TOP
Clinical
Use of
tPA in Acute Ischemic Stroke
Introduction
It
has been over 5 years since the NINDS clinical trials showed the efficacy
of tPA in the treatment of patients with acute ischemic stroke.
1
Since that time, the opportunity to use this treatment modality
in the clinical setting has resulted in nine studies of its clinical
use.
2-10
This
discussion addresses the important clinical issues that must be considered
by the Emergency Medicine physician when considering the use of tPA
in patients with acute ischemic stroke, based on both the results
of the NINDS clinical trials and these subsequent nine clinical reports.
This discussion will focus on the most important issues regarding
tPA use, especially what these published studies tell us about how
it should be used, with what expected outcomes, as well as how decisions
regarding its use must be documented in the medical record.
The
NINDS Efficacy Trials of tPA in Stroke:
The
NINDS study was the landmark clinical trials that examined the efficacy
of tPA in acute ischemic stroke.
1
Patients who received tPA within
180 minutes of symptom onset had a 30% greater chance of a good three
month neurologic outcome than did those treated with standard therapy.
These data support the use of tPA in the emergency management
of patients with an acute ischemic stroke.
This thrombolytic therapy is the only specific therapy that
has shown to be effective in improving the outcome of patients with
ischemic stroke.
The
overall rate of intra-cranial hemorrhage
(ICH) at 36 hours was 3x greater (10.9 vs. 3.5%, p<.001),
and symptomatic ICH rates were 10x greater in patients treated with
tPA (6.4 vs. 0.6%, p<.001).
Although the risk of ICH was greater in tPA-treated patients,
mortality in the treatment group was similar at three months to those
treated with our standard therapy (17 vs. 21%, p=.30).
This comparable mortality rate in tPA treated patients, despite
a higher symptomatic hemorrhagic rate, underscores the significant
mortality risk of ischemic stroke patients who are treated with our
best standard emergent therapy.
Several
important design issues in this NINDS tPA stroke trials should be
considered when analyzing outcome, symptomatic ICH and mortality rates,
and the apparent treatment errors from the nine subsequent clinical
use reports. These study
design issues include:
1.
All patients with a blood pressure above 185/110 mm Hg were
excluded.
2.
All patients requiring aggressive treatment to reduce BP
were excluded.
3.
All patients anti-coagulated within the prior 48 hours were
excluded.
4.
No anticoagulants or antiplatelet drugs were to be given for
24 hours after infusion.
5.
Blood pressure was to be maintained within pre-specified values
after infusion.
Documentation
of the Use of tPA in the Acute Setting
Given
the complexity of the data from the NINDS study and the significant
risk of ICH associated with the use of tPA, the Emergency Physician
must be careful when considering the use of this therapy.
For each ischemic stroke patient, it must be documented that
tPA therapy was considered, and why it was or was not used.
Whether or not tPA is ultimately used, it must be documented
that the inclusion and exclusion criteria were considered, as was
the overall risk/benefit profile of its use for the individual patient
for whom it is being considered.
In addition, the Emergency Physician must also document that
the facts about tPA use were explained to the patient and/or family
members. These facts
include:
Whether
or not a consent document must be completed by the Emergency Physician
prior to tPA use is dependent on the policies and practices of each
institution. Suffice
it to say that the thought process behind its use and the discussion
regarding its risk/benefit profile must be documented for each patient
who presents to the Emergency Department with an acute ischemic stroke.
Clinical
Efficacy and the Effective Use of tPA in Acute Ischemic Stroke Patients
When
the FDA approved the use of tPA for acute ischemic stroke, it was
based on the efficacy demonstrated in the NINDS phase III clinical
trials. Efficacy is defined
as the power or capacity to produce a desired effect.
Once a therapy has been approved based on demonstrated efficacy,
the next question is whether or not the drug can be used effectively
in clinical practice. When
asking if a drug can be effectively used, the issues regarding the
use of a drug outside of the rigid controls of a phase III clinical
trial are taken into account. In other words, is it reasonable to expect that the front line
Emergency Physician can use tPA safely and effectively, given the
need to appropriately identify patients who are eligible for its use,
and manage these patients in a way that is consistent with the efficacy
trials, in the hopes of safely achieving a similar improvement in
outcome?
