INTRODUCTION
Landmark randomized clinical trials have shown that mechanical revascularization is more effective than drug therapy for certain stroke patients, leading to its recommendation as standard care for those with proximal artery occlusions [1-4]. Faster recanalization improves patient outcomes and reduces costs, prompting the exploration of various thrombectomy techniques: stent retriever (SR) only, SR and balloon guide catheter (BGC), SR in combination with aspiration catheter (AC), and BGC or AC only [5]. However, success rates still vary (53–83%) and often require multiple passes, increasing risks and worsening outcomes [6]. The importance of achieving a first pass effect (FPE) has been emphasized in several studies [7,8]. Initially a rescue approach, the double SR (DSR) technique is now emerging as a potential first-line strategy [9-12]. Limited evidence suggests DSR achieves FPE rates between 69% and 77%, higher than conventional methods [9,10].
Encouraged by these results, we adopted DSR as a first-line treatment for anterior circulation large vessel occlusions, specifically in isolated or combined M1 and internal carotid artery (ICA) occlusions (excluding posterior circulation). Here, we present our initial experience, challenges, and insights from using DSR in 31 of 141 thrombectomies performed at our institution in 2024.
TECHNICAL CONSIDERATIONS
Catheters
The main parameter, that will determine the method used (BGC and DSR or BGC, AC and DSR combined), is the caliber of the BGC and consequently the inner diameter (ID) of AC. We try to avoid the use of AC and do a direct approach with BGC with DSR. We use a BGC that can accommodate 2 microcatheters with an inner lumen of 0.021” (21MC) inside. In this scenario, we try to bring the BGC up to the skull base. It can sometime be challenging to bring the BGC distally because of vessel tortuosity. If the BGC cannot be navigated to the skull base, we opt for the usage of an AC, preferably a 6 Fr variant. This can allow for a distal positioning of BGC after DSR deployment (a very stable anchoring effect) in the later phase of the intervention. This application of an AC in this scenario has some direct consequences for the choice of microcatheters and SRs depending on the devices available at your institution. Most of the regular 6 Fr ACs do not allow for parallel use of 2 21MCs. Although many 21MCs can fit into the proximal part of 6 Fr ACs, they get stuck upon further advancement due to their tapered configuration. For example, the Prowler Select Plus, with an ID of 0.021” (Cerenovus), has a distal outer diameter (OD) of 2.3 Fr, but proximally it is 2.8 Fr. Even though some combinations should work based on the calculations, the Sofia Plus 6 Fr (MicroVention) has an ID of 0.070” and 2x Phenom 021 (Medtronic) with a proximal OD of 0.034”. They seem to be compatible on case-by-case basis– probably due to marginal diameter differences due to the manufacturing process. If a regular 6 Fr AC has been used, it necessitates the use of 2 0.017” ID microcatheters (17MC) or a sequential implementation of DSR that are only compatible with 21MCs. There are some 6 Fr ACs available, such as React 71 (Medtronic) that, due to a slightly larger ID of 0.071”, are compatible with 2 21MCs. Our internally tested compatibilities have been summarized in Table 1, but we strongly recommend testing it ex vivo on the materials available at your institution due to slight manufacturing differences between material batches.
Selecting Intracranial Vessels
If 2 MCs are used in parallel, we always try to select 2 separate vessels distal to the thrombus: M2 branches in middle cerebral artery (MCA) or M1 and A1 segment of anterior cerebral artery in distal, intracranial ICA occlusions. Since both time and complexity of exchange maneuvers (especially in DSR) in mechanical thrombectomy are a factor, it would be advisable (from a technical and not economic standpoint of view) to use 2 MCs, each with its own microwire (MW), thus allowing for rapid distal vessel selection. It is possible to first select 1 distal branch with a MW and MC and to retract MW to reuse it with a second MC to select a different distal branch or to retract MW, place a SR. and retract MC/MW to reuse them in a second branch. However, this add complexity, prolongs the procedure time, and, above all, increases the time to definitive reperfusion. Potential cost savings (only 1 MW or MW/MC used) are variable and dependent on pricing available at your institution.
The MW configuration for vessel selection is up to the neurointerventionist but most of have used an inverted wire for the selection of the first distal branch and 45–90° configuration to select the second vessel. If those attempts are futile, we will still allow the second MW to follow the first MC into the already selected branch. It is also possible to select 1 branch, deploy a SR, and to select the second branch through it. Different MC placements result in different SR positions and interactions, which are discussed in the next segment.
