• Population: Patients 18 years of age or older presenting within 4.5 hours of AIS symptoms with a cerebral vascular occlusion on CT angiography of the intracranial internal carotid artery or middle cerebral artery (first and/or second segments) and an NIHSS >1, and if endovascular thrombectomy was intended to be performed.
• Exclusions: Disability from a previous stroke or contraindication to IV alteplase, plus any contraindication for thrombolysis according to American Heart Association guidelines.
• Intervention: Endovascular thrombectomy.
• Comparison: Endovascular thrombectomy plus systemic tPA (0.9 mg/kg).
• Outcomes: 
  • Primary Outcome: Modified Rankin Scale (mRS) score assessed at 90 days.
  • Secondary Outcomes: Death from any cause at 90 days, successful reperfusion before thrombectomy, recanalization at 24–72 hours, NIHSS score at 24 hours and five-to-seven days, and final lesion volume on CT and mRS comparisons.
  • Safety Outcomes: All hemorrhages and symptomatic intracranial hemorrhages, occurrence of pseudoaneurysm and groin hematoma at the site of arterial puncture used for thrombectomy, cerebral infarction in a new vascular territory at five to seven days, and mortality within 90 days.

Authors’ Conclusions

“In Chinese patients with acute ischemic stroke from large-vessel occlusion, endovascular thrombectomy alone was noninferior with regard to functional outcome, within a 20 percent margin of confidence, to endovascular thrombectomy preceded by intravenous alteplase administered within 4.5 hours after symptom onset.”

Table 1: Secondary Outcomes

Key Results

The study included 656 patients with a median age of 69 years, and slightly more were male. The median NIHSS score was 17.

• Primary Outcome: Adjusted odds ratio (aOR) for the mRS.
  • aOR = 1.07 (95% CI, 0.81–1.40), demonstrating noninferiority because the lower limit of noninferiority was set at 0.80.
• Secondary Outcomes: see Table 1.

Evidence-Based Commentary

1) Consecutive Patients: The publication did not explicitly say patients were recruited consecutively. Without this information, it is hard to comment on whether there was selection bias.

2) Blinding: There was a lack of blinding of the treating physicians and study participants. This could have biased the study toward the EVT alone if that hypothesis was known to these two groups.

3) Intention-to-Treat (ITT): Using an ITT analysis is a quality indicator for superiority designs. However, for noninferiority trials, a per-protocol analysis is the more conservative approach to minimize bias (see Figure 1). Using an ITT can bias the results toward the null hypothesis (ie, finding noninferiority). The authors did perform a per-protocol analysis, but this could only be found in the supplemental material.

4) Patient Outcomes: These data were obtained by interviews either performed in person or by telephone. The authors did not provide information on how many assessments were performed using each method. Telephone interviews are suboptimal for assessing functionality. It would have been helpful in interpreting the data to see if there were any differences that could be attributed to the way outcomes were assessed. Wider confidence intervals around the point estimate of efficacy could have been introduced into the data set by telephone interviews and biased the results toward finding noninferiority.

5) External Validity: The standard AIS care in China may contain differences compared to the United States, limiting the external validity of these results.