tPA in ischemic stroke, “answers”


Check out our own Dr. Anand Swaminathan discussing this topic and more on ischemic stroke on ER Cast here and here!

1. How do you control blood pressure (BP) in patients who will be or/are receiving tPA?

For patients not receiving tPA for acute ischemic stroke, allowing autoregulation of blood pressure has long been the norm. tPA has muddied the waters, somewhat, for management of blood pressure for acute ischemic stroke. In the pilot NINDS study (Haley, 1993) of tPA in acute stroke, and in previous tPA and myocardial infarction studies, there was a higher association with intracranial hemorrhage (ICH) in patients with a blood pressure greater than 185 mm Hg systolic, 110 mm Hg diastolic, or who underwent aggressive treatment to reach these levels. Thus, in the randomized NINDS tPA Stroke Study (NINDS, 1995), patients were excluded if their blood pressure did not reach these goals. In the NINDS study, the term “aggressive treatment,” was not defined prospectively in the protocol. However, it has been thought to mean by expert consensus: intravenous nitroprusside, repeated doses of IV labetalol, enalaprilat, or nifedipine.

EML tPA answers

Specifically, if patients have a diastolic blood pressure >140 on two readings, they should be started on a continuous IV infusion of an antihypertensive agent, and they are not candidates for tPA therapy (Broderick, 1996). Patients who require more than two doses of labetalol or other antihypertensive agents to decrease blood pressure to <185 systolic or 110 diastolic are typically not appropriate for thrombolytic therapy (Broderick, 1996). This is a relative contraindication to thrombolytic therapy. Some stroke physicians, however, will still treat an acute ischemic stroke with tPA after two labetolol doses have been used, and a nicardipine drip has been started despite the lack of significant evidence for this practice.

Agents used to treat blood pressure in ischemic stroke should be easily titratable, have a quick onset of action, and limited risk of excessive or sudden onset of action.

In patients who require blood pressure treatment to be at an appropriate range for thrombolytic therapy, IV labetalol is a popular first line agent. Labetalol is easily titratable and is commonly started at 10 mg IV over 1-2 minutes. This can be repeated or doubled every 10 to 20 minutes. Another choice for blood pressure control is enalaprilat, which can be given in 1.25 mg increments. Nicardipine is commonly used as a titratable continuous infusion. Agents such as nitroglycerin or sublingual nifedipine may have effects that are more unpredictable like rapid drops in pressure with reflex tachycardia, so are considered second line, and are rarely used. Any patient who receives antihypertensives for ischemic stroke requires serial neurologic exams to look for signs of deterioration (Broderick, 1996).

2. Do you treat patients with tPA up to 4.5 hours after onset of symptoms. If so, which ones?

In February 2013, the American College of Emergency Physicians (ACEP) released a clinical policy statement on acute ischemic stroke, stating that they were giving treatment between 3 and 4.5 hours a Class B recommendation (ACEP, 2013). However, at this time, the use of tPA for stroke patients between 3 and 4.5 hours is not yet FDA approved, and remains widely and passionately debated.

Evidence for thrombolytics between 3 and 4.5 hours largely comes from the ECASS III trial. The benefit of tPA between 3 and 4.5 hours was directly tested in the ECASS III randomized controlled trial (Hacke, 2008). This trial used the same dosing as well as the same inclusion/exclusion criteria as the NINDS trial. This trial also excluded patients greater than age 80, those with a baseline NIH stroke scale (NIHSS) of 25 or greater, any oral anticoagulant use, and those with the combination of a previous stroke and diabetes mellitus. The number needed to treat in ECASS III was 14 patients, a more modest number than that found in NINDS, where the NNT was 8. Although the ICH rate in ECASS III was 27% in treated patients compared to 17% in untreated patients, there was no significant difference in overall mortality at 90 days. The incidence of symptomatic intracranial hemorrhage was 2.4% in treated patients compared to 0.2% in untreated patients.  Furthermore, the rate of symptomatic ICH was not higher in ECASS III compared to that in the NINDS trial (Hacke, 2008).  These trials defined a symptomatic ICH as one in which a new ICH was found on head CT in a patient with clinical deterioration following an acute ischemic stroke (NINDS, 1995).

