Why don’t more patients receive intravenous rt-PA for acute stroke?

Summary

Intravenous rt-PA was proven safe and effective for acute ischemic stroke in 1995, approved by FDA in 1996, and endorsed by the American Heart Association, American Academy of Neurology, and National Stroke Association in 1997. The treatment is remarkably cost effective, despite the high cost of the drug itself and the stroke teams to give it. Community-based practicing neurologists can use t-PA for acute stroke without the need for specialized expertise. The benefit is durable over long-term follow up and no particular subgroups, such as the elderly or those with very large strokes, should be excluded from treatment. Several additional studies have now confirmed the beneficial effects of thrombolytic therapy for stroke in de novo samples. So why isn’t the drug used more? Some troubling mis-understandings in the literature seem persistent and influential among clinicians. Considerable data supports the use of rt-PA for acute ischemic stroke, which should remove remaining doubts.

In 1995, the National Institutes of Neurologic Disorders and Stroke (NINDS) study of t-PA for acute stroke provided solid evidence that thrombolysis was safe and effective for treating acute ischemic stroke [1]. The therapy was approved by FDA in 1996, and endorsed by the American Heart Association, American Academy of Neurology, and National Stroke Association in 1997 [2, 3]. Almost immediately, some authorities opposed the approval, arguing that thrombolytic therapy with t-PA for stroke was not ready for general use [4]. More data was collected and published, and today all rational physicians agree that thrombolytic therapy is safe and effective for acute stroke patients. Use of intravenous rt-PA accelerates yearly, but hesitation remains, in part because of a few self-appointed ‘expert critics’ who remain steadfast in their opposition to this therapy [5–7]. Also, there are a few remaining myths about IV rt-PA, promulgated by physicians who doggedly refuse to acknowledge the considerable weight of evidence supporting thrombolytic therapy. For example, some of the reasons cited in opposition to using IV rt-PA include that the drug is too expensive; only highly trained Vascular Neurologists can give it; the drug must only be used after appropriate vascular imaging; it does not work in all subgroups; the drug has not been proven effective in more than the original study; and the drug is too risky. None of these myths withstand a rigorous review of the available data.

The available data suggest that t-PA for stroke therapy is highly cost effective, with rehabilitation savings far in excess of the cost of the drug itself and the stroke teams to give it [8] [9]. Intravenous rt-PA is not “expensive” and is one of the very few health care interventions that saves money for the entire healthcare system. Community-based practicing neurologists worry that they lack the expertise to use t-PA for acute stroke and obtain the same good results seen in the original research studies. It is false that the NINDS investigators were a select group of neurologists with special expertise, and that the study hospitals were unusual places where t-PA could be used safely. All the investigators in the original studies were trained to follow the study protocol properly, but it is important to note that a majority of patients were enrolled at community hospitals, not academic medical centers. Practicing clinicians can use the drug safely in non-academic medical centers. In fact, the recently published IST-3 trial confirmed that the magnitude of benefit due to thrombolysis is the same in multiple countries and multiple health care delivery systems, and regardless of experience [10]. The beneficial effects of IV rt-PA were confirmed in the original study up to one year later [11]. In addition the volume of infarction seen on CT scans 3 months after thrombolytic therapy was significantly reduced by IV rt-PA. median stroke volume was 25.5cm³ in placebo treated, vs. 15.5cm³ in t-PA treated, patients (p=0.039) [12]. Thus, there is considerable concordance of multiple analyses of the original data.

Some opposed the use of IV rt-PA without extensive vascular imaging first [4]. Subsequent data suggest that angiography is not necessary prior to administering t-PA, because the results do not alter therapy. In the original NINDS thrombolysis study, very few patients were treated “needlessly”, as evidenced by the observation that only 2% of placebo treated patients exhibited no neurologic deficit (NIHSS = 0) 24 hours after treatment [13]. These patients must have included some TIA and perhaps other non-stroke etiologies, but it is enlightening that so few patients were included in the trial who spontaneously recovered. No hemorrhages or other thrombolytic side effects occurred in these patients, suggesting that vascular imaging would not improve patient selection [13]; there is only a 2% chance of treating a non-stroke etiology with thrombolysis if the protocol is followed correctly. Further, evidence suggests that patients lacking any large vessel occlusion responded to therapy as well as or better than other subgroups [14]. Safety of IV rt-PA in this subgroup of “angio-negative” strokes is proven [15].

