Expecting the Best in Pregnancy and MS

 Pavle Repovic, MD, Ph.D, Neurologist, Multiple Sclerosis, Swedish Neuroscience Institute

Considering that multiple sclerosis (MS) affects primarily women of childbearing age, it comes as no surprise that for many patients MS and pregnancy often occur together. The issues to consider when discussing pregnancy and MS include:

• How pregnancy affects MS

• How MS affects pregnancy

• How MS treatment should be managed throughout pregnancy

The Pregnancy in MS (PRIMS) study of 254 patients revealed that pregnancy is generally protective against MS relapses, in particular during the third trimester. In contrast, the same study found a rebound of relapses during three months post delivery, with 30 percent of women experiencing a relapse within three months after delivery. Several strategies have been proposed to avert the risk of postpartum relapse, including the use of prophylactic IVIG or corticosteroids. More recently, exclusive breast-feeding has been found to offer some protection against postpartum MS activity; however, this finding was disputed in a subsequent study.

There is no evidence that MS impairs fertility or leads to an increased number of spontaneous abortions, stillbirths or congenital malformations. MS also does not increase a woman’s risk of preeclampsia or premature rupture of membranes. Pregnant women with MS are 1.3 times more likely to undergo antenatal hospitalization and to have a Cesarean delivery, and they are 1.7 times more likely to have infants who are small for gestational age 6.

Except for glatiramer acetate, all MS disease-modifying treatments (DMT) have documented in utero harmful effects in animal studies and are therefore FDA pregnancy category C agents. Glatiramer acetate is a category B agent and is not known to have harmful effects in animal studies, although human studies are lacking.

For these reasons, the National MS Society and most MS specialists advise women who intend to become pregnant to discontinue therapy. Given their pharmacokinetics, we suggest the following schedule based on the type of therapy: one month (glatiramer), two months (fingolimod) or three months (interferons, natalizumab) prior to anticipated conception. It is less clear when to resume the therapy following the delivery.

Because only a minuscule amount of medications is excreted in mother’s milk, some MS specialists advise patients to resume therapy – with the exception of fingolimod or natalizumab – as soon as possible, even in women who intend to breast-feed. In the event of an MS relapse during or after the pregnancy, treatment with high dose intravenous methylprednisolone is generally considered safe for both mother and baby.

Questionable Hope for CCSVI

James Bowen, MD, Neurology, Multiple Sclerosis, Swedish Neuroscience Institute 

Once again, multiple sclerosis patients’ area buzz over a new theory and treatment for the disease. The theory is called chronic cerebrospinal venous insufficiency (CCSVI); and, this time, social media is driving the patient excitement.

CCSVI is based on a controversial idea that impaired venous drainage of the brain due to blockage in venous structures causes MS. Increase in venous pressure promotes leakage of blood across capillaries, with inflammation resulting from the iron deposition into the brain. In 2009 Paolo Zamboni, M.D., reported that virtually all MS patients in a study had abnormalities in the jugular or azygous veins, whereas no control patients had such findings. The Zamboni, or Liberation, procedure involves either angioplasty or stenting of the abnormal vein. Many MS patients are understandably enthusiastic about this theory and treatment.

There are, however, a number of problems with the CCSVI theory that patients and MS neurologists should consider.

• Diagnostic criteria for CCSVI are not standardized or accepted

• CCSVI does not seem to explain the distribution of white matter lesions or the relapsing and remitting course that most patients experience

• CCSVI does not explain the presence of inflammation in MS lesions

• Iron in MS lesions is contained within macrophages, not erythrocytes or free, as predicted by the CCSVI theory

• Other diseases with increased venous pressure do not resemble MS

• Venous drainage is highly redundant, so stricture of a vein usually does not increase venous pressure in the brain

• Changes in muscle tone or posture from neurological disease may explain some venous blockages; and neurological disease can lead to lower venous blood flow due to circulatory auto regulation