Even
if a drug efficacy is shown in a randomized, placebo-controlled clinical
trial that includes specific management guidelines, it may or may
not be effectively used in broad clinical practice by the many different
clinicians who practice outside of the constraints imposed by a clinical
trial. If differences
in outcome and safety are observed, they may occur because of differences
in the following:
1)
Patient selection: the
patients identified and enrolled in the clinical trial cant be consistently
identified in clinical practice.
2)
Intervention administration: the drug cannot be administered as it was in the clinical trials.
3)
Concomitant therapy administration: the use of other therapies
cannot be adequately standardized in order to achieve a consistently
good outcome.
4)
Outcome measurement:
problems in outcome assessment lead to different results of
the intervention in clinical practice as compared to the clinical
trials.
An
additional consideration for the use of a therapy in clinical practice
is the frequency with which patients who can receive the therapy occur
in clinical practice. If
the frequency with which these patients are identified is significantly
low, then the individual practitioner may not be adequately skilled
to administer the therapy in an effective way.
This must be considered when examining the results of the nine
trials that have described the clinical use of tPA in ischemic stroke.
If
reports subsequent to an initial clinical efficacy trial suggest that
similar outcomes cannot be achieved in clinical practice, it suggests
that there are clinically important factors that are influencing outcome,
such as difficulties in identifying suitable patients or in managing
the patients who receive the therapy.
If these clinical factors adversely influence the safety or
efficacy results that were seen in the initial efficacy trials, a
narrow therapeutic window for the therapy is suggested.
That is to say, if the therapy cannot be used in clinical practice
as successfully as it was in a clinical trial, then the strength of
the therapys effect is relatively weak, or narrow, as compared to
the other clinical factors that influence patient selection, management
and outcome. This concept
will be discussed in more detail when considering the clinical reports
of the use of tPA in acute ischemic stroke.
The
NINDS Efficacy Trials: Issues Relevant to the Phase IV Clinical Reports of tPA Use
In
order to best understand the significance of the clinical reports
of tPA use, it is important to further examine the results reported
in the NINDS phase III efficacy trials.
The first issue to consider is stroke severity in the patients
treated in these clinical trials.
To put this in perspective, the authors state in the methods,
under outcome measures, the NIHSS, a serial measure of neurologic
deficit, is a 42-point scale that quantifies neurologic deficits in
11 categories. For example,
a mild facial paralysis is given a score of 1, and a complete right
hemiplegia with aphasia, gaze deviation, visual field deficit, dysarthria,
and sensory loss is given a score of 25.
The median NIH stroke scale (NIHSS) score was 14, somewhere
in between these two stroke severity levels.
This is but one issue to consider when examining the patients
who ultimately are chosen for infusion with tPA for stroke; others
which will be looked at in the phase IV studies are age and the presence
of stroke with greater than one third involvement of the cerebral
cortex supplied by the middle cerebral artery as determined by brain
computed tomography.
The
next issue to consider is how many patients had to be screened in
order to find enough patients who were eligible for tPA stroke therapy.
This information is important in that it suggests how often
this therapy can be utilized, which in turn suggests a favorable risk/benefit
profile and the potential for broad application in clinical practice. The NINDS study does not specifically state what number of
patients was screened in order to successfully randomize those patients
who were treated and included in the analysis.
The results do state that between 90 and 95% of patients were
successfully infused in each group, but this information does not
tell the reader what percentage of stroke patients could be reasonably
expected to receive tPA in clinical practice given the experience
in these phase III efficacy trials.
Another
relevant issue as we consider the clinical use of tPA is when the
tPA was administered in the efficacy trials.
The NINDS clinical trials discussion highlights the fact that
48% of patients were successfully treated within 90 minutes.
Does this mean that in order to achieve similar efficacy and
safety results in clinical practice, that at least half should be
treated within 90 minutes? The
data reported in these efficacy trials do not suggest improved outcome
with therapy within 90 minutes (as compared to 90-180 minutes), even
though the pathophysiology of acute cerebral vessel occlusion would
suggest that patients treated early after a stroke would have a better
outcome than those treated in a more delayed manner.
More importantly, given data that suggests that tPA therapy
three to five hours after symptom onset is associated with a higher
complication rate
11
, is it necessary to treat at least
half of all stroke patients within 90 minutes in order not to have
a higher symptomatic ICH rate than that seen in the NINDS clinical
trials?
Lastly,
how were the patients managed in the clinical trials?
Most importantly, how was blood pressure managed, and to what
effect? Given that patients
who required blood pressure management did less well than those without
the need for blood pressure management
12
, how is the practitioner in clinical
practice able to judge the need for blood pressure management in patients
who are eligible for tPA therapy?