Stent Retriever Sizing, Placement, and Interaction
We have experience with using Solitaire X (6x40 mm and 4x40 mm; Medtronic), Tiger (Tiger 17 and 21; Rapid Medical), and Aperio (6x40 mm and 4.5x40 mm; Accandis) with DSR. Two SRs can be either placed in parallel, with distal tips in 2 distinct vessels of the bifurcation (Fig. 1A, B), or partially overlapping, where 1 stent is introduced and unsheathed and the other is deployed through the first stent but with its distal tip placed in the other vessel (Fig. 1C, D). We usually opt to introduce a bigger SR (6 mm or Tiger 21) into the dominant M2 branch and a smaller SR (4 mm, 4.5 mm, or Tiger 21) into the smaller M2 branch. In ICA occlusions, we opt to deploy 2 bigger devices (6 mm or Tiger 21) to better accommodate for the size of target vessel and longer thrombus.
In both of deployment scenarios (parallel or stent-in-stent placement), distal parts of the stent in separate branches are thought to limit the possibility of emboli to new territories and to offer a more secure interaction with the thrombotic material in saddle emboli that are situated in the main branch and protrude into both distal vessels at the bifurcation. The proximal parts of the SRs placed in parallel are thought to interact with the thrombus by pinching it between both stents (“sandwiching” or “chopsticks effect”). In our in vivo experience, it is usually hard to assess if the stents are encapsulating embolic material, thus securing it between the 2 devices. More often, we had the impression that during vessel selection, the second MW and microcatheter follows the path of the first system around the proximal part of thrombus and thus displaces (especially in case of hard and/or calcified) thrombi to a peripheral position, at least in the segment of the vessel proximal to bifurcation.
If the goal is to place 1 stent through another (Fig. 1C, D), the physician must not only navigate the second MW and MC through the first stent but also select the lumen of the other branching vessel exiting the first stent through its mesh. This adds another level of complexity to the procedure, especially if the distal branches are not visualized by backflow through collaterals or if most of the thrombus is situated in or over the second branch. In this section, it is important to mention that passing a MW and MC through the mesh of the first SRs can prove to be difficult. The laser cut SRs have a wide-open mesh, and in many of our in vitro attempts, especially in curved vessels, we realized that the MW/MC goes through the mesh at an undesirable spot or completely passes between the first SR and the wall of the artery. This potentially leads to a parallel SR placement, but also severely limits the ability to select a second branch. Even if the MW can be passed through first SR into a distal vessel, sometimes upon MC advancement we would realize, that MW, at least partially, would go out and in again to the lumen of first stent, thus limiting the possibility to advance a second microcatheter. Another scenario that we have encountered while attempting a stent-in-stent configuration is that even when the MW was in the second branch, a second MC would get stuck when trying to advance it out of the main branch into a distal vessel– probably due to interaction with the mesh and/or thrombus (Fig. 2). It is worth mentioning that this technique is even harder if the first SR placed is a Tiger due to its high density, adjustable mesh. On the other hand, if a branching stent-in-stent configuration can be achieved, especially with 2 SRs of different construction, this configuration seems to offer some advantages.
Despite these considerations, the effect of all these techniques is based on the same mechanism– an interaction of a bigger surface area of SRs with an embolus combined with a protective “fishing net” effect of distal parts as well as a potential pinching effect.11 One could argue that the accumulated radial force due to overlapping or parallel stents leads to a better penetration of the thrombus, but this has not been addressed in the bench tests, and has only been mentioned in some publications [10].
The partial overlap (stent-in-stent) method of placement results in an interlocked, very stable construct (Supplementary Video 1), but this should not obviate the need to retrieval by pulling on both wires. On the other hand, we have observed a parallel configuration interlock only in some cases, and it is unclear whether this happens already at the site of the deployment or during the pull– either by different interaction with curvature of vessels between both SRs or with the aspiration or guide catheter. It needs to be addressed that we have never observed an interlocking when using 2 Tiger SRs if placed in parallel, probably due to their closed cell design. If Tiger was used in combination with a different SR, especially if using a stent-in-stent method, it would very often result in interlocking but at the same time limit the ability to control the radial force of Tiger due to interaction with the second SR.
It is unclear if interlocking is a desirable effect. On one hand, it is a stable configuration, potentially allowing for removing hard thrombi by creating a basket, and allowing for more security. In parallel placement, if 1 of the SRs is pulled asynchronously (Supplementary Video 2) it could potentially lead to a “rolling” of the thrombus and losing it. Potential disadvantages include additional damage to the vessel wall due to the rigidity of the system and damaging the SRs themselves. We have observed that interlocked constructs can sometimes be hard if not impossible to untangle. If an entanglement occurs, especially if a Tiger is used in combination with a conventionally designed SR, it may very well lead to complete failure of at least 1 of the devices (i.e., Tiger). If another mechanical thrombectomy (MTE) is required in this scenario, we usually follow up with a single SR thrombectomy (if it is undamaged). Then we reuse the available device or use direct aspiration, especially if untangling the SRs can potentially be achieved with more time (at our institution, MTEs are usually performed by 1 neurointerventionist without assistance). If both SRs are damaged and the occlusion persists, it might be necessary to deploy a third SR, leading to rising procedure costs. Due to these factors (and more complicated vessel selection through an already deployed first SR) we usually try to achieve a parallel DSR, but if this proves to not be possible, only then attempt a stent-in-stent placement.
