To Lyse or Not to Lyse: The Great tPA Debate
It’s been a few months since I’ve posted (a little over six, whoops), but as they say – new year new me. I’m planning on trying to stick to a once a month or so posting schedule so hopefully that goes well!
I’m kicking 2024 off with a real doozy – does thrombolysis really work in stroke?
Got your attention, maybe? Yes, this debate is withered and old and the same arguments from both sides get circulated routinely about the putative risks and benefits of providing a guideline-supported therapy with decades of supporting evidence to patients who are experiencing a neurologic emergency which has the potential to permanently alter their entire life. Ok, ok – that is the only bit of sass I’ll allow myself.
Key Points
|
Every few months or so, a variant of a particular post appears on Twitter. This is how it typically goes down: someone asks what everyone’s “hot take” in medicine is, someone inevitably states that either thrombolytics don’t work or there is “no evidence” supporting their use in stroke, and a back and forth occurs on differing interpretations of the same three to four studies and a debate on the value of different medical specialties, and rarely anything productive is accomplished. I have historically not engaged in these debates, mainly because my opinion is that it is just not a debate worth having, and engaging reminds me too much of this classic xkcd comic:
Thrombolysis is
effective in acute ischemic stroke, and this fact is supported by numerous
international treatment guidelines.1–3 Of course, the decision to administer
thrombolysis is not straightforward, and concluding that "thrombolysis
works for all strokes" without considering indications and
exclusions obviously misses the nuance of the treatment. Like any therapy,
there are risks associated with the potential benefit, and this is magnified by
the fact that the main risks associated with
administration can produce devastating consequences that would negate its
potential benefit. However, these considerations form the basis of who is and
who is not eligible to receive treatment. Over the last 30 years, the
population who is most likely to benefit and least likely to be harmed has been
refined and with refinement, a clear criteria list has been formed. Despite
this, this debate continues to come up, suggesting clinician opinion fall
somewhere on this spectrum:
It's important
to recognize that this debate is largely artificial, and in the real world, stroke guidelines are applied appropriately and eligible
patients are treated with thrombolysis in a thoughtful manner. Clinician
opinion on thrombolysis, like many things, falls on a bell curve and most
people fall in the middle:
How Did We Get Here?
It is worth briefly revisiting the history of thrombolysis in stroke and explore how we got to where we are today. If you're not interested in a history lesson, feel free to jump to the discussion of Where Were Now.
Since at least the late 1930s, the primary
pathophysiologic model of acute ischemic stroke has been known to be some form
of thrombosis of the cerebral circulation.4,5 Interest in developing therapies that
could dissolve these thrombi thus emerged as early as the 1950s, predating much
of the understanding of the nuance of stroke pathophysiology and predating
imaging technologies which would allow better stroke characterization and risk
stratification. Even without the guide of CT or MRI, early stroke neurologists
recognized the critical relationship between time from symptom onset and
initiation of treatment. Bernard Sussman and Thomas Fitch, describing three
patients with angiographically observed cerebral thrombi, stated “what has been
needed is an agent, known to be effective in reversing thrombosis, which can be
administered immediately on the appearance of symptoms by the first physician
on the scene.”6 In the case of their three patients,
this involved treatment with fibrinolysin, a bovine-derived plasmin derivative,
of which one patient with a documented MCA occlusion experienced reperfusion.
Fibrinolysin
specifically never quite gained steam and its subsequent use was limited mostly
to case reports,7–9 but when streptokinase was demonstrated
to be effective in coronary ischemia, interest grew in using it for cerebral ischemia. Early trials were less than encouraging, however – a
1963 RCT of 40 patients with anterior cerebral thrombosis randomized to
streptokinase and SQ heparin plus warfarin vs placebo and SQ heparin plus
warfarin within 72 hours of symptom onset revealed no significant difference in
functional improvement (after 10 days) but, somewhat surprisingly, no
differences in adverse effects.10 The authors did, however, note that
patients who experienced clinical improvement tended to exhibit markers of
enhanced fibrinolytic activity, regardless of the arm they were randomized to,
suggesting that identifying a dosing regimen of streptokinase which could
reliably induce a state of fibrinolysis could replicate the improvement. A
larger trial was thus initiated and published in 1964 using a higher dose of
streptokinase, but unfortunately analysis of the 73 randomized patients
demonstrated a significantly higher rate of early death (35.1% vs 11.1%) which
dampened enthusiasm to the strategy.11 Notably, these trials predated the existence
of CT, so while they did perform an LP to “rule out” ICH (and then gave a
lytic! Yikes!), it is not possible to know for sure whether some of the
randomized patients presented with hemorrhagic strokes.
The 1970s
introduced CT technology to vascular neurology12 and allowed neurologists to easily
identify one of the issues with thrombolytic administration (this time with
urokinase) - hemorrhagic conversion. A pilot study of urokinase in 31 patients
with ischemic stroke presenting within 36 hours of last known well demonstrated
urokinase could sustainably induce fibrinolysis, but no unequivocal early
neurologic improvement was noted.13 Furthermore, at least four patients
experienced early hemorrhagic conversion, with the cases distinctly discussed
in a separate paper.14 A 1981 review article on stroke
therapeutics notes that thrombolysis remains an attractive theory, but notes
the limitations of the evidence discussed above. The authors conclude that “the
ability to make a more accurate diagnosis… and improved technology to
administer and measure thrombolytic agents might make this form of treatment
feasible in the future.”15
Well, the
technology came – first, incorporation of CT imaging in the routine evaluation
of stroke revealed that nearly a quarter of patients with primary ICH may have
been incorrectly diagnosed as primary ischemic stroke prior to the availability
of CT imaging.16 Second, improvements in the rapidity of
diagnosis, identification of target lesions, and selection of dosing given more
extensive experience with thrombolysis in MI and PE led to some of the first
definitive reports of improved outcomes with thrombolytic therapy. A 1988
retrospective review of 65 cases, 43 of whom received intra-arterial
thrombolysis (urokinase or streptokinase) and 22 of whom received conventional
therapy (antiplatelets and anticoagulants) for vertebrobasilar occlusion
revealed patients who experienced recanalization with thrombolysis (44% of
treated patients) had significantly improved outcomes compared to conventional
therapy – 31.6% mortality vs 86.3% mortality (p = 0.0005) and 52.6% vs 13.6%
favorable outcome (p = 0.017). While thrombolysis without recanalization had
dismal outcomes (100% mortality, remember these were vertebrobasilar strokes so this is not terribly unexpected), this was not found to be statistically
significantly different than conventional therapy (p = 0.10 for both mortality
and functional outcomes).17 A pilot study from the same year found
similar results – patients treated with intra-arterial thrombolysis within 8
hours of last known well (note the time window of eligibility has been
gradually narrowing) tended to experience neurologic improvement when vessel
recanalization occurred, and neurologic demise was only observed in those
without recanalization. Interestingly, 4 of the 15 patients in this trial
experienced hemorrhagic conversion but did not experience neurologic decline
(in fact 2 of the 4 completely or nearly completely recovered), providing
an early example of how hemorrhagic conversion is not a universally poor
outcome.18
In sum, the
field of stroke neurology spent decades continuously refining the theoretical
basis for who may benefit from thrombolysis. Up until the late 1980s, the
strategy had yet to be proven effective in large scale trials, but the lessons
learned included that universal application of thrombolysis (especially
in combination with anticoagulation) was generally harmful, earlier initiation
was likely better than late, and that neurologic improvement seemed closely
linked with vessel recanalization. A comprehensive review published in 1987
outlined the theoretical basis for thrombolysis in cerebrovascular disease and
reviewed all available clinical and experimental data published to date,
ultimately concluding that with advances in imaging techniques, understanding
of fibrinolysis, and improves selection of cases, thrombolysis will likely
become a viable treatment option for stroke.19
The viability
of thrombolysis for stroke treatment was demonstrated perhaps for the first
time in a landmark meta-analysis of early clinical data in 1992. By narrowing
in to just the effects of thrombolysis revealed in studies conducted when CT
technology was available, the overall effect of thrombolysis on death and
neurologic deterioration was reported to be about a 6% improvement (p = 0.007).20
 |
| The first meta-analysis to demonstrate improvements in outcomes with thrombolytics, but only when data from the pre-CT era were excluded. |
This, in
combination with an improved understanding that long term functional
outcomes, rather than just death or short-term improvement, was most meaningful
to patients, prompted the initiation of the lytic study design which we know
today.21 The MAST-I,22 MAST-E,23 and ASK24 trials were the first major streptokinase
trials conducted using “traditional” thrombolysis methods – patients were
enrolled within 4 (ASK) or 6 (MAST-I and MAST-E) hours of last known well to
receive streptokinase or placebo, excluding patients with evidence of ICH, recent
anticoagulant therapy, and patients with only mild deficits or very severe
deficits. Overall, the three trials demonstrated the high risk of
bleeding and early mortality with streptokinase. Over 1200 patients were
randomized across the three trials and rates of symptomatic hemorrhagic
conversion and early mortality were higher in treated patients across all three
trials. The silver lining, however, was that in patients that survived,
patients treated with streptokinase had better functional outcomes, especially
if they were treated earlier. Now, this could easily be confounded by the fact
that more severe patients likely expired earlier and thus were excluded from
the 6-month outcome analysis, but it suggested that a more selective agent, one
with a lower likelihood of systemic and intracerebral bleeding, might be able
to capture that functional outcome benefit without the overwhelming risk of
bleeding and poor early outcomes.
