Many Factors Contributed Toward Periprocedural Stroke in SAMMPRIS Trial

Download this article's Factoid in PDF (& PPT for Gold Subscribers)

The randomized SAMMPRIS trial came to an early end when it picked up on markedly higher rates of early stroke or death in patients receiving intracranial stents plus medical therapy than in those treated with drugs alone. Now, a year after publication of the trial’s main results, investigators have identified multiple procedural and clinical reasons for the excess events. The new findings were published online September 13, 2012, ahead of print in Stroke.

SAMMPRIS, published last September in the New England Journal of Medicine, randomly assigned 451 patients who had recently experienced TIA or stroke attributed to a major intracranial artery (70 to 99% stenosis) to aggressive medical therapy with or without angioplasty and stenting (Wingspan; Boston Scientific, Natick, MA). Invasively treated patients were more than twice as likely to experience periprocedural stroke or death (≤ 30 days) as those randomized to medical therapy alone (14.7% vs. 5.8%; P = 0.002).

A Focused Look at Stroke

For the current study, David Fiorella, MD, PhD, of the State University of New York, Stony Brook (Stony Brook, NY), analyzed data from 224 patients randomized to the invasive arm of SAMMPRIS. Among them, 5 patients ultimately underwent angioplasty alone and 208 received both angioplasty and stenting.

A total of 33 periprocedural strokes occurred: 11 were deemed hemorrhagic and 19 ischemic (with an additional 3 ischemic strokes occurring in nonstented patients). The time between qualifying event and stenting appeared to have no effect on rates of either type of stroke.

On multivariate analysis, hemorrhagic stroke was associated with several factors (table 1).

Table 1. Factors Associated with Hemorrhagic Stroke



95% CI

P Value

Higher Percent Stenosis




Lower Modified Rankin Score (0 vs. 1-3)




Clopidogrel Loading with High Activated Clotting Time (> 300 sec)




On logistic regression analysis, several other factors were associated with ischemic stroke (table 2).

Table 2. Factors Associated with Ischemic Stroke



95% CI

P Value

Smoking (never vs. previous/current)




Symptomatic Artery (basilar vs. others)








Age (per 10-year increase)




Of the 19 periprocedural ischemic strokes, 12 were classified as predominantly perforator occlusions; 3 were embolic events, 2 were mixed perforator and embolic, and 2 were delayed stent occlusions. Six of the 11 symptomatic hemorrhagic strokes were parenchymal brain hemorrhages and 5 were subarachnoid hemorrhages.

Dr. Fiorella and colleagues draw attention to the strong influence of smoking status, with never smokers showing far higher rates of ischemic events than current or former smokers (18.1% vs. 4.7%). “This ‘smokers paradox,’ which has also been observed after coronary interventions, may be related to the fact that nonsmokers in SAMMPRIS were more likely to be female, diabetic, and hypertensive,” they write, adding that smokers also tend to respond better to clopidogrel.

Still a Role for Intracranial Stenting?

According to the authors, SAMMPRIS is the only large prospective cohort of patients undergoing stenting for intracranial stenosis by credentialed interventionalists at multiple sites with independently adjudicated endpoints. “As such,” they write, “the trial provides unique data to evaluate the technical success of [stenting] as well as the causes and risk factors for the cerebrovascular complications.”

Nevertheless, they add, “given the limitations of the present analysis, particularly the small number of events, the results should be viewed with caution and not used as the basis for clinical treatment without further confirmatory studies.”

Christopher J. White, MD, of the Ochsner Heart and Vascular Institute (New Orleans, LA), agreed that the risk factors identified in the analysis do not make patient selection easier. “The higher-risk features are broadly distributed in the population, making segregation into a lower-risk ‘treatment group’ very difficult,” he told TCTMD in an e-mail communication, adding that it is likely that the device used in the trial was not optimal

“Finally, there remains significant concern about how to treat patients who have ‘failed’ medical therapy, those that have recurrent TIA or minor strokes while on best medical therapy,” Dr. White noted. “It would seem that intervention should remain a viable option for them.”

SAMMPRIS as a Starting Point

In editorial accompanying the paper, Michael P. Marks, MD, of Stanford University Medical Center (Stanford, CA), also questioned the Wingspan system. “If we accept the current case made by the SAMMPRIS investigators that higher rates of stroke and death do not reflect inexperience and do not decline over time,” he said, “this suggests that there is an inherently high risk to the procedures with the device used in the trial.”

He cites a several “characteristics that could lead to increased complications,” including the relatively large 6-Fr guide catheter and the challenges presented by a 2-step procedure.

“Approaches besides the Wingspan may have intrinsic advantages,” Dr. Marks notes. For example, he suggests, “Angioplasty balloons are less rigid than self-expanding stents or balloon-expandable stents and are more easily trackable and less traumatic,” and they overcome the procedural difficulties specific to Wingspan.

Given the possible device limitations and less than ideal lesion morphology patients enrolled in SAMMPRIS, however, the results of that “trial should not stop further investigation of endovascular therapy for severe symptomatic intracranial stenosis,” Dr. Marks concludes, adding, “The SAMMPRIS trial was successful in 1 important way: it set a higher bar for the investigation of endovascular therapy for symptomatic intracranial stenosis. . . . The future for endovascular therapy in the management of this disease lies in its ability to demonstrate a benefit for those patients still at high risk for stroke, even with aggressive medical therapy.”

Study Details

Stenting was successful in 92% of the 214 patients in whom it was attempted. Before the procedure, the mean percent stenosis was 79.7%; this number dropped to 40.2% post angioplasty, and to 27.0% after the full procedure was completed.


1. Fiorella D, Derdeyn CP, Lynn MJ, et al. Detailed analysis of periprocedural strokes in patients undergoing intracranial stenting in stenting and aggressive medical management for preventing recurrent stroke in intracranial stenosis (SAMMPRIS). Stroke. 2012;Epub ahead of print.

2. Marks MP. Is there a future for endovascular treatment of intracranial atherosclerotic disease after stenting and aggressive medical management for preventing recurrent stroke and intracranial stenosis (SAMMPRIS)? Stroke. 2012;43:Epub ahead of print.



Related Stories:

  • The SAMMPRIS trial received research support from AstraZeneca, Nationwide Better Health-INTERVENT, and Stryker Neurovascular.
  • Dr. Fiorella reports receiving research support from Siemens Medical Imaging and Microvention; consulting fees from Codman/Johnson & Johnson, EV3/Covidien, NFocus, and WL Gore; honoraria from Scientia; and royalties from Codman/Johnson & Johnson. He also has ownership interests in CVSL and Vascular Simulations.
  • Dr. Marks reports serving as a consultant to Codman and Shurtleff.
  • Dr. White reports having served as principal investigator of the CABANA study, which was sponsored by Boston Scientific.

We Recommend