Thrombolytic Treatment of Acute Ischaemic Stroke

The enormous success of treating coronary thrombosis with thrombolytic therapy makes the treatment of ischaemic stroke the obvious next step but the situation is much more complex.

• Only 80% of strokes are ischaemic and giving thrombolysis for a haemorrhagic stroke would be disastrous.
• Stroke tends to be more gradual in onset and may even start during sleep. Hence patients tend to present later.
• The window of opportunity for effective thrombolysis is 3 hours from the onset of the stroke and in that time a firm diagnosis of ischaemic stroke must be made.
• As a result, only about 4% of patients fulfil the criteria.

• The incidence of stroke doubles for every decade after 45 years.
• It rises from 104 per 100,000 per year for aged 45 to 54 years to 1113 per 100,000 per year between age 75 and 84 years.
• 70% of strokes occur in the over 65s.
• There is no age limit to the use of thrombolytic therapy although there are few RCTs of treatment in the over 80s.

The following suggest haemorrhagic stroke or other reasons to avoid thrombolysis:

• Seizure at onset of stroke
• Symptoms suggestive of subarachnoid haemorrhage
• Stroke or serious head injury in last 3 months
• Major surgery or serious trauma within 2 weeks
• Previous intracranial haemorrhage
• Intracranial neoplasm
• Arteriovenous malformation or aneurysm
• GI or urinary tract haemorrhage in last 3 weeks
• Lumbar puncture in last week
• Current anticoagulation (INR>1.7)

Physical examination:

• Rapidly improving neurological signs. There is no benefit from thrombolysis in a transient ischaemic attack.
• Marked hypertension with systolic BP>185 mm Hg or diastolic BP>110 mm Hg or aggressive (continuous intravenous) treatment required to lower BP to this range
• Suspected acute pericarditis

Laboratory results

• Platelets <100 x 10⁹/L
• INR >1.7
• Glucose <50 mg/dL or >400 mg/dL
• Positive pregnancy test (very unusual in this group)

Blood should be sent for group and cross-match in case transfusion is required.

• Significant neurological deficit. This means more than minimal weakness, isolated ataxia, isolated sensory deficits, or isolated dysarthria.
• Time of onset <3 hours.
• No scan evidence of intracranial haemorrhage.

The important differential diagnoses are haemorrhagic stroke, including intracranial haemorrhage and subarachnoid haemorrhage, and transient ischaemic attack.

• It is essential to have a CT or MRI scan to differentiate the type of stroke before commencing treatment. MRI is as accurate as CT to detect acute haemorrhage in patients with focal stroke symptoms and is more accurate than CT for the detection of chronic intracerebral haemorrhage.¹
• Access to these investigations must be available 24 hours a day. Interpreting the scan requires experience but it is possible to train people to an acceptable level of competence in a short time.²
• Even if the window for thrombolysis has been missed, the Royal College of Physicians states that all patients should have CT or MRI within 24 hours.
• An ECG is useful in diagnosing pericarditis and possible causes of stroke including myocardial infarction and atrial fibrillation. An ECG is not required before starting tPA.

The interpretation of a CT or MRI scan to exclude haemorrhage is not easy and usually requires a senior radiologist or an experiecned geriatrician. It should not be delegated to a junior doctor in the small hours of the morning. The availablity of such expertise 24 hours a day, 7 days a week, adds further to the requirments of a stroke service.

There have been a number of RCTs published from North America, Europe and Australia examining the role of thrombolysis in stroke. Many have used intravenous tPA but some have used intra-arterial infusion, especially for vertebro-basilar strokes. Studies include:

• MAST-E Multicentre Acute Stroke Trial-Europe
• MAST-I Multicentre Acute Stroke Trial - Italy
• ASK Australian Streptokinase
• NINDS National Institute of Neurological Disorders and Stroke (USA)
• ECASS European Cooperative Acute Stroke Study
• ECASSS II European and Australian Cooperative Acute Stroke Study II
• ATLANTIS Alteplase Thrombolysis for Acute Noninterventional Treatment in Ischaemic Stroke
• STAT Stroke Treatment with Ancrod Trial

The results are very encouraging but only if treatment can be started within 3 hours of the onset of symptoms. Complications from treatment can mean that patients who receive treatment more than 3 hours after onset of stroke have more deaths than controls.³ The sooner tPA is given to patients, the greater the benefit.

