This site contains articles and commentary regarding stroke treatment and prevention. The articles can be accessed by clicking 'ARCHIVES' to your right, either ischemic, hemorrhagic, or both. The definitions of these terms is beneath that section.
In addition, it contains many links regarding stroke useful to both the layperson or stroke expert.
Two recent studies promote caution in the use of carotid artery stenting.
Summary:
(Information originally reported by Dan Hurley in the December 19, 2006 issue of Neurology Today)
Carotid arteries, the largest arteries in the neck, often develop atherosclerosis which can lead to ischemic strokes. A well-accepted surgical treatment to reduce this risk is called carotid endarterectomy (CEA). It essentially involves opening the artery and cutting out the atherosclerotic plaque. A newer technique, to treat the same disease, involves expanding the arterial wall with a small balloon and then placing a cylindrical wire mesh to hold it open - this is referred to as stent placement.
The EVA-3S trial was a multi-center trial performed in France which compared stenting versus CEA procedures, in 527 patients who were asymptomatic. Asymptomatic patients had a significant degree of atherosclerotic narrowing in the carotid artery, but hadn't yet had a stroke or stroke-like symptom. The trial was stopped early, since thirty days after the procedure, 3.9 percent of those randomized to CEA either had a stroke or died, whereas 9.6 percent of those stented had a stroke or died. The difference was statistically significant.
In a similar trial, acronymed SPACE, also performed in Europe, the 30 day rate of stroke or death was 6.34 percent among those receiving the CEA procedure and 6.84 percent among those randomized to the stenting procedure. The results were not statistically different.
Commentary:
Carotid artery stenting has been gaining momentum over the last 10 years. It has been assumed that with improvements in technology, this less invasive procedure would eventually supplant CEA surgery. The EVA-3S trial indicates that we may not be there yet. In this trail, there were more than twice the number of bad outcomes after thirty days, compared to CEA. Critics of the trail comment that the experience of those performing the procedures is important. In the EVA-3S trial, they note, the surgeons had far more experience with CEA than did the interventionalists with stenting.
This is a valid point and really is the take-home message of this trial. Just looking at surgery alone, we know that the more experienced the surgeon, the fewer complications (i.e. - strokes and death) are likely to occur. (Indeed, in terms of CEA surgery, I always recommend that patients seek out the surgeon in his/her area with the most experience with that surgery.) By comparing experienced surgeons to relatively inexperienced interventionalists, the trail has only served to emphasize the importance of physician knowledge and experience.
Observe that the similar trial, SPACE, had roughly equivalent outcomes after 30 days - therefore the higher complication rate in the EVA-3S trial for stenting was, again, likely an issue of inexperience. Critics have also pointed out that stenting in the SPACE trial was not performed with important, newer, technology. Technology, for example, that allows for the capture of debris downstream, while the procedure is ongoing. Such innovations, when they are finally tested in a trial, might be able to demonstrate superiority of stenting. Until then, a reasonable course of action, for someone who needs correction of carotid atherosclerosis, is to pursue an experienced surgeon first - considering stenting only if surgery is deemed too risky by the surgeon. If stenting must be done, the most experienced physician should be sought.
A recent study finds that sickle cell patients who have been undergoing transfusions to reduce their risk of stroke are at high risk of having a stroke if transfusions are stopped.
Summary:
(As reported by David Gillis in the February 7 issue of Neurology Today. Based on N Engl J Med 2005;353:2769-2778.)
Children with sickle cell disease are at high risk of suffering from strokes, mainly due to the development of narrowing of the largest blood vessels of the brain. It was previously found in the STOP trial (N Engl J Med 1998;339(1):5-11) that if patients with evidence of arterial narrowing are transfused once a month, their risk of stroke drops from about 10% per year to less than 1% per year. In the trial, arterial narrowing was identified using special ultrasound that allows penetration through the skull (transcranial doppler or TCD).
