Why are stroke patients excluded from TPA therapy?

Methods.

The study was performed using a prospective registry of stroke patients, together with a chart review at Foothills Medical Center (FMC) from October 1996 to December 1999. Patients with ischemic stroke admitted to other Calgary hospitals were identified by hospital discharge ICD-9CM codes (433, 434, and 436) for the same period of time. Patients admitted to more than one medical institution during the same period of hospitalization were traced and therefore not double-counted. The Calgary Stroke Program has been described previously.6 Patients with suspected acute stroke within 6 hours of onset are triaged urgently by the Emergency Medical Services (EMS) to the university teaching hospital (FMC), which has a catchment area population of 1.4 million.6 Patients not transported by the EMS directive or presenting after 6 hours may be admitted to one of the three Calgary hospitals.

We first reviewed stroke admissions at FMC. Patients admitted via the emergency department at FMC were identified by prospective daily review of emergency department logbooks and entered into a computerized database. A standardized study form was filled out by the stroke nurse practitioner on all stroke patients presenting to the emergency department. Where this could not be performed, the information was extracted from the hospital charts. Inclusion criteria consisted of an admission diagnosis of acute ischemic stroke, TIA, intracerebral hemorrhage, or subarachnoid hemorrhage. Medical records from these patients were reviewed. Patients were excluded if they had a final discharge diagnosis other than stroke.

“Eligible” patients were defined as those presenting with ischemic stroke to hospital within 3 hours of symptom onset. The type of stroke was determined from a combination of the neurologic examination and neuroimaging tests (CT or MRI). A diagnosis of stroke was made if there were rapidly developing clinical symptoms or signs of focal or global loss of cerebral function, with symptoms that lasted >24 hours or that led to death, with no apparent cause other than that of vascular origin.7 Focal neurologic deficits with causes other than stroke were excluded by appropriate investigations. Either CT or MR brain imaging differentiated intracerebral or subarachnoid hemorrhage from cerebral infarction. In cases where subarachnoid hemorrhage was suspected but could not be excluded by CT, a lumbar puncture was performed. All patients with stroke were reviewed by either a specialist stroke neurologist, a general neurologist, or a neurosurgeon. Patients with ischemic stroke were allocated to one of four groups according to the Oxford Community Stroke Program (OCSP) classification.8 We hypothesized that stroke type and lateralization might be associated with time to presentation.5

Time at stroke onset was defined as the time at which the patient or a witness first noted a definite neurologic abnormality. If the symptoms were first noted on awakening or if unwitnessed, the time at onset was recorded as when the patient was last seen well. The time of stroke was categorized into six periods: <3 hours, 3–6 hours, 6–12 hours, 12–24 hours, >24 hours, and onset unknown. For those patients admitted with acute stroke within 3 hours of symptom onset, the reasons for exclusion from alteplase therapy were identified using the standardized data collection forms. Stroke severity (National Institutes of Health Stroke Scale score [NIHSS]) was retrospectively9 recorded from the chart, and discharge outcomes were identified from the discharge summary. When neuroimaging abnormalities were documented to exclude the patient from treatment, the CT scans were retrospectively reviewed by a neuroradiologist, independently of the patient details. For patients admitted after 3 hours of stroke, reasons for delay in seeking medical attention were similarly recorded.

Using the combined database, we estimated the public health impact of thrombolysis on stroke in Calgary. We have previously published our experience with alteplase.10 The analysis was performed for the best available efficacy data for alteplase in patients experiencing acute ischemic stroke arriving at hospital under 3 hours (16% absolute increase in independent survival).1,11⇓ Independent survival is defined as a modified Rankin score of 0 to 2. Symptomatic intracerebral hemorrhage constituted a serious adverse event.

The major reasons for exclusion and time delay are presented in frequency tables. The χ² test or Fisher exact test was used to assess the association between presentation time and stroke classification and between presentation time and stroke side.

Results.

