Key words
ischemic stroke - large vessel occlusion - endovascular treatment - stent retriever
- thrombectomy
Introduction
Acute occlusion of large brain-supplying arteries was recognized as a frequent cause
of ischemic stroke by 1658 [1]. Intravenous thrombolysis (IVT) was the first therapeutic procedure established
for this condition [2]
[3]
[4]
[5]. However, its ability to recanalize large thrombi was low, and embolic occlusion
of large vessels remained associated with poor outcomes, often permanent disability
or death, despite IVT [6]. Over the past 65 years, interventional (neuro)radiologists have sought methods
to reopen acutely occluded intracranial vessels in an image-guided manner by using
catheters [7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]. While individual success cases have been published, none of these methods was proven
to be sufficiently simple, safe, and effective to become established as a first-line
treatment.
The misappropriated use for thrombectomy of a detachable stent originally intended
for assisted coil occlusion of intracranial aneurysms changed many things [40]
[41]. After initial setbacks arising from methodologically flawed studies, randomized
controlled trials (RCTs) confirmed the safety and high efficacy of stent retriever
thrombectomy [42].
Mechanical thrombectomy has revolutionized both stroke treatment and interventional
neuroradiology more than any other prior therapeutic concept. Intracranial thrombectomy
became the most common procedure performed in many interventional departments between
2013 and 2023 [43]. Here, we describe the historical development from local intra-arterial fibrinolysis
to stent retriever thrombectomy as a history of ideas.
Local Intra-arterial Fibrinolysis
Local Intra-arterial Fibrinolysis
The first attempt to induce intracranial recanalization by catheter intervention is
attributed to Sussman and Fitch (1958) [7], who described a patient receiving intra-arterial thrombolysis (IAT) for an occluded
internal carotid artery. Sussman and Fitch and, 24 years later, Zeumer (in patients
with basilar artery thrombosis via proximal catheters in the vertebral artery) infused
fibrinolysin and streptokinase, respectively, into distal embolically occluded brain-supplying
vessels from a supraaortic position [7]
[8]. In a further development, toward the end of the 1980 s, microcatheters were inserted
further distally, reaching near the actual site of occlusion, thus enabling local
application of the thrombolytic agent [9]
[10]. A phase 2 trial (Prolysis in Acute Cerebral Thromboembolism, PROACT) demonstrated
the feasibility and efficacy of local IAT (recombinant pro-urokinase in combination
with intravenous heparin) with concomitantly increased rates of intracerebral hemorrhage
[11]. This effect was independent of the choice of thrombolytic agent [12]. In 1999, PROACT II demonstrated the superiority of IAT compared to heparin in terms
of functional independence (modified Rankin Scale [mRS] 0–2) in patients with acute
middle cerebral artery (MCA) occlusion. However, the risk of intracerebral hemorrhage
persisted [13]. Retrospective data on vertebrobasilar occlusions also yielded promising results,
including a significant decrease in previously high mortality [12]
[14]
[15]. Problems and obstacles such as low adoption of the method, the timing and duration
of the intervention, and a lack of training facilities hindered global application
[15]. Moreover, superiority compared to intravenous thrombolysis could not be demonstrated
[16].
Latency Phase
Initially, no significant advances in endovascular procedures occurred in the 1990 s.
One reason was the focus on IVT as a promising recanalizing therapeutic option for
ischemic stroke [3]. Randomized controlled trials indicated the superiority of intravenous thrombolysis
compared to placebo in terms of patient outcomes (mRS 0–2; good functional outcome)
in a time window as long as 4.5 hours after symptom onset (for acute anterior circulation
ischemic stroke) [2]
[3]
[4]
[5]. A major factor determining outcomes is early recanalization of the acutely occluded
vessel [3]. However, the ability of IVT to recanalize acute occlusions of large cerebral vessels
remained limited, at approximately 40 %, and morbidity and mortality in patients with
acute stroke remained high [3]
[5].
Starting in the late 1990 s, innovative new endovascular therapy options were increasingly
developed for patients with severe stroke and poor response to IVT. Nakano et al.
(1998) described primary percutaneous transluminal angioplasty for acute MCA occlusion
by using a Stealth balloon catheter (Target Therapeutics, Fremont, CA, USA) in ten
patients with stenosis and thromboembolic MCA occlusion [17] ([Fig. 1]).
Fig. 1 Stealth balloon catheter originally developed for the treatment of cerebral vasospasm.