Dr. Robert Adams has now completed the STOP 2 study, which asked the question: Can transfusions in high risk patients be safely stopped at some point, and patients again monitored with TCD to see if they redevelop arterial narrowing? The study was discontinued early as it was found that 14 of the 41 patients in the transfusion-halted group had developed evidence of significant arterial narrowing. This is in contrast to none of the 38 patients who had continued transfusions. In addition, two of the patients in the transfusion-halted group suffered from a stroke - while none in the transfusion-continued group did.
Commentary:
The initial STOP study made it clear that patients with sickle cell disease need to be monitored with TCD for evidence of arterial narrowing. Those with evidence of arterial narrowing benefited greatly, in terms of stroke risk reduction, by undergoing transfusions. Unfortunately, transfusions are expensive and are not without their own long-term risks, and so the STOP 2 trial was an attempt to determine if transfuions could at some point be halted in these high risk patients. The results suggest that transfusions, once started in this high-risk group, cannot be safely halted.
A study of ischemic stroke in rats finds that umbilical cord blood markedly improves stroke recovery.
Summary:
(Presented by Dr. Willing at the 35th Annual Meeting of the Society for Neuroscience, November 2005)
In this study, researchers administered human umbilical cord blood cells to rats 48 hours after they were made to suffer from a large ischemic stroke. This corresponds to the typical time period of maximal inflammation and swelling that occurs after an ischemic stroke. After IV infusion of the cord blood, some animals were found to have little to no evidence of poor functioning - despite having undergone an ischemic insult which typically causes nearly half of the animal's brain to suffer a stroke. Researchers attributed the beneficial effect to a reduction in brain inflammation and apoptosis (programmed cell death) in the injured region.
Commentary:
While I don't normally cite studies involving animals, this study intrigued me both for the apparent strength of the effect of the treatment as well as the method by which this treatment was discovered.
By my reading of the prior similar studies (involving rats), researchers were attempting to find a way for stem cells to replace neurons in the brain which had been lost after an ischemic stroke. It's a great idea: the neurons in the brain have been destroyed, so let's inject cells which have the potential to become any type of cell (stem cells) and see if they can replace the neurons that were lost.
Researchers have, in fact, been able to do this, either through directly implanting stem cells into the injured portions of the brain or through injecting those cells into the fluid that surrounds the brain. The result was that animals had improved function and the stem cells did in fact survive in the brain. But undergoing a procedure of this kind would be very difficult for humans and we don't all have stem cells of our own that are available for injection!
An attempt was made, therefore, to inject human cord blood into the blood of the rats and hope they they would migrate to the injured portion of the brain. This did not work very well because the brain is surrounded by a tight barrier (the blood brain barrier) which prevents most material from crossing into it. Researchers then injected mannitol along with the cord blood. Mannitol has the effect of opening up the blood brain barrier. Using this combination of treatments, researchers discovered indeed that a significant benefit occurred; more neurons in the rats were surviving the ischemic insult. But was there evidence that the injected stem cells had become neurons to replace those that had died?
The answer was no. Something different seems to have occurred. While the injected cells did in fact migrate and survive within the injured region of the brain, they were not becoming neurons to replace those that had been lost or injured. Instead, the stem cells appeared to be producing cell-signaling proteins which have the effect of reducing the amount of inflammation in their vicinity and of 'telling' neurons to remain alive.
Why would a cell need to be 'told' to remain alive? This relates to the process of apoptosis. While many times, in the body, cells die due to their injuries; other times they die because they essentially commit suicide. This mechanism for dying is called apoptosis. The stem cells appear to be signaling neurons NOT to undergo this process.
So overall, with this study and others prior to it, while the intention of having stem cells become neurons was not realized, a very positive effect, and potential therapy, may emerge. I will be looking with great interest for the first human studies of this therapy.
An article reviews the many barriers to cholesterol lowering, both among patients and physicians.
Summary:
(Based upon 'Barriers to effective implementation of guideline recommendations', Leif Erhardt, M.D., The American Journal of Medicine [2005], Vol 118 (12A), 36-41 Supplement.)
