Transcript Definitions of Stent Thrombosis

Reconciling Disparate Results in Analyses of
Stent Thrombosis in the use of Drug-Eluting
Stents
“Per Protocol” Definitions of Stent Thrombosis
Early Stent Thrombosis (≤ 30 days)
• Thrombotic occlusion of the stented segment observed
at the time of a clinically-driven angiographic restudy
• Q-wave MI in territory of stented vessel
• Any death not attributed to a non-cardiac cause
Late Stent Thrombosis (>30 days)
• MI attributable to the target vessel, with angiographic
documentation of thrombus or total occlusion of the
target site, and freedom from an interim
revascularization of the target vessel
www. Clinical trial results.org
CM Gibson 2006
ARC Definitions of Stent Thrombosis
Definite/Confirmed
• Acute coronary syndrome AND
• Angiographic confirmation of stent thrombus or occlusion
OR
• Pathologic confirmation of acute stent thrombosis
Probable
• Unexplained death within 30 days
• Target vessel MI without angiographic confirmation of stent thrombosis or
other identified culprit lesion
Possible
• Unexplained death after 30 days
• Note: Patients who have a TLR prior to stent thrombosis are included by
this set of definitions, as opposed to the “Per Protocol” definition
www. Clinical trial results.org
CM Gibson 2006
Potential Issues Arising From Different Definitions:
Stent Thrombosis
• Stent thrombosis may be angiographically confirmed
in some analyses but not in other analyses.
• If stent thrombosis is angiographically confirmed, the
patient, by definition, survived until the time of
angiography. It is therefore no surprise that
angiographically confirmed stent thrombosis has a
weak association if any with mortality.
www. Clinical trial results.org
CM Gibson 2006
Potential Issues Arising from Different Definitions:
Recurrent MI
• If recurrent myocardial infarction is used in the definition, the
definition of myocardial infarction should be provided.
• Some analyses define MI as a Q wave MI. It should be
realized that some patients with ST elevation will, and others
will not, evolve a Q wave. Likewise, some patients with a nonSTEMI elevation will go on to develop a Q wave.
• More detailed data is needed to understand if the event was
the result of recurrent ST elevation MI, or if the patient
presented with a positive troponin without ST elevation. It is
unlikely that restenosis presents with ST elevation MI, but it
may present in some cases with a troponin elevation.
• If ST elevation MI and troponin elevation are lumped together
in the definition of recurrent MI this should be specified as they
may carry quite different prognoses.
www. Clinical trial results.org
CM Gibson 2006
Potential Issues Arising from Different Definitions:
Mortality
• It should also be specified if mortality was included
in the definition, and if cardiac and non-cardiac
mortality are being evaluated separately or
together.
• It is also often quite difficult to ascertain the true
cause of death. The presence of a terminal illness
such as cancer does not exclude the possibility that
a patient may have sustained sudden death from
stent thrombosis, particularly if the illness is
associated with a prothrombotic state.
www. Clinical trial results.org
CM Gibson 2006
Other Complications in DES Analysis
• The analysis of stent thrombosis data can be
complicated by other factors:
– Both DES and BMS may be placed in the same patient.
– DES from different manufacturers may be placed in the
same patient.
– A DES may be placed inside a BMS for restenosis.
– Adjustments must be made for the concordance in the
behavior of the stents.
– Finally, thrombosis can occur outside of the stented
segment.
www. Clinical trial results.org
CM Gibson 2006
Randomized Trials vs Registry Data
•Randomized trial data and registry data will provide
different insights into the issue of stent thrombosis:
– Randomized trial data does reduce the potential for
confounding with respect to the procedural characteristics
that led an operator to place either a bare metal stent or a
drug eluting stent.