 |
| Functional outcomes from the MAST-E trial. Note that outcomes excluding mortality were significantly improved with streptokinase. The Rankin disability score would later be modified to include death to account for this limitation. |
This is where
alteplase enters the story. Alteplase, as a true “tissue” plasminogen
activator, provides a number of theoretical advantages over the far less
specific fibrinolytics streptokinase and urokinase and had already been
demonstrated to be a viable option for stroke in animal models by the time the
streptokinase trials were underway.25,26 Alteplase had also recently been
demonstrated to be superior to streptokinase for MI in the GUSTO trial,27 further supporting the notion that not
all lytic agents could be considered equal and that the pharmacodynamic
advantages of alteplase over streptokinase could lead to measurable clinical
differences. The first meaningful trials of alteplase in stroke, by Brott and
Haley, revealed two important findings – stroke patients could be rapidly
triaged and randomized to acute therapies, and that both the risk and benefit
of thrombolytic therapy was a function of time to treatment and the dose of the
agent.28,29 These findings, demonstrating that
alteplase at a dose of 0.9 mg/kg given within 180 minutes of last known well
appeared to balance clinical improvement and risk of hemorrhage, formed the
basis of the NINDS trial which went on to be the first large-scale positive
trial of thrombolysis in acute ischemic stroke.30 This trial prompted the FDA to approve
alteplase for the treatment of AIS within 3 hours of last known well and the
debate of exactly how beneficial alteplase is has raged on every since.
Where We Are Now
Since the
publication of the NINDS trial, alteplase has been evaluated in at least 8 RCTs
for stroke, which I have briefly summarized below.
|
Trial |
Inclusion Criteria and
Treatment |
Functional Outcomes |
Safety |
Notes |
|
NINDS30 |
Stroke onset within 3
hours
Standard exclusion
criteria
Alteplase 0.9 mg/kg vs
placebo |
624 patients
randomized across two arms
Global OR for improved
functional outcome: OR 1.7, 95% CI 1.2-2.6
(favoring tPA) |
Alteplase vs placebo sICH: 6.4% vs 0.1%,
P<0.001
Mortality: 17% vs 21%, P=0.30 |
Arm 1 designed to
assess early improvement, arm 2 designed to assess functional outcomes |
|
ECASS31 |
Moderate to severe
stroke onset within 6 hours
Key exclusion
criteria: very severe stroke, improving symptoms, bleeding diathesis, recent
trauma or surgery, ICH present on CT, malignant MCA ischemia
Alteplase 1.1 mg/kg vs
placebo |
620 patients randomized
Median mRS at 90 days: Intention to treat: 3 (tPA) vs 3
(placebo), P=0.41 Per protocol: 2 (tPA) vs 3
(placebo), P=0.035 |
Alteplase vs placebo Parenchymal hematoma: Intention to treat: 19.8% vs 6.5%,
P<0.001 Per protocol: 19.4% vs 6.8%,
P<0.001
Mortality: Intention to treat: 22.4% vs 15.8%, P=0.04 Per protocol: 19.4% vs 14.8%, P=0.17 |
Protocol violations contributed
to findings significantly
Most violations
related to degree of ischemia seen on initial CT
tPA dose higher than
current recommendation |
|
ECASS-II32 |
Moderate to severe
stroke onset within 6 hours
Standard exclusion
criteria
Alteplase 0.9 mg/kg vs
placebo |
800 patients randomized
mRS 0-1 at 90 days 40.4% (tPA) vs 36.6%
(placebo), P=0.277
mRS 0-2 at 90 days 54.3% (tPA) vs 46.1%
(placebo), P=0.024 |
Alteplase vs placebo Parenchymal hematoma: 8.1% vs 0.8% (no P
value provided)
Mortality: 10.3% vs 10.5%,
P=0.816 |
Outcomes in 0-3 and
3-6 hours strata not different, but n very small for 0-3 hours
8.3% reduction in
death or dependency (mRS 3-6) noted |
|
ATLANTIS-A33 |
Stroke onset within 6
hours
Standard exclusion
criteria
Alteplase 0.9 mg/kg vs
placebo
Stopped early due to
safety concerns in the 5-6 hour group |
142 patients
randomized
Median mRS at 90 days: 5 (tPA) vs 2
(placebo), P=0.05 |
Alteplase vs placebo sICH: 11.3% vs 0%, P=0.003
Mortality: 22.5% vs 7.0%, P=0.009 |
Poor outcomes mostly
limited to patients randomized between 5-6 hours from LKW and with baseline
NIHSS of >20
Randomization
continued to part B, restricting to LKW <5 hours |
|
ATLANTIS-B34 |
Stroke onset within 5
hours
Standard exclusion
criteria
Alteplase 0.9 mg/kg vs
placebo |
613 patients
randomized
mRS 0-1 at 90 days: 41.7% (tPA) vs 40.5%
(placebo), P=0.77 |
Alteplase vs placebo sICH: 6.7% vs 1.3%,
P<0.001
Mortality: 10.9% vs 6.9%, P=0.08 |
Analysis restricted to
patients enrolled within 3 hours of LKW demonstrated significant benefit with
tPA: mRS 0-1 61% vs 26%, P=0.0135 |
|
ECASS-III36 |
Stroke onset within
3-4.5 hours
Standard exclusion
criteria, additionally excluded patients with a history of stroke and
diabetes
Alteplase 0.9 mg/kg vs
placebo |
821 patients
randomized
mRS 0-1 at 90 days: 52.4% (tPA) vs 45.2%
(placebo), P=0.04 |
Alteplase vs placebo sICH: 2.4% vs 0.2%, P=0.008
Mortality: 7.7% vs 8.4%, P=0.68 |
sICH definition more
restrictive than in prior publications |
|
EPITHET37 |
Stroke onset within
3-6 hours and could undergo baseline MRI
Standard exclusion
criteria
Alteplase 0.9 mg/kg vs
placebo |
101 patients
randomized
mRS 0-1 at 90 days: 35% vs 24%, P=0.265 |
Alteplase vs placebo sICH: 7.7% vs 0% (no P value
provided)
Mortality: 25% vs 14%, P=0.102 |
Primary endpoint was
percent infarct growth; functional outcomes were a secondary outcome |
|
IST-338 |
Stroke onset within 6
hours and general uncertainty about whether the patient would benefit
Standard exclusion
criteria
Alteplase 0.9 mg/kg vs
placebo
Only pilot study was
double blind |
3035 patients
randomized
OHS 0-1 (similar to
mRS) at 180 days: 24% (tPA) vs 21%
(placebo), P=0.018
Ordinal shift in OHS: OR 1.17 (95% CI
1.03-1.33) |
Alteplase vs placebo sICH: 7% vs 1%, P<0.001
Mortality (at 6mo): 27% vs 27%, P=0.672
|
Several protocol
changes and lack of double blinding may limit external generalizability
OHS 0-1 significant
only after adjustment for confounding; OHS 0-2 not significant in unadjusted
or adjusted analyses |
|
EXTEND39 |
Stroke onset between
4.5 and 9 hours and MRI or CT perfusion demonstrating a penumbra:core ratio
of at least 1.2 with an absolute difference between penumbra and core of 10
mL and an absolute core volume of <70 mL
Standard exclusion
criteria plus patients undergoing thrombectomy were excluded
Alteplase 0.9 mg/kg vs
placebo |
225 patients enrolled
mRS 0-1 at 90 days: 35.4% (tPA) vs 29.5%
(placebo), P=0.04
Common odds for shift
towards improved outcomes: OR 1.55, 95% CI
0.96-2.49 |
Alteplase vs placebo sICH: 6.2% vs 0.9%, P=0.053
Mortality: 11.5% vs 8.9%, P=0.67 |
Primary outcome
significant after adjustment for confounders
|
|
WAKE-UP40 |
Stroke onset unknown,
but at least greater than 4.5 hours, and MRI demonstrating no FLAIR change
within the area of DWI restriction
Standard exclusion
criteria plus patients undergoing thrombectomy and whose baseline NIHSS were
>25 were excluded
Alteplase 0.9 mg/kg vs
placebo
Enrollment halted due
to lack of funding |
503 patients
randomized
mRS 0-1 at 90 days: 53.3% (tPA) vs 41.8%
(placebo), P=0.02 |
Alteplase vs placebo sICH: 2.4% vs 0.4%
(ECASSIII), P=0.21
Mortality: 4.1% vs 1.2%, P=0.07 |
Primary outcome
significant after adjustment for confounders
Rates of sICH were not
significantly different between arms regardless of definition used |
There is,
understandably, a lot going on in all of these trials, and not all of it can be
captured in a broad overview table. Some of the nuance I will discuss
throughout this post, but attempting to distill the cumulative knowledge
produced from over 20 years of clinical trials is no small feat. Rather, to
focus the rest of the text, I will try to identify the main objections tPA
skeptics make when interpreting these trials, and provide some commentary on a
more nuanced way of thinking about thrombolysis in stroke.