In this review article, Dr. Erhardt takes a wide-ranging view of the treatment of risk factors that relate to coronary artery (heart) disease. He establishes that there is a considerable treatment gap. For example, in a study of 1,252 survivors of heart attacks, 46% of patients still had not achieved recommended goals regarding cholesterol control. Also, more than half (53%) had not achieved recommended blood pressure goals. Dr. Erhardt also examines surveys which point out mismatches between actual risk and patient's perceived risk. These indicated that even in patients with two or more risk factors for heart attacks, which places them in a high risk category, only about half the patients identified themselves as being in a high risk category. Physicians have also completed surveys (REACT survey) which reveal that more than half (59%) do not use national, international or recognized cholesterol guidelines, instead using their "own practice guidelines" to manage cholesterol. In this same survey, physicians were also asked what they felt were the most common barriers to implementation of guidelines regarding risk factor management. Results were as follows: Lack of (physician) time (38%), Prescribing costs (30%), Patient adherence (17%), Too many guidelines (10%), Poor awareness of guidelines (10%), Lack of motivation (8%), Bureaucratic limitations/restrictions (6%).
Commentary:
This article is very relevant to stroke patients, since the risk factors for coronary artery disease are much the same as the risk factors for ischemic stroke. The review highlights that many people are not aware of their high risk for heart disease (and presumably stroke too) and that many of them are not achieving recommended cholesterol and blood pressure goals. Most interesting is the physician survey, which finds that the majority of physicians do not even use official guidelines to manage their patients' cholesterol and blood pressure. Interestingly, while physicians cited lack of patient adherence and medication costs as barriers to patients achieving recommended prevention goals, a significant number of them also cited their own lack of time and confusion regarding which guideline to use, as major issues. Eight percent of physicians even cited a lack of physician motivation to achieving these goals!
All of this should be a wakeup call for many patients. Clearly you cannot always rely on your physician to implement recommended guidelines. While only your physician can take into account your special circumstances, and make appropriate decisions in this context, being aware yourself of what the guidelines recommend would be helpful. Guidelines regarding appropriate goals for control of risk factors for heart disease and stroke exist and are available to everyone online (AHA/ACC Guidelines and JNC7 Blood pressure guidelines or even in my own book). It would not be unreasonable to show up to your doctor's appointment with these guidelines in hand, in order to focus you both on the essentials of stroke prevention.
A study finds that mild elevations of glucose, that do not meet the criteria for diabetes, still confer increased stroke risk in those patients who have suffered from an ischemic stroke or TIA.
Summary:
(Published in Stroke. 2006;37:1413.)
Patients with impaired glucose tolerance have moderate elevations of blood glucose, but the elevations are not severe enough for them to be diagnosed with diabetes mellitus. Dr. Vermeer et al. studied 3,127 TIA or minor stroke patients who participated in the Dutch TIA Trial. Baseline non-fasting glucose levels were obtained in all patients. After a 2.6 year follow-up, 9% of the patients experienced a stroke and 6% a heart attack or cardiac death. Those whose blood sugars diagnosed them with diabetes had a nearly tripled risk of stroke, compared with those with normal blood sugars. Patients with moderate levels of glucose elevation (impaired glucose tolerance) had nearly double the risk of stroke as those with normal glucose levels. Interestingly, patients with lower than normal blood glucose levels had about a 50% increased risk of stroke compared to those with normal glucose levels. There was no increased risk of heart attack or cardiac death related to glucose levels.
Commentary:
Diabetes has been a well-established risk factor for ischemic stroke for many years, and it was surely suspected that lesser degrees of glucose elevation would confer some increased risk. This study demonstrates that the risk of impaired glucose tolerance is significant and deserves to be evaluated in patients who have suffered from an ischemic stroke or TIA. For my part, I will begin to routinely test my non-diabetic stroke patients for impaired glucose tolerance.
This situation is similar to what has occurred regarding glucose levels and peripheral neuropathy. Patients with diabetes are known to suffer from nerve damage, often causing numbness of the feet and sometimes associated pain. It was later noted that even lesser glucose elevations (impaired glucose tolerance) could cause this nerve damage. Now neurologists routinely screen for impaired glucose tolerance in those patients with otherwise unexplained neuropathy.
