The current and future state of stroke treatment

Insight from UCSF's Management of the Hospitalized Patient conference.

Much research and discussion have focused on the right time window for giving thrombolysis to patients with acute ischemic stroke. Yet a day may come when the debate about a cutoff of 3 or 4.5 hours will seem rather quaint, according to S. Andrew Josephson, MD, medical director of inpatient neurology and chair of the neurohospitalist program at the University of California, San Francisco (UCSF).

Within a decade, thrombolysis decisions will perhaps be based on tissue, not time, said Dr. Josephson during a session at UCSF's Management of the Hospitalized Patient conference, held in San Francisco last October. Perfusion-weighted imaging has the ability to show whether a patient has so much infarcted tissue that IV tissue plasminogen activator (tPA) is inadvisable due to risk of hemorrhage or whether blood still could be restored to the tissue via tPA.

“We are no longer going to say that at 4.5 hours everyone can't get tPA. There have to be some people who can get tPA after that, and others you wouldn't want to give it to even at an earlier period,” Dr. Josephson said, adding that, at the moment, perfusion imaging still “isn't ready for prime time” since trials have not yet supported its use.

Another possible change in practice is that physicians will remove a clot if tPA fails, not just in those ineligible for IV thrombolysis. While a trial of this combination approach (published in the New England Journal of Medicine in March 2013) was stopped because no benefit was found, different results may emerge when the approach is tested with newer devices, Dr. Josephson said.

“We now have newer devices, which have been approved in the last 24 months, that increase the odds up to 90% of getting a clot out—whereas before, it was around 60%,” he said, adding that trials testing these new devices in a combination approach are ongoing and will soon be published. “For now, we should continue to use these devices if patients are ineligible for tPA, but not after using IV tPA, with the exception of basilar thrombosis, which was not tested thoroughly in these recent trials.”

Stroke treatment today

In addition to discussing the future, Dr. Josephson offered tips on best practices now, based on research that has been published in the last year or so.

For starters, every person with a stroke should have a non-contrast head CT scan, which is 100% sensitive for the presence of blood, he said.

“Even the best vascular neurologist, based on clinical presentation, labs, vitals, whatever, can't tell the difference between the 85% of people with an ischemic stroke and the approximately 15% with hemorrhagic stroke; a distinction that has obvious important therapeutic implications,” he said. “If there is no blood on the CT, it's an ischemic stroke.”

Since acute ischemic stroke treatment is still based on time windows, it's crucial to ask the right questions about symptom onset, he said. For example, if a man woke up at 6 a.m. unable to move his right arm and the physician later asked him what time his stroke started, the man likely would say 6 a.m., which may not be accurate.

“We want to ask, ‘When was the last time you were normal?’ Or, you can ask the family, ‘When was the last time this person was seen well?’ And if it was 9 p.m. when [the patient] went to bed, or 3 a.m. when he got up to use the bathroom, then that is when the clock begins so we do not misestimate these crucial time windows,” Dr. Josephson said.

A “game-changer” article, published in the Journal of the American Medical Association in 2013, illustrated just how important it is to deliver tPA as quickly as possible, he added. The study of approximately 1,400 hospitals found an association between faster administration of tPA, in 15-minute increments, and significantly lower rates of mortality and intracerebral hemorrhage, plus higher rates of discharge to home instead of nursing homes.

An August 2014 meta-analysis in The Lancet found that patients who receive tPA quickly, even within 90 minutes of stroke onset have “incredibly positive results—these are the people for whom the efficacy is greatest and the risk of hemorrhage is lowest,” Dr. Josephson noted. “So we are really looking for systems-based approaches at individual hospitals to speed up delivery.”

What's more, 4.5 hours was undoubtedly the point where the benefit started to wane in the study, he added. “I'm not coming back next year to say, ‘Now it's 6 hours' . . . 4.5 hours is really when we cross the null hypothesis” using time-based treatment.

Though tPA is not approved by the FDA for use beyond 3 hours, “compelling” randomized, controlled trials support its use up to 4.5 hours, and it's approved for this time frame in other countries, Dr. Josephson said, adding that most stroke centers in the U.S. have adopted this time frame as well.