Objection #1: There is No Positive Data for Thrombolysis
A common and
somewhat baffling argument against thrombolysis in stroke is that there is “no positive prospective data” for thrombolysis in stroke. At face value, this is a
bit perplexing – just a glance at the table above suggests there are at least 4
trials with positive top-line results with alteplase vs placebo. Two of these
are the EXTEND and WAKE-UP trials, which realistically can be excluded from
most general thrombolytic discussions because they require advanced imaging
modalities and neuroradiology expertise to interpret them, which
is often only available at comprehensive stroke centers. That leaves NINDS and
ECASS-3 which form the primary basis of thrombolytic recommendations (we’ll get
to the 2014 Lancet meta-analysis in a moment). The reason skeptics are doubtful
even these are positive is because their findings are particularly “fragile”41 and two specific re-analyses, one of
the NINDS trial42 and one of the ECASS-3 trial,43 have suggested the findings of the
primary analysis, when viewed through a different lens, may not be positive
after all. Both have different limitations than the trials they aim to refute,
however, and both miss the big picture of thrombolysis in acute ischemic
stroke.
The NINDS Re-Analyses
Hoffman and
Schriger approached the re-analysis of the NINDS trial aiming to more fully
characterize the trial dataset using a wide number of data visualizations.42 The central thesis is valid in that the
original trial contained some 60,000 variables and the original publication was
physically limited by journal space and could not fully represent the dataset. A
2004 re-analysis conducted after concerns of baseline imbalance were raised had
confirmed (and strengthened) the primary findings of the original publication,44 but the lack of detailed visualizations
left additional questions about the distribution of the data. The 2009
re-analysis plotted several key variables against each other, including
baseline NIHSS vs 90-day improvement in NIHSS, Bathel index, and mRS, as well
as plots evaluating time to treatment. In all plots, the effects favored
alteplase, but the authors note that effects were less consistent when baseline
NIHSS was <5 or >22, which only confirms current practice to generally
avoid thrombolysis in minor stroke and consider the risks in severe stroke.
Notably, conclusions were only gained by visually assessing the shape of
the curve, which is naturally subject to bias from axis scaling and the degree
of jitter introduced to scatterplots. The authors also conclude that no
evidence of a time-is-brain effect was present, but do not visualize the effect
of door-to-needle time on functional outcomes, only on change in 90-day NIHSS.
 |
| The only figure in the Hoffman re-analysis which places mRS, rather than NIHSS, on the Y-axis |
In response to this re-analysis, a third re-analysis of the trial data was conducted in 2010. Saver and colleagues (who practice at the same center as Hoffman and Schriger; I can only imagine what those workplace conversations were like…) point to notable limitations in plotting the NIHSS continuously and as the primary marker of functional improvement.45 The NIHSS itself is a flawed scale, with biases towards anterior circulation strokes and strokes affecting the left (dominant) hemisphere.46,47
 |
| A graphical representation of the expected benefits and harms of alteplase by Saver and colleagues |
The ECASS-III Re-analysis
The next
article skeptics point to when suggesting no positive data exist for
thrombolysis is a 2020 re-analysis of the ECASS-III data. In this article,
Alper and colleagues re-analyze the ECASS-III dataset using a variety of
imputation and categorization methods and come to conclusions which differ
modestly from the original publication.43 The primary thesis of the authors here
is valid – following the exact statistical methods described in the body of the
text with the data provided to the re-analysis authors does not produce
precisely the same results. This, of course, is problematic. The entire purpose
of reporting methods is to ensure reproducibility, and a lack of
reproducibility jeopardizes the confidence in the original publication’s
methods. The authors clarified several points with the original study sponsor
and ultimately discover a few errors – one, the modified NIHSS used in the
study (ranging from 0-46) which is described in the study protocol was
mis-transcribed to the original NIHSS (ranging from 0-42) in the manuscript,
and two, several ordinal variables were analyzed categorically rather than
continuously. Changing the methods of adjustment yielded an odds ratio with a
95% confidence interval that crosses 1, casting doubt on the confidence of the
ECASS-III findings.
 |
| The 16 models trained by Alper and colleagues. Significant results in bold. |
This itself,
however, is also problematic – repeatedly running the primary analysis with
different types of confounders is bound to result in findings which fail to
reject the null. There is no correct model here and any of the models included
in this table could be perfectly valid on its own. As with the Hoffman NINDS
analysis, considering NIHSS as purely continuous is likely problematic, as it
assumes each 1-point change in the scale represents an equal degree of
neurologic change, which it does not. Using door-to-needle time as categorical
is probably also problematic, but is not invalid per se. Regardless, one thing
is universal to all models – the effect size for all favors alteplase, and when
considered in the context of the overall body of evidence (as the authors
suggest), it is consistent with benefit with alteplase when used in the
3-to-4.5-hour window.
The Overall Body of Evidence
The conclusion
I just made above may seem a bit cheap, but I would argue the context is
important here. Much of the skepticism of the evidence supporting thrombolysis
emerges from concerns at the trial level, meaning that concerns over baseline
imbalances, selection bias, or reporting methods may yield inconclusive or
incorrect results. While we could journal club each of these trials to death, focusing
on the individual trials runs a substantial risk of missing the forest from the
trees given the body of evidence available to date. Time and time again, improvements in the implementation, efficiency, and
selection process of thrombolysis has been associated with improvements in patient
outcomes and safety. Early observational data aimed to assess the safety of
increased thrombolysis implementation found consistent functional outcome improvements
in 6483 treated patients compared to untreated patients from clinical trials48 and analyses from the Get With the Guidelines
registry has demonstrated a clear correlation between time to treatment and functional
outcomes in over 50,000 treated patients.49,50 It is astoundingly unlikely that the
clear improvements in patient outcomes related to time to treatment
with thrombolysis can be entirely ascribed to confounding; in the real world,
thrombolysis works, and there is data to prove it.
Of course, guideline-level
treatment recommendations cannot be made on observational evidence alone.
Luckily, the same pattern seen in the real-world evidence is seen when patient-level
trial data is pooled. A landmark 2014 patient-level meta-analysis of alteplase
trials published in Lancet evaluated the cumulative effect of alteplase on
functional outcomes across all the major alteplase trials published to date.51 This publication illustrates a few
crucial points. First, by combining the baseline and treatment data from 9
clinical trials including nearly 7,000 patients, it substantially increases the
precision of treatment effects. Second, because of the size of the study, effect
can be illustrated across the continuous axis of time, providing a clearer
method of adjusting assumed benefits to patient-specific characteristics. This
figure from the paper nicely demonstrates how the likelihood of attaining an
mRS of 0-1 at 90 days is strongly correlated with door-to-needle time, a
relationship consistently seen in nearly all trials and all real-world evidence
to date.
 |
| The odds of achieving an mRS of 0-1 at 90 days is linearly associated with time to treatment from symptom onset which provides strong evidence for the beneficial effect of alteplase. |
Now, there is a
small elephant in the room here worth discussing – exactly what does “benefit” mean
to the average patient? On a philosophical basis, the purpose of clinical
trials is to inform our understanding of what the likelihood of benefit
is for an individual patient. Thrombolysis in stroke is an enormously complex decision
and cannot be captured by a simple benefit/no benefit decision. The likelihood
thrombolysis will benefit an individual patient experiencing stroke depends on
at least these variables:
- The time from stroke symptom onset to treatment initiation
- The severity of the stroke symptoms at baseline
- The patient’s baseline level of function or disability
- The degree/extent of ischemia present at the time of evaluation
- More, potentially undiscovered, effect modifiers
Additionally,
the use of the term “benefit” also makes several assumptions. Most clinical
trials dichotomize the mRS into “excellent” (0-1) or “good” (0-2) as their
primary outcome which substantially limits the power of detecting differences
across the ordinal scale.52,53 Dichotomization makes interpreting
study results easier, but misses important clinical changes which represent
meaningful outcomes for patients, such as improvement from an mRS of 4 (the
patient needs assistance for most ADLs) to an mRS of 3 (the patient needs some
assistance for daily tasks, but can walk independently).54 Not only is shift analysis a more
efficient clinical trial design method, consideration of the entire ordinal
scale provides a more holistic and accurate assessment of patient outcomes and
cost of treatment.55 The point being – point estimates from
clinical trials that only report the difference in rates of attainment of
dichotomous lumps of mRS may incompletely represent the benefit a patient may
expert to see.
