BAD: a challenging diagnosis on conventional imaging
The concept of branch atheromatous disease, described by Caplan in 1989 [1], remains largely underdiagnosed in routine practice. The culprit plaque lies at the very origin of the perforating artery—too small to be resolved by conventional MR or CT angiography—so these examinations are most often reported as normal, wrongly suggesting simple small vessel lipohyalinosis. In our first case, initial MR angiography showed no significant stenosis, and it was the clinical course—early neurological deterioration despite appropriately administered dual antiplatelet therapy—that raised the hypothesis of BAD. In the second case, MR angiography and transcranial Doppler showed a focal M1 stenosis; HR-VWI nevertheless established its atheromatous nature, characterized its activity (contrast enhancement), and linked it to the infarcted lenticulostriate territory. It should be emphasized that, in this case, the presence of significant stenosis of the parent artery brings the presentation closer to a large-artery atherosclerotic mechanism with involvement of the perforator origin than to BAD in the strict sense, depending on the definitions used; the boundary between these entities remains debated.

Contribution of high-resolution vessel wall imaging (HR-VWI)
High-resolution vessel wall imaging overcomes this limitation by directly imaging the wall of the parent artery rather than its lumen alone. Khan et al. recently showed that HR-VWI could differentiate BAD from small vessel disease by demonstrating a mural plaque abutting the origin of the perforator, with contrast enhancement observed in approximately half of patients with confirmed BAD [2]. Such mural enhancement is considered a marker of inflammatory activity and plaque vulnerability: Zhao et al. showed that it was an independent predictor of ischemic recurrence in patients with intracranial large-artery atherosclerosis [3]. In our second case, the marked enhancement of the left M1 plaque is consistent with these data and, had this information been available earlier, might have strengthened the rationale for more intensive monitoring and treatment from the initial phase.

In Cases 1 and 2, HR-VWI confirmation was obtained only after the acute episode (approximately two and seven months after the index event, respectively) and therefore did not directly guide therapeutic decision-making in the hyperacute phase, which relied solely on the clinical evidence and conventional imaging. It did, however, retrospectively confirm the suspected atherothrombotic mechanism, allow reclassification of the etiology according to TOAST, and prompt corresponding adjustment of the intensity of secondary prevention. This argues for earlier integration of HR-VWI into the etiological work-up of deep infarcts with a clinical course suggestive of BAD, wherever the technique is accessible, rather than its use as a delayed second-line examination.

Our third case illustrates another contribution of HR-VWI: in the setting of multiple vertebrobasilar infarcts, the demonstration of multifocal stenoses (right V4, left P2) points toward diffuse intracranial atherosclerosis of the posterior circulation, which may influence the intensity of secondary prevention. This case, which does not represent BAD, highlights the value of the technique in distinguishing the mechanisms underlying multifocal infarcts.

Early neurological deterioration and the role of tirofiban
BAD is associated with a high risk of early neurological deterioration, estimated at up to 40% of cases, because oral dual antiplatelet therapy does not effectively block the final common pathway of platelet aggregation (fibrinogen binding to the glycoprotein IIb/IIIa receptor) that drives progressive microthrombosis within the perforating artery [4]. In a recent analysis combining propensity score matching and multivariable regression, Wen et al. showed that intravenous tirofiban significantly reduced the risk of END compared with oral antiplatelet therapy alone [4]. In addition, a post hoc analysis of the TREND trial suggested that the benefit of tirofiban was particularly marked in patients with severe intracranial stenosis [5], which matches the profile of the patient in Case 2 (severe M1 stenosis). The ongoing multicenter randomized TiMIS trial should help define the role of tirofiban followed by sequential dual antiplatelet therapy in mild-to-moderate ischemic stroke [6].

These data must nevertheless be weighed against the bleeding risk of tirofiban, which has been reported to be increased in some studies, although no significant increase in symptomatic intracranial hemorrhage was observed in the TREND trial. In our second case, the occurrence of hematuria on dual antiplatelet therapy (ticagrelor)—although not attributable to tirofiban itself—illustrates the need for close bleeding surveillance and individualized adjustment of long-term antithrombotic therapy.

Implications for management and limitations
Beyond the acute phase, identification of an atheromatous plaque in the parent artery modifies the secondary prevention strategy: TOAST reclassification as large-artery atherosclerosis, intensification of lipid-lowering therapy, optimization of vascular risk factor control, and, for severe symptomatic stenosis despite optimal medical therapy, discussion of intracranial angioplasty and stenting—an issue raised but not resolved in our second case.

This work has limitations inherent to the case report format: it describes three isolated cases interpreted descriptively, without histopathological confirmation of the plaque, and plaque imaging was performed only after the acute episode in Cases 1 and 2, which limits the conclusions that can be drawn about the contribution of HR-VWI to therapeutic decision-making in the hyperacute phase itself. Larger prospective studies incorporating HR-VWI into the initial etiological work-up are needed to establish its precise place in the diagnostic and therapeutic algorithm for deep infarcts suggestive of BAD.