Article Text
PDF
Abstract
Introduction The Woven EndoBridge (WEB) is a novel device for the treatment of wide-necked intracranial bifurcation aneurysms. The present series demonstrates our ‘real-world experience’ in the use of all iterations of WEB devices (available in Europe) in ruptured and unruptured aneurysms.
Methods We analyzed our all-inclusive cerebrovascular database for patients treated with the WEB device between October 2010 and May 2015. Anatomic and clinical results are reported for all patients.
Results One hundred and eight patients with 114 intracranial aneurysms were included in the series. Forty-seven aneurysms (41.2%) were ruptured. Eighty-six patients received angiographic and clinical follow-up after a mean of 13.4 months. One hundred and ten of 114 WEB devices (96.5%) were deployed successfully. Thromboembolic complications occurred in 11 of 110 interventions (10.0%), with a new permanent deficit in one patient. Re-rupture after WEB treatment was detected in two aneurysms (4.3%), which had both initially presented with subarachnoid hemorrhage. Angiographic follow-up revealed adequate occlusion in 68 of 90 aneurysms (75.6%). Fifteen aneurysms required retreatment.
Conclusions This series confirms a high level of safety and efficacy of the WEB device for the treatment of wide-necked intracranial aneurysms.
- Aneurysm
- Device
- Flow Diverter
- Hemorrhage
Statistics from Altmetric.com
Introduction
Endovascular coil embolization has been established as state-of-the-art therapy for unruptured as well as ruptured intracranial saccular aneurysms in most anatomical locations.1–3 Temporary and permanent assist devices such as balloons and stents facilitate endovascular therapy even in complex aneurysm geometries including fusiform, wide-necked, and giant aneurysms. Aneurysms located in vessel bifurcations frequently require more advanced techniques such as dual balloon remodeling or Y-stenting.
Intravascular stents and flow diverters necessitate lifetime antiplatelet therapy. Even with these therapies, a risk for in-stent thrombosis in intraluminal flow diverters may persist even months after treatment.4 ,5 The Matrix and Platinum Science Trial (MAPS) demonstrated recanalization in 37.1% of treated wide-necked unruptured aneurysms.6 Moreover, the neurosurgical approach to middle cerebral artery (MCA) aneurysms is favorable and, as such, clipping is still the therapy of choice in some centers for wide-necked bifurcation aneurysms.7
In 2010 the approach of intrasaccular flow disruption was introduced with the Woven Endobridge (WEB; Sequent Medical, Aliso Viejo, California, USA). Using the WEB device, rapid stasis of aneurysmal blood flow results from the microbraided mesh structure of the intrasaccular implant. Unlike stent-assisted coiling, the device has the ability to embolize wide-necked bifurcation aneurysms with a single-step procedure. Based on the first reported clinical experiences, further iterations of the device improved its ease of delivery, deployment, and conformability.8–11 Because of the present ease of delivery, the duration of treatment can be reduced significantly for experienced users, with lower doses of contrast agent and radiation exposure.
Numerous multicenter series have confirmed the feasibility and safety of the WEB device, predominantly in unruptured aneurysms.10 ,12 ,13 More recent studies have also included greater numbers of ruptured aneurysms.11 ,14 ,15 We have been using the WEB device since 2010 and here we publish the largest single-center experience to date.
Materials and methods
Between October 2010 and May 2015, 108 patients with 114 saccular aneurysms treated with the WEB were consecutively and prospectively enrolled. Clinical and angiographic data were retrospectively analyzed (table 1).
Patient/aneurysm characteristics
Device characteristics
The WEB device is a self-expanding intra-aneurysmal flow disruptor consisting of compliant braided nitinol mesh. Initially the WEB device was designed as a DL flow disruptor, composed of an inner and an outer braid. Barrel-shaped (WEB-SL) and spherical (WEB-SLS) SL devices became commercially available in 2013. The SL devices are of considerably lower profile and are compatible with microcatheters with inner diameters of either 0.027 inch or 0.033 inch depending on the diameter of the device. A more recent iteration, introduced in 2015, has allowed WEB-SL and WEB-SLS devices of 4–7 mm diameter to be deployed through a 0.021 inch microcatheter.