How should impaired glucose tolerance be treated? Some endocrinologists would recommend adding an oral hypoglycemic agent, typically used to treat diabetes, to keep a patient's blood sugars at a lower overall level. The risk, though, is that at some points in the day, blood sugar levels may go very low, which can be dangerous. Another approach is simply to have patients follow a diabetic diet, low in carbohydrates, to avoid peaks in blood sugar. Certainly, in the long-term, patients need to be advised to achieve their ideal weight, since impaired glucose tolerance might be reversible by this means.
A large study confirms that Aggrenox (the combination of aspirin with dipyrimadole) is superior to aspirin alone for ischemic stroke prevention.
Summary:
(From the Lancet, 2006 May 20;367[9523]:1638-9.)
The ESPRIT study group (European/Australasian Stroke Prevention in Reversible Ischaemia Trial) organized a comparison study of aspirin alone versus the combination of extended-release dipyrimadole with aspirin for ischemic stroke prevention and the prevention of other forms of arterial occlusion. Patients were entered into the trial within 6 months of having experienced a transient ischemic attack or small stroke of probable arterial origin and were followed for up to 5 years. A randomized trial design assigned patients to either aspirin alone (30-325mg daily; 1,376 people) or the same dosing of aspirin combined with long-acting dipyrimadole (200mg twice a day; 1,363 people). Results were that 16% of those taking aspirin alone had a primary event (death from a blood vessel occlusion cause, non-fatal stroke, non-fatal heart attack, or major bleeding event) whereas only 13% of those on the combination therapy had such an event. When the data from this trial was added to a meta-analysis of similar studies, the combination therapy was found to have an 18% reduction, compared to aspirin alone, for the primary outcomes listed above.
Commentary:
Platelets, small flat blood cells, can adhere to each other and contribute to blockage of an artery. Aspirin is the most commonly used anti-platelet drug and has been demonstrated to reduce stroke and heart attack risk. Dipyrimadole is a unique anti-platelet drug which, in this study, appears to work safely and effectively in combination with aspirin.
I have been an advocate of Aggrenox (aspirin plus extended release dipyrimadole) for the prevention of ischemic strokes, based on the results of a previous trial: ESPS-2, a European trial of similar design. Some, though, were critical of this trial since it essentially stood alone, there being multiple prior trials of the combination of aspirin and dipyrimadole that did not demonstrate superiority to aspirin alone. One important difference in the ESPS-2 trial, from prior trials, was the use of extended-release dipyrimadole, as opposed to short-acting dipyrimadole. Extended-release dipyrimadole allows for more consistent blood levels throughout the day. The ESPRIT trial is the second study to show that the combination of aspirin with extended-release dipyrimadole is superior to aspirin alone. Critics may be more willing to accept the results of this study given now that the original result has been confirmed.
Another criticism of the older ESPS-2 trial was that the dose of aspirin (50mg a day) was not adequate for a proper comparison. Many neurologists and cardiologists felt that aspirin dosing should at least be as high as 75mg, for maximal stroke and heart attack prevention. While the aspirin dose in the current trial was as low as 30mg in some patients, hopefully the fact that the median dose of aspirin for the overall trial was 75mg will put this criticism to rest.
A physician puts into context the data that implicates Vioxx and other COX-2 inhibitors as causing strokes and heart attacks.
Summary:
(As found in the May, 2006 issue [Volume 9;No 5] of CNS News)
In this article, Dr. Ivan Oransky briefly reviewed the data regarding use of Vioxx (rofecoxib) and stroke and heart attack risk. The first study which found a possible association between Vioxx use and heart attack risk came from the VIGOR trial (Vioxx Gastrointestinal Outcome Research, published 2000). In this study, those taking Vioxx had a 0.4% risk of stroke. Those taking another anti-inflammatory medication (naproxen) had only a 0.1% risk; which was found to be a statistically significant difference. It was unclear, though, whether this represented a protective effect of naproxen, or a detrimental effect of Vioxx. Later studies, involving use of Vioxx to prevent colonic polyps, were reviewed carefully with regard to increased risk of heart attacks and strokes. In one such study, a thrombotic event (generally an ischemic stroke or heart attack) occurred at a rate of 1.5 events per 100 patient-years in the Vioxx group and at a rate of 0.78 events per 100 patient-years in the placebo group. This represented essentially a doubling of the risk of an event in those taking Vioxx. The author puts this information into more meaningful terms as follows: if you were of similar risk as those who were enrolled in the study, and did not take Vioxx, your risk would be 16% of an ischemic stroke or heart attack over 20 years. If you took Vioxx over those 20 years, your risk of having an ischemic stroke or heart attack in that time would increase to 30%.
