Dear Sir:
Posterior circulation ischemic stroke generally has a poor prognosis. Compared with ischemic stroke affecting the anterior circulation, few studies have discussed its management. However, the role of the perforating vessels in the vertebrobasilar territory is not well understood. In this letter, we discuss the potential significance of perforating vessels in this cerebrovascular territory in patients with acute ischemic stroke (AIS) who were treated successfully with mechanical thrombectomy (MT), with a particular focus on high-resolution cone-beam computed tomography (CT) (CBCT) imaging findings, brief aspects of clinical management, and eventual outcomes. Relevant literature was also reviewed. As an illustration, we reviewed two patients who presented to our center within 24 hours of symptom onset. Both patients were noted to have basilar artery occlusion (BAO) on pretreatment imaging and underwent MT. Despite successful revascularization in both cases, different clinical outcomes were observed, which we postulate may relate to the pre- and post-MT findings within the brainstem on high-resolution intra-arterial CBCT.
The first patient was a 58-year-old woman with no known medical history who was brought by her family members to an outside hospital with a complaint of reduced consciousness. Upon arrival at the hospital, the National Institutes of Health Stroke Scale (NIHSS) score was 24/42. However, in the emergency room, her consciousness deteriorated, prompting emergency intubation. A preliminary CT scan at another center revealed a suspicious acute infarct involving the brainstem; she was then referred to us approximately 5 hours after symptom onset. Her prestroke modified Rankin Scale (mRS) score was zero. Upon arrival at our center, an urgent magnetic resonance imaging (MRI) stroke protocol was performed, which demonstrated an acute pontomedullary infarct with associated BAO (
Figure 1A-
C, white arrows). After discussion with the family members, a decision was made to pursue MT. The onset-to-inguinal puncture time was 6 hours. Prior to MT, high-resolution intra-arterial CBCT showed a paucity of vertebrobasilar perforators (
Figure 1D). Nevertheless, aspiration (A Direct Aspiration First Pass Technique [ADAPT]) of the thrombus was performed via a 5-French aspiration catheter (single pass), with restoration of distal blood flow 30 minutes after inguinal puncture. Post-revascularization, intra-arterial CBCT was immediately performed for evaluation, which showed no change in the status of the vertebrobasilar perforators (
Figure 1E). This paucity of brainstem perforators was also demonstrated on post-revascularization angiography (
Figure 1F, white circle). Repeat MRI performed a few days post-MT showed patency of the vertebrobasilar vessels (
Figure 1G). Despite this, the patient remained intubated and eventually succumbed to the illness.
The second patient was a 79-year-old man with underlying hypertension, dyslipidemia, a history of mild strokes, and prostate malignancy who was referred to our center after presenting to an outside center approximately 4 hours after symptom onset with a complaint of generalized weakness. Intravenous thrombolysis with tenecteplase was administered at the outside center after a suspicious acute infarct involving the brainstem was noted, with evidence of basilar artery steno-occlusion. His prestroke mRS score was zero. Upon arrival at our center 6 hours after symptom onset, the assessed NIHSS score was 8/42. The urgent MRI stroke protocol performed upon arrival showed evidence of BAO with an acute pontine infarct (
Figure 2A-
C, white arrows). After discussions with the family members, it was decided that MT should be attempted. The onset-to-inguinal puncture time was 7 hours. No clot was retrieved despite both ADAPT and combined techniques. We decided to perform angioplasty for suspected intracranial atherosclerotic disease. Post-angioplasty, distal flow was restored approximately 1.5 hours after inguinal puncture. Pre-MT/angioplasty high-resolution intra-arterial CBCT was not performed; however, post-intervention CBCT and angiography showed well-visualized brainstem perforating vessels (
Figure 2D and
E). Surveillance MRI performed a few days after MT revealed patency of the vertebrobasilar vessels (
Figure 2F). The patient was eventually discharged with an NIHSS score of 2/42.