Studies that have been performed to date have not supported the CCSVI theory. Preliminary results from the Buffalo Neuroimaging Analysis Center at the University at Buffalo, The State University of New York, found 56 percent of patients with MS, 42 percent of those with neurological diseases other than MS, and 23 percent of controls met criteria for CCSVI. This group found a decrease in cerebral venous volume in MS patients compared to normal, whereas blockage in venous flow would be expected to induce vascular dilation and increased venous volume. At the recent ECTRIMS (European Committee for the Treatment and Research of Multiple Sclerosis) meeting in Sweden, groups from Gotteborg, Berlin, Padua, Amsterdam and London found no evidence for venous obstruction in MS patients compared to controls. A study from London, Ontario, found that venous blockage increased with age in both MS and controls. A study from Beirut, Lebanon, found venous changes in only 9 percent of patients after their first MS attack, increasing to 92 percent with advanced MS, possibly suggesting a late finding that is unrelated to the cause of the disease. Nevertheless, as anecdotal reports continue to circulate of individual patients with dramatic responses to treatment of CCSVI, the theory has been of great interest among MS patients and has become the subject of a significant number of studies.

Before CCSVI can be considered as contributing to MS, three criteria should be required.

1. Venous blockage must be shown to be increased in MS patients relative to healthy controls and other neurological diseases. This would demonstrate an association between CCSVI and MS, but not prove causation.

2. Treatment of CCSVI should stop the progression of MS symptoms in placebo controlled, blinded, multi-center studies. This would prove that CCSVI contributes to MS, but not prove causation.

3. MS should develop in humans or animals with venous blockage.

The scientific community is taking CCSVI very seriously. Most of the studies currently under way are testing whether MS patients have venous blockage. The National MS Society and the MS Society of Canada funded seven 24-month studies in June 2010, totaling $2,400,000 of ultrasound, MRI and angiographic techniques to determine whether venous blockage is specific to MS patients in both adult and pediatric populations. The MS Society of Italy is contributing £900,000 (nearly $1.3 million) to study the question. The Buffalo Neuroimaging Analysis Center is conducting a study of venous imaging, as well as a small treatment trial. Saskatchewan and MS Research Australia have also initiated large studies.

The risks of the Liberation procedure in MS are not entirely understood, though stent migration and fatal intracerebral bleeding have already been described. Only controlled studies will teach us the true complication rate. In the meantime, most MS neurologists are recommending that patients await the results of current studies before proceeding with this controversial treatment. 

SNI Communication Tools

John W. Henson, MD, Swedish Neuroscience Institute     

 

SNI is leveraging communication tools that deliver information to patients, referring physicians and the public as a crucial part of providing care at the advancing edge of neu­rological knowledge. The goals of these tools are two-fold. One goal is to update established patients and their doctors regarding the latest developments in our programs and centers. The other is to lower the barrier for patients and physicians who are facing a new neuro­logical problem to discover tertiary subspe­cialty care.

A new SNI communication tool

Dan Rizzuto, Ph.D., director of SNI research, and John Henson, M.D., recently launched SNIblog.com to complement other commu­nication efforts and to provide a communication outlet for the staff of SNI. SNIblog.com offers brief notes about advances in neurological care provided in SNI’s centers, as well as news items about the institute that are of interest to our patients and referring physicians.

Blog content is more dynamic than Web content. Search engines are able to detect targeted key words within each entry, which helps direct highly relevant Web traffic to the blog. This aids in the dissemination of infor­mation to patients and physicians. Viewers also can subscribe to an e-mail notification system that will alert them to newly posted material.

As is the case with any communication tool, however, blogs also have limitations. For example, blogs are not able to incorpo­rate the interactive features of social media that exponentiate information transfer to a selected audience. Therefore, SNI will build upon its existing menu of communication tools by launching social media outlets in the near future.

Other SNI communication tools

At SNI we provide print and electronic options to meet the needs of various audi­ences.

BrainWaves. This print news letter is de­signed to be an educa­tional resource about neurological topics for physicians in thePacific Northwest. Each issue focuses on conditions treated at SNI. Staff members au­thor the articles.

Physicians Practice. Swedish has the exclu­sive contract with this practice management journal to customize content in six issues annually. Although the journal is not exclu­sively used for SNI in­formation, many neu­rological topics have been covered in recent issues. Swedish maintains an online library of past content in the Health Professionals sec­tion of its website at www.swedish.org.

Swedish Neuroscience Institute website. Earlier this year the SNI website took on a new look as part of a project to redesign the entireSwedishMedicalCenter website. SNI members are now able to directly update information about their programs, and are actively developing online patient resources for each of SNI’s programs. The SNI website is a repository of information about our neurological services, research, clin­ical trials and professional expertise.