Antiplatelet and anticoagulation therapy
Patients with unruptured as well as ruptured aneurysms, treated early in our series, did not receive antiplatelet or anticoagulation therapy prior to the procedure. Later on, patients with wide-necked unruptured aneurysms received 100 mg aspirin and 75 mg clopidogrel 7 days prior to treatment. Periprocedurally, patients with unruptured aneurysms were heparinized with 5000 international units (IU) and received a further single dose of 500 mg aspirin intravenously. Weight-adapted low molecular weight heparin was continued for 3 days after the procedure. Aspirin 100 mg per day or clopidogrel 75 mg per day was continued for 6 months in patients with unruptured aneurysms. In cases of ruptured aneurysms, antiplatelet medication and anticoagulation were only administered immediately after WEB deployment. Patients with subarachnoid hemorrhage (SAH) also received weight-adapted low molecular weight heparin for 3 days. Furthermore, patients either did not receive aspirin after the intervention or they received it for only 12 weeks, depending on the aneurysm configuration and initial grade of occlusion. Additional antiplatelet therapy (ie, the glycoprotein IIb/IIIa inhibitor eptifibatide) was used in ruptured as well as unruptured aneurysm cases where thrombus formation was visualized.
Procedure
Treatment decisions were made based on an interdisciplinary assessment of each case conducted by a senior neuroradiologist in concert with an experienced neurosurgeon. Cognitively intact patients were thoroughly acquainted with the risks and benefits as well as alternative endovascular and surgical therapeutic options to the WEB procedure and, ultimately, provided informed consent.
Access was typically via a standard 6 F guiding catheter. Catheterization of the aneurysms was achieved through microcatheters with an internal diameter of ≥0.027 inch or, since April 2015, by use of the VIA 21 microcatheter (Sequent Medical, Aliso Viejo, California, USA), depending on the size of WEB required, manipulated over standard microwires. The WEB devices were delivered through the microcatheter and unsheathed using a technique similar to stent delivery. All patients were treated with the CE-marked WEB-DL, WEB-SL, or WEB-SLS devices.
Following WEB placement, post-interventional control angiography was performed to evaluate the degree of initial aneurysm occlusion. Patients' clinical status was assessed immediately post-procedure, at discharge, and at the time of follow-up using the modified Rankin Scale (mRS). According to the WEB Clinical Assessment of Intrasaccular Aneurysm Therapy (WEBCAST) trial, morbidity was defined as an mRS >2 in case of ruptured aneurysms or when the pre-interventional mRS was ≤2, and any increase of at least one point when the pre-interventional mRS was >2.16 Follow-up digital subtraction angiography (DSA), magnetic resonance angiography (MRA), or CT angiography (CTA) and clinical examinations have been—and still need to be—performed at 6 months, 12 months, and 5 years post-intervention. Aneurysm occlusion was graded as either ‘adequate’ (complete occlusion, neck remnant) or ‘aneurysm remnant’ (if filling of the aneurysm beyond the neck was noted). Two independent radiologists evaluated the grade of occlusion.
Statistical analyses were carried out using the non-parametric Mann–Whitney U test.
Results
Patient data
One hundred of the 114 saccular aneurysms (87.7%) were wide-necked, defined as a dome to neck ratio <2 or a neck width ≥4 mm. Seventy-eight patients (72.2%) were female, with a mean age of 55.6 years (range 26–85 years). Forty-seven of the 108 patients (43.5%) were treated within the context of acute SAH (Hunt and Hess (H&H) grade I, n=14; H&H grade II, n=9; H&H grade III, n=14; H&H grade IV, n=5; H&H grade V, n=5). Thirty-eight of these ruptured aneurysms (80.9%) were small and nine (19.1%) were large with a width of 2.9–19.0 mm (mean 6.8 mm) and height of 2.4–19.6 mm (mean 7.3 mm). Thirty-eight of the ruptured aneurysms (80.9%) were wide-necked. Fifty-eight of 67 unruptured aneurysms (86.6%) were small and nine (13.4%) were large with a width of 3.4–11.0 mm (mean 6.2 mm) and a height of 3.0–13.0 mm (mean 6.4 mm). Sixty-two of these unruptured aneurysms (92.5%) were wide-necked. Most of the aneurysms in the series were located in the anterior circulation (n=97 (85.1%); internal carotid artery (ICA), n=7; ICA-T, n=10; anterior cerebral artery (ACA), n=7; anterior communicating artery (ACom), n=27; middle cerebral artery (MCA), n=39; posterior communicating artery (PCom), n=7). Seventeen posterior circulation aneurysms (14.9%; vertebral artery (VA), n=1; basilar artery (BA), n=15; posterior cerebral artery (PCA), n=1) were also included.
