Article Text

Download PDFPDF

Original research
Longer procedural times are independently associated with symptomatic intracranial hemorrhage in patients with large vessel occlusion stroke undergoing thrombectomy
  1. Tareq Kass-Hout1,
  2. Omar Kass-Hout1,2,
  3. Chung-Huan Johnny Sun1,3,
  4. Taha A Kass-Hout4,5,
  5. Raul Nogueira1,
  6. Rishi Gupta1,6
  1. 1Department of Neurology, Emory University School of Medicine, Atlanta, Georgia, USA
  2. 2Catholic Health system, Department of Neuroscience, Buffalo, New York, USA
  3. 3Columbia University, Department of Neurology New York, New York, USA
  4. 4Humanitarian Tracker, Washington DC, USA
  5. 5Rochester Regional Health System, Department of Neuroscience, Rochester, New York, USA
  6. 6Department of Wellstar Neurosurgery, Kennestone Hospital, Marietta, Georgia, USA
  1. Correspondence to Dr Rishi Gupta, Department of Wellstar Neurosurgery, Kennestone Hospital, 61 Whitcher Street, Suite 3110, Marietta, GA 30060, USA; Guptar31{at}


Background Time to reperfusion is an essential factor in determination of outcomes in acute ischemic stroke (AIS).

Objective To establish the effect of the procedural time on the clinical outcomes of patients with AIS.

Methods Data from all consecutive patients who underwent mechanical thrombectomy between September 2010 and July 2012 were analysed retrospectively. The variable of interest was procedural time (defined as time from groin puncture to final recanalization time). Outcome measures included the rates of symptomatic intracranial hemorrhage (sICH, defined as any parenchymal hematoma—eg, PH-1/PH-2), final infarct volume, 90-day mortality, and independent functional outcomes (modified Rankin Scale 0–2) at 90 days.

Results The cohort included 242 patients with a mean age of 65.5±14.2 and median baseline National Institutes of Health Stroke Scale score 20. 51% of the patients were female. The mean procedure time was significantly shorter in patients with a good outcome (86.7 vs 73.1 min, respectively, p=0.0228). Patients with SICH had significantly higher mean procedure time than patients without SICH (79.67 vs 104.5 min, respectively; p=0.0319), which remained significant when controlling for the previous factors (OR=0.974, 95% CI 0.957 to 0.991). No correlation was found between the volume of infarction and the procedure time (r=0.10996, p=0.0984). No association was seen between procedure time and 90-day mortality (77.8 vs 88.2 min in survivals vs deaths, respectively; p=0.0958).

Conclusions Our data support an association between the risk of SICH and a longer procedure time, but no association between procedural times and the final infarction volume or long-term functional outcomes was found.

  • Stroke
  • Stent
  • Hemorrhage
  • Thrombectomy

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.


Time is one of the most important aspects of the management of acute ischemic stroke (AIS) and any delays in starting treatment will result in a worse outcome.1 ,2 Even though the American Heart Association/American Stroke Association (AHA/ASA) guidelines for stroke therapy recommend initiation of thrombolysis with tissue plasminogen activator (t-PA) within 60 min of patient arrival at the emergency department (ED), this is accomplished in <27% of patients.3 ,4 Endovascular thrombectomy has been a common practice in most major comprehensive stroke centers (CSCs) in the USA for acute large vessel occlusion (LVO) since the results of the Prolyse in Acute Cerebral Thromboembolism (PROACT) II trial were published in 1999.5 However, the procedure has not been approved as better than medical treatment until recently, with multiple positive trials showing an overall positive outcome increase from 35% to 60% with endovascular therapy.6–10

Reminiscent of the paradigm shift in treatment of coronary artery disease that was introduced by percutaneous coronary intervention for patients with ST elevation myocardial infarction (STEMI), endovascular therapy for stroke has major implications for systems of stroke care. This includes a stroke infrastructure, which has been partly implemented by the Joint Commission in the USA for intravenous thrombolysis and needs to be adapted to meet endovascular therapy requirements. The CathPCI Registry demonstrated the importance of ‘door-to-balloon’ (D2B) times on patient mortality,11 achieving D2B times of <90 min in most hospitals with percutaneous coronary intervention capabilities, and reflecting successful implementation of best practice health systems for patients with STEMI.12 Similarly, the Rapid Reperfusion Registry was developed to standardize data collection and to provide a real-time assessment of the pre-procedural metrics of ‘last known normal-to-puncture’ and “door-to-puncture” (D2P).13 The Rapid Reperfusion Registry carried out a retrospective observational study of nine major stroke centers in the USA and showed an association of improved patient outcomes with reduced D2P times. Another important measure that was not previously studied is the intraprocedural time or ‘puncture-to-recanalization’ (P2R).