– However, randomized trial data provides no guidance
regarding clinical outcomes in more complex lesions and in
patients with co-morbid conditions as these high risk
lesions/patients are often excluded from randomized trials.
www. Clinical trial results.org
CM Gibson 2006
Randomized Trials and Registry Data
• While registry data does provide insight into “real
world” practice patterns and outcomes, these data
do not adjust for the fact that bare metal stents and
drug eluting stents may be placed in different
lesion/patient subsets.
• For instance, in the DEScover Registry, a greater
proportion of bare metal stents were placed in
STEMI patients, and this patient population is at
higher risk of mortality.
www. Clinical trial results.org
CM Gibson 2006
Public Health Considerations
• In placing these data in context from a public health
perspective, it should be borne in mind that
approximately 1 million stents are inserted each year.
• It will be important to determine if the excess risk of
stent thrombosis is confined to a limited period of time,
or if the excess risk of stent thrombosis persists in
each subsequent year after a drug eluting stent is
placed.
www. Clinical trial results.org
CM Gibson 2006
Lesion Revascularization and Stent Thrombosis over
20 years
Speculative model comparing Lesion Revascularization
• The model is based
vs Stent Thrombosis over 20 years
0.12
on a steady 0.5% rate
of stent thrombosis
0.1
each year
• If a cohort of 1 million
0.08
patients from 2006 is
followed for 20 years
0.06
with a stent
thrombosis rate of
0.04
0.5% each year, this
would yield over
0.02
95,000 stent
0
thromboses within
that cohort over a 20
Lesion Revascularization
Stent Thrombosis
year period
www. Clinical trial results.org
CM Gibson 2006
Lesion Revascularization and Stent Thrombosis over
20 years
• Even with the conservative estimate provided by the
Cardiovascular Research Foundation group (1 thrombosis for
500 patient years of exposure), if a cohort of 1 million patients
from 2006 is followed for 20 years, this would yield roughly
40,000 stent thromboses
• At this conservative rate, the total number of stent thromboses
over a 20 year period would be roughly 800,000.
www. Clinical trial results.org
CM Gibson 2006
Thienopyridine Therapy
• If there is in fact an excess of stent thromboses associated
with DES use, prolonged thienopyridine therapy has been
proposed as one solution to reduce the risk of thrombosis.
• The potential benefit of prolonged thienopyridine therapy in
reducing stent thrombosis, however, has not been evaluated in
prospective randomized trials.
• Prolonged thienopyridine therapy is not without risk and may
be associated with a hazard of bleeding.
• While there is a higher cost associated with DES vs BMS
devices, the societal cost of treating millions of patients with
prolonged thienopyridine therapy could be quite high as well.
• The cost of prolonged thienopyridine therapy, the cost of stent
thromboses and the excess cost of a DES vs a BMS must be
weighed against the cost of repeat revascularization with BMS
as well.
www. Clinical trial results.org
CM Gibson 2006
Questions and Considerations
• These data raise important questions regarding the adoption of
new device technologies:
–Should a new device (DES) replace a current device (BMS) if short
term efficacy has been demonstrated in a surrogate endpoint
(restenosis), but if long term safety data are not yet available?
–How are we as a community to deal with the evaluation of
catastrophic but infrequent adverse events?
–Should the level of certainty surrounding safety be higher for a
permanently implanted device than a drug which can be
discontinued?
–How many patient years of exposure in randomized trials is
necessary before a permanently implanted device can be assumed
to be safe?
–What are the relative risks and benefits of drug eluting stents over
aggressive medical therapy in the treatment of low risk patients
with stable angina, given that the AVERT trial showed no benefit of
percutaneous intervention in these patients (N Eng J Med 1999,
341: 70-76)?
www. Clinical trial results.org
CM Gibson 2006
Conclusions
• Future randomized trials of 8,000 to 30,000 patients
have been proposed to evaluate the relative risks and
benefits of DES vs BMS.
• In parallel, efforts are also underway to develop
standardized definitions of stent thrombosis to facilitate
the comparison of trial data.
www. Clinical trial results.org
CM Gibson 2006