In 2009, a metanalysis was published to specifically look at the efficacy and safety of tPA in the 3 to 4.5 hour time frame (Landsberg, 2009). They evaluated pooled data from patients in this time frame from four fairly homogenous studies: ECASS-I (n = 234), ECASS-II (n = 265), ECASS-III (n = 821) and The Alteplase Thrombolysis for Acute Noninterventional Therapy in Ischemic Stroke (ATLANTIS) (n = 302). These authors concluded that patients treated in this time window have an increased rate of favorable outcome without adversely affecting mortality (Landsberg, 2009).

One of the concerns physicians expressed behind allowing tPA administration up to 4.5 hours is that patients and treating physicians might feel as though they had more time to treat acute ischemic strokes, which could lead to a decreased benefit of the drug. Another controversy behind administering tPA up to 4.5 hours out in acute ischemic stroke was that the ECASS III study used a slightly different definition of symptomatic ICH (sICH) than the NINDS trial did. However, even once the more conservative NINDS definition of sICH was applied in the ECASS III trial, the percent of sICH remained the same in both of the studies.

Based on these studies, tPA may be safe when administered 3 to 4.5 hours after symptom onset, as long as the specific safety criteria from the ECASS III and NINDS trials are met. However, patients still have the best outcome when tPA is administered as early as possible. The final line in the ECASS III trial states: “Having more time does not mean we should be allowed to take more time.”

3. How do you determine if an acute ischemic stroke is improving enough to not give tPA to a patient?

One of the most common reasons to withhold tPA in ischemic stroke is with mild or rapidly improving stroke symptoms (Nedeltchev, 2007).  For example, the American Heart Association stroke guidelines state that eligibility for tPA requires that “neurological signs should not be minor and isolated” (Adams, 2007). The reason behind this is that patients with rapidly improving symptoms are likely having a TIA, rather than a CVA .

The NINDS recombinant tPA Stroke Trial (NINDS, 1995) included RISS (rapidly improving stroke symptoms) as an exclusion criterion to avoid treatment of transient ischemic attacks which would have recovered completely without treatment. The NINDS trial included a very small number of RISS patients, which they defined as an NIHSS of 5 or less (58 patients with RISS were included, but 2971 were excluded due to mild symptoms). Thus, conclusions about this subgroup of stroke patients cannot be drawn from the NINDS study (Khatri, 2010).

When the FDA approved tPA in 1996, all eligibility criteria from the NINDS recombinant tPA Stroke Study were adopted (NINDS, 1995). The package insert gave contraindications and warnings directly from the study protocol, including excluding those with RISS.  The TREAT task force attempted to clarify this exclusion criteria (TREAT, 2013). They held an in-person “RISS Summit” to obtain a better understanding of this phenomenon.

The results of the TREAT task force were that, in the absence of other contraindications, patients who experience improvement of any degree, but have a persisting neurologic deficit that is potentially disabling, should be treated with IV tPA (TREAT, 2013). Improvement should be monitored for the time needed to prepare and administer the IV tPA. There was also consensus in the TREAT task force that all neurologic deficits present at the time of the treatment decision should be considered in the patient’s individual risk and benefit, as well as the patient’s baseline functional status (TREAT, 2013).

Many studies, however, have suggested that the outcome of patients with MRIS (mild and rapidly improving symptoms) who do not receive tPA is not always benign. A large study from Canada found that 32% of patients considered “too good to treat” were dependent at hospital discharge or had died (Barber, 2001). A separate study from Massachusetts General Hospital reported that patients with a high initial NIHSS, but with RISS, had a four times greater chance of neurologic worsening than patients presenting with initial mild symptoms (Adams, 2007). A third study, from UCLA, demonstrated that 10% of patients who were excluded from thrombolysis only because of their RISS status showed early neurological deterioration. Twenty percent showed a poor outcome at discharge as defined by a modified Rankin score of 3 or greater (Rajajee, 2006).

