Stroke in the 21st Century: A Tale of Two Studies

What follows is a review of two important studies pertaining to stroke, both published around the same time. Each had significant implications for stroke care, but only one has made an impression on the mainstream medical community. I’ll leave it to you to draw your conclusions about why this might be.

Study 1: The Kitava Stroke Study

In 1993, Stefan Lindeberg and colleagues published a study entitled “Apparent absence of stroke and ischaemic heart disease in a traditional Melanesian island: a clinical study in Kitava” in the Journal of Internal Medicine. Kitava is an island in Papua New Guinea, its inhabitants are the Kitavans. They are one of the last remaining populations of hunter-gatherers whose diet is largely untainted by the products of modern agriculture and technology. As such, they offer a rare opportunity to examine the health consequences of the modern lifestyle, including our post-agricultural diets.

Prior to the publication of the study, Lindeberg and his colleagues had performed a systematic examination of the Kitavan’s health. As part of this effort, they attempted to search and record cases of stroke.

They didn’t find any.

Not only did they not find any cases amongst the living residents, they didn’t find any inhabitants who could recall a deceased resident having experienced anything resembling a stroke – this in spite of their detailed knowledge of prior generations. In fact, the notion that a person could be suddenly stricken with an inability to speak or move was altogether foreign to them. Amongst the physicians on the Trobriand Islands and in other parts of New Guinea, Lindeberg’s findings weren’t surprising. They’d considered the absence of stroke in the aboriginal populations an established fact (and one that couldn’t be explained by a lack of elderly residents).

As remarkable as this sounds, these findings were not unprecedented. On the contrary, an absence of stroke amongst native populations in various parts of the world had been reported many times before. A 1944 study reviewing the records of 269 consecutive patients treated for neurological disease at a clinic in Kampala, Uganda found not a single case of stroke. Relatedly, a review of 3,000 autopsies performed in Uganda during the 1930s and 1940s showed not one ischemic stroke (and only four cerebral hemorrhages). As the Uganda natives transitioned to modern life and diet, however, things began to change for the worse. By 1968, stroke had gone from non-existent to the most common neurological diagnosis — a story familiar to anyone acquainted with the literature on diseases of civilization.

Lindeberg’s Kitava study, then, was not an isolated finding. It corroborated the experience of so many others in the medical field who had spent time caring for hunter-gatherer populations. But it was one of the most systematic and rigorous accountings of stroke in a native society, and had the benefit of being published in a widely circulated, peer-reviewed medical journal.

Surely, then, a finding this improbable — an entire civilization free of stroke — would ripple through the medical community. After all, the potential public health implications are enormous — it’s hard to even fathom a world without stroke.

But there were no ripples. I entered medical school in 1997, four years after the publication of Lindeberg’s study. And in the next eight years of my medical training, I heard nothing of the Kitavans. Moreover, there was no mention of the diseases of civilization literature, showing entire classes of disease like diabetes, heart disease, stroke, obesity, cancer that were non-existent in native populations. To this day, this omission is not one I fully understand.

What I did hear a LOT about, and what continues to impact my daily life as a neurologist, was another study that would be published two years later…

Study 2: The NINDS t-PA Trial

In 1995, the National Institute of Neurological Disorders and Stroke (NINDS) rt-PA study group published a study entitled “Tissue Plasminogen Activator for Ischemic Stroke”. It was phase III clinical trial of the drug Alteplase for use in acute ischemic stroke. Alteplase, more commonly known as t-PA, is a direct thrombin inhibitor, classified as a thrombolytic agent. It interferes with the blood clotting process, disrupting the formation of the thrombi that lead to strokes and heart attacks. 312 patients with an acute stroke were enrolled in the trial, half received a placebo, half received the active drug. The drug had to be given less than three hours from the onset of the stroke (prior trials had shown a significant increase in serious bleeding when thrombolytics were given after this time frame).

Those who received the drug did better than those who didn’t. This was a significant and important finding, as this was the first positive trial of a drug given to improve the outcome in those suffering from an acute stroke. Following publication of the study, there was a massive effort — spearheaded by the American Heart Association (AHA) — to raise public awareness about stroke in order to maximize t-PA utilization. Stroke was rebranded as a “brain attack” to help convey the same sense of urgency as “heart attack” (the name never stuck). Hospitals around the country developed “acute stroke protocols”, their primary purpose to ensure rapid delivery of t-PA to eligible patients. Committees were formed, faster CT scanners were purchased, ER procedures were streamlined. In other words, all available resources were marshaled to maximize the potential benefit of this new drug. Hospital competency in treating stroke is now judged largely on the single measure of how many eligible stroke patients receive this drug.

It should be noted that the NINDS t-PA trial itself and the AHA response were not without their critics. After all, this wasn’t the first trial of a thrombolytic agent. Fifteen trials of thrombolytic drugs had been conducted prior to this one, each failing to show benefit. Faithful adherence to the scientific method would normally mandate a result like this be replicated before its conclusions are accepted.

Once hospital ERs around the country began using t-PA, the chorus of dissension grew a bit louder. To the physicians administering the drug, it seemed like patients receiving it were bleeding into their brains more than anticipated. Real-world studies of actual t-PA usage in community hospitals confirmed these suspicions, revealing hemorrhage rates substantially higher than those reported in the NINDS trial – high enough to shift the benefit to risk ratio away from t-PA.

