CREST Subanalysis Stirs Controversy Regarding Possible Gender Differences

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

A sex-specific subgroup analysis from the CREST trial suggests women are more than twice as likely to suffer a postprocedural stroke following carotid artery stenting (CAS) than after carotid endarterectomy (CEA). The analysis, published online May 9, 2011, ahead of print in Lancet Neurology, found no such difference between the 2 treatments in men and no sex-interaction in the primary endpoint of the trial, raising numerous questions and causing the study investigators and others to urge caution in interpreting the data.

The Carotid Revascularization Endarterectomy vs. Stenting Trial (CREST), originally published in the July 1, 2010, issue of the New England Journal of Medicine, randomized 2,502 patients with either symptomatic or asymptomatic disease to undergo CEA or CAS. Overall, the trial found no significant difference in the estimated 4-year rates of the primary endpoint (composite of any periprocedural stroke, MI, or death, or the incidence of ipsilateral stroke ≤ 4 years) between the 2 groups. While there was a higher risk of stroke with stenting and a higher risk of MI with surgery, there was no evidence of a difference between sexes in rates of stroke (P = 0.65) or stroke or death (P = 0.79).

Although the CREST trial set a recruitment goal of 40% women to provide reasonable power to detect potential treatment differences, the final female enrollment rate was only 34.9%. For the new subanalysis, CREST researchers led by Thomas G. Brott, MD, of the Mayo Clinic (Jacksonville, FL), examined sex-specific differences in the primary endpoint as well as its individual components.

Higher Periprocedural Stroke Rate in Women

Although the primary endpoint was similar between the sexes at 4 years, during the periprocedural period a difference was observed in women but not seen in men (table 1).

Table 1. Periprocedural Rates of Primary Endpoint by Sex

 
This difference was explained by a more than doubling of the stroke rate in women receiving CAS compared with CEA (table 2).

Table 2. Periprocedural Stroke Rate by Sex

The test for a treatment-by-sex interaction was not statistically significant for the periprocedural period (P = 0.33 for symptomatic status and 0.52 for age) or at 4 years (P = 0.35 for symptomatic status and 0.45 for age).

Sex Should Factor Into Decision-Making Process

Several of the study authors were investigators in ACAS (Asymptomatic Carotid Atherosclerosis Study), which was the first CEA trial to report the post-hoc finding that women had a higher perioperative stroke and death rate than men; however, this finding was not statistically significant.

They say this additional information from the CREST subanalysis supports the notion that sex should be taken into account in decisions regarding treatment of carotid disease.

“Additional pooled analyses of data including CREST might provide a partial answer to the question of whether the patient’s sex is an important consideration in selecting revascularization technique,” Dr. Brott and colleagues write.

One possible explanation is that women have internal carotid arteries that are on average about 40% smaller than those of men. In addition, women in CREST were more likely than men to be hypertensive and to have a higher mean systolic blood pressure, lower mean diastolic blood pressure, and lower weight. For patients assigned to CAS who received a procedural angiogram, lesion length was significantly shorter for women than for men. Further analyses of procedural angiograms are underway, Dr. Brott and colleagues note.

But the study authors note that the findings also may be “a result of spurious relations introduced as a natural product of the large number of associations that were assessed.” For this reason, they urge caution in the interpretation of the results.

This sentiment is echoed by Martin M. Brown, MD, and Rosalind Raine, MBBS, PhD, of University College London (London, United Kingdom), in an accompanying commentary.

Difficult to Form Conclusions Based on Single Subanalysis

According to Drs. Brown and Raine, the findings are not consistent with other trial data, specifically the Carotid Stenting Trialists Collaboration (CSTC) analysis of individual data from 3 European-based randomized trials of CAS vs. CEA. In that trial, “the differences in risk go in the opposite direction to CREST, with risks of any stroke or death after carotid artery stenting of 8.5% in women and 9.0% in men and a lower carotid artery stenting to carotid endarterectomy risk ratio in women than men (1.22 vs. 1.68),” they write.

In their own meta-analysis of available data from CREST, CSTC, and CAVATAS, no effect of sex on the risks of either treatment was evident.

The editorialists also point out that CREST included both patients with asymptomatic and symptomatic stenosis, the latter being a much more powerful determinant of the risks of treatment than sex. In all, 466 women and 855 men had symptomatic stenosis, while 406 women and 775 men had asymptomatic stenosis.

“Within CREST, the subgroups are too small to form conclusions on the influence of sex within the symptomatic and asymptomatic groups separately,” Drs. Brown and Raine write. “Nevertheless, one could argue that the data from CREST do not justify either carotid artery stenting or carotid endarterectomy for asymptomatic [stenosis] . . . given the present evidence that the rate of stroke in patients with asymptomatic stenosis treated with optimum medical therapy is only about 1% per year.”

May Not Reflect Contemporary Stenting Practice

In a telephone interview with TCTMD, William A. Gray, MD, of Columbia University Medical Center (New York, NY), said the findings definitely come as a bit of a surprise given the initial CREST paper, which found no interaction for sex or symptomatic status in relation to the comparative differences between CAS and CEA.

“Now along comes this paper saying there may be an influence of sex,” he said. “Much of that difference is driven by symptomatic women, who account for 10% of the patient population. I just worry that the subset of women who drove this difference is so small as to be potentially confounded by statistical error or sampling error.”

Still, the higher stroke rate in symptomatic women is troubling and needs further clarification, Dr. Gray added. One clue may come from the learning curve of operators in the CREST trial. Unpublished data, he said, indicate that the temporal improvement in the last half of CREST was remarkable compared with the first half.

“Right now we don’t understand when these strokes occurred. If the majority occurred early in CREST, we need to recognize that the contemporary practice of carotid stenting may not be reflected in these data,” Dr. Gray said.

Michael R. Jaff, DO, of Massachusetts General Hospital (Boston, MA), told TCTMD in a telephone interview that he agreed with that assessment. “[The impact of the learning curve] is an excellent point,” he said. “That certainly could be, but it’s not reported by these investigators, so we just don’t know.”

Dr. Jaff said the take-home message is that the jury is still out on whether sex matters in carotid revascularization.

“This is a curious finding,” he said. “Obviously, you will figure this into your decision making, but it is just one of many factors. It’s not a deal breaker. It’s just another piece of information to put in your hat.”

 

---

Sources:
1. Howard VJ, Lutsep HL, Mackey A, et al. Influence of sex on outcomes of stenting versus endarterectomy: a subgroup analysis of the Carotid Revascularization Endarterectomy versus Stenting Trial (CREST). Lancet Neurol. 2011;Epub ahead of print.

2. Brown MM, Raine R. Should sex influence the choice between carotid stenting and carotid endarterectomy? Lancet Neurol. 2011;Epub ahead of print.

3. Brott TG, Hobson RW II, Howard G, et al. Stenting versus endarterectomy for treatment of carotid-artery stenosis. N Engl J Med. 2010;363:11–23.

 

 

Related Stories:

• CREST Analysis Breaks Down Results in Symptomatic, Asymptomatic Patients
• CREST Trial Results Continue to Cause Controversy
• CREST Published: Long-Term Outcomes Equivalent for Carotid Endarterectomy, Stenting