A small balloon diameter (2–2.5 mm) and very low inflation pressure (2–3 atm) prevented
vascular injury. However, the recanalization was based on thrombus fragmentation rather
than removal. From a current perspective, the resultant distal occlusion of many small
vessels is unacceptable. However, percutaneous transluminal angioplasty remains an
important option when an atherosclerotic stenosis underlies an intracranial vessel
occlusion. The dedicated balloon catheters (e. g., pITA, phenox, Bochum, Germany)
available for this purpose are technically adapted derivatives of originally coronary
devices for intracranial use.
Microsnares were developed to remove foreign bodies (e. g., coils) from cerebral vessels
([Fig. 2]). Wikholm (1998) and Chopko et al. (2000) described the use of microsnares for endovascular
embolectomy [32]
[33]. The technical procedure is analogous to foreign body removal. The occluded vessel
is catheterized with a microcatheter distal to the occlusion plane. Here, the microsnare
is deployed by slight retraction of the microcatheter. Because of its shape memory
features, the microsnare opens and bends slightly against the longitudinal axis of
the catheterized vessel. Slow retraction of the microcatheter and microsnare is performed
to encircle the thrombus. The microcatheter is then advanced slightly, and the loop
is partially closed, sufficient to grasp but not dissect the thrombus. This method
thus requires extensive experience and great skill. These technical challenges have
prevented this method’s widespread adoption. Recanalization with a microsnare is now
considered when a vessel is occluded by an organized or heavily calcified thrombus
that cannot be captured with a stent retriever.
Fig. 2 “Goose neck” microsnare (Microvena, White Bear Lake, MN, USA).
The first case reports of direct thrombus aspiration were published in 2002. Lutsep
et al. (2002) were able to remove large amounts of thrombus from the internal carotid
artery by using commercially available guiding catheters [18]. For the posterior circulation, Chapot et al. (2002) successfully used 4F or 5F
catheters for diagnostic angiography [19]. At that time, because no dedicated aspiration catheters were available, aspiration
thrombectomy was limited to select patients with favorable anatomical conditions ([Fig. 3]).
Fig. 3 Recanalization of an internal carotid artery (ICA) with proximal atherosclerotic
stenosis and distal thrombus formation. Contrast medium injection of the common carotid
artery shows the proximal ICA stenosis (A). An undersized (2.5 mm) balloon angioplasty (B) allows insertion of an 8F guide catheter into the proximal ICA (C). Contrast injection of the ICA after aspiration with a 50 cc syringe confirmed the
patency of this vessel (D). (E) depicts the removed thrombus. The proximal stenosis was dilated to 4 mm and sealed
with a self-expanding stent (F).
Corkscrews, Brushes, and Baskets: Mechanical Thrombectomy
Corkscrews, Brushes, and Baskets: Mechanical Thrombectomy
In the early 2000 s, the first dedicated devices based on direct interaction with
the blood clot were developed. With these devices, the blood clot is grasped and then
retrieved from the occluded vessel. Unlike microsnares and balloon catheters, these
devices were designed for this purpose. This time period marked the beginning of mechanical
thrombectomy.
Wensel and Gobin had filed a patent application for the concept of using a coil for
endovascular removal of intracranial thrombi in 1996 (later US Patent No. 6,530,935
B2). They claimed the use of a wire with shape memory features, which is inserted
in a stretched state through a microcatheter into the target vessel under X-ray fluoroscopy.
Then the distal end of the wire exits into the thrombus and takes the shape of a spiral.
This spiral “grasps” the thrombus, which is then removed from the body together with
the microcatheter and the “retriever” [20] ([Fig. 4]).
Fig. 4 MERCI retriever (Concentric Medical, Mountain View, CA, USA).
The “Mechanical Embolus Removal in Cerebral Ischemia” (MERCI) device (Concentric Medical,
Mountain View, CA, USA) was first used clinically in 2001 and received FDA approval
for the US market in 2004 [23]. Fragmentation of the thrombus during thrombus passage or retraction, associated
with distal embolization, has been reported to occur in as many as 35 % of cases [21]. Retrospective cohort studies have indicated recanalization rates of 50–65 %, a
32 % rate of good functional outcomes (mRS 0–2), and a mortality rate of 35.2 % [22].
The Neuronet (Guidant, Santa Clara, CA, USA) from 2002 was a self-expanding basket
firmly connected to a guidewire, which was advanced via a microcatheter and used to
retract the thrombus in toto [24]. The Neuronet was never commercially available.
The CATCH device, described and initiated by Chapot in 2005 (Balt, Montmorency, France)
[25]
[26], works in a similar manner and is a stent-like, distally closed wire structure ([Fig. 5]).
Fig. 5 CATCH device (Balt Extrusion).
It is also advanced via a microcatheter distal to the thrombus, where it is released.
