Bandolier 92 reported a meta-analysis of trials looking at a comparison of stents
versus balloon angioplasty indicating stents to be better. With stents the death rate
was 3.8% and reinfarction rate 2.1% over six months in randomised trials up to 1999.
The interesting question is whether this retrospective analysis represents what is
happening in today's clinical practice, with technical advances and improved
anticoagulation. An analysis of ongoing studies suggests that death and reinfarction
rates are lower now than they were [1].
Study
Six major clinical trials of coronary stenting were coordinated from Boston, all
using similar inclusion criteria and protocols, but with different stents. Protocols
used routine high-pressure postdilation, and with aspirin 325 mg daily and
ticlopidine 250 mg twice daily for four weeks, but with glycoprotein IIb/IIIa
inhibitors discouraged.
All used stent thrombosis as a clinical endpoint, and with the same definition of
clinical and angiographic stent thrombosis and myocardial infarction. From the paper
it appears that some, but not all, of the trials were randomised.
Results
In all there were 6,186 patients who received a stent. Thrombosis occurred in 53 of
them (0.9%), and thrombosis was confirmed angiographically in 45 of the 53. Most
stent thromboses occurred within the first two days after the procedure (Figure 1),
and events after one week were rare. With thrombosis mortality was 19% at one month
and 21% at six months. Mortality or myocardial infarction with thrombosis was 70% at
one or six months. Overall mortality was 10 in 6,186 (0.2%) at one month.
Figure 1: Thrombosis after stenting
|
Predictors for clinical stent thrombosis were persistent dissection after
stenting, longer stent length and final minimal lumen diameter. Unstable angina,
use of glycoprotein IIb/IIIa inhibitors, nonrandomised versus randomised status
nor stent type were not associated with any significant stent thrombosis.
Comment
Bandolier
occasionally sees surgeons getting a bashing for not being "evidence-based"
about their work. While that may occasionally be true, there are lots of
systematic reviews for surgery and the perioperative period, and lots of
randomised trials. But the challenges are different and sometimes different
approaches are needed.
Here we see the benefits of a registry of ongoing trials, some randomised, some
not, but all benefiting from common outcomes and design that allows mature
overview. In industrial manufacture this would be called quality control, a sort
of high-level "kaizen" working to continually improve quality and
competitiveness.
The bottom line for the moment is that the prognosis with stent thrombosis is
dire. But we know when it occurs, and we know some of the predictors, so that
although it is uncommon, that already low rate can be reduced even further.
References:
- 1 DE Cutlip et al. Stent thrombosis in the modern era. A pooled analysis of
multicentre coronary stent clinical trials. Circulation 2001 103:
1967-1971.
|
previous
or
next
story in this issue
