Robert W. Regenhardt, MD, PhD
@rwregen
With the 2015 trials irrefutably showing the superiority of endovascular thrombectomy (ET) over intravenous tPA alone for the treatment of stroke secondary to large vessel occlusion (LVO), and the 2018 trials showing it may be effective for up to 24 hours from symptom onset, current research efforts focus on expanding the number of patients who may be eligible for this highly effective treatment (e.g., larger core, more distal occlusions) and optimizing protocols for more complex cases. The latter is exemplified by questions that remain about the best approach to treating tandem lesions, which involve both the cervical internal carotid artery (ICA) and an intracranial artery. The most common etiology is cervical ICA atherosclerosis, but tandem lesions can also result from cervical ICA dissection.
Perhaps the biggest conundrum in the management of tandem lesions is whether or not to stent the cervical ICA in the acute setting. Given the risk of dual antiplatelet therapy, especially in patients who received tPA and have larger cores, some interventionalists choose to defer in the acute setting and offer stenting versus endarterectomy later. If stenting is offered in the acute setting, it is unclear whether cervical ICA stenting should be done before or after intracranial ET. Furthermore, the role of angioplasty and the optimum antithrombotic regimen have yet to be determined. There is limited data available to help guide these decisions. While many of the ET trials included patients with tandem lesions, the management was highly variable. Tandem lesions were present in 32% of MR CLEAN, 18% of REVASCAT, and 17% of ESCAPE, while they were excluded from SWIFT PRIME and EXTEND IA. An analysis of the 30 patients with tandem lesions that were treated with ET in ESCAPE showed 17 underwent cervical ICA stenting, 10 before and 7 after intracranial ET. Of the 13 for which stenting was deferred in the acute setting, only 4 underwent ICA revascularization later.
This recent study in the February 2019 issue of Stroke analyzed patients from the STRATIS registry. STRATIS included 984 patients treated at 55 sites with the Solitaire device within 8 hours of symptom onset. There were no restrictions on the management of tandem lesions; 147 (15%) were identified, of which 80 were acutely stented and 67 were not (angioplasty alone or medical management). The stented vs non-stented groups were well balanced for most variables, including age, ASPECTS, collaterals, and tPA administration (Table 1). However, patients in the stented group had lower initial NIHSS (16 vs 18), shorter onset to arterial puncture (134 vs 163 min), and lower rates of atrial fibrillation (AF, 6% vs 25%). At 90 days, the stented group had significantly more patients with mRS 0-2 (Table 2, 69% vs 42%, P=0.003). Furthermore, there was no difference in reperfusion rates and no difference in complications, including mortality and symptomatic hemorrhage (Table 2).
To control for confounding variables and confirm stenting was an independent predictor of 90-day mRS 0-2, the authors performed a multivariable analysis (Table 3). Significant predictors of 90-day mRS 0-2 were initial NIHSS (OR 0.9, 0.83-0.97), ASPECTS (OR 1.40, 1.03-1.89), and stenting (OR 2.41, 1.09-5.32). Nonsignificant variables included age, AF, intracranial occlusion site, tPA administration, and onset to arterial puncture time. To explore predictors of stent placement, another multivariable analysis was utilized (Table 4). Only history of AF was associated with a lower likelihood of stenting (OR 0.20, 0.06-0.67). Initial NIHSS, ASPECTS, and tPA administration were not associated.
The authors also conducted analyses of stenting before vs after intracranial ET and the role of angioplasty alone. Of the patients stented, there were no differences in reperfusion rates or 90-day mRS when comparing stenting before or after intracranial ET. There was a non-statistically significant shorter puncture to reperfusion time in those undergoing ET first. Of those not stented, the angioplasty alone subset had similar reperfusion rates, symptomatic hemorrhage, and mortality compared to the stented group, but less patients with mRS 0-2 at 90 days (Table 2). While complete details of periprocedural and postprocedural antiplatelet use were not available, the authors do report some adjunctive pharmacotherapy. In the non-stented group, 2 patients received intra-arterial (IA) tPA. In the stented group, 8 received IA eptifibatide, 3 received IA tPA, 3 received IA abciximab, and 1 intravenous abciximab.
The authors conclude that stenting is as safe as deferring in the acute setting. They further state that stenting may lead to improved outcomes, acknowledging that a randomized trial is necessary to confirm this benefit. Indeed, lack of randomization is the biggest limitation of this study. Selection bias may exist, as the treatment decisions were left to the interventionalist, even though there were no differences in tPA administration, ASPECTS, or collaterals between those stented vs not. While there were also some baseline differences between these groups, the multivariable analysis (including initial NIHSS, onset to arterial puncture, and AF) showed stenting was an independent predictor of mRS 0-2 at 90 days. Another limitation is lack of follow-up data regarding the ICA status after the acute period. How many remained patent? For those not stented acutely, how many ultimately underwent revascularization later? These data would aid interpretation of the 90-day mRS difference found in this study. As the authors note, there were no standard treatment protocols and no details of periprocedural and postprocedural antiplatelet use available (medications, doses, timing). They also acknowledge that patients with poor outcomes may be under-represented given patients could be enrolled in STRATIS for up to 7 days, though this likely would have affected both groups similarly.
Overall, this study adds further support that acute stenting may be safe and effective in certain patients, but a randomized trial with a standardized treatment protocol is the next step.