Nedelchev, et al., also found that patients with persisting proximal vessel occlusions and RISS were 7 times (95% CI: 1.1 to 45.5; P0.038) more likely to have an unfavorable outcome at three months (2007). They defined proximal occlusions as those of the internal carotid artery, M1 and M2 segments of the middle cerebral artery, A1 segment of the anterior cerebral artery, V4 segment of the vertebral artery, basilar artery, and P1 segment of the posterior cerebral artery. They also found that rapidly improving but still severe symptoms (NIHSS greater than or equal to 10 points on admission) increased the odds of unfavorable outcome 17-fold (95% – CI: 1.8 – 159.5; P = 0.013). These findings suggest that patients with persistent large-vessel occlusions and those with a NIHSS score greater than or equal to 10 points at onset of symptoms might benefit from thrombolysis despite resultant mild symptoms or rapidly improving symptoms at presentation (Nedeltchev, 2007). This study demonstrated that 75% of patients with mild or rapidly improving symptoms who were not treated had a favorable outcome at 3 months, defined by a modified Rankin score of 0 or 1, without treatment.

4. Do you use a specific age cutoff when determining whether or not a patient should or should not receive tPA?

Elderly patients with acute ischemic stroke have historically been challenging for neurologists and other stroke physicians to treat. Physicians have typically feared a higher incidence of symptomatic ICH in this group of patients. They are often excluded from trials on tPA and ischemic stroke for this reason. Thus, little data exists on the safety and efficacy of treating elderly patients with tPA for acute ischemic stroke.  This is an important topic to study, since 30% of strokes occur in patients over the age of 80 (Mishra, 2010).

In all the ECASS studies, the age restriction was set at 80. In the NINDS trial, only 44 patients older than 80 were randomized.  There had been an initial age limit of 80 years or older, but this was removed, so that some patients 80 and older were ultimately included. Their outcomes at three months were not significantly improved compared to those who did not receive tPA (NINDS, 1995). In a later subgroup analysis of this patient population, it was found that 25 of the 44 patients older than 80 had been given tPA. This group of patients were 2.87 times more likely than their younger counterparts in the study to experience a symptomatic ICH (Longstreth, 2009).

However, other studies have demonstrated more positive results of treating elderly patients with tPA. In fact, a meta-analysis of the SITS-ISTR and VISTA data (n = 29,228) revealed that increasing age is associated with a poorer outcome in general in acute ischemic stroke, but that this association was found regardless of whether or not patients were treated with tPA (Mishra, 2010). This study compared outcomes at 90 days in patients who received tPA and controls. Specifically, they examined the association of thrombolysis treatment with outcome between various age groups, with 3,439 patients aged over 80. The number needed to treat for a favorable outcome (score of 0-2 on a modified Rankin scale) was 8.2 patients. They stated that poorer outcomes are more likely to occur in the elderly due to other comorbidities rather than an increase in symptomatic ICH (Mishra, 2010). Furthermore, the tPA stroke survey experience, published in Stroke, concluded that there was no evidence to withhold tPA in patients greater than 80, as long as they were appropriately selected (Tanne, 2000). A third study, looking at stroke patients in three German stroke centers, found similar results in 228 patients, 38 of whom were 80 years or older. This study found a higher mortality in older patients (21.1% versus 5.3% at 90 days), but no difference in the rate of ICH between younger and older patients, with the authors also concluding that there is no evidence to exclude ischemic stroke patients from thrombolysis based on a predefined age threshold (Berrouschot, 2005).

At EM Lyceum we love debate, and know this is an area of particular controversy for EPs. Although our aim this month is not to rehash the controversies, we hope to add some more data to your thinking about this topic.  Even amongst our group of writers and editors we differ greatly in how we approach these questions.  We would love to hear your thoughts.

Thanks to Dr. William Knight of the University of Cincinnati for his expert thoughts on this topic.



This entry was posted in Uncategorized and tagged , , , , , . Bookmark the permalink.

13 Responses to tPA in ischemic stroke, “answers”

  1. This topic probably had more debate around it than any other topic we have done. There was a lot of internal debate within the EM Lyceum team on how these answers should be worded, particularly number 2 regarding the use of tPA in the 3-4.5 hour window. Ultimately, we went with the current form of the answers and decided to let the debate play out on line. Here are my thoughts regarding question #2:

    I definitely come from the Hoffman/Newman school of thought. I think tPA is a good drug in AMI and there may be a group of stroke patients that it benefits but I don’t know who that group is. I don’t think the studies shed a ton of light on this either. Instead of rehashing the whole argument, I would refer you to to check out Newman’s two podcasts on the topic. It’s a long listen but completely worth it. I also highly recommend reading the cited articles after listening to it if you really want to grasp the topic. The gist is that there are far more trials showing harm or no benefit for tPA in ischemic stroke than there are trials that show benefit.

