Introduction
Aneurysmal subarachnoid hemorrhage (aSAH) is a rare event that affects primarily young patients and therefore has a major social influence. Large-vessel vasospasm (i.e., vessel narrowing) had long been considered the sole or major cause of poor outcomes after aSAH. Recently, the focus of clinical research has shifted to other causes. Evidence suggests that patients with aSAH can develop DCI without angiographic vasospasm of large vessels [
1‐
4]. Since 2004, when this phenomenon was first described by Kusaka [
5], many studies have reported early brain injury. The sudden increase in intracranial pressure as a result of aSAH may lead to a decrease in cerebral blood flow [
6] or to a breakdown in the blood–brain barrier and subsequent edema [
7], thus leading to transient or persistent ischemia. Furthermore, several studies have reported the possible role of spreading depolarization after SAB in the development of cytotoxic, ionic, and vasogenic edema [
8‐
11]. Distal cerebral vasospasm (CVS) after aSAH may also lead to microvascular dysfunction and microthrombosis [
12]. Despite the above-mentioned reasons, CVS remains a leading cause of neurological deterioration after aSAH. CVS is responsible for a morbidity and mortality incidence of approximately 20% [
13,
14]. Among patients with aSAH, 50–70% develop CVS, and approximately 30% present neurologic deficits [
15,
16]. Despite limited accuracy, transcranial Doppler sonography (TCD) and transcranial color-coded Doppler (TCCD) remain preferred diagnostic methods in intensive care units for the detection of vasospasm, because of their logistic simplicity [
17].
Computed tomography angiography (CTA) with computed tomography perfusion (CTP), and magnetic resonance angiography (MRA) are more reliable in CVS detection. Digital subtraction angiography (DSA) remains the gold standard for detecting and quantifying CVS [
18]. Conservative management by intravenous or oral administration of calcium channel blockers has yielded sobering results [
19]. Endovascular management strategies allow for effective treatment of CVS. These strategies include short-term intra-arterial administration of nimodipine [
20], milrinone[
21,
22], or verapamil [
23]; long-term intra-arterial selective infusion of nimodipine [
24]; and mechanical dilatation of large arteries with non-compliant balloons, compliant balloons [
25‐
29], or stent-retrievers [
30‐
34].
The substantial logistic efforts required for diagnosis and endovascular treatment outside the intensive care unit, and the complex cardiovascular and respiratory monitoring required before and during treatment, may lead to treatment delays and consequently poor outcomes. Despite the recent development of pharmacological and endovascular treatment options, the management of post-SAH CVS remains a challenge because of its frequent occurrence and the limited efficacy of the available treatment concepts.
Here, we report our anecdotal experience in the use of intracranial stenting for the treatment of patients with post-SAH CVS.
Discussion
Although the endovascular and microsurgical treatment of ruptured aneurysms has substantially improved during the past three decades, similar progress in the treatment of CVS is lacking. The past 20 years have seen a mindset shift from a fatalistic acceptance of poor outcomes due to CVS toward active attempts to address these issues. The search for a safe and efficacious treatment for CVS has been ongoing since the 1970s [
35]. Jabbar et al. [
36] have compared two cohorts including 1057 patients. All patients underwent daily TCD ultrasonography to detect CVS. Patients in group A were treated immediately after any suspicion of CVS, regardless of the TCD results. Patients in group B were treated by endovascular means only after persistent CVS despite induced hypertension for 2–4 h or when a mean flow velocity above 160 cm/sec was gradually exceeded for two consecutive days. In group A, 24.4% of patients underwent the first endovascular treatment on day 6 ± 3.64, and in group B, 14.4% underwent the first endovascular treatment on day 8.9 ± 4.78. The rate of DCI was lower in cohort A (20.8% vs. 29%,
p = 0.0023, OR 0.64, 95% CI 0.48–0.85), as confirmed by multivariate analysis (
p = 0.001, adjusted OR 0.59, 95% CI 0.44–0.8). The rates of DCI were higher in patients undergoing endovascular treatment for delayed ischemic neurologic deficit than in those undergoing endovascular treatment solely because of TCD measurements (64% vs. 44.7%,
p = 0.0277). The rate of unfavorable outcomes after SAH was also lower in cohort A (44% vs. 56%,
p = 0.0404. This study demonstrated that early identification and aggressive treatment might result in better functional outcomes.
Balloon dilatation of proximal arteries (mainly the distal ICA and the proximal MCA) is an acceptable option, but the risk of vessel dissection and even rupture remains a concern.
In recent years, endovascular mechanical vessel dilatation using stent retrievers (e.g., Solitaire, Medtronic; pRESET, phenox) has been established.
