Background: The majority of patients enrolled in SWIFT PRIME had CT perfusion imaging immediately prior to randomization; 34 patients were randomized following a diffusion and perfusion MRI.
Methods: Patients with MCA (M1) and distal carotid occlusions were randomized to treatment with tPA alone vs. tPA + Solitaire stent retriever. The primary outcome was the distribution of the mRS at 90 days. The same automated software (RAPID) was used for both MRI and CTP patients to identify patients with the Target mismatch profile for enrollment.
Results: MRI selection was performed in 34 patients; CTP in 139. Mean age was 71 years (MRI) vs. 68 years (CTP), p=0.08; baseline NIHSS was 17 in both groups. Target mismatch profile was present in 95% (MRI) vs. 83% (CTP), p=0.3. Median time from ER arrival to randomization was 69 min (MRI) vs. 67 min (CTP), p=0.61. A higher percentage of the MRI group was transferred from an outside hospital (59% vs. 35%, p=0.02) and therefore the time from stroke onset to randomization was longer in the MRI group (236 minutes vs. 179 minutes, p=0.003.) Baseline median ischemic core volumes were similar in the MRI vs. CTP group (7 ml vs. 5 ml); baseline perfusion lesions were larger in the MRI group (133 ml vs. 97 ml, P=0.01)
Reperfusion rates (> 90%/TICI 3) did not differ in the Solitaire treated patients in the MRI (94%) vs. CTP (85%) groups, p=0.7. The primary efficacy analysis (distribution of mRS at 90 days) demonstrated a statistically significant benefit in both the MRI (p=0.02) and CTP subgroups (p=0.01). Among MRI selected patients, mRS 0-2 at 90 days occurred in 63% of the Solitaire group vs. 33% of the tPA alone group (ARR 30%, p=0.17) and infarct growth was reduced (17 ml vs. 50 ml, p=0.089). CTP selected patients had mRS 0-2 rates at 90 days of 60% in the Solitaire group vs. 40% tPA alone (ARR 20%, p=0.025) and a reduction in infarct growth (14 vs. 27 ml, p=0.047).
Conclusions: Although MRI selected patients in SWIFT PRIME were slightly older and treated longer after symptom onset, they demonstrated a statistically significant benefit on the primary efficacy endpoint. Reductions in infarct growth and rates of favorable clinical outcomes in the MR subgroup were also comparable to those seen in the CT perfusion subgroup.