A braided nitinol basket is proximally attached at two points to an insertion wire
and is distally occluded. The device “captures” the thrombus and is withdrawn together
with the microcatheter under aspiration. Proximal occlusion with BGC has not been
used [25]
[26]. The CATCH device was approved in the European Union (Conformité Européenne, CE
mark). Mourand et al. (2011) have achieved sufficient recanalization in 65 % of patients
with the CATCH device together with other treatment modalities, such as intra-arterial
and intravenous fibrinolysis [26].
The phenox clot retriever (pCR, CE mark in 2006; phenox GmbH, Bochum, Germany) consisted
of a core wire compound surrounded by a palisade of perpendicularly oriented stiff
polyamide microfilaments [27]. The BONnet (CE mark in 2010; phenox GmbH, Bochum, Germany) was a braided nitinol
basket eccentrically connected to an insertion wire. Its handling was similar to that
of the other mechanical thrombectomy instruments ([Fig. 6]). Both devices were introduced through a microcatheter, deployed distally to the
thrombus, and then pulled back under continuous aspiration.
Fig. 6 pCR and BONnet (phenox GmbH).
All the above thrombectomy devices of the pre-Solitaire era were able to remove intracranial
thrombi, had good safety profiles, were sometimes complicated to use, and did not
achieve a rate of adequate recanalization > 60 %.
Fragmentation, Thrombus Destruction, and Aspiration
Fragmentation, Thrombus Destruction, and Aspiration
The Penumbra system (Penumbra, Alameda, CA, USA), patent by Bose et al. (2004, US
609 028 P), uses a different approach to thrombus extraction [28]. An aspiration catheter is passed over a guide catheter just proximal to the blood clot. An aspiration pump is connected to this aspiration catheter. A
thin wire with a teardrop-shaped distal distension (“separator”) is inserted through
this catheter via a hemostatic valve. With the distal end of the aspiration catheter
in front of the thrombus, the thrombus is fragmented under continuous aspiration by
gentle movement of the separator back and forth through the thrombus. Consequently,
the thrombus is fragmented, and the resulting fragments are aspirated ([Fig. 7]).
Fig. 7 Penumbra system (Penumbra, Alameda, CA, USA): Aspiration catheter and separator.
In cases of persistent residual thrombus, direct thrombus extraction can be performed
via a “thrombus removal ring” under flow arrest through a proximal balloon guide catheter
[28]. With the Penumbra system, complete recanalization (TIMI 3 [Thrombolysis in Myocardial
Infarction] and TICI 3 [Thrombolysis in Cerebral Infarction]) has been achieved in
as many as 70 % of cases [29]
[30]. The CE mark was granted in 2007, and FDA approval followed in 2008. The Penumbra
system is indicated for endovascular stroke therapy within 8 hours after symptom onset
for patients with occlusion of the MCA (M1 and M2 segment), as well as the basilar
artery or a vertebral artery [31].
The AngioJet system (Possis Medical, Minneapolis, MN, USA) is a rheolytic thrombectomy
device using high-pressure saline jets that generate clot fragments, which are consecutively
sucked into the access catheter [34]. Complications have been reported to include dissection and difficult intracranial
navigation. Consequently, the AngioJet has not been further investigated for ischemic
stroke [34]
[35].
Endovascular photoacoustic recanalization (EPAR; Endovasix Inc, Belmont, CA, USA),
a laser- and catheter-based procedure for thrombus fragmentation (mechanical thrombolysis),
was introduced in 2001 [36]. Here, the emulsification of the thrombus appears to be attributable to a conversion
of photonic energy to acoustic energy at the fiberoptic tip of the EPAR microcatheter.
The EPAR microcatheter is pulled (in a guiding catheter) from distal to the occlusion,
through the thrombus, to proximal to the occlusion [37]. Retrospective data have indicated recanalization rates between 40 % and 60 % [37]. In direct laser procedures (LaTIS, Minneapolis, MA, USA), contrast agent is intended
to act as a light guide to transport energy from the catheter in situ to the thrombus
[38]. Neither approach was pursued further.
The EKOS device (Bothell, Washington, USA) was introduced as an ultrasound-assisted
procedure with the goal of augmenting locally applied IAT (via EKOS catheter) [39]. A single-lumen microcatheter with a piezoelectric ultrasound element at the distal
end was designed to locally enhance the interaction of IVT with thrombi (by passaging
fibrin separation). The positive results of the pilot study could not be confirmed
[39]
[44].
Temporary Bypass
Kelley et al. (2008) have presented the concept of a temporary bypass [45]. After unsuccessful IAT (MCA, M1 segment), an enterprise stent (Codman Neurovascular,
Raynham, MA, USA) was placed via a microcatheter within and distal to the thrombus.