The non-selective IA injection of vasodilators can lead to an steal phenomenon due to a significantly increased inflow in the non-spastic arteries [
37]. To avoid this steal phenomenon, distal vasospasm was suggested to be best treated with medication after initial mechanical treatment of the proximal vasospasm. Balloon angioplasty [
13‐
17] and “stentoplasty” [
18‐
22] have shown therapeutic efficacy and sustained improvement. Although balloon angioplasty was initially believed to damage the extracellular matrix, evidence has indicated that balloon angioplasty of an arterial segment instead induces paralysis of the vessel without necessarily damaging the underlying extracellular matrix [
32,
38]. The induced vessel paralysis after angioplasty is affected by the contractile state of the vessel. Contracted (vasospastic) vessels, require less dilatation to become paralyzed. Therefore, vessels with CVS are predisposed to paralysis after mechanical dilation and thus can be treated with devices with lower radial force [
32]. Damage to the underlying smooth muscle cells, which stiffen when contracted and may potentially be more prone to mechanical disruption, has been suggested to explain this phenomenon [
39]. This finding suggests that mechanical angioplasty should be performed before chemical angioplasty. Simultaneously, lower forces might be required, thus potentially explaining why stent retrievers have shown some success in treating CVS despite having much lower radial force than balloons. Kwon et al. [
32] compared patients in whom the stent was deployed initially followed by injection of a vasodilator (nicardipine) and those in whom a vasodilator was injected first, and stentoplasty was performed second. In the vasodilator-first group, 71.4% of treated vessel segments (10/14) showed vasodilation after stentoplasty, but 60% of patients (3/5) developed recurrent vasospasm requiring repeated angioplasty. In the stentriever-first group, 82.1% of segments (32/39) showed vasodilatation after stentoplasty, but none of the patients developed radiological or clinical evidence of recurrent vasospasm. This small clinical study corroborated the initial hypothesis suggested by Bhogal et al. [
33], building on the work of Fischell et al. [
38,
39].
Bhogal et al. [
33] have noted the following benefits of treating CVS with recheatable stent retrievers, as specifically compared with vasodilator medications or balloon angioplasty:
-
No flow arrest
-
Ability to track stents into the distal vessel segments
-
Familiarity with the use of stents from mechanical thrombectomy
-
A likely lower risk of perforation because of operator-independent stent expansion
-
Long-lasting and durable vasodilation
-
Short procedure time
-
Ability to inject vasodilating drugs at the same time
Similarly, Su et al. [
34] have evaluated the safety and efficacy of treatment of CVS with stent retriever-assisted angioplasty. In this study, 14 vessels from six patients with resistant CVS showed improvements in vessel diameter, and all patients had improved neurological examination findings.
Intracranial stenting offers the same above-mentioned advantages because of the similar release concept of the two devices in the distal vessels. In extremely resistant CVS cases, intracranial stenting offers a continuous effect as a bail-out option, mainly because of the permanent dilatation of the stented vessels and the better circulation of the vasodilating drugs in the affected distal segments that cannot be reached by angioplasty.
Stent implantation has also been reported. Andic et al. [
40] have analyzed data from 15 consecutive patients with 18 aneurysms, eight of whom underwent stent-assisted coiling. In most cases (
n = 6), an LVIS Jr. (MicoVention) stent was implanted, with a Solitaire (Medtronic) or Acclino (Acandis) stent used in the remainder. The authors observed moderate to complete dilatation of the spastic parent arteries after deployment of the stents in patients treated with stent-assisted coiling. In one case, refractory vasospasm occurred after the treatment of two aneurysms and the authors found no evidence of recurrent vasospasm in the stented segments but observed widespread recurrent vasospasm in the segments previously treated with chemical angioplasty.
Although these findings were based on only a small series, this article highlights the potential for braided and laser-cut stents to effectively treat and prevent recurrent CVS. Subsequently, Bhambri et al. [
41] described a novel method of using drug-eluting stents to treat CVS. The authors developed polymer-coated laser-cut stents and used various methods to coat the stent. The laser-cut and polymer-coated stents were also loaded with different doses of verapamil with varying drug release pharmacokinetics. In all cases, the combined stent demonstrated an initial burst phase of drug release followed by sustained drug release. Varying the concentration of the verapamil changed these different phases by altering the construction of the polymer coating. This preliminary in vitro study suggests the potential promise of further developing dedicated stents for vasospasm that release vasodilating drugs.
