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

Since the 2003 US approval of drug-eluting stents (DES) to treat coronary artery disease, clinical attitudes toward and adoption of the devices has zigzagged from early enthusiasm to safety scare to measured recovery, a course that is reflected in registry data published online May 9, 2014, ahead of print in the International Journal of Cardiology.

Evidence suggests that in recent years DES use has become more discriminating, directed toward high-risk subgroups but also underused in the elderly, African Americans, and Medicaid patients. 

From Peak to Nadir in 2 Years

Shortly after US Food and Drug Administration (FDA) approval of DES 11 years ago, their use increased rapidly, reaching as high as 89% of all stent implantations in 2005 (P for trend < .001 for DES vs BMS). However, following concerns raised in 2006 about increased risk of late stent thrombosis and other adverse events associated with DES, deployment of the new devices fell precipitously to about 66% in 2007. Thereafter, DES use underwent a modest recovery, leveling off at about a 73% share of PCIs from 2009 through 2011.

With the year 2003 as a reference standard, the age- and sex-adjusted odds ratio of DES use reached a peak in 2006 at 15.56 (95% CI 15.44-15.69), with a reduction in 2011 to 5.15 (95% CI 5.11-5.19; both P < .0001). In contrast, compared to 2001, the adjusted odds ratio of BMS use dropped 95% by 2011.

Over the study period, DES followed similar patterns of use in patients at high risk of restenosis, including those with diabetes, chronic kidney disease, and prior PCI.

Trends among ACS patients, who carry a high risk of stent thrombosis, were somewhat more variable. In STEMI patients, DES use fell dramatically from a high in 2005 to a nadir in 2008, and then recovered modestly by 2011. The patterns for NSTEMI and unstable angina patients followed that of the overall population, although the post-decline recovery was stronger in the latter cohort (P < .001). In elective patients, on the other hand, DES use peaked at 91% in 2005 and declined to 71% in 2008, with a slight recovery seen by 2011.The trend in high bleeding risk patients likewise mirrored those of the general population (table 1).

Table 1. Trends in DES Use in High-Risk Patients

 
Among cohorts with a potential for disparity of care, no difference in usage patterns was seen between men and women. However, in recent years black patients tended to receive more BMS than white patients (in 2011, 29.73% vs 25.77%). Differences also emerged based on payer status, with highest DES use seen in patients with private insurance (79.0%), decreasing in those with Medicare (72.1%) or Medicaid (68.3%), and self-payers (53.7%). In addition, patients at least 75 years of age were less likely to receive DES than younger patients (in 2011: 66.8% vs 74.4%). In all cases, the differences arose after comparable DES use between the dichotomous groups during the peak year of 2005.

The Story Behind the Numbers

The authors identify 5 phases in the understanding of DES efficacy and safety, which likely account for the variable usage over the years:

1. Unbridled enthusiasm. The early phase, starting about 2003, was fueled by trial data showing a 91% reduction in in-stent restenosis (SIRIUS) and a 73% reduction in TLR (TAXUS-IV) with DES compared with BMS.
2. False sense of security. DES use rose rapidly to about 90% across all patient subgroups.
3. Rude awakening. In 2006, a ‘firestorm’ struck in the form of a presentation at the European Society of Cardiology congress suggesting a higher rate of stent thrombosis, death, and MI with DES compared with BMS. In addition, the BASKET-LATE trial showed a 3- to 4-fold increase in late death or MI and a rise in late stent thrombosis compared with BMS after clopidogrel discontinuation. Similarly, a meta-analysis of randomized trials found an increase in very late stent thrombosis with DES vs BMS.
4. Resetting of expectations. In 2007 after reviewing available data, an FDA advisory panel recognized a small but significant risk of late stent thrombosis but endorsed DES as safe and effective when used ‘on label.’ Moreover, a number of trials of SES and PES suggested that the short- and long-term risks of stent thrombosis were no greater for DES than BMS.
5. Persistent mistrust. In recent years, DES use appears to have become more discriminating, with relatively more implantations in patients at risk of restenosis and less in those at risk of stent thrombosis or bleeding.

In addition, Dr. Bangalore and colleagues note a “disturbing pattern” of decreased DES use in certain patient groups at risk for suboptimal care: the elderly, African-Americans, and those whose procedures are paid for by Medicaid or themselves.

With regard to STEMI patients, the investigators suspect that the 61% rate of DES use in 2011 represents a ‘legacy’ effect of first-generation devices, which carry an increased risk of very late stent thrombosis and reinfarction, adding that it is not surprising that ACC/AHA guidelines on STEMI do not strongly endorse DES.

But device design has improved significantly since the early days, with thinner struts and thinner, more biocompatible polymers, which have reduced the risk of these events, they assert. An analysis of 28 randomized trials with more than 34,000 patient-years showed that all FDA-approved DES were associated with substantial decrease in the risk of TVR compared with BMS. Moreover, cobalt-chromium everolimus-eluting stents have the added advantage of substantially reducing in stent thrombosis compared with first-generation DES.

“Given these newer data, it is uncertain whether the intermediate DES usage rates seen in 2011 are justifiable,” the authors conclude.

In terms of limitations, the investigators acknowledge that the administrative data on which the study is based do not differentiate between first- and second-generation DES, and it is unclear whether usage patterns may have differed between the 2 generations of devices. Moreover, since data were only examined through 2011, the findings cannot be considered ‘contemporary.’

Related Stories:

• DES: The Year in Review
• The COURAGE Effect: Sustained Decline Seen in PCI for Stable Angina