    MAST-I – no benefit
    ECASS 1 – no benefit
    NINDS-1 – no benefit
    NINDS-2 – NNT = 8
    MAST-Europe – harmful, stopped early
    ASK – harmful, stopped early
    ECASS-II – no benefit
    ATLANTIS-B – harmful, stopped early
    ATLANTIS-A – harmful, stopped early
    DIAS-2 – no benefit
    ECASS-III – NNT = 15
    IST3 – no benefit

    Cochrane Meta-Analysis (2009) – 6% favorable outcome, 5% increase in ICH, 1% increased mortality. No difference across time window

    Cochrane Meta-Analysis (2013) – No evidence that one thrombolytic agent is better than another thrombolytic agent.

    So overall, 10 studies and a Cochrane review say there is no benefit or harm and 2 studies show a benefit. That doesn’t seem like definitive benefit to me for any time window. NINDS was plugged as “the one that mattered” because of the time window and the exclusion/inclusion criteria and the agent used (alteplase). However, the Cochrane reviews debunk the myth that “time is brain” as they found no difference in time windows. In 2013, the group showed that the agent makes no difference either. I think we already new this from the GUSTO trial. In that trial, they found that mortality was slightly lower (1%) with tPA in AMI but there was a higher ICH rate.

    What does this all mean for the question asked? Do you treat patients with tPA up to 4.5 hours after onset of symptoms? Well clearly, my answer is no. It’s not just the time that really matters to me, though. The studies we have don’t tell me who should get the drug and who shouldn’t. IST-3 actually showed a small benefit < 3 hours, harm in the 3-4.5 hour group and small benefit in 4.5 – 6. What?? How does that make any sense if “time is brain?” And that study was well done (just had asinine conclusions based on their data). The new ACEP policy says we should offer the drug to patients with ischemic CVA up to 3 hours and consider in for patients in the 3-4.5 hour window. I do this. I tell people the up and downside of the therapy in the 3-hour window and how there is a chance it’ll help, and there’s a chance you’ll bleed around the brain. Then, I let them decide (okay, my talk is longer than that but you get the idea). In the 3 – 4.5 window, I consider it but I advise against in almost all cases.

    Hopefully, this will stir more debate and discussion. It’s really what EM Lyceum is all about.

    • JH says:

      Dr. Swaminathan,
      I am an ER doc in Maine. Could you specifically let me know what specific Cochran review you are referring to above. Is the 2009 one: Wardlaw JM, Murray V, Berge E, et al. Thrombolysis for acute ischaemic stroke (review). Cochrane Collaboration 2009. Could you let me know the reference for 2013 as well. My medical director recently sent an email stating that ACEP may be rethinking its policy statement and we are being asked to submit our opinions. Are you aware of this? Thanks for all the hard work on EM Lyceum!!

  2. Pingback: Acute Ischemic Stroke -

  3. Bf says:

    Not much talk about nicardipene in Q1 that is my usual go to. Any reason why not in stroke?

  4. Anand Swaminathan says:

    I’m a big nicardipine fan as well. We don’t talk much about nicardipine here because the question is specifically looking at blood pressure control when you are going to give tPA. In NINDS, if aggressive BP lowering treatment was needed to get below 185/110, the patient was excluded from getting tPA. Most clinicians feel that starting a drip of a blood pressure lowering medication would qualify as aggressive treatment. I do, however, use nicardipine in hypertensive encephalopathy or in CVA where I’m not going to think about tPA.
    Also, check out the ERCast podcast this month where Rob and I talk about this in a bit more depth:

  5. Anand Swaminathan says:

    For those of you following the answers here regarding blood pressure control, please follow this link to ERCast where Scott Weingart, Pik Mukherji and I continue the debate

  6. Pingback: ED Briefs: Open Season on ACEP tPA Guidelines | BriPay News Aggregator

  7. Pingback: ED Briefs: Open Season on ACEP tPA Guidelines | Medical Device Articles

  8. Pingback: ER Briefs: Open Season on ACEP tPA Guidelines | Medical Device Articles

  9. Pingback: The Debate on tPA for Ischemic Stroke at EMCrit – What Does the Research Really Say? | Rogue Medic

  10. Pingback: WDYS: Talking about tPA – Expert and Community Commentary - Ed in the ED

  11. Pingback: The LITFL Review 123 - LITFL

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s