Disability, or more specifically how disability is
defined, is also a key assumption in the assessment of presumed benefit from
thrombolysis. Benefit in stroke trials has traditionally been limited to attainment
of specific mRS thresholds in patients who present with symptoms exceeding
certain NIHSS thresholds. This is conceptually restricting, as not all
meaningful (read: “disabling”) symptoms are equally represented on the NIHSS
(as discussed above), and the mRS, while reasonably reflective of function,
does not fully capture impairment.56 The ultimate point I am
building to is making the decision to administer thrombolysis to a patient
needs to be based on a continuous understanding of the risks and benefits of
treatment, rather than static effect sizes provided by clinical trials. The clear time-to-treatment benefit provides not only evidence that there is (broadly speaking) "positive" data for thrombolysis, but gives a continuous estimate of just how much benefit an individual patient might expect to see, whatever that benefit might mean to that patient.
Objection #2: The Harms are Undersold to Patients
This very well
may be true, depending on the specific conversation that is had with patients
and surrogates, but there is no compelling evidence that the risk of
thrombolysis is underreported to patients. In fact, evidence suggests it may be
the opposite – because emergency medicine guidelines provide a level B
recommendation for thrombolysis and a level C recommendation for utilizing a
shared decision model with patients regarding the use of thrombolysis, there
may be a perception that the risks and benefits are equivocal, and thus no
“best” decision exists.57 One semi-structured interview study of
medical professionals in a safety net hospital revealed that clinicians focused
heavily on the risks, and the emphasis on the potential for harm may have
confused or scared patients and posed a barrier to treatment.58 While patients absolutely retain the
right to refuse treatment, it is very challenging to accurately communicate the
short term risks in the context of the long term benefits, especially in the
setting of acute ischemic stroke. While patient refusals do occur, it is rare
and has decreased over time – one single center experience reported 30 patient
refusals over an 8-year period, representing a 4.2% refusal rate. However,
refusal steadily declined over the study period from 2004 to 2011 – each
additional year in the study was associated with a 37% reduction in the odds of
patient refusal. While the nature of the discussions with these patients was
not described, patients ultimately diagnosed with stroke mimics and with lower
baseline NIHSS were significantly more likely to refuse treatment, suggesting patients
with milder symptoms may be less willing to accept the risk of treatment.59
Now, there are realistically
two issues worth discussing here – one, the content of the conversation had
with patients or their surrogates, and two, the nature of the
conversation had with patients or their surrogates, namely whether the
conversation should involve “informed” consent, assent, shared
decision making, or simply informed refusal.
The Content of the Conversation
It is
challenging to characterize the content of real discussions had with patients
about thrombolysis, both because the discussions occur in time sensitive
environments and observation of the conversation will suffer from the Hawthorne
effect. Undoubtedly, however, there is substantial variation in exactly what
conversation is had. A survey of 101 resident and attending neurologists in the
Netherlands revealed that only 33% of residents and 21% of attendings “always”
ask for explicit consent prior to alteplase administration, and 62% of
respondents reported they spent less than a minute on the conversation, if they
had one at all.60 Similar findings were reported in a
survey of 199 neurologists and emergency medicine physicians in the US, with only
36.7% of respondents reporting universal informed consent and 54.8% spent less
than 5 minutes on the conversation. Over 90% of respondents noted that the risk
of intracranial hemorrhage was a dominant theme in the conversation, while the
time-dependent benefit was discussed by 76.4% of respondents and that a third
of patients will see significant benefit by 61.3% of respondents.61
Semi-structured
interviews of clinicians who have these conversations with patients and their
surrogates confirm that the risk of treatment is an intensive focus of the
consent process. One interview study of 13 clinicians in the UK revealed risk
communication to be a dominant theme of discussions clinicians had with
patients. One participant focused on the challenge of risk communication and the
time-dependent nature of the intervention, stating “you’ve just given them this
whole raft of information about risks and possible side-effects…and you’re
almost pushing them into a decision…because you know you’re under time
pressure.”62 The enormous pressure to communicate both
the benefits and the risks of treatment to patients or their surrogates makes
it challenging to universally satisfy the acceptable criteria of consent. A
study of the adequacy of consent of 14 tele-stroke consent conversations found
that acceptable patient or family understanding was observed in 78.6% of cases
as rated by a neurologist, an emergency medicine physician, a layperson, an
ethicist, and a lawyer, but agreement about the degree of adequacy was poor
across raters despite the use of a relatively consistent consent script which contained
information about the risk of the treatment.63
Ultimately,
though, any assessment of the “adequacy” of risk discussion assumes what level
of risk is acceptable to patients. A semi-structured interview study of stroke
survivors and their caregivers revealed participants preferred a balanced
presentation of risks and benefits, and that decision-making aids would be
beneficial in the context of making a consent decision. When clinicians were
interviewed in the same study, though, clinicians reported that risks dominated their
conversations. One nurse practitioner stated “I usually spend a fair amount of
time going into the negatives about it. Then [I] go over the percentages of
[the] likelihood of getting better.”64 Patients may not value a statistics-heavy
approach to discussing risk, however – one survey study of a convenience sample
of 184 patients, visitors, and employees recruited from a hospital’s cafeteria
revealed that after reading three sample scripts on alteplase’s risks and benefits,
a technical approach to discussing risks and benefits was the preferred by only
29% of respondents. A “parental” approach, where general statements about how while
risks existed, the benefits outweighed risks but the clinician would give this
medication to their loved one, was preferred by 55% of respondents.65
Perhaps the
largest challenge with assessing these conversations is that as a species, we
place much greater emphasis on shorter term risk than we do on longer term
risks and benefits. While any conversation with patients or their family should
contain a fair and balanced discussion of both the risks and benefits of thrombolysis,
the abstract perception of “benefit” is very difficult to grasp while the immediate
risks of hemorrhage and angioedema can be felt intrinsically. This is likely
why this particular concern is so real to so many individuals who are skeptical
of the benefits of thrombolysis – seeing and managing the oftentimes catastrophic
hazard of thrombolysis is an immediate experience after administering a
thrombolytic, while the benefits are not seen for months and only seen by the
individuals caring for the patient at home or in clinic. This is likely why the
bulk of individuals skeptical of thrombolytics are from the emergency medicine
specialty – these individuals only see the consequences, and never see
the benefits. Granted, it’s not possible to “see” the benefits at an individual
level, but those of us who work with stroke survivors during their recovery can
find it very easy to ascribe at least part of the recovery process to the
benefits of thrombolysis. I have the opportunity to work in an
interdisciplinary neurorecovery clinic which sees patients months to years
after their injury, and the benefits of thrombolysis can feel very real.
The Nature of the Conversation
I will preface
this further discussion with an obvious caveat – I, as a neurocritical care pharmacist,
have never and likely never will have the direct conversation with a patient or
their loved one about the risks and benefits of thrombolysis. I have not felt
the anxiety about helping a patient or family member make the right individual
decision. I can, though, highlight some of the evidence which has evaluated how
exactly this conversation could be had, and whether the way many institutions
currently conduct this conversation is best for patients. The 2022 AAN Policy
Statement on consent issues in the management of acute ischemic stroke states
that when patients retain decision making capacity or when surrogates
are available, they should be informed about the stroke diagnosis and rationale
for therapy, the prospects for a good functional recovery with and without treatment,
and the risks of treatment. However, when surrogates are not present, the
authors state “it may be reasonable to proceed with treatment on the
presumption of consent when patients’ cases fit extant inclusion/exclusion
criteria, contraindications (particularly, absolute contraindications) are
absent, and the overall balance of risks and benefits strongly favors
intervention” (emphasis mine).66 The statement suggests verbal consent
or refusal should be documented, but does not mandate any particular structure
to the consent process.
