Article Text

Download PDFPDF
O-001 Pre-Treatment Imaging is Critical in Identifying Stroke Patients with Small and Medium Infarcts Who are Likely to Benefit from IA Therapy
  1. D Frei1,
  2. A Yoo2,
  3. D Heck3,
  4. F Hellinger II4,
  5. V McCollom5,
  6. D Fiorella6,
  7. A Turk III7,
  8. T Malisch8,
  9. O Zaidat9,
  10. M Alexander10,
  11. Z Chaudhry2,
  12. R Gonzalez2,
  13. L Barraza11,
  14. A Bose11,
  15. S Sit11
  1. 1Swedish Medical Center, Englewood, CO
  2. 2Massachusetts General Hospital, Boston, MA
  3. 3Forsyth Medical Center, Winston-Salem, NC
  4. 4Florida Hospital, Orlando, FL
  5. 5Mercy Health System of Oklahoma, Oklahoma City, OK
  6. 6Stony Brook University Hospital, Stony Brook, NY
  7. 7Medical University of South Carolina, Charleston, NC
  8. 8Alexian Brothers Medical Center, Elk Grove Village, IL
  9. 9Medical College of Wisconsin, Milwaukee, WI
  10. 10Cedars-Sinai Medical Center, Los Angeles, CA
  11. 11Penumbra, Inc., Alameda, CA


Introduction/Purpose The Penumbra START Trial is a multicentre, prospective, single-arm trial with a primary aim of determining whether core infarct size on pre-treatment neuroimaging predicts clinical response to intra-arterial acute stroke therapy.

Materials and Methods For this study, major inclusion criteria include presence of proximal artery occlusion in the anterior circulation, baseline NIHSS score of 10 or greater, evaluable pre-treatment neuroimaging [noncontrast CT (NCCT), CTA source imaging, CT perfusion or MRI DWI], and treatment with the Penumbra System within 8 hours. Core infarct size was evaluated by a blinded imaging Core Laboratory. As pre-specified, infarcts were trichotomised into small [ASPECTS 8–10 (NCCT, CTA-SI) or lesion volume <50 cc (CTP or DWI)], medium (ASPECTS 5-7 or volume 50–100 cc), or large (ASPECTS 0–4 or volume >100 cc). In total, 146 patients were enrolled at 27 centres, including 104 patients with Core Lab review and evaluation at 90 days. Good clinical outcome is defined as mRS 0–2 at 90 days. Case review is still ongoing, and statistical analysis is from the 104 patients with complete information. We will update results at the time of the meeting.

Results Mean age is 66 ± 14, including 56% females. Median baseline NIHSS score is 19 (IQR14–22), and 58% of patients had occlusions located on their left side. Overall, TIMI 2–3 revascularisation was achieved in 85%. Forty-eight patients had a good outcome at 90 days (46%); 28 patients died (27%). The number of evaluable scans for each imaging method was: 31 CTP, 82 CTA-SI, 79 NCCT, and 6 DWI. After pooling all modalities in an aggregate analysis, a statistically significant relationship was found between core infarct size and good outcome, such that worse outcomes were found only in the large infarct group. The rate of 90-day good outcome was 54% in small, 56% in medium, and 19% in large infarcts (p=0.0010), despite similar rates of recanalisation (79% small, 93% medium, 84% large). Independent predictors of good outcome were age, baseline NIHSS score, time from onset to recanalisation, and infarct volume. Thirteen patients experienced symptomatic ICH (13%).

Conclusion Pre-treatment neuroimaging is important in identifying patients with large infarcts who are unlikely to respond to mechanical thrombectomy. These data support the use of rigorous imaging criteria in patient selection.

Disclosures D. Frei: 3; C; Penumbra, Inc. A. Yoo: 1; C; Penumbra, Inc. D. Heck: None. F. Hellinger II: None. V. McCollom: None. D. Fiorella: None. A. Turk III: 3; C; Penumbra, Inc. T. Malisch: None. O. Zaidat: 1; C; Penumbra, Inc. 3; C; Penumbra, Inc. M. Alexander: None. Z. Chaudhry: None. R. Gonzalez: None. L. Barraza: 5; C; Penumbra, Inc. A. Bose: 4; C; Penumbra, Inc. 5; C; Penumbra, Inc. S. Sit: 5; C; Penumbra, Inc. 6; C; Penumbra, Inc.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.