As the stent expanded, the thrombus was displaced, and blood flow was restored. IA
thrombolytics were applied during and after the procedure (as the stent was reconstrained
and removed after 20 minutes). The thrombus material between the stent and the vessel
wall had dissolved or migrated distally but was not removed from the body [45]. Ferrera has filed a patent application for a stent-like structure (Iriis Plus)
designed to be used as a temporary bypass (2012, U.S. 8,088,140 B2) ([Fig. 8]). This device was a particularly dense stent ([Fig. 6]). Neither the temporary bypass procedure nor the Iriis Plus device (MindFrame, Irvine,
CA, USA) could prevail over the stent retriever thrombectomy.
Fig. 8 Iriis Plus (Mindframe Inc., Irvine, CA, USA).
Stent Retriever Thrombectomy
Stent Retriever Thrombectomy
In three RCTs from 2013, first-generation thrombectomy devices (MERCI retriever; Penumbra
system; or IAT in the setting of EKOS or with a standard microcatheter) did not gain
acceptance [44]
[46]
[47]. No superiority compared to IVT was demonstrated, although modern thrombectomy devices
were also used in two of the three studies (e. g., Solitaire stent retriever; Covidien/now
Medtronic, Dublin, Ireland), in small numbers (IMS III: 1.5 %; SYNTHESIS: 11 %).The
Solitaire stent was originally designed for assisted coil occlusion of intracranial
aneurysms ([Fig. 9]).
Fig. 9 Solitaire stent (developed at Dendron GmbH, Bochum, Germany; today Medtronic, Dublin,
Ireland).
Technical development was guided by experience in the use of the Neuroform stent (Stryker,
Kalamazoo, MI, USA) and Enterprise stent (Cerenovus, Irvine, CA, USA) in aneurysm
treatment. The Solitaire stent was designed to be retrievable after complete deployment,
and its open cell design enables superior position correction compared to that with
the Neuroform stent. For easier passage of the microcatheter through the stent into
the aneurysm, the cells of the Solitaire were designed to be larger than those of
the Enterprise stent. For more reliable coil retention in the aneurysm than that with
the Neuroform stent, the stent structure was designed to be more stable. These properties
ultimately determined the suitability of the Solitaire stent also for mechanical thrombectomy:
-
The stent retriever needed to be retrievable at least into the guide or aspiration
catheter.
-
Sufficiently large cells were required to allow the thrombus to migrate into the deployed
stent.
-
A stable stent structure was required to “peel off” the thrombus adherent to the vessel
wall.
The suitability of the Solitaire stent for foreign body and thrombus removal was discussed
when this device was first described in 2003 [40]. Between 2002 and 2007, after the acquisition of Dendron by MTI/ev3, the Solitaire
stent was not produced for internal company reasons. In 2008, small quantities of
the stent were available with the CE mark still valid. Use of the Solitaire stent
for thrombectomy was rumored to have been considered or attempted with unknown results.
Subsequent investigations by Covidien (the new owner, later acquired by Medtronic)
revealed that no stent retriever thrombectomy had been successful with a Solitaire
stent before March 8, 2008. With the first technically and clinically successful stent
retriever thrombectomy that day, all parties involved were aware that a long-awaited
breakthrough had been achieved [41] ([Fig. 10]).
Fig. 10 The first thrombus removed from a middle cerebral artery with a Solitaire stent.
In subsequent years, numerous other stent retrievers were developed to the stage of
market readiness (e. g., Trevo, Concentric Medical, Mountain View, CA, USA; EmboTrap,
Neuravi, Galway, Ireland) [48]
[49]. In 2015, several RCTs and subsequent meta-analyses demonstrated significant superiority
of stent retriever thrombectomy compared to IVT for acute ischemic stroke due to the
occlusion of large intracranial vessels [42]. pRESET (phenox GmbH, Bochum, Germany) was the first stent retriever approved by
the FDA on the basis of comparison with the Solitaire stent in a randomized controlled
trial (ClinicalTrials.gov: NCT03994822) [50].
The successful use of the Solitaire stent in the first stent retriever thrombectomy
was performed in an otherwise hopeless situation outside of the approval of this medical
device at that time as compassionate use. A major aspect of the “discovery” of stent
retriever thrombectomy was serendipitous. In specialized neurovascular centers, a
rate of > 90 % sufficient recanalization with < 3 % complications is currently expected
(2023). Despite this unprecedented technical success rate, poor clinical outcomes
have been observed. Out-of-study and “all comers” treatment will each have one-third
of patients receiving thrombectomy survive independently, with disability, or still
die [43]. Closing the gap between the technical success of thrombectomy and the poor clinical
outcomes will be the next challenge in stroke treatment. In a separate article, we
will present new technical developments and concepts beyond stent retriever thrombectomy.