This vagueness
has led to substantial variation in exactly how consent/assent is obtained. One
survey of New York stroke centers found that written consent was required at
27.9% of centers when patients present within 3 hours of last known well but in
51.4% when patients presented after 3 hours, while documentation of consent was
not required at all at 18.4% and 7.2% of centers for patients presenting within
or after 3 hours, respectively.67 The variation in consent practices stems
from the challenging bioethical nature of thrombolysis in acute ischemic
stroke. The time sensitivity of acute therapy, uncertainty regarding patient-specific
preferences in the moment particularly when patients are decisionally impaired,
potential gaps in the understanding of thrombolysis in special populations, and
potentially devastating consequences of both treating or not treating with thrombolysis
make a universal approach to consent challenging.68 Several groups have proposed informed
refusal as a way to balance the time-sensitive nature of treatment with the
right of patient autonomy. This appears to be reasonable from a patient perspective,
as surveys including representative samples of the US population have revealed
that 76.2% of adults would want thrombolysis if they experienced an ischemic
stroke, which is a similar rate to the desire to receive CPR (75.9%), another
time-sensitive intervention where a type of informed refusal is used
(presumption of consent unless DNR status is known).69,70 The rates of acceptance were similar
when participants were asked what their decision would be if they could not
make the decision for themselves – when asked whether they would want
thrombolysis in a hypothetical scenario when they or a surrogate was not available
to make the decision, 78.1% of respondents stated they would want thrombolysis.70 Advance directives regarding
thrombolysis are also feasible, with one study of 121 patients in Ireland at risk
for stroke revealing that 82.6 and 89.3% of patients would opt for thrombolysis
if presenting with a stroke within 4.5 and 3 hours of last known well,
respectively. Over 75% of respondents preferred their physician make the final
decision to treat if they were decisionally impaired, with one participant
stating “I don’t think it’s fair to put the responsibility on someone who’s had
a stroke.”71
“Getting the
gist across” and allowing patients to refuse therapy if not within their goals
of care may be a safe and more efficient way of approaching the decision to
treat.58,72 Patients only remember about 55% of the
information shared with them about thrombolysis after 90 minutes, so the confidence
in a patient’s ability to make an informed decision at the time of stroke is
questionable.73 A verbal informed refusal method may
also be significantly more efficient as well and could reduce door to needle
time, which might translate to distinct clinical benefits. After four hospitals
in Singapore changed from a written consent to verbal consent process, door-to-needle
times declined by 5.6 minutes and the percent of patients who received
thrombolysis within 60 minutes of arrival increased from 35.6% to 66.1% (p =
0.01).74 Incremental improvements in time to
thrombolysis are associated with measurable improvements in patient outcomes,75 and if patients are given the
opportunity to refuse treatment after a reasonable discussion of the risks and
benefits rather than a lengthy written consent process, this may allow for
those improvements to occur.
Objection #3: Unacceptable Excess Harm is Caused After Misdiagnosis
Stroke is a
challenging diagnosis and is generally made based on clinical criteria alone. While
advanced imaging techniques like MRI and CT perfusion can produce more
definitive results, delaying treatment to obtain these results particularly at
centers who do not have them readily available can cause unacceptable delays to
treatment and thus the pressure to treat can be overwhelming even in the face
of diagnostic uncertainty. Because of this, there is sometimes a concern that a
more liberal approach to thrombolysis in borderline cases causes excess harm.
Of course, any adverse effect of a medication given to a patient who ultimately
is found to have not been indicated for it is unacceptable, but not
giving a time-sensitive acute therapy for a potentially devastating diagnosis
when some uncertainty exists is equally unacceptable. Luckily, patients who
present with stroke mimics ultimately have a degree of built in safety – the
most concerning adverse effect, hemorrhagic conversion, has a risk profile
directly linked to the severity of the underlying stroke. Because patients who
are ultimately found to have stroke mimics have no ischemic brain tissue, they
are at a significantly lower risk of hemorrhagic conversion than a traditional
stroke patient.
It's important
to frame this discussion from the perspective of making the decision to administer
a thrombolytic in the moment, rather than retrospectively wondering if the
decision that was made was the right one. In other words, if faced with a
patient where the diagnosis could either be ischemic stroke or a mimic (complex
migraine, Todd’s paralysis, functional neurologic disorder, etc), what is the risk-to-benefit
ratio of thrombolysis for that individual patient? First, it is helpful to
understand how often thrombolysis is given to stroke mimics. This depends
partly on how a stroke mimic is defined and reported, but registry studies have
estimated around 1.4-3.5% of thrombolytic administrations are given to patients
with an ultimate diagnosis of a stroke mimic.76–79 One single-center cohort study reported
a stroke mimic rate of 27.9%, but a third of “stroke mimics” were classified as
having transient ischemic attack, which is arguably disingenuous to refer to as
a stroke mimic and thus this center's true stroke mimic rate is likely lower
than reported.80 Other cohorts have reported rates of
stroke mimics up to 7-10.4%, but the higher rates of stroke mimics are likely a
function of smaller sample sizes than multicenter registry studies.81,82 Another group of patients that is often
grouped with stroke mimics are those with ischemia-negative imaging, where no
definitive ischemic lesion is identified on MRI or follow-up CT. It is more
difficult to disentangle whether these patients represent true stroke mimics or
an “aborted” stroke, but one cohort reported a 7% rate of “neuroimaging negative
cerebral ischemia” in all patients who received thrombolysis at their center.83
|
Pooled Data76,78–85 |
N |
% |
|
Patients who Received Thrombolysis |
81,675 |
|
|
Patients Diagnosed with Stroke Mimic |
2,933 |
3.6 |
|
Complex Migraine |
564 |
19.2 |
|
Functional
neurologic disorder |
343 |
11.7 |
|
Seizure |
335 |
11.4 |
|
Other |
1,691 |
57.6 |
Pooled incidence of stroke mimics who receive thrombolysis and their ultimate diagnosis
Putting this together, lysing stroke mimics is a relatively uncommon occurrence and fear of misdiagnosing a stroke mimic should not prevent one from administering a thrombolytic because the harms of not treating a true ischemic stroke far outweigh the risks of inappropriately treating a stroke mimic. Because patients with stroke mimics do not have any fresh ischemic cerebral tissue, they are somewhat protected from the risks of thrombolysis as well. Patients who present as stroke mimics who receive a thrombolytic experience symptomatic intracranial hemorrhage at significantly lower rates than patients with stroke (rates of 0%,76 0.4%,77 and 2%78 have been reported, all significantly lower than patients with stroke) and experience angioedema vanishingly rarely (0% in the two studies that distinctly report it).78,83 This of course does mean that everyone who presents with symptoms that could be vaguely suggestive of stroke should receive thrombolysis, of course, but it does suggest that if there is a lack of a clear alternate explanation for stroke-like symptoms, thrombolysis is reasonable more often than not if the patient otherwise meets criteria for thrombolysis. Interestingly, patient intuition may play a role here, too – studies of informed refusal of thrombolysis have found that patients with stroke mimics are significantly more likely to refuse treatment than those with actual strokes.59,86,87
Objection #4: Legal Risk of Adverse Effects
Unfortunately,
an additional hesitation about thrombolysis in AIS stems from concerns over
litigation because of the risk of harm caused by thrombolysis. Thrombolysis is
of course high risk, and with it comes the risk of potentially life-threatening
hemorrhagic conversion. While this is a known and expected risk, there is the
concern that families and caregivers could view the harms caused by
thrombolysis as the fault of the physician who prescribed the treatment. One
semi-structed interview of emergency medicine physicians in the UK highlighted the
cognitive dissonance faced by providers discussing the risks and benefits of
thrombolysis, with one participant noting a memorable case: “I told them about
risk of bleeding…and despite the fact that 2 CTs showed he didn’t have any
bleeding at all… he just took a massive brain swelling [sic]. When he died the
family turned around and said that we had killed him basically.”62 Undoubtedly, many clinicians may have
experienced similar sentiments and could feel that extreme caution to avoid
liability in the case complications occur.
Interestingly,
this has not been observed to be the case in the US medicolegal system.
Multiple evaluations of available legal cases have shown that litigation
related to acute stroke care is uncommon, and providers are more likely to be
sued for not administering thrombolysis rather than for complications of
therapy. At least four analyses of legal cases have found this consistent
result, and patterns of litigation have remained consistent over the decades as
the strength of the evidence supporting thrombolysis has grown. A summary of
the four publications is provided below:
|
Study |
Legal Cases Reviewed |
Cases Related to tPA Identified |
Findings |
|
Liang 200888 |
290,000 |
33 |
66.7% of cases involved delay in
diagnosis 87.9% related to failure to treat
with tPA 9.1% related to harms caused by tPA
Of cases ruled in favor of the
plaintiff (n=12), 83.3% were related to failure to administer tPA |
|
Bhatt 201389 |
18,113 |
40 |
70% related to failure to treat
with tPA 5% related to harms caused by tPA
Factors associated with a verdict
favoring the defendant included consent documentation, expert witness
testimony, patients falling out of the time window, or lack of tPA
availability
Factors associated with a verdict
favoring the plaintiff included failure to treat, failure to diagnose,
failure to transfer, or failure to document consent |
|
Haslett 201990 |
1.3 million |
246 (AIS) Included all cases related to
acute stroke, not just tPA |
26% (of all cases) related to failure
to treat with tPA, of which 62% ruled in favor of defendant 0.4% (n=1, of all cases) related to harms
caused by tPA, which was ruled in favor of the defendant |
|
Ganti 202291 |
40,000 |
66 |
100% related to failure to treat
with tPA |
Ultimately,
while litigation is rare, it is much more likely to be because of a lack of
treatment rather than because of complication of treatment. In cases where
litigation is pursued as a result of adverse effects, the decision if most
often in favor of the defendant. Clinicians should treat patients according to
the most up to date literature rather than practice defensively to avoid
litigation to begin with, and in this case avoiding thrombolysis as a method of
defensive medicine is a lose-lose situation – not only is it a disservice to
the patient, the avoidance itself is a risk of legal action by not follow
evidence based care.