If treatment is initiated between 4.5 and 6 hours, hospitalists can use an angiogram to find the clot, then use a microcatheter to deliver intra-arterial tPA directly into it. If it's been less than 8 hours since stroke onset, removing the clot via mechanical embolectomy is an option, Dr. Josephson said.

It's important to remember that a person who comes into the hospital on one of the oral direct thrombin or Xa inhibitors for anticoagulation is ineligible for tPA regardless of the time of stroke onset but likely can receive mechanical embolectomy, he noted.

Of interest, although intra-arterial tPA has proven to be effective in multiple trials, it is not approved by the FDA because the trials used a different lytic medication than is used now, Dr. Josephson noted. Meanwhile, many devices to perform mechanical embolectomy—”which is really the rage nowadays”—have been FDA-approved, despite having a lower level of evidence to support their use.

“It once again illustrates how much easier it is to get a device approved through the FDA compared to a drug,” he noted.

Beyond 8 hours

If the last known time a patient was seen well is beyond 8 hours, available acute treatment options narrow, Dr. Josephson said.

A common scenario in this case is the patient in his 60s who shows up at the hospital with 3 days of right arm weakness, a mild right facial droop, a mild right pronator drift, and slowed movements of the right hand. He takes aspirin and a blood pressure medication. Your first order of business would be to order a CT or MRI with diffusion-weighted imaging, Dr. Josephson said.

I don't use a lot of MRI for acute stroke because it's time-consuming and slows down administration of tPA. But an MRI [shows evidence of infarct] from seconds after a stroke and lasts for about 7 days,” he said. “So not only can it tell you it is a stroke, but it can be helpful in figuring out if it's an acute stroke or a chronic stroke.”

After MRI, the standard workup to identify the cause of the stroke includes an echocardiogram, a lipid panel, carotid evaluation, and extended cardiac telemetry, Dr. Josephson said. What you shouldn't order are the so-called “stroke labs”: vitamin B12, thyroid-stimulating hormone (TSH), rapid plasma reagin (RPR), and erythrocyte sedimentation rate (ESR) which are very low yield in most patients.

“I am not really sure where these ‘stroke labs' began, but I can't get our residents to stop ordering them,” Dr. Josephson said. “Please call me or e-mail me if you ever find a stroke related to hypothyroidism.”

As for which echocardiogram to order, research published in Stroke in 2006 found that transesophageal echocardiogram (TEE) was better than transthoracic echocardiogram (TTE) at spotting a cardiac source for emboli, as well as major risk factors that would lead to anticoagulation such as a left atrial thrombus.

At UCSF TEE is used for people under age 55 without an obvious cause of stroke, since cardioembolic sources are high in this age group. For those age 55 and older, TTE is used first followed by a TEE if necessary.

“I freely admit we may be missing a number of sources of stroke, and probably the best evidence-based way is to do TEE on everyone, but that's not so simple and is resource-intense,” Dr. Josephson said. “The bottom line is, you've got to think about TEE in someone who doesn't have another obvious cause of infarction.”

Cardiac telemetry should involve 21 to 30 days of wearing a Holter monitor to check for atrial fibrillation. “I understand the folks who say that's a lot of cardiac telemetry, but this was a big headline in 2014, how to look for afib in patients who have had a stroke,” he said.

He noted that in 2 New England Journal of Medicine studies published in 2014, an extended Holter monitoring showed atrial fibrillation in 15% to 20% of patients who had a stroke with an unknown cause (i.e., the carotid arteries looked OK without a known history of atrial fibrillation).


Secondary prevention options comprise anticoagulation and antiplatelets.

Many trials have shown that antiplatelets are at least as good as anticoagulants for secondary stroke prevention in most situations, Dr. Josephson noted. “So when do we use anticoagulants? First and foremost, in afib. And if that's all you remember, that's good, because 95% of our indications for anticoagulants following stroke will be in those with afib,” he said.

While neurologists are excited about direct thrombin and Xa inhibitors in general for primary prevention, there aren't yet great data for using these newer drugs in the setting of acute stroke, Dr. Josephson said.

“If someone has nonvalvular afib with a stroke, I generally start that patient on warfarin without a bridge, then consider switching over to one of these [newer] agents in the coming weeks,” he said. “These [newer] agents are nonreversible, so if someone gets into hemorrhagic complications you're in trouble.”