In this study we sought to establish the effect of procedural time (P2R) on the clinical and radiographic outcomes of patients with AIS undergoing IA endovascular therapy.


Data collection

A retrospective study involving all patients with ischemic stroke treated with mechanical thrombectomy devices at our institution over a 2-year period (September 2010–July 2012) was carried out. Approval for the study was granted by the Emory University institutional review board. Data were then obtained through chart reviews of electronic medical records and review of radiographic images. All information pertaining to age, baseline demographics, past medical history, admission glucose, admission National Institutes of Health Stroke Scale (NIHSS) score, and time from last known normal to arterial access were collated for this analysis.

Imaging interpretation

All patients undergoing endovascular therapy had pretreatment head CT scans evaluated for signs of early ischemic changes, as defined by the Alberta Stroke Program Early CT Score (ASPECTS).14 Procedure times were calculated as the duration between the start of femoral access and the time of final device pass. Reperfusion rates were graded based on the modified Thrombolysis in Cerebral Infarction (mTICI) scoring system. mTICI scores of 2b (perfusion with >50% distal branch filling) and 3 (perfusion with filling in all distal branches) were considered successful, whereas mTICI 2a or less were considered unsuccessful. All intracranial hemorrhages were classified using the European Cooperative Acute Stroke Study (ECASS) definition,15 and evaluated on CT images obtained within 24 h of the procedure. Symptomatic hemorrhages were defined as parenchymal hematomas type 1 or type 2 with space-occupying effects within the infarcted area. Postprocedural infarct volumes were calculated by summating the total area of infarct on each image slice (diffusion-weighted imaging MRI or 48 h head CT) and multiplying by the slice thickness.

Patient selection

All patients with acute stroke were selected for mechanical thrombectomy based on previously described protocols.16 Depending on the time period, device availability, and operator preference, patients were treated with one of three reperfusion options: the Merci retrievers (Stryker Neurovascular, Mountain View, California, USA), the Penumbra aspiration system (Penumbra Inc, Alameda, California, USA), or a stent-retriever device with either the Solitaire FR revascularization device (Covidien, Plymouth, Minnesota, USA) or the Trevo pro retrieval system (Stryker, Kalamazoo, Michigan, USA). Some patients received more than one thrombectomy device. All patients treated with intracranial stenting, extracranial stenting, angioplasty, and/or IA t-PA alone were excluded from the analysis.

Outcome measures

The modified Rankin Scale (mRS) score was used to evaluate each patient's functional independence at 90 days after treatment with endovascular therapy. The primary endpoint of ‘good outcomes’ was defined as an mRS of 0–2, and determined by an independent reviewer blinded to the procedure results. All evaluations were performed during follow-up examination visits, or by phone interviews with the aid of pre-set questionnaire.


IBM SPSS software V.20 was used to perform the statistical analysis. A two-sample t test and Mann–Whitney U test was performed to compare each normally and non-normally distributed continuous variable among the three devices. A Fisher’s exact test was also used to compare categorical variables and Spearman rank testing was performed to assess colinearity of the baseline variables. Non-colinear variables with p<0.20 on univariate testing were then included in the binary logistic regression model for interpretation.


Clinical and treatment characteristics

A total of 279 consecutive patients underwent AIS endovascular intervention during the study period. Of those, 37 patients were excluded (eight patients received intracranial stenting only, 16 patients received IA t-PA only, 12 patients received angioplasty only, and one patient did not require mechanical thrombectomy owing to spontaneous recanalization) to yield 242 patients for analysis. The mean age was 65.5±14.2, median NIHSS 20, and 51% were female. The median ASPECT upon arrival at hospital was 8. Site of occlusion included the middle cerebral artery (MCA) in 147 patients (61%) (MCA-M1 segment, n=125 (52%); MCA-M2 segment, n=22 (9%)), the internal carotid artery terminus in 77 patients (32%), and the basilar artery in 16 patients (7%).