Deployment and occlusion rates
In 110 of 114 treatments (96.5%), the WEB was successfully deployed. There was no difference in the ease of delivery between double-layer and single-layer WEB devices as such. Of course, smaller devices delivered via small inner diameter microcatheters were easier to maneuver. In four cases WEB deployment was impossible either due to aneurysm size or difficult access into the aneurysm. Two of these aneurysms were treated by stent-assisted coiling, one by coiling alone, and one by surgical clipping. Forty-nine WEB-DL devices were implanted (from October 2010 to February 2014), 44 WEB-SL (from June 2013 to May 2015), and 17 WEB-SLS (from August 2013 to May 2015). To deliver WEB-DL devices we used 29 0.027 inch VIA or Rebar catheters (VIA 27, Sequent Medical; Rebar 27, ev3, Irvine, California, USA). In four cases 0.033 inch VIA 33 catheters (Sequent Medical) were used, in 15 cases 0.038 inch catheters (DAC, Stryker Neurovascular, Fremont, California, USA), and in one case a 0.041 inch Penumbra catheter (Penumbra, Alameda, California, USA). The deployment of the WEB-SL/-SLS devices was achieved by use of five 0.021 inch VIA catheters, 46 0.027 inch VIA or Rebar catheters, eight 0.033 inch VIA catheters, one 0.038 inch DAC catheter, and one 0.041 inch Penumbra catheter.
One patient with a complete dislocation of the WEB into a distal M2 branch of the MCA as well as the four aneurysms in patients with failed WEB implantations were excluded from the immediate occlusion results. Evaluating the final control angiograms of all WEB-treated aneurysms, sufficient placement of the WEB and progressive stasis of contrast agent within the aneurysm was demonstrated in 87 of 109 aneurysms (79.8%). Thirty-five of 49 aneurysms treated with WEB-DL (71.4%) and 52 of 60 aneurysms treated with WEB-SL/-SLS (86.7%) showed an adequate result. A residual neck remained in 10 aneurysms (9.2%) (six (12.2%) treated with WEB-DL and four (6.7%) treated with WEB-SL/-SLS). Twelve aneurysms (11.0%) exhibited residual aneurysm filling (eight (16.3%) treated with WEB-DL and four (6.7%) treated with WEB-SL/-SLS).
Additional coil embolization was required to complete the treatment in six aneurysms (four ruptured, two unruptured). In three cases (one ruptured, two unruptured), stent-assisted coil embolization was performed. In four aneurysms (3.6%, three ruptured, one unruptured) a Solitaire AB stent (3×20 mm, n=1; 4×20 mm, n=2; 6×20 mm, n=1) had to be placed into the parent vessel due to a protrusion of the WEB into the parent artery. A protrusion means that the device juts out at the base of the aneurysm and affects part of the parent vessel. Stent deployment was technically successful in correcting the prolapse and none of these cases was associated with neurological sequelae. In total, adjunctive devices were used in 11.8% of the aneurysms treated with a WEB device.
Adverse events
Thromboembolic complications occurred in 11 of 110 WEB implantations (10.0%), in seven cases intra-procedurally and in four cases during the immediate peri-interventional period prior to discharge. A persisting neurological deficit due to a retinal ischemia occurred in one case. New ischemia on MRI was identified in three patients. Nine of these thromboembolic events occurred in the 49 cases in which the WEB-DL device was used compared with only two in the 61 cases in which WEB-SL/-SLS was implanted.
In two cases (1.8%) a small amount of extravasation was detected immediately after one WEB-SL and one WEB-SLS deployment, without clinical sequelae. Rebleeding occurred in two patients treated within the context of an acute SAH (4.3%). One patient sustained rebleeding after the treatment of a ruptured multilobulated ACom aneurysm without associated neurological deficits. Evaluation of the post-deployment imaging demonstrated that the WEB-SL device selected for this case was too small and did not fill out one of the aneurysm lobules sufficiently. In the second patient three aneurysms were treated, two with WEB-DL devices and the third by coiling. The patient died from the sequelae of the SAH, but it is not clear which of the aneurysms was responsible for the rebleed.