Review of the trial design also reveals a major flaw – though it was billed as a double blinded study, outcome raters were not blinded to the presence (or absence) of hemorrhage on the post-tPA CT scans (the presence of bleeding on the scan indicating the patient most likely received active drug), an oversight that could easily tip the scales in favor of t-PA. By 2003, nearly ten years after publication of the NINDS trial, the Cochrane collaboration (an independent nonprofit group that reviews randomized controlled trials) was still unwilling to give t-PA an endorsement for widespread use.

Other critics cited concerns over potential conflicts of interest, alarmed by what appeared to be a disturbingly cozy relationship between Genentech, the manufacturer of t-PA, and the AHA – the former donating $2.5 million dollars towards the construction of the association’s Dallas headquarters. The relationship between Genentech and the AHA remains a close one to this day.

In spite of these protestations, t-PA has remained the “standard of care” treatment for acute stroke since the publication of the NINDS trial. And thanks to the massive PR campaign, many in the general public are now aware that there’s a “clot-busting” drug for stroke. The common perception of t-PA – a perception still perpetuated by many in the medical community – is that if you’re lucky enough to get the drug in time it’ll bust up the stroke-causing clot and reverse the effects of the stroke. The actual data from the NINDS trial, however, paints a more sobering picture.

The NINDS t-PA Data: Separating Fact from Hype

The NINDS t-PA trial consisted of two parts. Part one was designed to determine if patients who received t-PA had early improvement. 147 patients were randomized to the placebo arm, 144 to t-PA. An NIH stroke scale was performed at the time of drug administration and at 24 hours. In the final analysis of part one, there was no statistically significant difference in the two groups. In other words, contrary to the prevailing perception, t-PA doesn’t immediately reverse the effects of stroke. A stroke victim given a placebo infusion was just as likely to rapidly improve as one given t-PA (in clinical practice these days such dramatic recoveries are invariably attributed to t-PA when it’s been given, divine intervention when it hasn’t).

Part two was designed to investigate differences between the two groups at three months. In this part, 168 received t-PA and 165 placebo. Subjects were administered a variety of different functional outcome scales at day 0 and at 3 months. At the 3 month mark, differences between the two groups did emerge. Overall, there were 12% more patients with a favorable outcome in the t-PA group than in the placebo group.

So no immediate improvement and only 12% at 90 days. Sounds a bit underwhelming, right? The truth is, given what we know about brain physiology, it’s somewhat surprising that t-PA has any effect at all.

Stroke Physiology: The Inconvenient Truth

An ischemic stroke occurs when the flow of blood through a blood vessel is suddenly impeded by the formation of a clot. Blood can’t get through the vessel, and the brain tissue that vessel supplies loses its oxygen supply. A brain cell completely deprived of oxygen dies within five minutes.

That’s five minutes. As in 300 seconds. As in nowhere near enough time to recognize you’re having a stroke, call an ambulance, get to the hospital, be seen by a physician, get necessary testing, retrieve the drug from the pharmacy, start an IV, etc…

So right off the bat we realize that with stroke, trying to intervene after the fact leaves us little opportunity to modify the outcome.

Now, all hope is not lost for the would-be-stroke-heroes, thanks to a phenomenon known as the ischemic penumbra. The penumbra is a region of brain tissue surrounding the central core of the stroke whose oxygen delivery has only been partially compromised, thanks to collateral blood supply from non-occluded arteries. The size of this region may differ considerably from one stroke case to another, depending on individual anatomical variation and the location of the clot. In some cases it may not exist at all. Nonetheless, the penumbra affords a potential opportunity for meaningful intervention beyond the 5-minute mark. Saving this “tissue at risk” is the mechanism by which t-PA exerts its beneficial effects when it does help.

The Impact of t-PA

Now let’s consider the practical impact of t-PA. Presently, roughly 4% of patients who have a stroke will receive t-PA – this after over a decade long massive effort (discussed above) to improve t-PA utilization. The reason for this is that there’s a fairly low ceiling – for a number of reasons, only a fraction of patients with stroke will ever be eligible for treatment with t-PA.

Let’s do the math, then – if we have 4% of stroke victims receiving t-PA, and if 12% of those will receive at least some kind of benefit, then that means that with the current state-of-the-art technology for acute stroke we can expect to help 0.48% of all people with stroke achieve at least some improvement with a pharmacologic intervention. In other words, we can expect to help slightly less than 1 out of every 200 stroke victims.

Now, this is a fantastic thing if you’re that one out of 200.

But what of the other 199?

The Path to a Stroke Free Future

Two studies.

One a study of a drug with the potential to help 1 out of 200 stroke victims at least a little bit – a study whose publication has transformed the delivery of emergency medical care in the developed world.

The other points a way towards a future free of stroke, yet whose publication went virtually unnoticed.

Based on what we know of stroke pathophysiology, it is naïve to think we can make a large scale impact on stroke with after-the-fact interventions.  Prevention isn’t glamorous, and nobody stands to make wad of cash from it. But, if we ever want to do more than just fiddle at the margins of stroke recovery, it’s the only means by which to do so.

Yet, this will never happen if we continue to ignore findings, like the Kitava study, that can lead us there. Nor will it happen if we continue to advocate prevention strategies (e.g. “low fat, low cholesterol diets”) that have no foundation in basic science and that have been summarily refuted by the scientific evidence.

2 thoughts on “Stroke in the 21st Century: A Tale of Two Studies

  1. My father had a stroke several years ago at age 72 and passed away not long after. He was in otherwise good health, and eating what his doctors told him. In retrospect, it probably wasn’t the best stuff. As you say, we’ve overlooked the prevention side of things for too long.

Leave a Reply to Harry Cancel reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>