Purpose The importance of early reperfusion to functional recovery after mechanical thrombectomy is well established. Factors contributing to the delay of endovascular therapies remain unclear. Data was analyzed to evaluate the cause (s) for delaying the start of IA intervention in the Penumbra trials.

Methods A pooled analysis of 1028 patients was conducted from the prospective and retrospective Penumbra trials (Pivotal N = 124, PICS N = 261, START N = 133, POST N = 108, RetroSTART N = 191, Speed 054 N = 71). All obtained treatment at <8 h from stroke symptom onset. Included patients were either refractory or ineligible to IV rtPA, had a NIHSS score of 8 or greater, and were in the anterior and posterior circulations.

The univariate relationships between the outcomes and predictor variables were calculated using Spearman’s rank correlation for continuous variables and Wilcoxon rank sums for two-group variables. Predictors of hours from arrival to procedure start with a univariate significance of p < 0.20 were included in the multivariate model. The final multivariate logistic regression model of hours from arrival to procedure start was determined using a standard stepwise selection method. The level of significance was set at p < 0.05.

Results Among the 1028 patients pooled, 742 met study criteria. Mean age was 66 ± 15 and 53% were female. The median admission NIHSS score was 18 (IQR 13–21) and 49% were administered IV-tPA prior to mechanical thrombectomy. Baseline ASPECTS of 0–4, 5–7, 8–10 was 9.5%, 34.5% and 56%, respectively. The median time from symptom onset to hospital arrival was 128 (IQR 63–210) minutes. The median time from hospital arrival to procedure start was 121 (79–165) minutes.

In the univariate analysis, age, gender, target vessel location, baseline NIHSS, IV-tPA prior to IAT, and general anesthesia were not associated with longer arrival to puncture times. The strongest predictors of arrival time to puncture were transfer patient status (p < 0.0001) and onset to arrival (p < 0.0001). Furthermore, transfer patients on average took longer to arrival the hospital, but were treated quicker. However, transfer patients still had a longer overall time from symptom onset to start of procedure. In the multivariate model, a transfer patient’s time from door to procedure was lower by 26 min (p < 0.0001), while every hour increase in onset to arrival time, reduced time to procedure by 18 min (p < 0.0001).

Conclusion In this pooled analysis, transferred patients had a lower door to puncture time, but overall greater total time from onset to puncture longer than direct admit. Even with direct admission patients, as onset to arrival time increased, door to puncture times decreased. Our findings suggest that there is a need to conduct an intensive review of in-hospital triaging procedures for endovascular therapy.

Disclosures A. Yoo: 2; C; NIH, Penumbra, Inc., Remedy Pharmaceuticals. R. Gupta: None. B. Mehta: None. H. Buell: 5; C; Penumbra, Inc. K. Adamski: 5; C; Penumbra, Inc. S. Hak: 5; C; Penumbra, Inc. S. Kuo: 5; C; Penumbra, Inc. A. Bose: 4; C; Penumbra, Inc. 5; C; Penumbra, Inc. S. Sit: 4; C; Penumbra, Inc. 5; C; Penumbra, Inc.