In two of 110 WEB implantations (1.8%), device movement occurred immediately after detachment. In one case a kick-out of the WEB-SLS device resulted in the occlusion of an MCA branch. In the second case the WEB-DL shifted in position during the treatment of an ACom aneurysm, resulting in an occlusion of the A1/A2 junction of ACA branches, in both cases without clinical deterioration.
Last, in one of the cases without final WEB deployment, a dissection of the ICA occurred during an attempt to catheterize the tortuous vessel anatomy, again without clinical deterioration.
Outcome
Two patients with multiple aneurysms were treated during two hospitalizations. Patients were discharged in a good clinical status (mRS 0, n=60; mRS 1, n=17; mRS 2, n=7) after 84 interventions (76.4%); 19 patients (17.3%) were moderately to severely disabled (mRS 3, n=12; mRS 4, n=4; mRS 5, n=3) and seven patients (6.4%) died (table 2).
Comparison of clinical status at discharge and follow-up
Within the subgroup of unruptured aneurysms, 61 of 63 patients (96.8%) left hospital in good clinical status (mRS 0, n=48; mRS 1, n=11; mRS 2, n=2). One patient discharged with mRS 3 had experienced a prior SAH and was at neurological baseline at discharge and a second patient had a procedural thromboembolism to the distal ACA which was partially recanalized during the procedure. This patient was discharged with mRS 4. However, at follow-up the score improved to mRS 0 and, ultimately, the incident was not classified as a permanent neurological morbidity. Thus, temporary morbidity in interventions with unruptured aneurysms occurred in one of 63 interventions (1.6%) and there were no deaths.
Twenty-three patients with SAH (48.9%) were discharged in a good clinical status (mRS 0, n=12; mRS 1, n=6; mRS 2, n=5), 17 patients (36.2%) were moderately or severely disabled (mRS 3, n=11; mRS 4, n=3; mRS 5, n=3) and seven patients (14.9%) died.
At the time of discharge 17 of 47 patients with SAH (36.2%) experienced significant morbidity, most of them presenting initially with severe SAH (H&H grade I, n=2; H&H grade II, n=3; H&H grade III, n=6; H&H grade IV, n=3; H&H grade V, n=3). Additional procedural or periprocedural thromboembolic events occurred in three patients, all of which were successfully treated. In one of these patients, hemiparesis and dysphasia improved only slightly after thrombectomy.
Follow-up
Eighty-six of 108 patients with 90 aneurysms were re-evaluated 3–53 months (mean 13.4 months) after treatment. Sixty-nine aneurysms (76.7%) were controlled by DSA, 20 aneurysms (22.2%) by MRA, and one (1.1%) by CTA. Seven patients died in the acute state of SAH, one patient died from stomach cancer, and 10 patients refused follow-up angiography and were ultimately lost. Four patients were exempted from follow-up angiography: three patients who had not received a WEB implantation were excluded. A fourth patient with failed WEB implantation in a PCA aneurysm received a WEB device for a BA aneurysm within the same intervention. Consequently, this patient was included for follow-up. In the fourth excluded patient the WEB device completely dislocated into an M2 branch.
Adequate aneurysm occlusion occurred in 68 of 90 aneurysms (75.6%; figures 1 and 2; table 3), with complete occlusion in 52 aneurysms (57.8%) and a neck remnant in 16 aneurysms (17.8%). In 22 aneurysms (24.4%) residual aneurysm filling was noted. Fifteen of the 90 controlled aneurysms (16.7%) were retreated. Additional treatment was necessary in nine initially ruptured and six unruptured aneurysms. The aneurysms had been treated previously by a WEB-DL device in eight cases, by a WEB-SL device in five cases, and a WEB-SLS in two cases. Retreatment was accomplished with stent-assisted coiling in six cases, coiling alone in six cases, implantation of a Pipeline embolization device (PED) in one case, implantation of a second WEB device in one case, and both a second WEB device as well as stent-assisted coiling in one case.
Comparison of anatomic results at follow-up according to the device used and the initial status of the aneurysm
(A–C) A patient with subarachnoid hemorrhage (SAH) Hunt and Hess grade I. Digital subtraction angiography (DSA) detected a left-sided middle cerebral artery aneurysm as the cause of the hemorrhage. (D, E) The aneurysm was treated in the acute state of SAH with a WEB-SL 7×3 mm within 28 min. (F) Control angiogram at 6 months showing complete occlusion.