These are some of the important factors for consideration as
we look at the subsequently published un-controlled reports of tPA
use in ischemic stroke.
The Phase IV Reports of
tPA Use in Acute Ischemic Stroke:
An Overview
Since
the publication of the NINDS efficacy trials, there have been thirteen
publications in seven different journals regarding the clinical use
of tPA, between January 1998 and September 2002.2-14 Nine are based
on the US experience, two are from Canada, and three from Germany.
(Table 1) The studies vary from the report of one hospital to that
of 57 hospitals, including a mix of community and academic hospitals.
The number of patients treated with tPA ranges from 30 to 389 in these
case series (similar to the 312 patients treated in the efficacy trials),
and the treatment of between 1.8 and 47% of eligible patients. As
was seen in the NINDS efficacy trials, which included patients with
a mean age of 68 years, the mean age of patients treated in these
clinical reports was between 63 and 71 years old. (Table 2) The median
NIHSS scores seen in these phase IV studies ranged from 10-15, which
is similar to the median NIHSS scores in the NINDS efficacy trials,
which was 14. Although the median time to therapy was not stated in
the NINDS efficacy trials, the median time to treatment in the seven
reports that reported this statistic ranged from two hours and six
minutes to two hours and forty-five minutes. These data suggest that
in the subsequent reports, although the mean age and stroke severity
was comparable to that in the NINDS efficacy trials, the median time
to tPA therapy was greater in clinical use than in the well controlled
efficacy trials.
The NINDS clinical trials
reported a favorable clinical outcome in 31-54% of patients. (Table
3) Although it is difficult in the subsequent reports to determine
parity in the determination of this outcome measure, the reports suggest
a favorable outcome in 30 to 95% of patients. The overall mortality
rate observed at three months in the NINDS clinical trials was 17%.
In the subsequent reports, mortality, measured at a time period ranging
from hospital discharge to five months, was 5.3 to 25%. The overall
ICH rate was 10.9% in the NINDS efficacy trials, and ranged from 8
to 31% in the later published clinical reports. The symptomatic ICH
rate was 6.4% in the NINDS trials, and 2.7 to 15.7% in the subsequent
reports. This outcome data suggests that both the safety and efficacy
of tPA in clinical use is similar to that seen in the efficacy trials.
The most important observation from the clinical use trials is the
occurrence of symptomatic ICH in 10.8 and 15.7% of patients in two
of the subsequent reports.6;8 These studies did not, however, report
significantly higher mortality rates than were seen in the other phase
IV studies, suggesting, as did the NINDS studies, that ischemic stroke
itself carries with it a significant risk of death.
Eight of the thirteen clinical
reports also reported the overall frequency with which protocol deviations
occurred, ranging from 1.3 to 67% of patients treated with tPA. Between
0 and 18.9% of patients were treated outside of the three hour window
in ten of the reports, and anticoagulant use was reported in 2.2 to
37% of patients in five of the clinical tPA use reports. Other protocol
deviations included inadequate blood pressure control in 3 to 79%
of cases (four studies) and two studies reported that 1.5 and 4% of
patients had a baseline coagulopathy prior to infusion. CT findings
suggestive of a stroke larger than one third of the distribution of
the middle cerebral artery distribution (precluding tPA infusion)
was seen in 2 to 15% of patients (four studies), and two studies reported
cerebral edema in 2 to 10% of patients (comparable to the 3-5% rate
seen in the NINDS efficacy trials). These data suggest that there
are clear areas for improvement in the adherence to the strict tPA
protocol, which generated efficacy in the NINDS clinical trials.
Clinical
Considerations: The Clinical
Use of tPA in Acute Ischemic Stroke
The NINDS
efficacy trials showed that tPA can be used to improve outcome in
acute ischemic stroke patients if a rigid treatment protocol is followed.