Among the whole cohort, 17 (7%) patients had symptomatic intracranial hemorrhage (sICH). The baseline characteristics were similar between the sICH and the non-sICH groups (table 1), except for a lower pretreatment ASPECT score in the sICH group (sICH=6.5 vs non-sICH=7.66, p=0.012) and higher atrial fibrillation rate in the non-sICH group (sICH=5.9% vs non-sICH=31.56%, p=0.045) (table 1).

Table 1

Demographics and clinical characteristics

Clinical outcomes

At the time of the analysis, 100% of the patients had had their 90-day functional assessment. Table 2 summarizes the clinical outcome differences between the two groups. Patients with sICH had significantly longer procedure times (104.5 sICH vs 79.7 min non-sICH, p=0.032). This remained significant when controlled for ASPECT scores, type of retrieval device, TICI scores, volume of infarct, interval from symptoms onset to puncture, and other comorbidities (OR=0.974, 95% CI 0.957 to 0.991, p=0.003). The mean procedure time was significantly shorter in patients with good outcome (86.7 vs 73.1 min respectively, p=0.0228). This was not the case after controlling for the previously mentioned factors (p=0.0228) (table 2) (figure 1). There was no association between procedure time and 90-day mortality (77.8 vs 88.2 min for survivals vs deaths, respectively; p=0.0958) (table 2). Also, there was no correlation between the volume of infarction and the procedure time (r=0.10996, p=0.0984) (figure 2).

Table 2

Primary and secondary outcomes association with procedure time

Figure 1

The difference in mean procedure time between patients with and without symptomatic intracranial hemorrhage (sICH) (left) and patients with good and bad clinical outcomes.

Figure 2

Correlation between procedure time (min) and final infarct volume (cm3).


Although our analysis showed no association between longer procedure times and final infarct volume or long-term functional outcome, there was a significant association with sICH. This suggests another opportunity for improving the safety and efficiency of endovascular reperfusion procedures for LVO. It also introduced a new stroke metric—namely, the procedure time or P2R. The recent awareness of the importance of establishing CSCs, with the capability of rapid and efficient stroke reperfusion therapy for LVO, comes from understanding the negative effects of time delays on clinical outcome. Stroke metric compliance will lead to better clinical outcome.17 However, such metrics continue to change and a general agreement on standard stroke metrics for CSCs is yet to be reached. Multiple lessons should be learnt from the cardiac literature in establishing stroke metrics in view of the major similarities.

Similar to stroke reperfusion therapy, delaying primary angioplasty (percutaneous transluminal coronary angioplasty (PTCA)) in patients presenting with STEMI was associated with increased mortality.18 ,19 This led the AHA to publish its first performance guidelines in 1999, promoting the current standard of care metric—namely, D2B of <90 min.20 Since then, owing to an understanding of the effect of compliance with D2B <90 min on the mortality rate in patients with STEMI, more and more centers have achieved this metric. Reportedly, it was achieved in 52% of patients in the collaborative registry in 2005, and in 76% of patients in the CathPCI group in 2008.21 ,22

Existing data about time metrics for endovascular reperfusion therapy for stroke are still scant. Even though a procedure time of <60 min was previously reported as appropriate for predicting patient outcome,23 a debate still exists as to whether the successful recanalization time or the procedure initiation time, known as the puncture time, should be considered the time endpoint. The AHA/ASA in its recommendations in 2011 considered an ‘arrival-to-treatment’ time of <2 h to be acceptable for endovascular reperfusion therapy. Physiologically the recanalization time would be preferable as the endpoint provided that the procedure initiation time is reasonable. This is a challenge for the neurointerventionalist as the procedure time itself is not predictable. Another challenge is the definition of recanalization as it is not uncommon to achieve partial recanalization before complete recanalization; this is especially true with the new technology of stentrievers.