Conclusion
If you’ve stuck
it out all the way here, thank you for taking the time to read my thoughts on
thrombolysis in stroke and my opinion on the “debate” of its merits. Thrombolysis
is one of the most effective therapies available for acute ischemic stroke
along with thrombectomy, and endlessly debating the merits of the RCTs
supporting its use while ignoring the overwhelming body of evidence
demonstrating clear and measurable improvements in stroke outcomes with improving
utilization of thrombolysis ultimately leads to confusion and does nothing to
benefit patients. Thrombolysis has real risks and these need to be communicated
to patients to ensure therapy is within goals of care, but these risks are far
outweighed by the benefits, benefits that are far more meaningful to patients
than the risks.
Andrew Webb, PharmD, BCCCP
Clinical Pharmacist, Neurocritical Care
Massachusetts General Hospital
Ajwebb@mgh.harvard.edu
References
1. Powers WJ, Rabinstein AA, Ackerson
T, et al. Guidelines for the early management of patients with acute ischemic
stroke: 2019 update to the 2018 guidelines for the early management of acute
ischemic stroke a guideline for healthcare professionals from the American
Heart Association/American Stroke A. Stroke. 2019;50(49):e344-e418.
doi:10.1161/STR.0000000000000211
2. Berge
E, Whiteley W, Audebert H, et al. European Stroke Organisation (ESO) guidelines
on intravenous thrombolysis for acute ischaemic stroke. Eur Stroke
J. 2021;6(1):I-LXII. doi:10.1177/2396987321989865
3. Alamowitch
S, Turc G, Palaiodimou L, et al. European Stroke Organisation (ESO) expedited
recommendation on tenecteplase for acute ischaemic stroke. Eur Stroke
J. 2023;8(1):8-54. doi:10.1177/23969873221150022
4. Hyland
HH. Thrombosis of intracranial arteries: report of three cases involving,
respectively, the anterior cerebral, basilar and internal carotid arteries. Arch
Neurol Psychiatry. 1933;30(2):342-356.
doi:10.1001/archneurpsyc.1933.02240140106005
5. Elvidge
AR, Werner A. Hemiplegia and thrombosis of the internal carotid system. Arch
Neurol Psychiatry. 1951;66(6):752-782.
doi:10.1001/archneurpsyc.1951.02320120085011
6. Sussman
BJ, Fitch TS. Thrombolysis with fibrinolysin in cerebral arterial occlusion. JAMA.
1958;167(14):1705-1709. doi:10.1001/jama.1958.02990310011002
7. Herndon
RM, Meyer JS, Johnson JF, Landers J. Treatment of cerebrovascular thrombosis
with fibrinolysin preliminary report. Am Journal of Cardiol.
1960;6(2):540-545. doi:10.1016/0002-9149(60)90350-7
8. Atkin
N, Nitzberg S, Dorsey J. Lysis of Intracerebral Thromboembolism With
Fibrinolysin: Report of a Case. Angiology. 1964;15(10):436-439.
doi:10.1177/000331976401501003
9. McBroom
FP. Lysis of a Cerebral Thrombus with Fibrinolysin: A Case Report. J Natl
Med Assoc. 1963;55(4):274-276.
10. Meyer
JS, Gilroy J, Barnhart MI, Johnson JF. Therapeutic thrombolysis in cerebral
thromboembolism: Double-blind evaluation of intravenous plasmin therapy in
carotid and middle cerebral arterial occlusion. Neurology.
1963;13(11):927-927. doi:10.1212/WNL.13.11.927
11. Meyer
JS, Gilroy J, Barnhart MI, Johnson JF. Anticoagulants plus streptokinase
therapy in progressive stroke. JAMA. 1964;189:373.
doi:10.1001/jama.1964.03070050039011
12. Report
from the Joint Commitee for Stroke Facilities -- XII. Computed Tomography in
the Managemnt of Cerebrovascular Disease. Stroke. 1975;6(1):103-107.
doi:10.1161/01.STR.6.1.103
13. Fletcher
AP, Alkjaersig N, Lewis M, et al. A pilot study of urokinase therapy in
cerebral infarction. Stroke. 1976;7(2):135-142.
doi:10.1161/01.STR.7.2.135
14. Hanaway
J, Torack R, Fletcher AP, Landau WM. Intracranial bleeding associated with
urokinase therapy for acute ischemic hemispheral stroke. Stroke.
1976;7(2):143-146. doi:10.1161/01.STR.7.2.143
15. Millikan
CH, McDowell FH. Treatment of progressing stroke. Stroke.
1981;12(4):397-409. doi:10.1161/01.STR.12.4.397
16. Drury
I, Whisnant JP, Garraway WM. Primary intracerebral hemorrhage: Impact of CT on
incidence. Neurology. 1984;34(5):653-653. doi:10.1212/WNL.34.5.653
17. Hacke
W, Zeumer H, Ferbert A, Brückmann H, del Zoppo GJ. Intra-arterial thrombolytic
therapy improves outcome in patients with acute vertebrobasilar occlusive
disease. Stroke. 1988;19(10):1216-1222. doi:10.1161/01.STR.19.10.1216
18. del
Zoppo GJ, Ferbert A, Otis S, et al. Local intra-arterial fibrinolytic therapy
in acute carotid territory stroke. A pilot study. Stroke.
1988;19(3):307-313. doi:10.1161/01.STR.19.3.307
19. Sloan
MA. Thrombolysis and Stroke: Past and Future. Arch Neurol.
1987;44(7):748-768. doi:10.1001/archneur.1987.00520190052017
20. Wardlaw
JM, Warlow CP. Thrombolysis in acute ischemic stroke: does it work? Stroke.
1992;23(12):1826-1839. doi:10.1161/01.STR.23.12.1826
21. Solomon
NA, Glick HA, Russo CJ, Lee J, Schulman KA. Patient preferences for stroke
outcomes. Stroke. 1994;25(9):1721-1725. doi:10.1161/01.STR.25.9.1721
22. Multicentre
Acute Stroke Trial—Italy (MAST-I) Group. Randomised controlled trial of
streptokinase, aspirin, and combination of both in treatment of acute ischaemic
stroke. Lancet. 1995;346(8989):1509-1514.
doi:10.1016/S0140-6736(95)92049-8
23. Multicenter
Acute Stroke Trial--Europe Study Group, Hommel M, Cornu C, Boutitie F, Boissel
JP. Thrombolytic therapy with streptokinase in acute ischemic stroke. N Engl
J Med. 1996;335(3):145-150. doi:10.1056/NEJM199607183350301
24. Donnan
A, Davis M, Chambers BR, et al. Streptokinase for Acute Ischemic Stroke With
Relationship to Time of Administration. JAMA. 1996;276(12):961-966.
25. Zivin
JA, Fisher M, DeGirolami U, Hemenway CC, Stashak JA. Tissue Plasminogen
Activator Reduces Neurological Damage After Cerebral Embolism. Science.
1985;230(4731):1289-1292. doi:10.1126/science.3934754
26. Bednar
MM, McAuliffe T, Raymond S, Gross CE. Tissue plasminogen activator reduces
brain injury in a rabbit model of thromboembolic stroke. Stroke.
1990;21(12):1705-1709. doi:10.1161/01.STR.21.12.1705
27. GUSTO
Investigators. An International Randomized Trial Comparing Four Thrombolytic
Strategies for Acute Myocardial Infarction. N Engl J Med.
1993;329(10):673-682. doi:10.1056/NEJM199309023291001
28. Brott
TG, Haley ECJ, Levy DE, et al. Urgent therapy for stroke. Part I. Pilot study
of tissue plasminogen activator administered within 90 minutes. Stroke.
1992;23(5):632-640.
29. Haley
EC, Levy DE, Brott TG, et al. Urgent therapy for stroke. Part II. Pilot study
of tissue plasminogen activator administered 91-180 minutes from onset. Stroke.