A total of 2165 patients with acute stroke were admitted to FMC during the study period: 1168 (53.9%) ischemic strokes, 16.5% intracerebral hemorrhage, 9.6% subarachnoid hemorrhage, 5.7% subdural hemorrhage, and 13.9% TIA ( figure). The incidence of intracerebral and subarachnoid hemorrhage in this series is higher than in most community-based studies, probably because of bias toward admitting more patients with severe strokes to a neurosurgical referral center.12 Of the 1168 patients with ischemic stroke, the mean age was 69.7 ± 14.2 years and 51.9% were male. The frequency of stroke types was as follows: 20.1% total anterior circulation syndrome (TACS), 33.5% partial anterior circulation syndrome (PACS), 15.3% posterior occipital circulation syndrome (POCS), and 22.3% lacunar anterior circulation syndrome (LACS). OCSP stroke classification was either difficult to categorize or unavailable in 104 (8.7%) patients.

• Download figure
• Open in new tab
• Download powerpoint

Figure. Stroke types admitted to Foothills Medical Center (FMC) from October 1996 to December1999.

The percentages of patients with OCSP stroke types presenting under 3 hours were as follows: 41.4% TACS, 32.7% PACS, 22.2% POCS, and 23% LACS. There was a significantly higher proportion of patients with TACS (41%) presenting within 3 hours than at any other time period (p < 0.001). A higher proportion of patients with LACS or POCS (73%) sought medical attention after 3 hours (p < 0.05 and p < 0.01). There was no difference for PACS (31%; p = 0.34). We found no relationship between side of stroke and time of presentation (p = 0.35).

Twenty-seven percent (n = 314) of patients with acute ischemic stroke were admitted within 3 hours of symptom onset. A total of 84 (7.2%) patients received treatment with IV alteplase (2 at Calgary hospitals other than FMC). The median NIHSS of these treated cases was 15. Forty-six patients (54.8%) were independent (Rankin score of 0 to 2) at 3 months, and there were six symptomatic hemorrhages (7.1%). Assuming that 6 patients need to be treated with alteplase to produce 1 independent outcome, 14 individuals might have benefited from the therapy.11 A further 16 patients with ischemic stroke received intra-arterial thrombolysis according to a protocol for basilar artery thrombosis13 and severe middle cerebral artery stroke that would otherwise have been excluded from IV thrombolysis because the 3-hour time limit was exceeded or there were other contraindications such as recent surgery.14

The reasons for excluding eligible patients from alteplase therapy are reviewed in table 1. Thirty-two (32%) patients were excluded from treatment because they were judged to have made a significant improvement. Twenty-one patients were documented to have had major motor improvement, yet only one of these patients was discharged from hospital independent.

The major reasons for protocol exclusion included severe stroke with obtundation (n = 12), CT abnormalities (subacute stroke [n = 2], white matter disease [n = 3], early CT ischemic change [n = 6], large area of hypodensity >50% of middle cerebral artery territory [n = 2], an “old” stroke [n = 1], suspected tumor [n = 1]), recent surgery or myocardial infarction (n = 5), seizures with confirmed stroke (n = 2), hypertension (n = 3; highest readings were below 185/110 mm Hg), warfarin use with associated International Normalized Ratio (INR) >1.4 (n = 6), and aspirin prophylaxis (n = 1). Five of the patients with severe stroke underwent intra-arterial thrombolysis (all suspected basilar artery occlusion), together with a patient with recent anterior myocardial infarction and another who had recent coronary artery surgery. Of the 15 patients excluded because of neuroimaging abnormalities or early CT ischemic change, only 4 were considered to have contraindications to treatment when retrospectively reviewed. All three patients excluded because of hypertension had blood pressure readings below the threshold of 185/110 mm Hg. The Oxford Stoke Classification, NIH Stroke Severity scale, and outcome at hospital discharge of these patients are detailed in table 2. The major reasons for patients delaying medical attention were that they were uncertain of time at onset, ignored or waited to see if symptoms improved, and had organizational difficulties in getting from peripheral communities to hospital in time ( table 3).

We found that 638 patients with ischemic stroke were admitted to one of the three other Calgary hospitals during the study period, giving a total of 1806 patients with ischemic stroke admitted to Calgary hospitals. Therefore, 4.7% of the total received treatment with IV alteplase (5.5% received either IV or intra-arterial thrombolysis).

Discussion.

This study has identified some of the major reasons for excluding alteplase therapy. For the purpose of this discussion, we have divided some of the difficulties of appropriately identifying patients for interventional treatment into ones that can be conceivably put into place (organizational initiatives, education of medical staff and patients) and ones still under current investigation (neurovascular imaging, surrogate markers of ischemia).

