especially if the center has specific expertise in young-onset cerebrovascular amyloid disease.
Another limitation is that amyloid PET cannot cleanly separate vascular amyloid from neuritic plaque burden. This is especially relevant because some patients may have mixed pathology. The abstract also does not provide detailed distributions for tau biomarkers, cognitive status, hemorrhagic burden, or exact exposure histories, all of which could further refine interpretation. Finally, as with any observational biomarker study, residual confounding cannot be excluded.
How This Fits With the Existing Literature
The concept of iCAA has gained traction over the past decade through case reports, neuropathologic studies, and epidemiologic observations linking early-onset CAA to prior neurosurgical exposure. This study advances the field by moving beyond descriptive case narratives to a more systematic biomarker comparison with sporadic CAA.
Its main contribution is not simply that iCAA exists, but that it appears biomarker-distinct. That matters because it supports the broader principle that amyloid-related cerebrovascular disease can arise through more than one initiating pathway. The findings also resonate with ongoing efforts to disentangle vascular amyloid from Alzheimer-type amyloid, even though current PET tools cannot fully resolve that distinction in vivo.
Implications for Research and Practice
Several next steps follow naturally. Prospective multicenter studies are needed to validate the higher Centiloid burden in iCAA and determine whether it predicts hemorrhage risk, cognitive decline, or inflammatory complications. Longitudinal biomarker studies would be especially valuable to test whether PET signal rises faster in iCAA or whether CSF markers plateau early.
Future work should also integrate detailed MRI phenotyping, including lobar microbleed burden, cortical superficial siderosis, enlarged perivascular spaces, and white matter markers. If biomarker signatures map onto specific radiologic patterns, clinicians could gain a more practical framework for distinguishing iCAA from sporadic CAA during routine workup.
There is also a need for pathology-correlated imaging studies. These could help determine how much of the PET signal in iCAA is attributable to vascular versus parenchymal amyloid, and whether tracer retention patterns differ regionally from sporadic CAA or Alzheimer disease. Finally, harmonized registries of suspected iCAA cases could improve ascertainment of exposure history and clarify incidence in contemporary neurosurgical practice.
Conclusion
This Neurology study provides persuasive evidence that iatrogenic cerebral amyloid angiopathy is biomarker-distinct from sporadic CAA. Compared with sporadic CAA, iCAA was associated with substantially higher amyloid PET burden, absence of the expected inverse relationship between PET signal and CSF amyloid markers, and a strikingly lower frequency of APOE ε4. Collectively, these findings support a divergent, plausibly exogenous mechanism of amyloid propagation rather than a simple younger-onset variant of sporadic disease.
For clinicians, the message is practical: in younger patients with hemorrhagic or MRI features of CAA and a possible remote neurosurgical exposure, strong amyloid PET positivity should heighten suspicion for iCAA, even when CSF or genetic findings do not fit familiar sporadic patterns. For researchers, the study opens an important path toward more precise biologic classification of cerebral amyloid disorders.
Funding and Trial Registration
The study was linked to SENECA, ClinicalTrials.gov Identifier: NCT04204642, submitted on December 19, 2019. The abstract provided here does not specify additional funding details.
Reference
Storti B, Capozza A, Marinoni G, Strazzabosco C, Stanziano M, Rifino N, Boncoraglio GB, Scala I, Romoli M, Paccagnella A, Canavero I, Tagliabue L, Bersano A. Amyloid PET Burden, CSF Biomarkers, and APOE Genotype in People With Iatrogenic and Sporadic Cerebral Amyloid Angiopathy. Neurology. 2026-05-05;106(10):e214949. PMID: 42085649. https://pubmed.ncbi.nlm.nih.gov/42085649/