Carotid stenting still shown to increase stroke compared to CEA surgery

February 27, 2010

Two major studies were reported this week testing different procedures to prevent stroke: carotid endarterectomy (CEA) and carotid artery stenting (CAS). The CREST study received most of the headlines as a “long –awaited” study that finally vindicated stenting as being as safe as the open-incision neck surgery. The primary endpoint was a composite of: “any stroke, MI, or death within 30 days plus subsequent ipsilateral stroke”. In both the CAS and CEA groups, this composite endpoint was statistically equivalent. However, the pure stroke component was 78% greater for the stent cohort (4.1% v 2.3%), consistent with previous studies. Conversely, the MI component was 110% greater in the CEA group (2.3% v 1.1%).

The other major study released this week was the ICSS from Europe. These data were just from an interim analysis. Therefore, it received less press coverage. In contrast to the CREST study, the primary endpoint in the ICSS was a single clinical outcome of “serious” stroke, which is more clinically relevant. In the stent group, all forms of stroke were 88% greater (7.7% v 4.1%) in the stent group.

Most carotid stenting trials recognize the difference between small brain infarcts from emboli released during the procedure and “serious” strokes caused years later unrelated to the type of preventive procedure. In both the CREST and ICSS trial, the minor strokes caused by the procedure were nearly twice as common with stents. A subset of ICSS patient underwent MRI-imaging and three times the number of stent patients revealed small brain infarcts than the CEA group.

What is a “minor stroke”? A good example was seen in the world of sport a few years ago. The famous New England Patriot All-Pro linebacker Tedy Bruschi suffered a short-term “mild” stroke that forced him to retire. He was partially paralyzed on an entire side of his body.

Proponents of the CREST trial who believe that carotid stenting is a valid alternative to CEA surgery point out that the periprocedural heart attack (MI) rate offsets the risk of minor stroke. However, as was seen recently with Vice President Dick Cheney, modern “heart attacks” are often small infarcts diagnosed only by enzyme elevations. It is unknown at this time how many of the MI’s seen in the CEA group were true Q-wave serious “heart attacks” and how many were “Dick Cheney” attacks.

The HCC interviewed one of the investigators of the CREST trial, Dr. Nick Hopkins, about the issues raised above. Dr. Hopkins is Chairman of Neurosurgery and Professor of Radiology at the University at Buffalo, State University of New York. Regarding the rationale for using a composite endpoint in CREST rather than a single clinical endpoint as in the ICSS, he replied, “Stroke alone ignores one of the most common and important complications of CEA surgery: MI. The composite endpoint is the only “all in” way to evaluate these procedures so clinicians can make intelligent judgments….Doctors will now better understand the risks and benefits of both procedures for each patient so they can make the right choice for each individual. ICSS has many issues and is an interim analysis only.”

Dr. Hopkins declined to comment on the merits of ICSS using the single parameter of “serious” stroke as the primary endpoint. In previous interviews, Drs. Sanjay Kaul, Gordon Guyatt, and Nortin Hadler have explained how composite endpoints can lead to misleading outcomes favoring the drug or device under study.

Critics of the ICSS trial that showed much higher rates of MRI-image-verified new brain infarct lesions and clinically diagnosed stroke point out that the CREST trial used a single type of Abbott device. They also claim that the CREST trial doctors were better trained (see also Dr. White’s comments). Dr. Hopkins wrote, “Credentialing was significantly less rigorous in ICSS than in CREST. “Experienced” interventionalists only had to have done 10 CAS procedures to begin randomization…Telltale numbers…64 CAS procedures were aborted (8%) whereas only 2 CEA’s were aborted…and Experienced centers fared less well than supervised centers. In ICSS two centers were removed from the study but not until they randomized 11 patients resulting in 5 disabling/ fatal strokes. This might suggest CAS operators were overall on an earlier learning curve compared to CREST where we required committee review of data and if that looked good, 20 lead in cases prior to randomization and formal didactic and observational training before randomization.”

One could argue, however, that the ICSS mirrors the real world better than CREST and would likely be what the U.S. would see once Medicare began to reimburse for carotid stenting and let the Genie out of the bottle. Thousands of doctors with little to no experience at carotid stenting and neurovascular anatomy, using numerous different makes and models of stents and filters, would begin stenting millions of patients. The ICSS outcomes are precisely what CMS (Medicare) is hesitant to unleash with a national coverage decision.

The full data from CREST need to be evaluated before the medical community, FDA, and CMS officials can properly weigh the pros and cons of carotid stenting versus CEA surgery. If the MI’s seen in CREST were of the mild enzyme elevation variety, then the temporary paralysis and neurocognitive disability caused more commonly by stenting will not be warranted. On the other hand, permanent vocal cord damage from laryngeal nerve damage and serious Q-wave MI’s during CEA surgery would support stenting.

2 Comments

• 

  By Neil E Strickman MD, February 28, 2010 @ 2:00 am

  Finally data again release. i have been involved in 984
  caarotid stents since 1998 at the texas heart Institute..Our CVA rate is <1.0%….Minor strokes TIA’s are 3% which resolve < 24 hrs…How many surgeons have a neurologist see their CEA pre and post…how about 0%…How many surgeons have carotid dopplers follow-up for 5 years..how about 0%…No procedure in the last century has been more scrutinized than CAS..the only reason medicare has not sanctioned it is $$$$$$$$$$$$$$$$$$$$$$$$…also surgeons can never learn how to do it unless they have endovascular training ( some do )..until that day the fight will rage on

  Neil E Strickman MD

• 

  By Atul Gawande, MD, March 1, 2010 @ 5:43 pm

  You have it spot on. Many thanks

  Atul Gawande, MD
  Harvard Medical School

Other Links to this Post

Notice: comments_rss_link is deprecated since version 2.5! Use post_comments_feed_link() instead. in /data/30/1/121/80/1121895/user/1185656/htdocs/wordpress/wp-includes/functions.php on line 3406
RSS feed for comments on this post. TrackBack URI

Leave a comment

*Name

*Email (not published)

Website