Figures and Interpretation

• When tPA was given within 3 hours of onset of symptoms, the number needed to treat for 1 more patient to have a normal or near normal outcome was 8, and the number needed to treat for 1 more patient to have an improved outcome was 3. These NNT are very impressive.
• The single most important factor excluding patients from thrombolytic therapy is presentation after the 3 hours "golden window" had passed. Only 4% of the patients in the NINDS study met the criteria for tPA therapy.
• If a stroke did not wake up a sleeping person, tPA cannot be given.
• As cerebral infarction progresses it can undergo haemorrhagic transformation and this would account for the unfavourable outcome in those treated later.
• Cochrane is uncertain as to whether lower doses of thrombolytics are safer or if there is any advantage to intra-arterial over intravenous routes of administration.⁴ Intra-arterial treatment tends to be used for posterior stroke but it is still uncertain if it is superior.⁵ The important point is early recovery of flow and the vertebo-basilar system appears to take longer to recover.
• Meta-analysis does suggest that there may be some benefit from treatment up to 6 hours along with gains in quality of life.⁶ Between 3 and 6 hours the intra-arterial route may be preferable.⁷
• Only oral aspirin within 48 hours and tissue plasminogen activator within 3 hours of ischaemic stroke have been shown to have clinical benefit.⁸ Other "neuroprotectors" have been disappointing.

The Royal College of Physicians, along with other organisations, have produced evidence-based reviews of the management of stroke. They recommend:

• Everyone with a stroke should be admitted to a specialist stroke unit. Level of evidence A.
• Hospitals offering thrombolysis outside of a trial should have their staff specially trained. Evidence level B.
• Brain imaging should be undertaken within 24 hours. Evidence level B.
• Evidence from Phase IV trials in America suggest that if protocols for thrombolysis are not strictly adhered to the intervention can do more harm than good and that the benefits of intra-arterial therapy remain dubious.
• They are more enthusiastic about other measures including aspirin 300mg at onset unless there is to be thrombolysis when it is delayed 24 hours. Evidence level A.

In the NINDS trials, the rate of clinical deterioration from new intracranial haemorrhage, 24 to 36 hours after treatment, was 6.4% with tPA versus 0.6% with placebo. Mortality at three months was 17% in the t-PA group and 21% in the placebo group (P = 0.30).

A good outcome 24 hours after thrombolysis tends to predict a good outcome at 3 months.⁹

Neurological deficits

Three months after tPA therapy:

• approximately 30% of patients are neurologically normal or near normal
• 30% have mild to moderate neurological deficits
• 20% have moderate to severe neurological deficits
• 20% have died.

Functional disability

Three months after tPA therapy:

• approximately 50% of patients are completely or almost completely independent in activities of daily living
• 15% are moderately dependent on others
• 15% are completely dependent on others
• and 20% have died.

There is good evidence that thrombolysis is of benefit in acute stroke if haemorrhage can be excluded and treatment started within 3 hours of the onset of the stroke. Only a small number of people can meet those criteria and significant resources are required.
A stroke unit for intensive management of survivors of stroke represents far greater value for money.

Rapid Admission

Although the high technology of scanning and thrombolysis is delivered within a hospital system, there are still implications for primary care. The most obvious is that if there is reason to suspect a stroke the patient must be admitted to a suitable unit with the utmost speed. Simply being in hospital within 3 hours is not good enough. The scan has to have been taken, read and the infusion started within 3 hours of the onset of symptoms. Hospitals also need to look to their systems to reduce delay¹⁰ as has been done with acute myocardial infarction.
A recent study from Spain¹¹ showed a thrombolysis rate of 7.5% for acute ischaemic stroke. This is most commendable compared with more usual figures of 4 or 5%. 16% arrived in hospital on time but more than half of these were excluded. Of the exclusions, 75% were for "non-modifiable" reasons. They concluded that reducing the "door to needle" time was the most important aim in improving numbers treated. Their own figures ranged from 33 to 122 minutes with a mean of 51 minutes. This includes time for admission, history, examination, blood tests, scan and interpretation.
Even if the deadline has passed, patients should still be admitted to hospital as all should have a scan within 24 hours and the outcome for all is better in a stroke unit.
Not only do doctors and hospitals have to examine their systems to prevent delay in the process, but patients also need to be able to recognise what is happening and know that a rapid response is required.

Prophylactic Admission

Transient ischaemic attacks are discussed in their own article. They do not benefit from thrombolysis but once the symptoms have resolved, the RCP recommend that all should be started on aspirin at 300mg daily initially. All should be seen and investigated by a specialist service. Anyone who has had more than one TIA in a week needs to be admitted to hospital immediately.

Secondary Prevention

The RCP is also very enthusiastic about the role of primary care in the secondary prevention of stroke. This also has its own article and will be discussed no further here.