This systemic review and meta-analysis in patients presenting with stroke and ICAS demonstrated that primary end points in the AMM group were superior to PTAS group in entire follow-up and within 30 days, and there was no significant difference beyond 30 days. In addition, there was no significant difference in either disabling or fatal stroke or mortality between the two groups.
Our results showed that AMM provided benefits in entire follow-up and within 30 days compared with PTAS. This is similar to the results of several studies[8, 17] and confirms that AMM should be considered first for patients with ICAS rather than PTAS[18]. In entire follow-up and within 30 days, AMM is superior to PTAS perhaps associated with unstable plaque, which increases the risk of adverse events such as distal embolism after stenting[19, 20]. Chimowitz et al.[17] included patients with TIA or nondisabling stroke due to 70–99% stenosis of the diameter of the large intracranial arteries confirmed by angiography within 30 days before enrollment. Similar to the conclusions of this study, AMM provided more benefits for patients than PTAS within 30 days. The difference is that the AMM effect remains superior to PTAS beyond 30 days, and there were no statistically significant differences in any of the outcomes beyond 30 days in this study. The primary endpoint was also assessed in a multicenter, open-label RCT by Gao et al[8]. Similar to this study, there were no statistical differences in outcomes beyond 30 days. Differences in conclusions may be attributable to characteristics in patients, other biases are equally non-negligible. Multicenter studies imply a larger number of participants and operators, which may lead to inconsistencies in their experience with stenting. The importance of experience is suggested by the association of high-volume centers with a lower risk of complications[21–24]. Besides, the outcome of the PTAS group may be related to the quality and limitations of the stent device, and the iterations of stent may enhance the safety and success.
In addition, the selection of patients may also lead to different results. For example, Gao et al.[8] used magnetic resonance imaging (MRI) and computed tomography for screening patients in addition to applying angiography, thus excluding those with perforator stroke alone without artery-to-artery embolism or distal hypoperfusion. This appears to identify high-risk patients, select patients for intracranial PTAS and reduce the occurrence of perforator occlusion during stent implantation, to which several studies have attributed perioperative embolism after stenting procedures[25–29]. For the difference between short-term and long-term outcomes, this may be related to the timing of treatment. Early stenting is associated with a higher risk of complications, and longer time intervals have a reduced risk of complications[14, 17, 23]. In addition, the effect of PTAS may vary depending on the stage of the stroke.
A meta-analysis published in 2017 compared the effects of applying PTAS and drug therapy to patients with Symptomatic Intracranial Atherosclerotic Disease[30]. The results showed that medical therapy was superior to PTAS within 30 days, and there was no statistical difference between those beyond 30 days, which was similar to the results of this study. However, the authors included only 3 eligible RCTs with 581 participants, which may affect the stability of the conclusions. Although PTAS has a high short-term complication rate, based on the importance of extending the time window for endovascular embolization and improving reperfusion, PTAS might be used as a salvage treatment for failed mechanical embolization of large arterial occlusions in the anterior circulation[31–33].
This meta-analysis has the following limitations. Firstly, the number of studies included is limited because there are fewer RCTs published in English and related to our research topics, and we have excluded all retrospective studies. Secondly, heterogeneity is a difficult part of meta-analysis to ignore, with factors such as patient characteristics, inclusion and exclusion criteria contributing to heterogeneity. However, in the present study, the results of the heterogeneity test were acceptable. In addition, an included RCT had only 16 participants, so the risk of bias is existed.