Our goal is to continuously improve ac­cess to neurological information and the de­livery of that information to interested groups of patients and referring physicians.

Hugh Markus – 2011 Merrill P. Spencer Lecturer

 

Hugh Markus, B.M., B.Ch., D.M., FRCP
Featured Presenter: 6^th Annual Merrill P. Spencer, M.D. Endowed Lecture

 

Each spring, The Merrill P. Spencer, M.D. Endowed Lecture is presented in conjunction with the annual Swedish Neuroscience Institute Cerebrovascular Symposium. This year, we are pleased to welcome Dr. Hugh Markus, Professor of Neurology at St. George’s University of London.

Hugh Markus was educated in Medicine at Cambridge and Oxford Universities and then carried out medical jobs in Oxford, London and Nottingham before training in neurology in London. He was senior lecturer and subsequently, reader in neurology at Kings College London before moving to the chair of neurology at St George’s in 2000.

His clinical interests are in stroke, and he is clinical lead for stroke at St George’s Hospital. He is involved in both acute stroke care and outpatient stroke clinics, and runs specialist services for patients with sub cortical vascular disease and genetic forms of stroke.

His research interests are in applying molecular genetic and imaging techniques to investigate the pathogenesis of stroke. Genetic studies are primarily trying to identify genetic causes of sporadic stroke and he is the principal investigator for the Wellcome Trust Case Control Consortium 2 Ischemic Stroke Study, which is performing a large genome-wide association study in ischemic stroke. The imaging techniques he uses are Transcranial Doppler emboli detection and MRI.

His postdoctoral thesis was on emboli detection, which involved experimental studies validating the technique and early clinical studies applying it to patients with a variety of potential embolic sources. He has carried out a number of studies showing that embolic signals predict stroke in carotid artery stenosis, and pioneered the use of the technique to evaluate anti-platelet therapies. He was also principal investigator for the CARESS study. Recently, he finished the Asymptomatic Carotid Emboli Study (ACES) which demonstrated that embolic signals predict risk in asymptomatic carotid stenosis.

The first international conference which Dr. Markus attended was a Transcranial Doppler ultrasound workshop organized by Merrill Spencer, M.D. in the early 1990s.

To register for the 5^th Annual Cerebrovascular Symposium: New Therapeutics for Today’s Patient on May 12-13, visit www.swedish.org/cvdregister. Registration for the conference includes the Merrill P. Spencer, M.D. Endowed Lecture.

To attend only the reception and Merrill P. Spencer, M.D. Endowed Lecture on May 12: www.swedish.org/cvdspencer. This is a free CME program. However, pre-registration is required as space is limited.

 

 

 

Multiple Sclerosis Center 2nd Annual Art Show 2011

 

Bobbie Severson, ARNP, Multiple Sclerosis Center, Swedish Neuroscience Institute

 

The Multiple Sclerosis Center at Swedish Neuroscience Institute is hosting its Second Annual Multiple Sclerosis Center Art Show at the Bellevue Arts Museum on Saturday and Sunday, June 18 & 19, 2011 from 11:00am to 5:00pm.  There will be an ‘Artist Only Meet ‘n’ Greet, Sunday June 19^th from 3pm – 5pm

Entry Criteria:

1. Anyone living with, or touched by, MS can enter the MS art show.
2. All abilities welcome.
3. All art is accepted unless you are otherwise notified.
4. You do not need to be a patient of the MS Center, at Swedish, to submit your art.
5. Art may be: painting, photography, sculpture, metal, crafts, etc.
6. Art limit: 2 pieces per artist. Please mark each entry as primary and secondary.
7. Submission Deadline:  May 13, 2011 by midnight. 

Important information: 

Submit your name, name of artwork, biography of artist (optional, not to exceed 200 words), and a statement about the art piece (optional, not to exceed 200 words).

Indicate whether you want your email address and/or phone number to be disclosed with the biography as contact information for the artist.

This is an art show. There will be NO sale of art. Any sales must be conducted privately through the contact information provided.

Enrollment:  Email msartshow11@gmail.com for an entry form. Submit form by May 13, 2011.

For Questions:  Email msartshow11@gmail.com or call Chaz Gilbert 206 320 2552.

Misc:  Details regarding drop off and pick up of art will be emailed to you once you upon submission of your entry. We do not take responsibility for any damaged art. Any art not picked up after the show will be donated to the MS Center and/or other charitable organizations.