Anticoagulant therapy may also be indicated if a patient has the antiphospholipid antibody syndrome or an echocardiogram that shows a thrombus in the heart, Dr. Josephson said. Beyond that, the indications for anticoagulants become murkier.

Trials in the last few years have failed to demonstrate that anticoagulants are any better than antiplatelets for patients with cardiomyopathy and an ejection fraction less than 30%, as well as for patients with a patent foramen ovale, he said.

Anticoagulants may have a role in patients with vertebral artery dissection, he said. “[Vertebral dissection] is actually quite common as a source of stroke in young people. It's often accompanied by recent neck trauma including a recent spell of vomiting or severe coughing, or having been in a motor vehicle accident and gotten whiplash,” he said.

While there aren't good data on treating vertebral dissection patients, the general rule of thumb is to treat them for a short period of time—for example, 3 months—with anticoagulants, and then switch them to antiplatelets, he said.

Do not use a heparin bridge to anticoagulation in acute stroke, Dr. Josephson added, not even in patients with a high risk of recurrent ischemia and a low risk of hemorrhage. A 2013 study in The Lancet Neurology found no evidence to support it. “We should not use heparin in acute stroke except in very, very rare situations,” he said.

Antiplatelets and other treatment

The 3 options for antiplatelets in secondary stroke prevention are aspirin, clopidogrel (Plavix), and aspirin/dipyridamole (Aggrenox). Multiple trials have shown that the latter 2 are slightly better than aspirin and are basically identical to each other in terms of effectiveness, Dr. Josephson said.

“If someone has a stroke and isn't already on an antiplatelet med, I generally start them on clopidogrel or [aspirin/dipyridamole]. I use a lot more clopidogrel because it's once a day and generic, and because people don't get a headache on it, but either is perfectly acceptable,” he said.

If a patient cannot tolerate either of these medications—”including financially”—then aspirin is a fine alternative, he added. 81 mg is a perfectly acceptable dose and not inferior to 325 mg.

If a person is already on aspirin and has a stroke, switch them to clopidogrel or aspirin/dipyridamole. If he or she has a stroke while on clopidogrel or aspirin/dipyridamole, it's not clear what to do, as there are no data. “I often switch to the one they are not on, thinking they may be intolerant to the first. Do not add a second antiplatelet—there is no current proven benefit in the long term—and do not anticoagulate them,” he said.

Other things to do in acute stroke are to give a high-dose statin to anyone with an LDL cholesterol level higher than 100 mg/dL and to control glucose and fever tightly since both are deleterious in the setting of brain injury. It's also wise to give enoxaparin for deep venous thrombosis prophylaxis, except never within 24 hours of tPA, Dr. Josephson said.

One common mistake is to lower blood pressure unnecessarily in stroke patients, he added. National guidelines say to lower blood pressure in the setting of ischemic stroke only if the pressure is above 220 mm Hg systolic—a strategy called “permissive hypertension” except if tPA is given, where lower targets are recommended for the first 24 hours. “Remember, the higher the blood pressure [the more] blood flow is getting to areas of the brain that are starved and ischemic, but not yet infarcted,” he said.

When to stop permissive hypertension is controversial, and there are no good data on the subject, Dr. Josephson said. “At our place, we start a medication about 72 hours after the stroke or before discharge. We aim for normotension in the next month or so, using thiazides and [angiotensin-converting enzyme] inhibitors as our preferential drugs,” he said.

A word about TIAs

If a patient has a transient ischemic attack (TIA), perform exactly the same workup as you would for a stroke, and try to do it quickly, within 48 hours, Dr. Josephson said.

“Multiple trials show that if you take a person with TIA and quickly do everything—start them on an antiplatelet, perform an echo, look at the carotids, look for afib—you can reduce the risk of subsequent stroke by 75% to 80%,” he said.

The workup can be done either in the hospital or as an outpatient, if possible. “If you have the capability to get these done as an outpatient, great, but don't tell your patients to follow up with a primary care physician in a week or two to get all this done,” he said. “There is a high risk of recurrent events with TIA, and you have an opportunity to help these folks if you act quickly.”