Stroke caused by arteries that are diseased for reasons other than atherosclerosis has long resisted tidy classification. A new narrative review in Frontiers in Neurology proposes a way out of the tangle: instead of sorting these disorders by vessel size, angiographic appearance, or presumed cause, it groups them by the dominant biological process driving the vascular phenotype. The framework, developed by neurologists at the All India Institute of Medical Sciences in New Delhi, divides non-atherosclerotic vasculopathies, or NAV, into three compartments: structural disorders of arterial architecture, functional disorders of vascular tone and autoregulation, and inflammatory disorders driven by immune- or infection-mediated vascular injury.
The authors argue that the clinical question should not simply be what an angiogram resembles, but which biological process is most likely driving the picture and what should be tested or treated next. A structural hypothesis prioritises characterisation of mural architecture, dissection, collateral formation, systemic vascular involvement, and inherited arteriopathy. A functional hypothesis prioritises trigger identification, serial vascular imaging, and demonstration of reversibility. An inflammatory hypothesis prioritises cerebrospinal fluid examination, infectious and immune investigations, vessel wall imaging, and, in selected cases, biopsy.
The review defines NAV as cerebrovascular disorders in which a non-atherosclerotic abnormality of arterial structure, vascular tone or autoregulation, or immune- or infection-mediated vascular injury is central to the neurological phenotype. While atherosclerosis remains a leading cause of stroke worldwide, NAV are an important and distinct cause in young adults, defined in different studies as those under 45 to 50 years of age. Cervical artery dissection is the most common, accounting for approximately 20 to 25 percent of cases. Compared with atherosclerotic disease, NAV tend to occur in younger patients, are associated with fewer traditional vascular risk factors, and show lower markers of systemic atherosclerosis such as carotid intima-media thickness and homocysteine levels. The scope is not restricted by age, however: cerebral amyloid angiopathy predominantly affects older adults but remains relevant because the primary vascular pathology is non-atherosclerotic.
Existing classification systems, the authors note, have real limitations. The same disorder can fall into multiple categories depending on the framework used; moyamoya disease, for example, may be viewed as a genetic arteriopathy, a large-vessel steno-occlusive disorder, or an angiographic syndrome. A recurring diagnostic challenge is misclassification as atherosclerotic disease when interpretation relies primarily on luminal stenosis. Isolated middle cerebral artery disease may represent early or incomplete forms of moyamoya disease in younger patients. Many NAV are also dynamic processes, with temporal evolution such as the resolution of vasoconstriction in reversible cerebral vasoconstriction syndrome or the progression and healing of arterial dissection, which static classification systems fail to capture. Primary central nervous system vasculitis and RCVS may both present with multifocal arterial narrowing despite markedly different underlying biology.
Structural vasculopathies arise from intrinsic abnormalities of arterial wall architecture that compromise vascular integrity, producing stenosis, dissection, aneurysm formation, or altered remodelling. Susceptibility genes including RNF213, NOTCH3, COL4A1, ACTA2, CTSA, KRIT1, and TREX1 have been implicated in arterial wall pathology. Histopathology reinforces the concept of vessel wall vulnerability: cervical artery dissection is characterised by intramural hematoma, fibromuscular dysplasia by abnormal smooth muscle cell differentiation and medial fibroplasia, and CADASIL by vascular smooth muscle degeneration with granular osmiophilic material deposition. Arterial dissection may arise from an intimal disruption that permits luminal blood to enter the arterial wall, or from primary intramural haemorrhage potentially originating from the vasa vasorum. In many patients, cervical dissections follow trivial mechanical triggers such as minor neck rotation, coughing, or chiropractic manipulation, highlighting that underlying arterial wall fragility is central to pathogenesis.
The review assembles detailed cohort evidence on the relationship between fibromuscular dysplasia and dissection. In the international STOP-CAD cohort of 3,714 patients with spontaneous cervical artery dissection, FMD was identified in 196 patients, or 5.3 percent. Patients with FMD were less likely to have an isolated vertebral artery dissection, and FMD was associated with a higher 24-month risk of recurrent cervical artery dissection, 7.7 percent versus 2.8 percent, with an adjusted hazard ratio of 2.75. Conversely, among patients in the United States Registry for FMD, cervical artery dissection was a presenting manifestation in 12.1 percent, a different denominator from the prevalence of FMD among patients presenting with dissection. Carotid webs, thin shelf-like projections from the posterior wall of the carotid bulb, are increasingly implicated in cryptogenic ischaemic stroke, accounting for up to 13 percent of such cases in selected cohorts, with local flow disturbance and stasis promoting thrombus formation and artery-to-artery embolism.