Advances in thrombolysis

Bill Likosky, MD, FAAN, FAHA, Director for Stroke and Telestroke, Swedish Neuroscience Institute

 

 

Washington State has one of the high­est stroke mortality rates in the nation. To improve this situation, acute intervention­al therapies for stroke are being employed to restore circulation to ischemic brain tissue that surrounds areas of completed infraction, while avoiding risk of hemor­rhage due to reperfusion of large areas of infracted brain tissue.

Urgent thrombolysis with intrave­nous alteplase is the only therapy known to improve clinical outcomes following acute stroke. Unfortunately, alteplase has had limited usage because many patients arrive in an emergency department after the three-hour treatment window. The FDA has also approved two clot removal devices based on the ability to restore circulation. These devices are used up to eight hours after symptom onset. Several approaches to improved acute stroke care are now under way, including extension of the thrombolysis window to 4.5 hours, identification of safer thrombolytic agents and research identifying brain at risk of in­farction following a stroke.

A recent European study demonstrat­ed the efficacy of alteplase up to 4.5 hours after ischemic stroke in patients younger than age 80 years who have neither dia­betes mellitus or prior stroke. The safety profile during this longer window for these patients appears similar to that at three hours.

Another promising advance employs a new thrombolytic agent called des­moteplase. Derived from the saliva of the vampire bat, this agent has a longer half life than alteplase and does not break down basement membranes, leading to a lower risk of hemorrhagic complica­tions. The Swedish Stroke Program is part of an international effort to test this drug in a nine-hour window.

Todd Czartoski, M.D., and Bart Keogh, M.D., Ph.D., are collaborat­ing with the stroke team at Stanford University to identify patients with vi­able ischemic tissue regardless of time from onset of symptoms. Perfusion MRI identifies impaired blood flow in brain (the “penumbra”) surrounding an infarct. In cases where there is a large area at risk, the use of alteplase or clot retrieval may prove beneficial long after the three-hour window has elapsed.

Telestroke is another important development in acute stroke care. This program enables the timely alteplase treatment of patients in emergency rooms around the Pacific Northwest that lack onsite neurological expertise.

For more information about the Swedish Stroke Program, contact Sherene Schlegel, R.N., FAHA, at 206-320-3484. For information about telestroke, contact Tammy Cress, R.N., MSN, at 206-320-3112.

Detecting cerebral microemboli with transcranial doppler.

 

David W. Newell, MD, Cerebrovascular Surgery, Neurosurgery, Swedish Neuroscience Institute 

 

 

 

 

Colleen Douville, RVT, Director, Cerebrovascular Ultrasound, Swedish Neuroscience Institute

 

 

Since its introduction in 1982, transcranial doppler ultrasound (TCD) has evolved into a por­table, multimodality, noninvasive method for real-time imaging of intracranial vasculature.

The detection of cerebral microemboli is among the more remarkable capabilities of TCD. Emboli create countable signals in the ultrasound display due to the higher reflection of sound waves compared to the blood cells. Experimental mod­els have shown a high sensitivity and specificity for detection of a variety of substrates, including thrombotic, platelet and atheromatous emboli.

Microembolic signals (MES) within the in­tracranial vasculature are most frequently identi­fied in patients with large-vessel atherosclerotic disease, such as carotid stenosis. They have also been reported in intracranial arterial stenosis, ar­terial dissection, cardiac disease and atheroaortic plaque. Additionally, they have been seen in arter­ies distal to coiled aneurysms.

There is strong evidence that MES detection predicts future ipsilateral stroke risk in patients with symptomatic carotid stenosis (Markus HS, et al.; King A, et al.). A recent study of patients with asymptomatic carotid stenosis demonstrated that MES predicted subsequent ipsilateral stroke and TIA, and also ipsilateral stroke alone, and that it is helpful in selecting patients who will benefit from carotid endarterectomy (Markus, HS et al.).

Identification of active embolization provides crucial patho­physiological information to the neurologist and can also aid in the selection of tailored therapy aimed at reducing the risk of stroke. Emboli from different sources have unique compositions and re­quire specific therapy, such as antiplatelet agents for emboli from large artery atherosclerotic plaque and anticoagulants for cardiac emboli.

