Subscribe to eTOC


I agree with Drs. Anthony Furlan and Gary Roubin, who commented at a session at the American Stroke Association Annual Meeting, that neurologists who treat stroke could benefit from the strategies used by cardiologists to treat and research myocardial infarction (“Approach Stroke the Way Cardiologists Treat Acute Myocardial Infarction,” Neurology Today, March 2002, page 8).

Seven years ago, I coauthored an editorial on the (slow) pace of acute stroke treatment research with Dr. Furlan and Drs. Greg DelZoppo and Mark Fisher – and most of these issues have not yet been addressed (Stroke 1995; 26: 2216–2218).

There is only one approved treatment for stroke, tPA, and it was approved in 1996. Neuroprotection remains an elusive goal and the logical idea of “cocktail” combination therapy seems a distant dream.


The question is why? Are there not enough patients to prove new therapies? This is one part of the problem. There are at least two to three times as many myocardial infarctions (MI) as strokes each year in the US. Almost all of these MIs are driven by similar pathophysiology, whereas a large fraction of strokes are not infarctions and different ischemic subtypes may respond to different therapies.

Sorting out ischemic subtypes reliably in the ultra early phase has been almost impossible. Patients with myocardial infarction may be seen earlier due to pain, and often they do not have altered mental states, which makes getting informed consent easier. Still considering this, the number of acute stroke patients in treatment trials is easily less than 10 percent of those in recent cardiology trials; so there must be more to it than simple numbers.


How the patients are studied is part of the problem. The relatively healthy, acute ischemic stroke patient targeted for treatment trials is a rare asset, and when they are spread among small competing trials, sponsored typically by pharmaceutical companies, the chance we have to show benefit is reduced.

The appealing notion of factorial design, rarely used in stroke, will require considerably larger numbers than the typical single drug stroke trial, which enrolls 500 to 700 patients, and greater organization and cooperation between governmental and private sponsors may be needed. Greater use of “implied consent” should be explored; there is a precedent for its use in patients with severe neurologic injury.

Are there too few researchers to mount large trials? This is clearly evident in comparison with cardiology. There are more cardiologists than neurologists and probably more in private practice participate in research. The practice of cardiology is more focused on the acute “event,” myocardial infarction, than the neurologists who treat epilepsy (arrhythmias) or degenerative dementing diseases (congestive heart failure). So cardiologists have greater numbers and they are more accustomed to dealing with the acute manifestations of vascular disease in their organ groups than we neurologists are in ours.


There is still no formal subspecialty for stroke, although progress in this area is being made. Most neurologists are clinic focused and evaluating acute stroke is at best inconvenient and may carry with it a financial penalty by reducing other income. Neurologists who are particularly committed to acute stroke evaluation and treatment are few and far between.

In addition, neurologists get paid relatively little for evaluation and even treatment with tPA for stroke, in comparison with compensation that cardiologists get for treatment of myocardial infarction. There is widespread perception that the CPT code for tPA treatment of stroke may be billed, but is not reimbursed by Medicare.


Dr. Robert J. Adams: “I believe that like cardiologists, we should engage in a continuous stream of trials leading to serial refinements in therapy, and use special testing to sort out subsets rather than limit entry.”

So, there are fewer of us willing to do this work, we are focused in other areas, and we don't get paid to do it in relation to time or the “opportunity cost,” for example, of a missed EMG.


While it is true that no one can examine the nervous system like a neurologist, our limited numbers and availability should be faced squarely. More partnerships with emergency room physicians and perhaps cardiologists should be sought to increase enrollment in stroke trials. We should also look at how to involve more private practice neurologists in stroke research by looking at how cardiologists carry out the large trials that have been performed in recent years.

I'm not sure trial design is the main problem. When there are 20,000 patients in a trial, small differences in outcome can be detected regardless of the endpoint, but when there are 500, large outcome disparities are required. There are methods to measure brain infarct volume, and compare between treatment groups, with computed tomography (CT) and certainly with magnetic resonance technology (MR), but these have not been used in primary evaluation of outcome. We should admit that clinical assessment is not very sensitive, and “brain saved,” even if we lack the clinical tools to immediately measure or appreciate the benefit of its rescue, is ultimately a good thing. Surrogate endpoints based on imaging and infarct size should be adopted.


I believe that like cardiologists, we should engage in a continuous stream of trials leading to serial refinements in therapy, and use special testing to sort out subsets rather than limit entry. If MRI special studies were routine in the minutes after the start of therapy, the window of study would be expanded without delaying treatment. It is unlikely that any treatment we use will immediately reverse the ischemic cascade but MR study in the first minutes after treatment initiation with later repeat studies seems both practical and logical. When we have to evaluate large numbers of candidates to enter one patient, motivation of all members of the team is adversely affected.

The comparisons to cardiology particularly in terms of acute treatment research are useful to focus attention on the pace of progress in our field. Perhaps we need an in-depth comparison of all the factors involved and an aggressive blueprint for change. The optimism of the early 1980s that we would soon be using multiple drug therapy to treat stroke is now a disappointing memory. Maybe real changes in our approach are needed.