Heterogeneity of ICAD

ICAD is complex and may cause symptomatic stroke or transient ischemic attacks (TIA) in various ways. First, severe atherosclerotic disease may exhibit irregular or ruptured plaque which may lead to emboli distal to the stenosis. Second, severe ICAD in a relatively isolated vascular territory may lead to ischemic hypoperfusion in that territory. Finally, a feature that is distinct from coronary atherosclerotic disease, ICAD may lead to perforator strokes due to plaque occlusion of side perforator arteries. Based on these presentations, it would be reasonable to speculate that medical therapy may be best in some presentations and stenting better in others (in fact, stenting may be relatively contraindicated in the third group due to the possibility of the ‘snow plow’ effect worsening perforator occlusion). Unfortunately, all three of these patient groups were combined in the SAMMPRIS study, making analysis difficult.

Patients with the second type of presentation are often hemodynamically unstable with regard to cerebral blood flow, sometimes requiring admission to an intensive care unit with crescendo TIAs or strokes. These high-risk critical patients, who were deteriorating on maximum medical therapy, were largely not included in the SAMMPRIS study. In fact, most patients were evaluated in an outpatient clinic after they had come through their critical ischemic event. As a result, many patients who were deemed to be too critical for the trial were stented outside of the trial. Although the initial studies evaluating coiling for cerebral aneurysms and stenting for carotid disease4 were performed in patients considered to be high-risk, the SAMMPRIS study excluded the higher risk patients from the trial, possibly missing those who may have benefited most from stenting.

Off-label stent use from FDA approval

Some very unconventional techniques were used in the SAMMPRIS trial design that made the results dubious. For example, 35% of the patients stented in the trial were stented in an off-label use of the FDA-approved Humanitarian Device Exemption indication for the Wingspan stent. The FDA approval for this device was for patients who had failed medical therapy. Over a third of the patients who were enrolled in the stenting arm had a stroke or TIA while not on antiplatelet therapy. The predicate registries and trials used the Wingspan according to its FDA indication of patients who had failed medical therapy.5–7 The Clinical Outcomes Utilizing Revascularization and Aggressive DruG Evaluation (COURAGE) trial,8 which evaluated aggressive medical therapy for coronary artery disease, showed that medical therapy should be initiated first with symptoms and stenting should be considered when medical therapy fails. Why shouldn't this paradigm also hold true in ICAD?

Operator experience

Another issue plaguing the SAMMPRIS study is the effect on the results of operator experience. The study was initially designed to use only operators who had performed a minimum of 20 Wingspan stents; however, relatively few operators met this criterion. As a result, many operators were allowed to participate by substituting experience in other stent systems used for aneurysms, not ICAD. Moreover, intracranial angioplasty alone was permitted as a substitute as part of these required cases. As has been demonstrated quite clearly in cervical carotid angioplasty and stenting, operator experience is critical in reducing intraprocedural complications. In the EVA-3S carotid stenting trial,9 with minimal operator experience, clinical periprocedural data showed some of the worst results for a carotid stenting trial in the past decade. In contrast, the Carotid Revascularization Endarterectomy Versus Stenting Trial (CREST) trial10 used very experienced operators and produced excellent clinical results in the carotid stenting group compared with surgery.

It has been commented that there was no significant difference between the high and low enrolling sites with regard to stenting complications. However, this distinction between high and low enrolling sites may have been the difference between an operator performing five or eight stents in the trial. Both of these experiences are very low compared with other device trials. Studies from China (where ICAD is a considerably more prevalent cause of stroke) involving operators with much more experience than the US physicians have had superior results with intracranial stenting. One recent study demonstrated a 1-year stroke and death rate of 7.3% in a cohort of patients treated with the Wingspan stent for >70% symptomatic stenosis.11 This is superior to the results from the aggressive medical arm of the SAMMPRIS study (12.2%), indicating that operator experience is key in reducing periprocedural complications and events. Similarly, a recent multicenter study of over 600 stented patients with experienced operators showed an event rate of 6.1% during the 30 days post-procedure compared with an event rate of 14.7% in the SAMMPRIS study.12

Resistance to antiplatelet therapy

While the study investigated the idealized aggressive medical therapy in the medical arm, the idealized therapy for stenting was not implemented. In fact, some would argue that not even the standard medical therapy was performed in the stenting arm. Although every patient in the trial was treated with dual antiplatelet therapy with aspirin and clopidogrel, no assessment was made by any test to determine whether the patients were at a therapeutic level of these medications or whether they were resistant. The most common clinical complications associated with intracranial stenting for ICAD (or other pathologies) are thromboembolic events. In fact, 76% of the strokes in the stenting arm of the SAMMPRIS study occurred within 24 h, suggesting inadequate antiplatelet or anticoagulant therapy. We also know from a large clinical coronary study of over 1200 patients13 that approximately 14% of the study population were resistant to clopidogrel and 9.7% were resistant to aspirin. Furthermore, it was demonstrated in this trial that patients who were stented and found to have clopidogrel resistance had a statistically higher risk for thromboembolic complications.13 Patients with placement of a new intravascular stent are more likely to thrombose the stent acutely or develop thromboembolic events with subtherapeutic antiplatelet therapy than those with a chronic arterial plaque. Not testing patients for resistance to antiplatelet therapy puts those stented in the study at a clear disadvantage. While many top centers for ICAD treatment routinely perform these antiplatelet resistance tests for patient safety,14 this was not a requirement in the trial.

Finally, although aggressive medical therapy performed better than expected in the SAMMPRIS trial, it did not perform well. In the study, 12.2% of the patients in the medical therapy arm had a recurrent stroke or died within a year. Furthermore, in the SAMMPRIS trial there was a higher rate of stroke and death in the period between 30 days after enrollment and 1 year. There was a 1.1% absolute higher event rate in the medical arm, which represents a 20% increase in the RR in the medical arm in this period, suggesting that this therapy is not as durable as we would like.

Target artery issues

The diameter of the arteries stented may also have an impact on the study results. The Latin American Small Vessel Study (LASMAL)15 demonstrated statistically worse clinical results in patients who underwent stenting in arteries ≤2.7 mm compared with larger arteries (≥3.4 mm). In the SAMMPRIS trial, 41% of the patients underwent stenting of middle cerebral arteries in this small diameter range. Additionally, stenting was performed without regard for adjacent perforators. Since perforator strokes occurred in the trial in the M1 segment of the middle cerebral artery and the perforator risk segment of the basilar trunk, this should be taken into consideration for future trials.