Studies that specifically examine the significance of vertebro-basilar territory perforating vessels, particularly in relation to AIS successfully treated with MT, are scarce. We conducted a literature search for relevant studies, or at least studies that depict and discuss the evaluation of these perforating vessels via CBCT, considering their association with AIS and MT. We were able to find one study that discusses pontine perforators in the setting of vertebrobasilar dolichoectasia and their relation to ischemic stroke pathophysiology, assessed via CBCT [
1]. Because of the lack of studies discussing vertebrobasilar perforating vessel anatomy assessed using CBCT and its relation to AIS/MT, we broadened our search to include high-resolution cross-sectional imaging studies assessing perforating vessels regardless of intracranial arterial territory. We found another study that assessed perforating vessels (lenticulostriate arteries [LSAs]) in the middle cerebral artery territory using high-resolution magnetic resonance angiography (MRA) at 7 tesla [
2]. These two studies are discussed in the context of our findings.
Dobrocky et al. [
1], in their study of seven patients with vertebrobasilar dolichoectasia assessed via ultra-high-resolution CBCT, found an absence of pontine perforators at the fusiform segment of the aneurysm, with perforating vessels seen at the unaffected segments. However, in the absence of pontine perforators, collateralization can be seen draping the surface of the brainstem, termed brainstem vasocorona, which is postulated to develop gradually over time, in keeping with the pathophysiology of vertebrobasilar dolichoectasia.
This gradual progression also allows for the recruitment of collateral vessels from the anterior inferior cerebellar artery, posterior inferior cerebellar artery, and superior cerebellar artery. Despite conditions favoring collateralization, one of the seven patients demonstrated a lesion with restricted diffusion on diffusion-weighted imaging (DWI), which was indicative of acute ischemia. Hyperacute conditions, such as AIS, affecting the vertebrobasilar arterial territory in the absence of conditions favoring sufficient collateralization, may involve perforating brainstem vessels, which could influence the clinical outcome even with timely revascularization therapy.
One study [
2] evaluated the imaging findings indicative of recanalization in patients with AIS affecting the LSA territory who had spontaneous recanalization, which was assessed using highresolution MRA at 7 tesla. Of the 39 patients initially screened, 32 were eligible for analysis. Of these, eight patients (25%) showed evidence of recanalization of the LSAs when assessed at presentation and at the 1-month follow-up. No therapeutic interventions using MT were performed in this study. The study reported no difference in the clinical outcomes between the recanalized and non-recanalized groups. However, it is noteworthy to remember that the arterial territory involved is the anterior circulation, and the patients mentioned in the study had spontaneous recanalization of the initially occluded vessel segment. Therefore, it is difficult to postulate whether the perforating vessels (LSAs) in this territory, if affected by initial vessel occlusion (before spontaneous recanalization), are comparable to perforating vessels within the brainstem when affected by an occlusive disease process.
The review of our patients posed a few questions. In our first case, the patient was younger, had no known comorbidities, and presented to our center less than 5 hours after symptom onset. Intubation was performed because of reduced consciousness to protect the airways; despite successful recanalization via the ADAPT technique (single pass) and the vessels being patent on follow-up imaging, the patient succumbed to the illness. Intraarterial CBCT showed a paucity of brainstem perforating vessels pre- and post-MT. Our second patient was older, with comorbidities and an active malignancy under oncology follow-up. Intravenous thrombolysis with tenecteplase was administered 4 hours after symptom onset, after objective evidence of an acute brainstem infarct and BAO. The patient presented to our center slightly later, 6 hours after symptom onset. We had to resort to rescue angioplasty after failing both ADAPT and combined techniques. Intra-arterial CBCT showed evidence of good brainstem perforating vessels, and the patient was discharged with an NIHSS score of 2/42. Additionally, the DWI volume for the first patient was larger compared with that of the second patient. The MRA of the first patient showed no signal at the vertebrobasilar region, whereas in the second patient, a “truncal type” occlusion was seen; altogether, these findings suggest the likelihood of an embolic occlusion in the first patient and intracranial atherosclerotic occlusion in the second patient. Looking at the risk factors for both patients, age, pre-stroke mRS, and granting that the second patient received intravenous thrombolysis, the only difference was the status of the brainstem perforating vessels depicted on CBCT. Although only speculative, we suspect that perforating vessels, particularly in the vertebrobasilar territory, may play a more significant role than previously known. Reviewing both the CBCT and angiogram images, we found that the CBCT images better depicted the perforating vessels owing to the highresolution 3D volumetric images, depiction of both vessels and the adjacent parenchyma, and depiction of tiny perforators through better spatial resolution, albeit without the ability to depict flow, as in angiographic images.