Interestingly, referring to a separate study, the author also notes that Celebrex (celecoxib, a product still on the market) had similar effects. There was a 1% risk of thrombotic events in those on placebo and a 4% risk in those taking Celebrex (400mg twice a day). The risk was less dramatic in those taking Celebrex at only 200mg twice a day (2.3%).
Finally, a study analyzing effects of placebo versus valdecoxib (Bextra) versus parecobix (a pro-drug of valdecoxib) was conducted for the treatment of pain in patients who had just undergone coronary bypass grafting. In this group, there was a 0.5% risk of a cardiovascular event in the placebo group, but a 2.0% risk in the groups given drug.
Commentary:
This was a great, short article, attempting to state clearly the risks of these medications. Clearly Bextra, Celebrex, and Vioxx can all increase the risk of an ischemic stroke or heart attack. Three variables, though, must be considered in anyone who may be considering taking these medications. The first is baseline risk. A person with known coronary heart disease, history of a stroke, atherosclerosis in other regions of the body, diabetes, high blood pressure, or elevated cholesterol will have a high baseline risk of a stroke or heart attack. In these folks, a doubling of the risk (on a COX-2) may result in a much higher risk of an event over time. On the other hand, a person without such a high risk (the athelete with low cholesterol in his or her 20's) may experience the doubling of his baseline insignificant risk: which would still likely be an insignificant risk.
The second variable to consider is time. Taking these medications for 20 years (for chronic arthitis, for example) will be much more of a risk than taking them for three days (to treat an acute pain).
A final important variable is the degree of one's pain and what other options are available. Patients who cannot tolerate other medications, and who have severe pain from rheumatoid arthritis, may be willing to assume the longterm risks of the COX-2 inhibitors, in order to improve their quality of life.
It is with reference to these points that I feel all of these medications should still be available to patients. (Vioxx and Bextra are off the market, Celebrex remains on the market). It should be up to patients, in consultation with their physicians, whether or not to use these medications, determining for themselves whether the benefits outweigh the risks. But the FDA, unfortunately, adopts a one-size fits all approach to this issue - forever erring on the side of caution regarding potential risks. Also, given the current medicolegal climate, drug companies probably do not feel they can keep such products on the market, no matter how informed the patients and physicians are of potential risks.
A large study finds that low cholesterol levels are associated with an increased risk of hemorrhagic stroke.
Summary:
Tobias Kurth presented data from the Women's Health Study at the International Stroke Conference (2/06) analyzing cholesterol levels and risk of hemorrhagic stroke. The Women's Health Study is a large study of female health professionals. Of the 39,876 participants, 28,345 women provided baseline blood samples. Dr. Tobias et al. analyzed these baseline cholesterol levels as well as the number of hemorrhagic strokes occurring in these women an average of 10 years later. A total of 65 women experienced a hemorrhagic stroke in this time. Various cholesterol measures were analyzed, including total cholesterol, LDL cholesterol, HDL cholesterol, triglycerides, and total cholesterol/HDL ratio. (For an explanation of cholesterol types, click here). Results were complex, but it was overall found that low LDL, triglycerides, and total cholesterol/HDL ratio were associated with a higher risk of hemorrhagic stroke. A high HDL cholesterol level was also associated with high hemorrhagic stroke risk, although the relationship was complex (taking on a 'U'-shape with both very high and very low levels being bad, intermediate levels being beneficial).