The subsequently published reports of a series of patients treated
with tPA suggest that, despite its narrow therapeutic window, patients
can be managed effectively with this therapy outside of the rigid
guidelines that controlled its use in the NINDS efficacy trials. The
recently published article by Lewandowski and Barsan stresses the
need to closely adhere to the NINDS studies inclusion and exclusion
criteria when giving tPA, a recommendation that highlights the relatively
narrow therapeutic window of tPA use in acute ischemic stroke.17 This
review article states that the most difficult aspect of tPA use is
the ability to determine that symptom onset occurred less than three
hours prior to the time of tPA infusion onset. This article also states
that the clinical experience with tPA is favorable when considering
all of the published phase IV studies, and reiterates the need to
not deviate from established guidelines and to not treat patients
with CT findings suggestive of an acute CVA. Although this article
states that only eight patients need to be treated in order to return
one patient to full function, the clinical experience with tPA as
seen in the subsequently published cases series suggests that significant
complications can occur, often as a result of protocol violations
that occur in general clinical use. As such, the number of patients
needed to treat in order to achieve full recovery in one patient may
actually be higher in clinical practice than is suggested by the NINDS
data itself.
Conclusions
- Although
tPA is effective in improving long-term outcome in ischemic stroke,
the complications associated with its use, especially symptomatic
ICH, are significant, and occur not infrequently.
- The
occurrence of significant complications suggests a relatively narrow
therapeutic window, such that informed consent should be considered
prior to the use of tPA in ischemic stroke.
- A
limited number of stroke patients are treated with tPA in clinical
practice.
- Outcomes
similar to those seen in the NINDS clinical trials can be duplicated
in clinical practice.
- A
checklist of exclusion criteria should be used prior to tPA use
in ischemic stroke.
- Blood
pressure should be maintained below 185/110 mm Hg.
- Protocol
violations occur frequently in clinical practice, often leading
to higher ICH rates.
- Most
patients are treated late in the three-hour window, and many are
treated outside of this critical time frame in error.
Patients should be encouraged to present themselves to the
Emergency Department soon after the onset of stroke symptoms.
- CT
abnormalities that suggest a large infarct or cerebral edema should
increase the likelihood that tPA use be deferred because of a higher
risk of hemorrhagic complications.
- Measures
should be instituted to prevent anticoagulant use during the 24
hours after tPA use.
- Education
that standardizes the clinical use of tPA across institutions should
be provided.
- Documentation should
explicitly address the relevant issues in the use of tPA in stroke,
including what information is provided to the patient or family
members prior to obtaining informed consent and/or the administration
of the drug.
- Further
research should verify that the guidelines established by the NINDS
efficacy trials are, indeed, being followed in the clinical setting
of acute ischemic stroke.
Table
1. Institutional and
Patient Data from the Published Cased Series of tPA Use in Ischemic Stroke.
|
Author
|
Journal
|
Date
|
Location
|
#
Hospitals
|
#
Eligible Patients
|
#
Patients Receiving tPA, (%)
|
Hospital
Type
|
|
NINDS
|
NEJM
|
12/95
|
US
|
8
|
-
|
312
|
8
Academic
|
|
Chiu
|
Stroke
|
1/98
|
US
|
3
|
1035
|
30
(2.9%)
|
|
|
Grond
|
Stroke
|
8/98
|
Germany
|
1
|
453
|
100
(22%)
|
1
University
|
|
Smith
|
AEM
|
6/99
|
US
|
4
|
-
|
37
|
|
|
Tanne
|
Neurology
|
7/99
|
US
|
13
|
-
|
189
|
|
|
Wang
|
Stroke
|
1/00
|
US
|
20
|
900
|
57
(6.3%
|
-
|
|
Buchan
|
Neurology
|
2/00
|
Canada
|
1
|
1540
|
68
(4.4%)
|
1
Academic
|
|
Albers
|
JAMA
|
3/00
|
US
|
57
|
-
|
389
|
|
|
Katzan
|
JAMA
|
3/00
|
US
|
29
|
3948
|
70
(1.8%)
|
-
|
|
Koennecke
|
Srtoke
|
5/00
|
Germany
|
1
|
161
|
75
(47%)
|
1
Academic
|
|
Chapman
|
Stroke
|
12/00
|
Canada
|
1
|
2556
|
46
(1.8%)
|
1
Academic
|
|
Grotte
|
Arch
Neurology
|
12/01
|
US
|
4
|
1689
|
269
(16%)
|
1Academic
3 Community
|
|
Schmulling
|
Stroke
|
4/00
|
Germany
|
1
|
-
|
150
|
1
Academic
|
|
Bravata
|
Arch
Intern Med
|
9/02
|
US
|
16
|
-
|
63
|
16
Community
|
|
Totals
|
7
|
5
|
3
|
151
|
12,282
|
928
(5.8%)
|
37
Academic
65 Community
|