Several theories can explain the significance of perforating vessels and their roles in AIS. One such mechanism, known as the “no-reflow phenomenon,” postulates that in patients with vessel occlusion due to AIS, successful revascularization of the occluded large vessel (macrovascular reperfusion) is negated by failure of microvascular reperfusion at the level of the perforating vessels [
3]. This occurs in a few ways: damage to the vessel walls makes them prone to injury and blockage, and tissue edema could potentially compress the perforating vessels, disrupt blood flow, or block the perforating vessel lumen by debris or clots. Interestingly, the study also found that this phenomenon occurred in up to one-third of patients with successful macrovascular reperfusion.
Another plausible theory that supports the significance of perforating vessels in determining the eventual clinical outcome of successfully revascularized AIS is the structural and functional integrity of these vessels and, in turn, the microcirculation. One study [
4] examined the rate of futile recanalization in patients with various grades of leukoaraiosis, which is a radiological marker of white matter disease in the brain.
Leukoaraiosis has been found to correlate with the intrinsic ability of brain tissue to withstand ischemic injury; the greater the extent of leukoaraiosis, the poorer the cerebral microvascular reserve, which predisposes patients to faster progression of infarction. In their study of 68 patients with successful revascularization of an anterior circulation large-vessel occlusion, futile recanalization (poor clinical outcome despite a successful procedural outcome) occurred in 32.4% of cases. Moderate- to severe-grade leukoaraiosis is an independent predictor of futile recanalization.
Tsivgoulis et al. [
5] demonstrated in their study that patients with high pretreatment systolic blood pressure (SBP) showed poorer recanalization rates after receiving systemic tissue plasminogen activator treatment. This could be attributed to several factors. Higher pretreatment SBP levels are associated with increased blood-brain barrier permeability, which exacerbates cerebral edema in the first few hours after the initial injury. This contributes to increased resistance to blood flow, which impairs optimal reperfusion in both the macro- and microcirculation after recanalization therapy. Animal models have shown microcirculatory impedance within the core region of the infarct in cases of middle cerebral artery occlusion [
6]. Moreover, higher pretreatment SBP levels could markedly limit the activation of the endogenous fibrinolytic system, which is vital for ensuring effective tissue plasminogen activator activity. Although these studies primarily focused on patients receiving revascularization therapy via systemic thrombolysis, the importance of microcirculatory reperfusion— and therefore perforating vessels—in the eventual patient outcome should not be overlooked.
Additionally, the underlying pathophysiology of occlusion in relation to the development of vertebrobasilar perforating vessels could influence, at least in part, the outcomes of revascularization therapy. In patients with underlying atherosclerotic disease, a gradual and chronic steno-occlusive process promotes the development of collaterals because of ischemic preconditioning [
7]. When acute occlusion occurs, the presence of collateral vessels and perforators favors better outcomes when revascularization therapy is attempted. One study found that in patients with acute stroke secondary to cervical carotid atherosclerosis, better cerebral collaterals and slightly better median mRS scores at 90 days were observed than those in patients with cardioembolism as the mechanism of acute occlusion [
8]. In patients with cardioembolism or embolism of unknown origin as the source of acute occlusion, the sudden nature of the occlusion prevents adequate collateralization and the development of perforating vessels, which could partly explain the poorer outcomes observed in this group, even when revascularization therapy is successfully attempted. A recent study [
9] demonstrated that within 240 minutes from onset to recanalization, there was no significant difference in the clinical outcomes of patients with large-artery atherosclerosis and those with cardioembolism as the stroke etiology; however, beyond this time point, a stronger time dependency was observed in patients with cardioembolism in achieving a favorable outcome. These pathophysiological processes could partly explain the clinical outcomes observed in the two cases described previously.
We acknowledge that there is insufficient published literature discussing CBCT findings of perforating vessels in the vertebrobasilar arterial territory and their significance in relation to AIS eligible for revascularization therapy, specifically via MT. Larger and better-designed studies in more heterogeneous populations are needed to corroborate our theory.