Despite the development of endovascular techniques and their demonstrated effectiveness in treating recurrent vasospasm, some of these therapies remain unsuccessful. The possible reasons for the failure of endovascular treatment are as follows:
-
Steal phenomena, particularly with intra-arterial vasodilator treatment
-
Distal location: difficult navigation and dilatation of CVS in distal vessel segments for endovascular mechanical treatments
-
Tortuosity and curved vessels: complex dilatation of spastic vessels in a curved course, such as the transition from the A1 to the A2 segment, particularly for balloon angioplasty
Some endovascular treatments have a short-term effect, but CVS is a dynamic and potentially long-lasting pathology during the first 3 weeks after aSAH. Recurrent CVS may not be successfully treated in some patients despite multiple endovascular treatments sessions. Consequently, the search for a therapeutic solution for refractory CVS continues. Intracranial stenting of vessels with vasospasm is an ultima ratio treatment in patients with severe recurrent vasospasm. We have attempted to enforce permanent dilatation of vessels with resistant vasospasms by stent implantation. The implanted device is intended to avoid any steal phenomenon and enhance vasodilator drugs’ effects, particularly on the peripheral vessels. By improving perfusion in the proximal vessels, we have also observed a reduction of CVS in the distal vessels.
This level of improvement occurred only after stent implantation and can therefore not be explained by the previous IA administration of the vasodilator.
The concept of treating CVS with intracranial stenting arose from incidental experience during the management of complications from aneurysm coil occlusion. During coiling, impaired perfusion of the parent vessel and the dependent vasculature was observed, owing to inadvertent displacement of the implanted coils. Stents were implanted to maintain the coils inside the aneurysm and restore normal blood flow. The vessels where the stents were implanted had shown CVS before and during the coiling despite the IA vasodilator administration. After stenting, we observed decreased vasospasms with similar perfusion improvement in the distal supply territory. One of the patients was diagnosed with high-grade CVS after day 1, but not within or distal to the stented vessel. After our initial two incidental experiences and the excellent long-term outcome, we decided to perform this treatment in the four described cases after several unsuccessful chemical and mechanical treatments. In the first two patients, we implanted only one stent in the MCA in the M1-M2 junction. With experience, in the third patient, we implanted two stents in the middle and anterior cerebral artery ipsilaterally. In the last patient, four stents were implanted, two on each side. The stent implantations were always performed under dual antiplatelet therapy (DAPT). Overall, only one transient thromboembolic occlusion occurred as a periprocedural complication. None of the patients showed postprocedural hemorrhage, in-stent stenosis, or DCI of the entire vascular territories of the stented vessels. DAPT was maintained in patients for 6 months. Antiplatelet therapy was de-escalated to single antiplatelet therapy after the 6-month DSA follow-up.
The repetition of endovascular treatment of recurrent vasospasm increases the risk of periprocedural complications and radiation exposure [
42]. Endovascular treatment may not always be feasible, owing to the complicated clinical management of patients with aSAH, particularly respiratory and cardiovascular management. Nonetheless, we believe that it is a viable treatment option as a last resort in refractory vasospasm to avoid life-threatening DCI after the failure of medicinal and mechanical treatments.
The majority of neurovascular stents (e.g., Enterprise, Cerenovus; Solitaire, Medtronic; LVIS, MicroVention) require microcatheters with an 0.021″ ID. These stents have a significantly higher radial force than their counterparts with a reduced crossing profile. Neuroform Atlas (Stryker), LVIS Jr (MicroVention), and pEGASUS (phenox) can be implanted through 0.017″ ID microcatheters. Three of the implanted stents were Neuroform Atlas. The low profile 0.017″ ID microcatheter (e.g., Excelsior SL-10, Stryker) allows for atraumatic navigation of distal and narrow vessel segments. The relatively low radial force was not an issue.
Limitations
The technique of stenting as a treatment for CVS and the present study have several limitations.
The currently available stents are not designed for this purpose, and their use for this indication is off-label. Technical specifications such as radial force, cell structure (open versus closed cell), and cell dimensions have not been evaluated. Space certainly exists for an optimized implant. All stents that we have used to date were uncoated and require DAPT. The need for DAPT during the acute phase after aSAH is a matter of concern. pEGASUS (femtos) has a hydrophilic coating and can be implanted under SAPT. Patients in the acute phase after aSAH have elevated platelet activity and usually require increased dosages of antiplatelet medication for the intended platelet function inhibition.
The data presented herein come from anecdotal cases treated under emergency circumstances according to the knowledge and experience of the senior author. The available devices for the mechanical treatment of CVS (e.g., pRELAX, femtos) have a CE mark for this indication. A large scale randomized controlled trial is in preparation in Germany. In this trial, stenting may become an option if the temporary deployment of pRELAX does not prevent the recurrence of vasospasm. Stenting proximal arteries is only one option in the armamentarium for the treatment of post-SAH vasospasm.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.