The first studied time metric to enhance good clinical outcome in patients with AIS undergoing endovascular reperfusion therapy was described by Sun et al.24 This new time metric was the ‘picture-to-puncture’ time or P2P time, which is the time from obtaining the first image of the brain to the time of initiating the procedure or groin puncture. This report compared the P2P time between patients transferred to the facility for endovascular reperfusion therapy and patients presenting to the facility ED. The P2P was significantly lower in patients who presented locally to the facility ED, which had a positive effect on the clinical outcome. Moreover, in a logistic regression model, the P2P times were independently associated with good outcomes (OR=0.994; 95% CI 0.990 to 0.999; p=0.009).

This led to the development of a multicenter Rapid Reperfusion Registry that described another new time metric for endovascular reperfusion therapy—namely, D2P.13 Even though only 52% of the Registry achieved the goal metric of a D2P time of <2 h, short D2P times (OR=0.993, 95% CI 0.990 to 0.996; p<0.001) were associated with good outcomes. Based on this Rapid Reperfusion Registry data, the D2P time is considered clinically relevant and is the most studied time metric by which system processes can be targeted to streamline the delivery of endovascular reperfusion therapy. However, the cardiology universal and most usable time metric for PTCA is D2B time, which clearly involves the time from groin puncture to recanalization or the procedure time. Although, as mentioned above, physiologically the time to recanalization is the most relevant to good clinical outcome, this was not thoroughly studied and was not included in the previous new time metrics. Our small sample size failed to show a difference in functional outcome based on the procedure time, but since the procedure time was an independent risk factor for sICH it might have an impact on a larger scale and should be studied more comprehensively.

The correlation we found between a longer procedure time and the rate of sICH is consistent with the results of many recently published clinical trials that proved that endovascular reperfusion therapy is better than medical treatment alone in patients with LVO. The sICH rate was lower than previously reported in EXTEND IA (0%), SWIFT PRIME (0%), and ESCAPE (3.6%), where the procedure time median was reported to be <1 h (43, 28, and 33 min, respectively).7–9 The sICH rate in REVASCAT was higher than in the other trials (4.9%) and the median procedure time was significantly longer than an hour (86 min).10 Moreover, a recent analysis showed that a procedure time beyond 60 min increases complications, such as sICH, and device cost rates while worsening outcome.25 Although these numbers are worth mentioning, a firm conclusion cannot be reached until a systematic review adjusting for different variables is carried out.

The final outcome, including the functional outcome, the sICH rate, and the final infarct volume, might be dependent on other factors such as devices used during the thrombectomy procedure (long access sheath vs regular length sheath, balloon guide catheter vs regular guide catheter, stentrievers vs aspiration vs Merci retriever device, etc) the vascular territory, operator experience, and the difficulty of the individual anatomy. However, this was not examined in our cohort, owing to the retrospective nature of the analysis, These factors would be better compared individually in a randomized controlled trial.


Our initial data suggest that a longer procedure time is an independent risk factor for sICH in patients with LVO receiving endovascular reperfusion therapy. No definite association between procedural times and the final infarction volume or long-term functional outcomes was found after adjustment for multiple imbalances. This supports the need for a randomized clinical trial to study the effect of the procedure time on the functional clinical outcome in these patients and the need to include it in a justified time metric for CSCs.



  • Twitter Follow Taha Kass-Hout at @kasshout

  • Contributors Conceived and designed the research: RN, RG. Acquired the data: C-HJS, TK-H. Analyzed and interpreted the data; performed the statistical analysis: C-HJS, TK-H, TAK-H, OK-H. Handled funding and supervision: RN, RG. Drafted the manuscript: TK-H, RN, RG. Made critical revision of the manuscript for important intellectual content: all authors.

  • Competing interests RN: Stryker Neurovascular (Trevo-2 and DAWN Trials principal instigator); Covidien (SWIFT and SWIFT Prime Trials Steering Committee; STAR Trial Angiographic Core Laboratory), Penumbra (3D Separator Trial Executive Committee), Rapid Medical (Stroke Trial Data Safety Monitoring Board). RG: consultant: Stryker Neurovascular, Covidien, Rapid Medical; Steering Committee for Penumbra THERAPY trial; associate editor Journal of Neuroimaging; associate editor Interventional Neurology; royalties from UpToDate.

  • Provenance and peer review Not commissioned; externally peer reviewed.