(A–D) Incidental finding of a basilar artery aneurysm in a patient by a diagnostic angiogram. (C, D) Immediately after WEB deployment, rapid stasis of contrast agent occurred in the aneurysm sac (D). (E) Follow-up at 2, 8, and 14 months after treatment shows stable and complete occlusion of the basilar artery aneurysm.
Altogether, 16 patients showed improved clinical status at follow-up compared with the status at discharge (mRS 1–4, median 3; table 2).
Two patients with unruptured aneurysms exhibited a new morbidity, both with mRS 4. In one case, neurological deterioration occurred after resection of a craniocervical meningioma and, in the other patient, new cerebral metastatic spread from an endometrial carcinoma accounted for the decline. In the interim between discharge and follow-up, no delayed (re)bleed occurred.
In 15 patients with SAH the mRS score improved by a median of 3 points. Consistent morbidity was observed in two patients with initial SAH. One patient with initial SAH exhibited a new morbidity. This patient suffered from a pulmonary embolism and received lysis, which was complicated by an acute subdural hematoma treated with evacuation and hemicraniectomy. The patient was discharged at follow-up with mRS 4.
Finally, permanent morbidity and mortality in all patients who received a WEB device and follow-up angiography were 5.3% and 8.5% (five and eight patients of 94 patients involved), respectively. Three patients suffered from morbidity that was entirely unrelated to the treatment. The percentage of deaths is explained by the relatively large proportion of patients treated within the context of an acute SAH.
Discussion
Since its introduction in Europe in 2010, the WEB device has been evaluated in several multicenter studies including predominantly unruptured aneurysms.10 ,12 ,13 ,16–19 Even more than with standard coiling procedures, aneurysms treated with the WEB device are not immediately completely or nearly completely occluded but rather show intra-aneurysmal stasis (figure 2). More recent WEB studies with more ruptured aneurysms or only ruptured aneurysms have shown promising results.11 ,14 ,15 ,20
In our single-center study we enrolled more than 40% ruptured aneurysms, used large 0.027 inch microcatheters in 95.5% of cases, and deployed the first-generation dual-layer devices in 44.5%. The success rate of WEB implantation of 96.5% is comparable to that found in the existing literature (92.9–97.1%).11 ,13 ,16–18 ,20
The rate of adjunctive endovascular techniques (required in 11.8% of aneurysms in the present series) used in addition to the WEB was approximately the same in ruptured and in unruptured aneurysms and is comparable to that reported in the literature (10.5–23.8%).10 ,11 ,13 ,17 ,18 In 3.6% of the aneurysms in our series, intravascular stents were required due to a protrusion of the WEB into the parent artery or a regional branch vessel—a known event and in the literature presented in up to 14.3% of published cases.11 ,13 ,16 ,19
Access to PCom and paraophthalmic aneurysms was sometimes challenging. This difficulty was solved by the use of intermediate catheters and/or temporary stents and balloons if necessary.
Nine of 11 thromboembolic events (81.8%) appeared during the use of WEB-DL devices. Thus, thromboembolic events emerged significantly more frequently with the use of WEB-DL devices than with WEB-SL devices (p=0.0091; two-tailed Mann–Whitney U test) and in our early experience during which pre-interventional antiplatelet medications were not administered. Our current regimen of antiplatelet medication follows the recommendations of Spelle and Liebig.21 The anticipated thromboembolic rate in our series is lower than in the WEBCAST trial (9 of 51 cases, 17.6%),16 but higher than in the multicenter WEB-SL study of Caroff et al (6 of 98 cases, 6.1%).11 Now that WEB-SL devices represent the vast majority of those implanted, we expect lower rates of thromboembolic complications in the future.
SAH following treatment with WEB devices was only observed in patients who initially presented with ruptured aneurysms. These events were interpreted as early re-hemorrhage. Although rapid stasis within the aneurysm is often demonstrated after WEB deployment, the aneurysm is not immediately excluded (figure 2). Two cases with very small amounts of intraprocedural contrast extravasation were not graded as major periprocedural hemorrhages. The catheterization of aneurysms (both ruptured and unruptured) with larger 0.027 inch internal diameter microcatheters did not result in aneurysm perforation in the present series.