1992;23(5):641-645. doi:10.1161/01.STR.23.5.641
30. The
National Institute of Neurological Disorders and Stroke rt-PA Stroke Study
Group. Tissue plasminogen activator for acute ischemic stroke. N Engl Journal Med. 1995;333(24):1581-1588.
doi:10.1056/NEJM199512143332401
31. Hacke
W, Kaste M, Fieschi C, et al. Intravenous Thrombolysis With Recombinant Tissue
Plasminogen Activator for Acute Hemispheric Stroke: The European Cooperative
Acute Stroke Study (ECASS). JAMA. 1995;274(13):1017-1025.
doi:10.1001/jama.1995.03530130023023
32. Hacke
W, Kaste M, Fieschi C, et al. Randomised double-blind placebo-controlled trial
of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke
(ECASS II). Lancet. 1998;352(9136):1245-1251.
doi:10.1016/S0140-6736(98)08020-9
33. Clark
WM, Albers GW, Madden KP, Hamilton S. The rtPA (Alteplase) 0- to 6-Hour Acute
Stroke Trial, Part A (A0276g). Stroke. 2000;31(4):811-816.
doi:10.1161/01.STR.31.4.811
34. Clark
WM, Wissman S, Albers GW, Jhamandas JH, Madden KP, Hamilton S. Recombinant
tissue-type plasminogen activator (Alteplase) for ischemic stroke 3 to 5 hours
after symptom onset the ATLANTIS study: A randomized controlled trial. JAMA. 1999;282(21):2019-2026.
doi:10.1001/jama.282.21.2019
35. Albers
GW, Clark WM, Madden KP, Hamilton SA. ATLANTIS Trial: Results for Patients
Treated Within 3 Hours of Stroke Onset. Stroke. 2002;33(2):493-496.
doi:10.1161/hs0202.102599
36. Hacke
W, Kaste M, Bluhmki E, et al. Thrombolysis with Alteplase 3 to 4.5 Hours after
Acute Ischemic Stroke. N Engl J Med.
2008;359(13):1317-1329. doi:10.1056/NEJMoa0804656
37. Davis
SM, Donnan GA, Parsons MW, et al. Effects of alteplase beyond 3 h after stroke
in the Echoplanar Imaging Thrombolytic Evaluation Trial (EPITHET): a
placebo-controlled randomised trial. Lancet Neurol.
2008;7(4):299-309. doi:10.1016/S1474-4422(08)70044-9
38. Sandercock
P, Wardlaw JM, Lindley RI, et al. The benefits and harms of intravenous
thrombolysis with recombinant tissue plasminogen activator within 6 h of acute
ischaemic stroke (the third international stroke trial [IST-3]): A randomised
controlled trial. Lancet. 2012;379(9834):2352-2363.
doi:10.1016/S0140-6736(12)60768-5
39. Ma H,
Campbell BCV, Parsons MW, et al. Thrombolysis Guided by Perfusion Imaging up to
9 Hours after Onset of Stroke. N Engl J Med.
2019;380(19):1795-1803. doi:10.1056/NEJMoa1813046
40. Thomalla
G, Simonsen CZ, Boutitie F, et al. MRI-Guided thrombolysis for stroke with
unknown time of onset. N Engl J Med.
2018;379(7):611-622. doi:10.1056/NEJMoa1804355
41. Ridgeon
EE, Young PJ, Bellomo R, Mucchetti M, Lembo R, Landoni G. The Fragility Index
in Multicenter Randomized Controlled Critical Care Trials. Crit Care Med.
2016;44(7):1278-1284. doi:10.1097/CCM.0000000000001670
42. Hoffman
JR, Schriger DL. A Graphic Reanalysis of the NINDS Trial. Ann Emerg Medicine. 2009;54(3):329-336.e35.
doi:10.1016/j.annemergmed.2009.03.019
43. Alper
BS, Foster G, Thabane L, Rae-Grant A, Malone-Moses M, Manheimer E. Thrombolysis
with alteplase 3-4.5 hours after acute ischaemic stroke: trial reanalysis
adjusted for baseline imbalances. BMJ Evid Based Med.
2020;25(5):168-171. doi:10.1136/bmjebm-2020-111386
44. Ingall
TJ, O’Fallon WM, Asplund K, et al. Findings From the Reanalysis of the NINDS
Tissue Plasminogen Activator for Acute Ischemic Stroke Treatment Trial. Stroke.
2004;35(10):2418-2424. doi:10.1161/01.STR.0000140891.70547.56
45. Saver
JL, Gornbein J, Starkman S. Graphic Reanalysis of the Two NINDS-tPA Trials
Confirms Substantial Treatment Benefit. Stroke. 2010;41(10):2381-2390.
doi:10.1161/STROKEAHA.110.583807
46. Meyer
BC, Lyden PD. The Modified National Institutes of Health Stroke Scale (mNIHSS):
Its Time Has Come. Int J Stroke. 2009;4(4):267-273.
doi:10.1111/j.1747-4949.2009.00294.x
47. Martin-Schild
S, Albright KC, Tanksley J, et al. Zero on the NIHSS Does NOT Equal the Absence
of Stroke. Ann Emerg Med. 2011;57(1):42-45.
doi:10.1016/j.annemergmed.2010.06.564
48. Wahlgren
N, Ahmed N, Dávalos A, et al. Thrombolysis with alteplase for acute ischaemic
stroke in the Safe Implementation of Thrombolysis in Stroke-Monitoring Study
(SITS-MOST): an observational study. Lancet. 2007;369(9558):275-282.
doi:10.1016/S0140-6736(07)60149-4
49. Fonarow
GC, Zhao X, Smith EE, et al. Door-to-Needle Times for Tissue Plasminogen
Activator Administration and Clinical Outcomes in Acute Ischemic Stroke Before
and After a Quality Improvement Initiative. JAMA.
2014;311(16):1632-1640. doi:10.1001/jama.2014.3203
50. Saver
JL, Fonarow GC, Smith EE, et al. Time to Treatment With Intravenous Tissue
Plasminogen Activator and Outcome From Acute Ischemic Stroke. JAMA.
2013;309(23):2480-2488. doi:10.1001/jama.2013.6959
51. Emberson
J, Lees KR, Lyden P, et al. Effect of treatment delay, age, and stroke severity
on the effects of intravenous thrombolysis with alteplase for acute ischaemic
stroke: a meta-analysis of individual patient data from randomised trials. Lancet.
2014;384(9958):1929-1935. doi:10.1016/S0140-6736(14)60584-5
52. Altman
DG. Problems in dichotomizing continuous variables. Am J Epidemiol. 1994;139(4):442-442. doi:10.1093/oxfordjournals.aje.a117020
53. Altman
DG, Royston P. The cost of dichotomising continuous variables. BMJ.
2006;332(7549):1080.1. doi:10.1136/bmj.332.7549.1080
54. Saver
JL, Gornbein J. Treatment effects for which shift or binary analyses are
advantageous in acute stroke trials. Neurology. 2009;72(15):1310-1315.
doi:10.1212/01.wnl.0000341308.73506.b7
55. Ganesh
A, Luengo-Fernandez R, Wharton RM, Rothwell PM, on behalf of the Oxford
Vascular Study. Ordinal vs dichotomous analyses of modified Rankin Scale,
5-year outcome, and cost of stroke. Neurology. 2018;91(21):e1951-e1960.
doi:10.1212/WNL.0000000000006554
56. Erler
KS, Wu R, DiCarlo JA, et al. Association of Modified Rankin Scale With Recovery
Phenotypes in Patients With Upper Extremity Weakness After Stroke. Neurology.
2022;98(18):e1877-e1885. doi:10.1212/WNL.0000000000200154
57. Cantrill
SV, Brown MD, Brecher D, et al. Clinical Policy: Use of Intravenous Tissue
Plasminogen Activator for the Management of Acute Ischemic Stroke in the
Emergency Department. Ann of Emerg Med.
2015;66(3):322-333.e31. doi:10.1016/j.annemergmed.2015.06.031
58. Skolarus
LE, O’Brien A, Meurer WJ, Zikmund Fisher BJ. Getting the Gist Across Is Enough
for Informed Consent for Acute Stroke Thrombolytics. Stroke.
2019;50(6):1595-1597. doi:10.1161/STROKEAHA.119.024653
59. Vahidy
FS, Rahbar MH, Lal AP, Grotta JC, Savitz SI. Patient Refusal of Thrombolytic
Therapy for Suspected Acute Ischemic Stroke. Int J Stroke.
