Thrombolytic Therapy for Ischemic Stroke: Title and subTitle BreakFrom Clinical Trials to Clinical Practice

Intravenous tissue-type plasminogen activator (tPA) for ischemic stroke has broken a major therapeutic barrier, and much more progress is certain to follow. Steadily accumulating data continue to justify current guidelines for use of tPA in stroke,¹ and demonstrate that low complication rates can be achieved with proper use of this therapy. Renewed interest in hyperacute treatment of stroke may breathe new life into neuroprotective agents and acute interventional and surgical therapies, alone or with tPA.

Two studies^{2 - 3} in this issue of THE JOURNAL reflect 2 perspectives on the tPA experience. In the first, Albers and colleagues² present data from a prospective, multicenter study that provides further evidence that physicians with experience using tPA can replicate the favorable outcomes and similar complication rates reported in earlier studies.^{4 - 6} Of 389 enrolled patients, the 30-day mortality rate was 13% and the rate of intracerebral hemorrhage at 3 days was 10% (3% symptomatic, 7% asymptomatic). In this study, nearly all of the site principal investigators were neurologists who had previous experience treating stroke patients in a clinical trial. However, protocol violations, representing deviations from national treatment guidelines, occurred in 34.7% of patients and most commonly involved treatment beyond the 3-hour time window in 13% and use of other anticoagulants within 24 hours of tPA administration in 9%.

The second study, a brave report from Cleveland-area hospitals by Katzan and colleagues,³ conducted during the same time frame, documents the evolving community experience. Among 70 patients who received tPA, the in-hospital mortality rate was 15.7% (compared with 5.1% for patients with stroke who did not receive tPA) and the rate of intracerebral hemorrhage was 22% (15.7% symptomatic). The authors report that deviations from national treatment guidelines were identified in 50% of cases, including antithrombotic treatment in 37.1%, and treatment beyond the 3-hour time window in 12.9%. Although clinician experience with tPA may prove difficult to quantify by conventional statistics, the sobering findings of Katzan et al suggest community tPA use might not achieve the same outcomes as in the pivotal National Institute of Neurological Disorders and Stroke clinical trial.⁷

In the beginning phases of thrombolysis for stroke, there was hope that the acute patient, flattened by stroke, would begin to stir on receiving the drug, and would rise from the cot recovered, saying, "Thanks, I needed that." These hopes were soon dashed save in rare instances. Early studies, directed at angiographic proof of occlusion and treatment after recanalization, were disappointing considering the many hours consumed in patient recruitment and angiography, low rates either of occlusion or recanalization, and hemorrhagic complications.⁸ After pilot studies that evaluated earlier time frames for imaging,⁹ different scoring systems,¹⁰ and estimations of sample size¹¹ required for an anticipated modest treatment effect, intravenous tPA showed a benefit⁷ sufficient to obtain Food and Drug Administration approval and the field started to open up.

The goal of achieving favorable outcomes for patients with acute stroke is challenging and expanding the scope of health care systems nationwide. Never mind that only 1% to 2% of patients with stroke proved eligible for thrombolysis in the published trials. What was a trickle of patients could easily become a flood of potentially eligible patients, especially if public education efforts are successful in getting people to recognize stroke symptoms and seek medical care promptly after symptom onset. For clinicians on the front lines, with resources already taxed for treatment of hyperacute illnesses, the essential elements for tPA decisions are time, imaging data, and dose. With every tick of the clock up to the 3-hour limit, the chances for neurologic improvement fade and hemorrhagic complications increase.⁵ Following immediate stabilization, imaging takes precedence over time spent sorting out clinical details: should an acute cerebral infarction, hemorrhage, or other diagnosis be seen, current indications for tPA use are over. Readily available hospital staff must be prepared for rapid patient evaluation, be able to obtain and interpret the acute brain imaging study expeditiously, be knowledgeable about indications and contraindications for tPA, and be familiar with doses, to be able to respond to clinical and public expectations.