Retrieval of DSR is always performed with proximal aspiration with an inflated BGC in the ICA.
DSR in Combination with Aspiration Catheters
Combing DSR and distal aspiration is possible. As mentioned above, the limiting factor is the ID of common ACs that does not allow for introduction of 2 21MCs. This limitation can be either avoided by using newer generation ACs with a slightly larger internal diameter, such as React 71, or large bore ACs, such as Milliped 088 (Perfuze). A different possibility is sequential placement of SRs. First, the SR would be deployed and the microcatheter removed (“bare metal wire”) [13]. The second SR is introduced parallel to the pusher wire of the first SR. This approach can lead to some difficulties while selecting intracranial vessels (as discussed above), but this requires the use of only 1 MC and MW.
Independent of the approach used, it is highly advisable after deployment of the second SR to remove microcatheters to facilitate a better distal aspiration by maximizing the lumen of the catheter [13]. If no AC is used, we would be discouraged from using the bare metal wire technique due to the danger of vessel injury.
Another possibility is to use SRs that require 17MCs but this either severely limits the available sizes and lengths of applicable SRs or requires a use of a SR with adjustable radial force (i.e., Tiger 17) that, due to design, do not allow for removing of the microcatheter, and thus limits the available lumen AC and decreases the available suction.
Before retrieval of DSR, we usually advance the AC over the pusher wires of SRs, slightly over the anticipated proximal part of the thrombus or up to a build-up of resistance. In the next step, the whole system (with DSR fixed against distal AC) is removed. This maneuver is accompanied by proximal aspiration with an inflated BGC in the cervical ICA- a standard procedure at our institution. In rare cases, if the intracranial access is difficult and losing it could lead to a considerable loss of time, a retrieval of DSRs into the AC could be attempted, but due to the amount of device material and the danger of thrombus scraping off against the catheter is not advisable.
TIPS AND TRICKS
In DSR, the stents should be initially gently pulled on, separately at first, so that they both come under similar tension and then be removed simultaneously without introducing excessive movement between them. To facilitate the combined pulling maneuver, if the bare metal wire technique is used, a single torque device can be placed on both wires securing their position. This is also possible, even if one of the SR pusher wires is still in the microcatheter and the other is not. Due to the design of SRs with adjustable expansion force, this technique cannot be used. More attention is needed by the physician to pull on both SRs without leaving one behind. In scenarios, where a classical and an adjustable SR are used together, a torque device can be placed on the wire of the former one at a position where it is parallel to the handle of the latter one so both can be secured and pulled comfortably with one hand.
In the scenario of sequential SR placement, selecting a different branching vessel to the one that has already been secured with first SR in place can be challenging. It could often be achieved with an inverted wire, but this configuration does not allow for wire control and thus we must rely on anatomy guiding the system into a proper vessel. In our experience, in MCA bifurcations, if possible, it is better to place the first SR in the vessel that is branching off more caudally. The reason for that is that the inverted wire usually follows the medial and cranial aspect (outer curve) of the distal ICA and proximal M1 segment, and thus has higher chances of ending up in the vessel of MCA-bifurcation that branches cranially. We have also found that in these scenarios, using a smaller diameter of SR first and placing a bigger one as the second one elevates the chances to select a proper branch. A different possibility would be to use an adjustable stent as the first one, and if the vessel selection for the second SR is difficult, to reduce the expansion of the first device.
In the scenario depicted in Fig. 1D (partial stent-in-stent) we have found that the initial selection of the lumen of the first, already positioned, SR is both achievable with an inverted as well as with an angulated MW configuration. On the other hand, the attempt to select the other distal bifurcating vessel from the inside of the first stent is usually only possible with a normal MW configuration (slight bend). This thus requires the microcatheter to be pushed partially into the lumen of the first SR, and the MW needs to be retracted in order to resolve the inverted configuration if it was used for the first part of the maneuver. Still, finding the other branching vessel from the inside of the first SR, especially if the ostium is covered by the embolus often proves to be hard, if not impossible. In those scenarios we opt for complete overlap of 2 SRs in the same branch. It is worth mentioning that if an adjustable SR is used as the first one, it is possible to penetrate it with a MW, but we have never been able to follow with a microcatheter.
Despite the sequential technique having some advantages (less material to interact with, lower cost, higher probability of interlocking if a branching stent-in-stent configuration can be achieved), it is slower than using 2 MWs and MCs.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