Future advances in TCD technology will permit full automa­tion and better identification of the composition and size of circu­lating embolic materials, thus improving its value for patients with cerebrovascular disease.

Contact Colleen Douville, RVT, at colleen.douville@swedish.org or 206-320-4080, for more information about TCD for detec­tion of cerebral microemboli.

 

PFO closure for migraine

Mark Reisman, MD, Director of Cardiovascular Research and Education

 

Migraine is a primary headache dis­order that causes significant suffering in approximately 13 percent of the popula­tion of the United States. It accounts for an estimated $23 billion in annual cost to the economy through health-care expenses and lost productivity.

Two major features of migraine are migraine aura (MA) and headache. MA occurs in nearly one-third of migraine pa­tients and consists of one or more focal neurological symptoms that develop gradually over 5-20 minutes and persist for less than 60 minutes. MA typically precedes development of migraine headache.

Several years ago single-center retrospective analyses first reported an apparent association between partial or complete relief of migraine symptoms and transcatheter clo­sure of patent foramen ovale (PFO) for secondary stroke prevention (Reisman M, et al., 2005). The fora­men ovale normally serves as a one-way valve in the interatrial septum for physiologic right-to-left shunt in utero. Complete fusion of interatrial septae normally occurs by two years of age. When septae fail to fuse, how­ever, the PFO is a potential tunnel that can be opened by reversal of the interatrial pressure gradient. PFO is the most common form of right-to-left circulatory shunt (RLS).

Studies have shown that as many as 50 percent of individuals with MA will have a PFO, whereas PFO is present in about 25 percent of the general population and in migraineurs without aura (MO). In analy­ses performed by Swedish researchers, MA patients had a larger RLS than patients with MO, despite similar interatrial anat­omy (Jesurum JT, et al., 2007), and were about 4.5 times more likely to have greater than 50 percent reduction in migraine fre­quency following PFO closure (Jesurum JT, et al., 2008). These observations indicated a potential pathophysiological relationship between migraine and PFO.

The mechanism for this potential re­lationship is not understood, but investi­gators have focused on possible interatrial transit of vasoactive chemicals that bypass the pulmonary capillary bed, or on micro­emboli from the venous circulation which might trigger cortical spreading depres­sion and transient regional hypoperfu­sion. Migraineurs may have higher plate­let reactivity (Jesurum JT et al., 2010) or pro-coagulant state (e.g., protein C or S deficiency) than non-migraineurs, possibly resulting in greater load of microemboli in the arterial circula­tion. The brains of migraineurs may be more sensitive to circulatory changes than are the brains of those without migraine. The combination of potential triggers and susceptible neuronal substrate may result in an enhanced risk of MA among pa­tients with PFO.

The Migraine Intervention with STARFlex Technology (MIST) trial was a randomized trial of PFO clo­sure in migraine (Dowson A et al.). The failure of the trial to meet its primary endpoint (cessation of headache) and secondary endpoint (>50-percent re­duction in headache frequency and days) was surprising. Eligibility criteria for the trial may have excluded those patients who were most likely to benefit from PFO clo­sure. For instance, patients were excluded from MIST if they had a history of stroke or hypercoagulability, and subjects had to fit within a narrow range of headache fre­quency. If patients with a greater migraine burden or hypercoagulability were more likely to achieve meaningful reductions in headache frequency and severity, these exclusion cri­teria could have altered the study outcome.

Other trials are in progress or in the pipeline that may better elu­cidate the effect of PFO closure on migraine. The migraine-PFO asso­ciation offers opportunities for col­laboration between scientists and clinicians in both neurology and cardiology. The long-term goals of collaborative trials are improved quality of life and reduced cerebro­vascular sequelae for individuals who suffer from migraine.

 

Options widening for wide-necked aneurysms

 

Yince Loh, MD, Interventional Neuroradiology, Neurosurgery, Swedish Neuroscience Institute 

 

Intracranial aneurysms are present in up to 4 percent of the population. These potentially dangerous vascular lesions are being detected with increasing frequency in asymptomatic patients by advances in noninvasive imaging techniques, such as magnetic resonance angiography (MRA). Appearing like blisters on the wall of the brain’s blood vessels, aneurysms develop when the blood vessel’s native repair ability is exceeded by the mild, but constant, injury created by flowing blood under high pressure. The five most common risk factors for developing an aneurysm are: smoking, female gender, high blood pressure, middle age and family history.