Current guidelines for acute stroke treatment continue to be probed in several directions, with intra-arterial administration leading the list. In a study by Lewandowski et al,¹² angiography proved possible in 35 patients within 3 hours, approaching what is surely the lower time limit of reasonable responsiveness. Combined intravenous and intra-arterial therapy achieved more recanalizations than did intra-arterial tPA alone in this small study. Using another thrombolytic, the Pro-Urokinase study by Furlan et al¹³ treated patients by intra-arterial injection at a mean of 5.3 hours from stroke onset, and found a beneficial outcome (using a slightly different scale) at the same 3-month time interval as did intravenous tPA. The later treatment time also was associated with higher rates of intracerebral hemorrhage, both in those treated with tPA or receiving placebo, but showed that for some patients, benefit can occur from treatment administered even beyond the 3-hour limit.¹⁴

Examining the details of the many thrombolytic clinical trials reveals numerous item-by-item differences, trial by trial, some major, some minor. The unblinking demands of the modern clinical trial require bemused clinicians to make predictions about important patient outcomes that might cause experienced gamblers to balk. And, once made, the direction of results from intent-to-treat designs may hinge on small points.¹⁵ For instance, in European Australasian Cooperative Acute Stroke Study (ECASS) I,¹⁶ the outcome assessment used changes in median scores, whereas the National Institute of Neurological Disorders and Stroke trial⁷ searched for those with minimal residual disability. Further, unanticipated differences between centers and the central reading center for brain image evidence of cerebral infarction prior to therapy led to the inclusion of enough patients outside the target population that overall results of the ECASS-I trial were negative. ECASS claims for success in the target population were not uncritically accepted.¹⁶ In ECASS-II,¹⁷ the clinical scale outcomes selected by the study organizers were not met even though a single-digit difference in scores, done prospectively, would have made the trial positive. For some, these outcomes are not sufficient; for instance, use of tPA for acute stroke is still pending approval in Europe.

The emphasis on late outcomes has directed attention away from the mechanisms of action of thrombolytics. The guiding principle of such therapy has been the expectation that prompt recanalization of an occlusion may, at the least, limit the volume of ischemia, and at best, prevent infarction. That assumed, efforts heretofore have been directed toward a safe dose and demonstration of benefits by some easily understood clinical outcome measures. With more experience has come an opportunity to make some inquiries into the fine details.¹⁸ For example, in an early report, some residual occlusions were found in 27 of 54 patients treated within 90 minutes with intravenous tPA.¹¹ More recent reports from angiographic data collected within 3 hours showed clot in 22 of 34 patients.¹² Of 474 patients in the Pro-Urokinase study who had angiograms obtained at 3 to 6 hours, 38% had the clinically suspected middle cerebral occlusion, 30% had no or incomplete middle cerebral occlusion, and 27% had a nonatherosclerotic arteriopathy.¹³ Single institution studies extending treatment out to 6 hours continue to expand the data base,^{19 - 20} and also keep alive views that embolism leads the list of causes of ischemia, even when the cause is not easily found.²¹

Interest in the occurrence or timing of recanalization reflects decades-old hypotheses that it could lead to reperfusion but with consequent hemorrhagic infarction.²² In some patients, this complication is clinically benign. In others, it extends the volume of injured brain tissue, nullifying or worsening the results had the occlusion remained in place.²³ Hemorrhagic infarction also has been documented with persistent occlusion, mediated by retrograde flow from neighboring patent arteries supplying border zones. Whether there exists a time line after which efforts at recanalization are better avoided remains unsettled.

One theme through all this literature has been the hesitancy to treat when vessel patency is demonstrated. It might be assumed that the outcome is better for these patients, but little information on their course is available. The simple assumption may not be justified that intravenous tPA has its effects through recanalization of larger, angiographically visible arteries, nor that, when no data exist (as in most of the intravenous tPA trials), the outcome was related to recanalization anyway. Noninvasive duplex and transcranial Doppler technology is expanding the information about large-vessel occlusion or patency documentable within the 3-hour treatment window.²⁴ However, further data are necessary to understand the relationships between early large-artery occlusion, recanalization, or persistence occlusion, and hemorrhagic changes after treatment or after placebo.

Further development of clinical scales also should help account for those clinical syndromes currently not easily quantified. For outcome predictions, the National Institutes of Health Stroke Scale score appears to have a dichotomy on a value of 7, above which is considered a poor outcome.²⁵ However, the numerical correlation is only modest for total lesion volume, with considerable variation for nondominant hemisphere lesions.²⁶ Modified scales may provide insights into instances of dramatic improvement found in some syndromes inferred to reflect prompt distal migration of the occlusion²⁷ and how improvement also occurs in some cases despite persistence of the infarct.^{28 - 29}

Fortunately, the nihilistic attitude toward treatment of acute ischemic stroke is a thing of the past, and much of the progress is directly attributable to innovative, high-quality clinical trials. However, the important studies by Albers et al and Katzan et al provide equally compelling lessons and much needed cautions regarding extending thrombolytic therapy for stroke to routine clinical practice. It is only with careful patient selection, individualized therapy according to the causative lesion, strict adherence to treatment guidelines, and evidence-based therapeutic decision making and drug administration by physicians with experience using these new agents and interventions that patients will derive optimal benefit from hyperacute treatment of stroke.