Imaging is central to the framework, particularly high-resolution vessel wall magnetic resonance imaging, which enables direct visualisation of arterial wall pathology beyond luminal appearance. In dissection, fat-suppressed T1-weighted imaging can demonstrate a crescentic intramural hematoma, and classic luminal signs such as intimal flap or double lumen may be absent, especially intracranially. FMD typically shows a string-of-beads appearance and predominantly affects mid-to-distal vessel segments, in contrast to the proximal segments typical of atherosclerosis. Carotid webs require careful evaluation on sagittal and oblique CTA reconstructions, as axial images alone may miss the shelf-like projection. The authors caution, however, that vessel wall enhancement and thickening are not mechanism-specific: similar patterns may occur with atherosclerosis, dissection, moyamoya angiopathy, infection, radiation injury, recent infarction, or vascular intervention, and interpretation requires clinical correlation.
The functional compartment is exemplified by RCVS, characterised by recurrent thunderclap headaches over several days, often preceding neurological complications and offering a diagnostic window. Identifiable triggers, present in 50 to 60 percent of cases, include vasoactive drugs, the postpartum state, and illicit substances. In a contemporary systematic review of unselected RCVS cohorts, the pooled frequency of ischaemic stroke was 15.9 percent and that of intracranial haemorrhage was 22.1 percent, though frequencies varied widely across studies. Vasoconstriction usually peaks within one to two weeks and resolves over one to three months, making serial imaging more informative than a single static study. Distinguishing RCVS from primary central nervous system vasculitis is a key diagnostic challenge, since misdiagnosis may lead to inappropriate immunosuppressive therapy. The RCVS2 score has been proposed as a clinical tool for this distinction. On management, the review notes that in a prospective observational cohort of 82 patients in which all participants received nimodipine, 71 patients experienced remission of thunderclap headache after treatment initiation, but the absence of an untreated comparator precludes a causal inference, and there is no established evidence that nimodipine prevents ischaemic stroke or intracranial haemorrhage. In a retrospective cohort of 162 patients, glucocorticoid exposure was independently associated with clinical, radiological, and angiographic worsening, although the observational design does not establish causality.
On treatment, the review reports that an individual-patient-data meta-analysis of randomised trials involving 444 patients with arterial dissection found no significant difference between anticoagulation and antiplatelet therapy for the 90-day composite outcome of ischaemic stroke, death, or major bleeding, and the European Stroke Organisation guideline considers either strategy acceptable. For symptomatic carotid webs, antithrombotic therapy alone has been associated with a substantial risk of recurrent ischaemic stroke; in one cohort, 17 percent experienced recurrent stroke within two years despite predominantly antiplatelet-based management, while a systematic review of 158 patients reported no recurrent strokes among the 72 percent who underwent carotid revascularisation, and a meta-analysis found a substantially lower recurrence risk with carotid endarterectomy or stenting than with medical therapy alone. The available evidence is derived predominantly from observational studies, and randomised trials comparing medical therapy, endarterectomy, and stenting are lacking. In inherited arteriopathies such as CADASIL and COL4A1-related disorders, management remains largely supportive, and disease-modifying therapies targeting the underlying arterial wall pathology are currently lacking.
The authors emphasise that the compartments are neither mutually exclusive nor fixed disease labels, and that the framework should be regarded as a clinical reasoning heuristic rather than a formally validated diagnostic taxonomy. Sickle cell disease, DADA2, radiation vasculopathy, infection-associated arteriopathy, and moyamoya syndromes illustrate context-dependent assignment, since an initial endothelial or inflammatory injury may subsequently produce fixed fibrosis, remodelling, and structural stenosis. Cerebral amyloid angiopathy-related inflammation and amyloid-beta-related angiitis represent structural-inflammatory overlap syndromes that may resemble primary central nervous system vasculitis but are potentially responsive to immunosuppressive treatment. Because this was a narrative review rather than a formal systematic review, no formal evidence-grading framework was applied, and much of the available evidence is observational. The authors identify key gaps, including the need for prospective multicentre registries with harmonised phenotyping, validation of biomarker panels, pragmatic trials in RCVS and PCNSV, and standardisation of vessel wall imaging acquisition and reporting. The framework is intended to complement, rather than replace, existing aetiological and imaging-based classifications, by linking vascular phenotypes to mechanism-directed diagnostic testing and initial management.
Subject of Research: A pathobiology-based framework for classifying non-atherosclerotic vasculopathies that cause stroke
Article Title: Non-atherosclerotic vasculopathies in stroke: a pathobiology-based framework
Article References: Makharia, A., Garg, D., Srivastava, A. K., & Agarwal, A. (2026). Non-atherosclerotic vasculopathies in stroke: a pathobiology-based framework. Frontiers in Neurology, 17, Article 1908716. https://doi.org/10.3389/fneur.2026.1908716
Image Credits: AI Generated
DOI: 10.3389/fneur.2026.1908716
Keywords: stroke, non-atherosclerotic vasculopathy, vessel wall imaging, RCVS, vasculitis, arterial dissection, moyamoya, fibromuscular dysplasia, carotid web, young stroke, narrative review, diagnosis