The 3.7% rate of permanent morbidity (2 of 54 patients with follow-up) due to unrelated events within our subgroup of unruptured aneurysms is comparable to related morbidity rates in WEB studies evaluating mainly unruptured aneurysms (0.0–5.3%).10 ,13 ,16–18 This morbidity rate is also comparable to other studies of aneurysm treatment with conventional endovascular techniques (morbidity and mortality rates of 1.7% and 1.4% in the ATENA study).3
In the subgroup of ruptured aneurysms our rate of permanent dependency and mortality (10.3% and 17.9%, respectively) is comparable to that reported in other series of ruptured aneurysms treated with conventional endovascular techniques such as the CLARITY study (23.3%).2 Our overall morbidity rate of 5.3% is also similar to that reported by Caroff et al (4.4%) in a multicenter series of patients undergoing treatment with WEB.11 Like our study, Caroff et al included a relatively high proportion of patients presenting with SAH.
Consistent with the broader literature,10 ,13 ,16–18 our periprocedural mortality rate was 0.0% in the subgroup of unruptured aneurysms. Our mortality rate in the subgroup of ruptured aneurysms (14.9% with initial mRS 6) is comparable to that in the study by Liebig et al (17.0%).15
The rate of adequate occlusion at follow-up (75.6% in our center; table 4) is slightly lower than in the abovementioned series of mainly unruptured aneurysms, varying between 80.0% and 92.9%,10 ,13 ,17 ,18 as well as in the WEBCAST trial (85.4%).16
Comparison of anatomic results, morbidity, and mortality at follow-up with other studies
However, occlusion rates are slightly better than those reported by Behme et al20 who have recently published a smaller two-center series of 55 aneurysms with 66% favorable occlusions. In the multicenter WEB-SL study of Caroff et al,11 satisfactory occlusion at 3.3 months follow-up was achieved in 65%. In our series there was no difference in the satisfactory occlusion rate between ruptured and unruptured aneurysms.
In general, occlusion rates at follow-up and the rates of retreatment continually improved over the years, reflecting progressive improvements in the iterative versions of the WEB device as well as improved insight into case selection, optimal sizing, and delivery of the device. Small differences in the retreatment rates between aneurysms treated with WEB-DL and WEB-SL/-SLS devices (table 3) are explained by the previously mentioned issue that WEB-DL devices were mostly used in the early stage of our series. The lower rates of retreatment after WEB-SL/-SLS deployment should therefore be primarily related to the experience we gained as well as the iterative technical improvements in the devices. Furthermore, higher rates of retreatment after the treatment of unruptured aneurysms occurred due to a relatively high percentage of these aneurysms in our early cohort.
Study limitations
The present study has several important limitations. First, this is a non-randomized retrospective single-center study in which clinical outcomes and angiographic results were interpreted by unblinded radiologists (no core laboratory evaluation). Second, different versions of the WEB device, strategies for device sizing, and antiplatelet medication regimes were used during the course of the study. Thus, changes in occlusion rates and clinical outcomes cannot only be attributed to the type of implant but also to other confounding factors, making comparisons with existing studies of the WEB and other endovascular aneurysm treatment strategies more difficult. Finally, long-term results are pending.
Conclusion
We present the largest single-center series of aneurysms treated with the WEB device. The overall promising results in feasibility, safety, and effectiveness of the WEB reported in other studies is confirmed in our ‘real-world experience’. Intra-aneurysmal flow disruptors represent an emerging endovascular treatment strategy for wide-necked bifurcation aneurysms, with the main advantages of not requiring permanent antiplatelet therapy and allowing the treatment of many of these aneurysms during a technically straightforward, efficient, single-step procedure. Supporting techniques will continue to be necessary in a minority of cases. From the results of the WEBCAST, the FROB and our series,22 we think that the WEB device can become the treatment of choice for many cerebral aneurysms.
References
Footnotes
Contributors All authors made substantial contributions to conception and design, and interpretation of the data. CS, TF, VS, JK and CC performed acquisition and analysis of the data. CC performed statistical analysis. All authors participated in drafting the article or revising it critically for important intellectual content. All authors provided final approval of the version to be published and gave their agreement to be accountable for all aspects of the work.
Competing interests CC, CS, VS: grants from Sequent Medical during the conduct of the WEB-IT study. DF: other from Sequent Medical during the conduct of the study; grants from Siemens and Microvention; personal fees from Covidien ev3 and Cordis; other from Vascular Simulators, TDC Technologies, CVSL, and Codman & Shurtleff outside the submitted work. JK: grants from Sequent Medical during the conduct of the WEB-IT study; consultant for Sequent Medical.
Provenance and peer review Not commissioned; externally peer reviewed.