Intracranial aneurysms are complex lesions that require a highly specialized, multidisciplinary approach that is individualized for each patient. Key members of the care team for these patients include endovascular neuroradiologists, neurosurgeons with special expertise in aneurysm surgery and neuroanesthesiologists. Availability of dedicated neurocritical care units is an essential care component. A consensus recommendation by these specialists may include close observation, obliteration of the aneurysm with a surgical clip, or filling the vascular outpouching with filamentous coils that are introduced by endovascular microcatheters via an artery in the leg. This latter process is called “coiling.”

Those aneurysms that have a balloon-like opening, or neck, from the parent vessel are typically good candidates for coiling. Not infrequently, however, the aneurysm’s shape does not permit safe coiling. When the aneurysm’s neck is wide, it appears more like a molehill than a balloon. A molehill configuration is often referred to as a “wide-necked aneurysm.” The wide neck allows an unwanted protrusion of coils back into the artery. This can lead to a number of problems, including failure to obliterate the aneurysm and stroke. Thus, in situations where an aneurysm is not surgically accessible or the patient cannot undergo surgery, no therapeutic options can be offered.

Until recently, wide-neck aneurysms could not be treated by coiling. The U.S. Food and Drug Administration, however, has approved a tubular device called an intracranial stent to be used for such situations.

Once a stent is deployed across the neck of the aneurysm, coils are placed into the aneurysm through the stent wall. The stent struts prevent the coils from falling back into the artery by essentially creating a “chain link fence” across the neck of the aneurysm.

Stenting, however, produces another set of problems. A stent is a foreign body that can promote the formation of a blood clot inside the vessel, which is why patients are placed on two antiplatelet medications to thin the blood, usually aspirin and clopidogrel (Plavix®), after placement of a stent. The length of time required to thin the blood after stent placement is unclear, although stents may become incorporated into the vessel wall and covered with endothelium within weeks.

Emerging concepts in vascular neurology: TIA clinics help prevent strokes and unnecessary hospital admissions

Michael Fruin, ARNP, Swedish Neuroscience Institute

Tom Jaspee placed an anxious call to Dr. Lewis’s office at 9 a.m. sharp. He didn’t give many details, other than to say his wife was worried about problems he was having with his speech the previous night. Later that morning in Dr. Lewis’s office, Tom said he had trouble getting his thoughts out for a few minutes. He said he felt fine im­mediately afterwards and didn’t want to raise a ruckus. Tom’s wife added that his right face drooped and the episode took al­most 30 minutes to clear up. She was wor­ried that Tom had suffered a stroke.

Dr. Lewis was well aware of Tom’s high risk of stroke following his transient isch­emic attack (TIA). Realizing that he could not manage this urgent issue in his office, Dr. Lewis sent the patient to the emergency room and after a six-hour stay, Tom was admitted as an inpatient for a 24-hour ob­servation and evaluation.

This mock case study highlights the role a TIA clinic might have played in avoiding an emergency room visit and hos­pitalization, while still providing the TIA patient the necessary urgent care.

While hospital admission is appropri­ate for the subset of patients at high risk for having a stroke after TIA, significant num­bers of emergency room visits and admis­sions could be avoided by a recent advance in evaluating patients in a TIA clinic. TIA clinics are being pioneered in the United Kingdom, where patients with TIA can be seen by a stroke specialist in an urgent-care clinic setting in which a standardized pro­tocol of neurologic evaluation and diagnos­tic testing is administered.

The effectiveness of the TIA clinic is supported by findings from the EXPRESS Trial (Luengo-Fernandez R, et al.). In this trial, there was an 80 percent reduction of 90-day stroke risk when TIA and minor stroke patients received urgent evaluation and treatment in a standardized urgent-care clinic setting. Patients at high risk of stroke, such as those with high-grade ste­nosis of the internal carotid artery or with atrial fibrillation, are admitted to the neurology service as indicted. Patients at low risk of stroke receive patient education and a stroke prevention plan is implemented.

Reference

Luengo-Fernandez R, Gray AM, Rothwell PM. “Effect of urgent treatment for transient ischaemic attack and minor stroke on disability and hospital costs (EXPRESS study): a prospective population-based sequential comparison.” Lancet Neurology 8:235-243. 2009.

.