Multiple additional research groups have documented safety and benefit from IV rt-PA for acute stroke. For example, the European Cooperative Acute Stroke Study (ECASS-III) provided persuasive evidence that thrombolytic therapy is safe and effective for acute stroke patients [16]. The design of ECASS-III closely mirrored that of the original NINDS trial of rt-PA for Acute Stroke, but included only patients who presented between 3 and 4.5 hours after symptom onset. Patients were, on average, less severe, and less frequently diabetic, in ECASS-III so outcomes in the placebo group included more positive responders than in the NINDS trial. The Third International Stroke Study (IST-3) tested IV rt-PA in a larger number of European countries, focusing mainly on patients thought less likely to benefit [17]. The final sample size of 3035 patients makes IST-3 the largest randomized trial of iv rt-PA ever done. Overall, IST-III confirmed that IV rt-PA is useful when administered earlier and to appropriate patients [10]. But in addition, treatment benefited patients outside the traditional treatment protocol: Among 1007 (33%) patients treated after 4.5 hours, the benefit was confirmed but with a confidence interval overlapping 1. Aged patients were enrolled in IST-3: 1407 (46%) of the patients were aged 81 to 90 and 210 (7%) were over aged 90 years. A powerful and significant benefit of rt-PA was documented in very elderly patients. The IST-3 data also confirmed the low risk and considerable benefit of other so-called ‘off-label’ use, such as use in very mild patients. Significantly, the IST-3 results were similar across all sites and all levels of expertise, eliminating concern over the effect of inexperience on benefit and risk. A pooled meta-analysis of all available data demonstrates the clear and persuasive benefit of rt-PA across multiple trials and patient populations [18]. In fact, the pooled meta-analysis failed to identify any sub-group of patients from whom the drug could justifiably be withheld.

The first full year of drug approval was in 1997 since rt-PA received approval for use in stroke patients in June of 1996. The national body speaking for most ED physicians, the American College of Emergency Physicians (ACEP) issued a joint statement in 2012 with the American Academy of Neurology (AAN) endorsing the use of rt-PA for acute stroke [19]. During the years 1997 to 2012, newsletter writers among the ED physician community who were critical of IV rt-PA published opinions lacking full scientific foundation; some of these opinions found their way into reputable journals [20]. So between 1997 and 2012 the ED physicians in the US were unsure whether the drug should be used or not. We can estimate the impact of the published opinions on use of rt-PA. Although the incidence of stroke has varied over years, a reasonable estimate for the time period 1997 to 2012 is about 750,000 new strokes every year in the United States [21], of which 85% or about 635,000 were ischemic; we picked the 2006 estimate as it was the most commonly cited value in those years. Using updated statistics would perhaps over-represent the true rate of stroke during the 15 years following approval of rt-PA. About half of the US population lives within a reasonable distance of medical centers with 24/7 CT scan capability, i.e., a travel time short enough to allow thrombolytic therapy [22]. Therefore, about 317,500 stroke ischemic stroke patients each year are near enough to a stroke center to arrive in time to be considered for IV rt-PA. In well-organized stroke centers, about 25% of arriving patients get rt-PA [23], but it is also true that in some centers the treatment rate is not so high [24]. If we assume that adoption of IV thrombolytic therapy could have been greater than it was, it might be reasonable to estimate the use across centers of varying size and commitment. For example, from 2003 to 2011 medical centers dedicated to stroke treatment were able to almost double the rate of rt-PA use from about 4% to about 7% [25]. Absent objections from the ED leadership, perhaps the rate could have been as high as 10%, taken for illustration purposes. If so, then during these 15 years from drug approval until the ACEP statement, perhaps 31,700 patients per year, or 475,500 patients, could have been treated, assuming early adoption at stroke centers and rapid recognition by bystanders. In the NINDS trial, a complete cure (mRS 0,1) was noted in 39% of rt-PA treated patients, compared to 26% of placebo treated, for a net increase of 13%. Thus, perhaps around 61,815 Americans—perhaps a few less, perhaps quite a bit more—remain disabled today due to the confusion surrounding the efficacy and safety of rt-PA; the remainder would have been improved significantly [26]. The Centers for Disease Control estimates the life time cost of stroke to be $90,081, and thus the systematic under-utilization of IV rt-PA represents a cost to all of us of perhaps $5.6billion [27]. Of course, we can never know whether uptake of IV rt-PA could have been this high or not, but it is inescapable that a large number of patients did not get treated who could have; they remain disabled for life; and the cost to our society is staggering.

Why don’t more patients receive IV rt-PA for acute ischemic stroke? Perhaps ED physicians have given too much weight to the opinions of non-scientists; perhaps neurologists were too slow to support their ED colleagues with rapid, effective Code Stroke assessments and action. It should be noted that thrombolytic stroke therapy presented a novel situation to clinically active neurologists. For neurologists trained before 1997, the approach in neurology was to do no harm and allow the patient’s disease to run its course. A generation of neurologists needed to adjust to a therapy that has side effects but also has a net benefit: The first therapy for stroke was also the first therapy with difficult choices for the neurologist, the patients, and their families, in a time pressured decision making environment. In 20145, now that the promising initial data have been confirmed and supplemented more and more medical centers have initiated aggressive stroke treatment programs. Although increasing numbers of stroke patients receive thrombolytic therapy, a majority of eligible patients do not receive it. The dictum primum no nocere still applies: we must do no harm, either by actively committing an act or by withholding a proven therapy through inaction.