The prehospital and hospital infrastructure needs to be streamlined to handle ischemic stroke with high priority.6 Delay occurred in the emergency department during triage, contacting of the stroke team, obtaining a CT scan, and deciding on initiating therapy. Education of all emergency department personnel of the importance of a rapid “door-to-needle” treatment time as well as, perhaps, a separate fast track for patients with suspected acute ischemic stroke are essential.

Many of the issues of patient eligibility are even more relevant to rural communities. Over 10% of patients might have been eligible for therapy had they not presented to their local hospital first (where alteplase was unavailable) or had helicopter ambulance been available to cover large distances. Allowing EMS to bypass peripheral hospitals, bringing patients to stroke centers, should increase the number of eligible patients.15 Prehospital stroke screening by EMS has been problematic,16 but a new screening assessment, the Los Angeles Prehospital Stroke Screen,17 has shown very high positive and negative predictive value. Where transport distances are too great, telemedicine may allow greater accessibility to treatment.18

Another area of focus is clinicians’ conservative interpretation of eligibility criteria for thrombolysis, provoked by the fear that thrombolytic therapy will induce severe or fatal hemorrhagic complications.19 Recent major surgery, bleeding diathesis, and significant comorbidity may pose unacceptable risks, but most published contraindications are probably only relative.1 Currently, the only CT exclusion to treatment is intracranial hemorrhage. Early CT ischemic changes, small-vessel disease, and leukomalacia from previous strokes are not contraindications. Severe stroke and stroke with obtundation should not be considered contraindications unless the CT scan shows large hypodense areas.20 The risk of intracerebral hemorrhage in hypertensive patients remains controversial. Aspirin prophylaxis does not contravene treatment. There is no evidence for withholding alteplase treatment in the elderly for appropriately selected patients.21

One of the major findings in this study was that 32% of patients considered to have mild or significantly improving neurologic symptoms were either dependent at discharge or dead during hospital admission. A recurring reason was the improvement of motor symptoms, but resolution was not as complete as the treating physician had hoped, leaving the patient either profoundly dysphasic or with disabling neglect and/or inattention. Many of the patients excluded from treatment with alteplase because the treating physician perceived there to be contraindications might have benefited from the treatment and still been treated according to national guidelines. Efforts need to be made to identify patients that are unlikely to improve or may deteriorate without intervention. Clinical assessment alone may not be sufficient.22 Neurovascular imaging with either multimodal MRI or transcranial Doppler may show either “tissue at risk” or persisting vascular occlusion. Patients with persistent large intracerebral artery occlusion, for instance, are at risk of neurologic deterioration.23

Although several investigators have cited the need to improve stroke recognition,24 the influence of stroke education on presentation is largely unknown. We found in this study that 29% of patients presenting after 3 hours had recognized their symptoms but had chosen not to seek medical attention. Further widespread and repetitive public education is needed to increase awareness of warning signs, stroke risk factors, and availability of treatment. The latter factor should not be underestimated as stroke has been widely thought to be untreatable.

Uncertain time of symptom onset is a major cause for exclusion from treatment (32.9%). Many of these patients could potentially benefit from therapy if surrogate markers of cerebral ischemia could reliably be identified. The rigid time window is not representative of the pathophysiology of human neuronal ischemia. Clinical experience and experimental evidence point to the wide interindividual variation in potentially salvageable brain tissue even out to 8 hours from symptom onset.25

Increasing the proportion of patients with acute ischemic stroke who are treated with alteplase is an important challenge facing contemporary physicians. By positively addressing some of the reasons for excluding therapy and identifying patients at risk of death or dependency, many more patients could be treated. The necessity of accepting occasional “iatrogenic” problems to achieve an overall benefit has been difficult to adopt by many clinicians. The benefits of alteplase therapy in this community may appear modest, but the potential public health impact if 5% of the 60,000 ischemic stroke admissions in Canada per year received the therapy would be the saving of 500 patients from death or dependency per year, assuming that the treatment could be administered as safely and effectively as in the National Institute of Neurological Disorders and Stroke trial. With improved accessibility and higher treatment rates, many more could be saved. We do not propose the indiscriminate use of alteplase therapy in stroke patients, as it is becoming clear that some subgroups do not benefit from this therapy.11,26,27⇓⇓ The success of alteplase treatment is dependent on appropriate administration of treatment to as many stroke patients as feasibly possible.