Melanie R. F. Greenway, MD
@MRFGreenwayMD

Flottmann F, Brekenfeld C, Broocks G, Leischner H, McDonough R, Faizy T, Deb-Chatterji M, Alegiani A, Thomalla G, Mpotsaris A, et al. Good Clinical Outcome Decreases With Number of Retrieval Attempts in Stroke Thrombectomy: Beyond the First-Pass Effect. Stroke. 2021;52:482-490.

Between discussion of the “First Pass Effect”1,2 and wondering “When to Stop,”3 finding a magic number of retrieval attempts that optimizes potential good outcome and minimizes risk of adverse events continues to be an important clinical question. 

Utilizing the German Stroke Registry-Endovascular Treatment section, the authors focused on reviewing number of retrieval attempts as it correlates with final TICI score and modified Rankin Score (mRS) at 90 days.  They analyzed 2611 patients from 2015-2018, 1225 of which met inclusion and exclusion criteria. All patients included were 18 years of age or older, underwent endovascular therapy of the intracranial anterior circulation, and had a complete set of data needed for the study. A successful reperfusion was defined as TICI 2b or 3, and a “good clinical outcome” was defined as mRS 0-2.

The mean age was 73, and the median NIHSS was 15. Within their sample, the median number of retrieval attempts was 2. Overall, they reported mRS at 90 days of 0-2 in 37.5% of patients and greater than 2 in 62.5%. Figure 1A demonstrates the main findings of their report. While the cumulative proportion of patients with TICI2b/3 increases with the number of thrombectomy attempts, the proportion of a good clinical outcome did not increase at the same rate after 3 attempts. Figure 1B demonstrates the relationship between the number of retrieval attempts and mRS score at 90 days. Of the 1225 patients included in this study, 536 (43.8%) achieved successful reperfusion with 1 pass, and 257 (48%) of them had a good clinical outcome. Comparatively, 76 had 6 or more passes before successful reperfusion, and only 11.8% of them had a good clinical outcome.

In the primary multivariate analysis, successful reperfusion after the first, second, or third retrieval attempt had the highest odds ratio for a good clinical outcome of mRS 0-2, after adjusting for confounders of age, initial NIHSS, and ASPECTS. As expected, there was a negative association with increasing age and higher initial NIHSS. The table below demonstrates the odds ratio and 95% confidence interval according to the increasing number of retrieval attempts.

Interestingly, the relationship between number of retrieval attempts and a good clinical outcome was not linear after retrieval attempt 3.  The fluctuation in positive odds ratio with 4th, 5th, and 6th attempt may represent a statistical artifact given the relatively lower sample size of patients.  The authors speculate that variables including collateral flow may have played a role in the variability of outcomes in patients requiring more than 3 attempts. 

Overall, they suggest at least three retrieval attempts should be attempted, possibly up to 5 in selected patients. Keeping in mind that the analysis is limited to those with intracranial anterior circulation large vessel occlusions, this article adds to the body of literature designed to target the magic number of retrieval attempts.

Figure 1 from the article, left, represents the relationship between number of retrieval attempts and good outcome (mRS 0-2 at 90 days) and successful reperfusion (TICI 2b/3). The table, right, is adapted from Table 3 in the article and represents the odds ratio of successful reperfusion at each retrieval attempt with the corresponding odds ratio.
Figure 1 from the article, left, represents the relationship between number of retrieval attempts and good outcome (mRS 0-2 at 90 days) and successful reperfusion (TICI 2b/3). The table, right, is adapted from Table 3 in the article and represents the odds ratio of successful reperfusion at each retrieval attempt with the corresponding odds ratio.

References:

  1. Zaidat O, Castonguay A, Linfante I, Gupta R, Martin C, Holloway W, Mueller-Kronast N, English J, Dabus G, Malisch T, et al. First Pass Effect. Stroke. 2018;49:660-666.
  2. Nikoubashman O, Dekeyzer S, Riabikin A,  Keulers A, Reich A, Mpotsaris A, Wiesmann A. True First-Pass Effect. Stroke. 2019:50:2140-2146.
  3. García-Tornel A, Requena M, Rubiera M, Muchada M, Pagola J, Rodriguez-Luna D, Deck M, Juega J, Rodríguez-Villatoro N, Boned S, et al. When to Stop. Stroke. 2019;50:1781-1788.