Commentary:
Nearly everything you read regarding stroke and heart attack risk these days indicates that low total cholesterol reduces your chances of having an ischemic stroke or heart attack. This study does nothing to challenge this notion, but it begins to explore the other side of the coin. Is there a lower limit of cholesterol, below which one shouldn't go? The ischemic stroke literature supports bringing LDL (so called 'bad' cholesterol) down below 100mg/dL in any patient who has suffered an ischemic stroke. The cardiac literature supports pushing patients with coronary heart disease below 70mg/dL. But no one knows just how low you can go to continue to achieve an overall benefit. Clearly the lowest levels of LDL and total cholesterol in this study were associated with more hemorrhagic strokes, but if the overall risk of an ischemic stroke remains higher than that of a hemorrhagic stroke, the benefits of low cholesterol might still outweigh the risk.
Perhaps the researchers involved in the Women's Health Study have the answers at their fingertips. If they simply review the number of deaths (both overall, and due to vascular disease) stratified according to cholesterol level, they would be able to determine if people with the lowest cholesterol levels had a higher or lower chance of death overall and death due to vascular causes. This would begin to give us insight into how seriously to take the risk of hemorrhagic strokes due to low cholesterol levels.
A retrospective database study finds that stroke patients who were brought to the hospital by ambulance were treated more urgently than those who arrived by other means.
Summary:
(As reported by online Stroke Journal Report of the American Heart Association, 2/16/06)
Dr. Yousef Mohammad et al. conducted a study which analyzed a large emergency services database, indentifying 630,000 patients who were treated for stroke in the emergency room. The mode of arrival of stroke patients was categorized into one of three: (1) by ambulance, (2) walk-ins (arrival by their own means, whether walking, driving, etc.) or (3) by a public service other than ambulance (police, etc.) or unknown. About half of the patients in the database arrived by ambulance, 43 percent were walk-ins, and 4 percent arrived by public service or by unknown means. It was found that patients who arrived by ambulance were seen on average within 30 minutes, compared to 34 minutes for walk-ins and 55 minutes for those who arrived by public transportation or by unknown means. Analysis of testing found that 73 percent of patients who came by ambulance underwent appropriate diagnostic imaging (MRI or CT scan), while only 63 percent of walk-ins received this imaging. Only 60% of those arriving by public or unknown means received this imaging.
Commentary:
It is generally recommended that anyone who feels he/she is suffering from a stroke call 911, or have someone else make the call, to activate emergency services. This is the safest approach since the ambulance personnel are skilled in administering potentially helpful therapies such as oxygen and IV fluids, and they may need to treat concomitant medical disorders. The emergency room is also typically notified of the impending arrival of a stroke patient, which helps speed the evaluatory process once a patient arrives.
This study looked at the mode of arrival of patients, concluding that arrival by ambulance led to a (slight) reduction in time seen by physician, compared to arrival by your own transportation. There was also a significant increase in the use of diagnostic imaging in those arriving by ambulance. The results may be confounded, though, both by the severity of stroke and by whether stroke was the true diagnosis. First of all, the severity of one's symptoms often affects decision making. A patient who is experiencing mild facial numbness over the coarse of a few days may decide to be taken by family members to the emergency room, while someone with sudden weakness on one side and inability to speak will likely have an ambulance called for them. The first case is indeed less urgent than the second, and therefore may account for the difference in evaluation speed by physicians in the emergency room.
The second point is that in many cases, the final diagnosis may not be a stroke. Miscoding within the database, based on initial symptoms, but never corrected based on later information, may account for this. My evidence for this is the overall high number of patients who did not receive diagnostic imaging. I would venture to say that nearly 100 percent of patients suspected of having a stroke by ER physicians would undergo a CT scan or brain MRI. The fact that 27-40 percent of patients did not undergo such imaging indicates that stroke was not seriously being considered as a diagnosis. This is further demonstrated by the high percentage of patients who were sent home in the study (only 53 percent of walk-in patients were admitted to the hospital.)
Thus, while it is best to call 911 in cases of suspected stroke and to be transported by ambulance, the data in this study used to support this conclusion are somewhat suspect.