2015;10(6):882-886. doi:10.1111/j.1747-4949.2012.00945.x
60. Zonjee
VJ, Slenders JPL, de Beer F, et al. Practice variation in the informed consent
procedure for thrombolysis in acute ischemic stroke: a survey among
neurologists and neurology residents. BMC Med Ethics. 2021;22(1):114.
doi:10.1186/s12910-021-00684-6
61. Mendelson
SJ, Courtney DM, Gordon EJ, Thomas LF, Holl JL, Prabhakaran S. National
Practice Patterns of Obtaining Informed Consent for Stroke Thrombolysis. Stroke.
2018;49(3):765-767. doi:10.1161/STROKEAHA.117.020474
62. Lie
MLS, Murtagh MJ, Watson DB, et al. Risk communication in the hyperacute setting
of stroke thrombolysis: an interview study of clinicians. Emerg Med J.
2015;32(5):357-363. doi:10.1136/emermed-2014-203717
63. Thomas
L, Viswanathan A, Cochrane T, et al. Variability in the Perception of Informed
Consent for IV-tPA during Telestroke Consultation. Front Neurol.
2012;3:128. doi:10.3389/fneur.2012.00128
64. Decker
C, Chhatriwalla E, Gialde E, et al. Patient-Centered Decision Support in Acute
Ischemic Stroke. Circ Cardiovasc Qual Outcomes.
2015;8(6_suppl_3):S109-S116. doi:10.1161/CIRCOUTCOMES.115.002003
65. Murray
AM, Petrone AB, Adcock AK. Utilization of a Parental Approach to Informed
Consent in Intravenous Tissue Plasminogen Activator Administration
Decision-Making: Patient Preference and Ethical Considerations. Neurol Res Int. 2019;2019:e9240603. doi:10.1155/2019/9240603
66. Sattin
JA, Chiong W, Bonnie RJ, Kirschen MP, Russell JA, on behalf of the Ethics, Law,
and Humanities Committee, a joint committee of the American Academy of
Neurology, American Neurological Association, and Child Neurology Society.
Consent Issues in the Management of Acute Ischemic Stroke. Neurology.
2022;98(2):73-79. doi:10.1212/WNL.0000000000013040
67. Weintraub
MI, Colello AD, Johnson SA, et al. Informed Consent for Intravenous Tissue
Plasminogen Activator in New York State Designated Stroke Centers. J Stroke Cerebrovasc Dis. 2017;26(6):1274-1279.
doi:10.1016/j.jstrokecerebrovasdis.2017.01.021
68. Mbonde
A, Young MJ, Dmytriw AA, et al. Informed consent practices for acute stroke
therapy: principles, challenges and emerging opportunities. J Neurol.
2024;271(1):188-197. doi:10.1007/s00415-023-12028-y
69. Chiong
W, Kim AS, Huang IA, Farahany NA, Josephson SA. Testing the Presumption of
Consent to Emergency Treatment for Acute Ischemic Stroke. JAMA.
2014;311(16):1689-1691. doi:10.1001/jama.2014.3302
70. Chiong
W, Kim AS, Huang IA, Farahany NA, Josephson SA. Inability to consent does not
diminish the desirability of stroke thrombolysis. Ann Neurol.
2014;76(2):296-304. doi:10.1002/ana.24209
71. Ó
Flatharta T, Khan A, Walsh T, O’Donnell M, O’Keefe ST. Advance preferences
regarding thrombolysis in patients at risk for stroke: a cross-sectional study.
QJM. 2015;108(1):27-31.
doi:10.1093/qjmed/hcu142
72. Schwamm
LH. Acute Stroke: Shifting From Informed Consent to Informed Refusal of
Intravenous Tissue-Type Plasminogen Activator. Circ Cardiovas Qual Outcomes. 2015;8(6_suppl_3):S69-S72.
doi:10.1161/CIRCOUTCOMES.115.002282
73. Schuster
L, Essig F, Daneshkhah N, et al. Ability of patients with acute ischemic stroke
to recall given information on intravenous thrombolysis: Results of a
prospective multicenter study. Eur Stroke J.
2023;8(1):241-250. doi:10.1177/23969873221143856
74. Xu H,
Silva DAD, Woon FP, et al. Changes in Informed Consent Policy and Treatment
Delays in Stroke Thrombolysis. J Stroke Cerebrovasc Dis. 2021;30(3). doi:10.1016/j.jstrokecerebrovasdis.2020.105551
75. Meretoja
A, Keshtkaran M, Saver JL, et al. Stroke Thrombolysis: Save a Minute, Save a
Day. Stroke. 2014;45(4):1053-1058. doi:10.1161/STROKEAHA.113.002910
76. Artto
V, Putaala J, Strbian D, et al. Stroke Mimics and Intravenous Thrombolysis. Ann Emerg Med. 2012;59(1):27-32.
doi:10.1016/j.annemergmed.2011.09.011
77. Ali-Ahmed
F, Federspiel JJ, Liang L, et al. Intravenous Tissue Plasminogen Activator in
Stroke Mimics. Circ Cardiovasc Qual Outcomes.
2019;12(8):e005609. doi:10.1161/CIRCOUTCOMES.119.005609
78. Zinkstok
SM, Engelter ST, Gensicke H, et al. Safety of Thrombolysis in Stroke Mimics. Stroke.
2013;44(4):1080-1084. doi:10.1161/STROKEAHA.111.000126
79. Chen
Y, Bogosavljevic V, Leys D, Jovanovic D, Beslac-Bumbasirevic L, Lucas C.
Intravenous thrombolytic therapy in patients with stroke mimics: baseline
characteristics and safety profile. E J Neurol.
2011;18(10):1246-1250. doi:10.1111/j.1468-1331.2011.03367.x
80. Garg
R, Rech MA, Schneck M. Stroke Mimics: An Important Source of Bias in Acute
Ischemic Stroke Research. J Stroke Cerebrovasc Dis.
2019;28(9):2475-2480. doi:10.1016/j.jstrokecerebrovasdis.2019.06.019
81. Tsivgoulis
G, Alexandrov AV, Chang J, et al. Safety and Outcomes of Intravenous
Thrombolysis in Stroke Mimics. Stroke. 2011;42(6):1771-1774.
doi:10.1161/STROKEAHA.110.609339
82. Scott
PA, Silbergleit R. Misdiagnosis of stroke in tissue plasminogen
activator–treated patients: Characteristics and outcomes. Ann Emerg Med. 2003;42(5):611-618. doi:10.1016/S0196-0644(03)00443-8
83. Chernyshev
OY, Martin-Schild S, Albright KC, et al. Safety of tPA in stroke mimics and
neuroimaging-negative cerebral ischemia. Neurology.
2010;74(17):1340-1345. doi:10.1212/WNL.0b013e3181dad5a6
84. Förster
A, Griebe M, Wolf ME, Szabo K, Hennerici MG, Kern R. How to identify stroke
mimics in patients eligible for intravenous thrombolysis? J Neurol.
2012;259(7):1347-1353. doi:10.1007/s00415-011-6354-9
85. Giraldo
EA, Khalid A, Zand R. Safety of Intravenous Thrombolysis within 4.5 h of
Symptom Onset in Patients with Negative Post-Treatment Stroke Imaging for
Cerebral Infarction. Neurocrit Care. 2011;15(1):76-79.
doi:10.1007/s12028-011-9523-x
86. Zha
A, Rosero A, Malazarte R, et al. Thrombolytic Refusal Over Telestroke. Neurology
Clinical Practice. 2021;11(3):e287-e293. doi:10.1212/CPJ.0000000000000975
87. Kleist
T von, Meyer D, Rapp K, Meyer BC, Modir R. Characteristics and Outcomes of
Patients who Refuse Intravenous Thrombolysis for Acute Ischemic Stroke - The
San Diego Experience. J Stroke Cerebrovasc Dis.
2020;29(11). doi:10.1016/j.jstrokecerebrovasdis.2020.105137
88. Liang
BA, Zivin JA. Empirical Characteristics of Litigation Involving Tissue
Plasminogen Activator and Ischemic Stroke. Ann Emerg Med.
2008;52(2):160-164. doi:10.1016/j.annemergmed.2007.12.017
89. Bhatt
A, Safdar A, Chaudhari D, et al. Medicolegal Considerations with Intravenous
Tissue Plasminogen Activator in Stroke: A Systematic Review. Stroke Res
Treat. 2013;2013:562564. doi:10.1155/2013/562564
90. Haslett
JJ, Genadry L, Zhang X, et al. Systematic Review of Malpractice Litigation in
the Diagnosis and Treatment of Acute Stroke. Stroke.
2019;50(10):2858-2864. doi:10.1161/STROKEAHA.119.025352
91. Ganti
L, Kwon B, George A, Stead T, Plamoottil C, Banerjee P. Tissue plasminogen
activator and patients with acute ischemic stroke: The litigation landscape. J Am College of Emergency Phys Open. 2022;3(1):e12646.
doi:10.1002/emp2.12646
Comments
Post a Comment