Dr. Anne L. Abbott: What is the Role of Noninvasive Medical Intervention in Stroke Prevention?

Q: What are you doing that is not the typical treatment for asymptomatic carotid stenosis?

A:  I am advocating medical (non-invasive) intervention alone to reduce the risk of stroke associated with asymptomatic carotid stenosis (of any severity). Medical intervention, here, refers to combined non-invasive strategies to reduce risk of stroke (and other vascular disease complications). It comprises optimal diagnosis of heart and vascular disease risk factors and risk reduction using support of healthy lifestyles and appropriate drugs.

Q: Tell me about your asymptomatic carotid artery disease research. How did it begin?

A: It began officially when I started my PhD in 1998. My main PhD responsibility was an observational cohort study of 202 patients with 60%-99% asymptomatic carotid stenosis being managed with medical intervention alone. The average follow-up was 34 months. These people had carotid stenosis but no previous symptoms of stroke or TIA on the same side of the brain as the carotid artery lesion. We assessed them clinically (risk factors, stroke and TIA symptoms, other diagnoses and medication) at baseline and every 6 months. We were also using a new technique called transcranial Doppler (ultrasound) to detect tiny emboli traveling downstream from the carotid lesion into the brain to see if this may be a method to better risk stratify patients. That is to identify those with higher than average stroke risk who may benefit more than average from invasive carotid procedures (such as surgery).

Q:  Why do you believe medical intervention looks superior to surgical intervention?

A: The most surprising result from our study of patients with asymptomatic carotid stenosis was that the average annual rate of first, same-sided stroke (‘ipsilateral’ to the carotid lesion) in our patients was only 1.0%. This was about 2-3 times lower than expected based on results in the literature when our study was planned and commenced. Our rates of embolus detection with transcranial Doppler were also very low (about 0.16/hour on average). This meant we needed a much larger sample size than we originally thought to test our main hypothesis that the detection of embolic signals would identify a subgroup of patients with a significantly higher risk of same-sided stroke or TIA. Our study was never funded in the first place and we simply had no resources to continue. However, more important was this lower than expected stroke rate. It made me wonder that perhaps medical intervention had improved enough since the randomized trials of surgery versus medical intervention alone (Veterans’ Affairs Cooperative Study [VACS], Asymptomatic Carotid Atherosclerosis Study [ACAS] and the Asymptomatic Carotid Surgery Trial [ACST]) and that routine carotid surgery for asymptomatic carotid stenosis was now inappropriate.

This suspicion was confirmed during my post-doctoral studies by the results of my meta-analysis of stroke risk with medical intervention alone over time (Abbott, Stroke 2009). With the help of world-renown medical statistician, Dr. John Ludbrook, I demonstrated that the risk of same-sided or any stroke (and/or TIA) has fallen significantly since the 1980s (when the first reliable measurements were made) to at least match rates for randomized trial operated patients. These findings have been independently, internationally validated. Using most recent studies, we are now seeing an average annual rate of first ipsilateral stroke of only about 0.5%. This is about 5 times lower than non-operated randomized surgical trial asymptomatic patients, 3 times lower than randomized surgical trial operated asymptomatic patients, twice as low as operated asymptomatic patients in the recent Carotid Revascularization Endarterectomy versus Stenting Trial (CREST) and 3 times lower than CREST stented asymptomatic patients. In addition, what makes the choice for medical intervention alone in routine practice obvious is that medical intervention alone is much cheaper for stroke prevention; it prevents other complications (like heart attack, caused by invasive carotid procedures); and in routine practice procedural complications are not usually measured, or if measured, they are often higher than in randomized trials. The bottom line is that all patients with carotid stenosis should receive optimal current medical intervention to reduce their risk of stroke and other heart/vascular disease complications. Additional routine practice carotid surgery or stenting is now more likely to cause harm or, at best, be an expensive, ineffectual procedure that the community pays for.

Q:  Tell us more about the studies you included in your meta-analysis. How many patients total? What time period was covered?

A: The studies used in my meta-analysis were selected because they satisfied basic criteria for sound methodology (reliability). These included prospective design with at least 100 patients with directly imaged 50%-99% asymptomatic carotid stenosis. Asymptomatic meant no previous symptoms of same-sided stroke or TIA. Only 11 studies were identified satisfying these criteria (including VACS and ACAS but not ACST). They were published between 1985 and 2007). The total number of patients involved in the 11 studies was 3724. With the latest studies (published since 2009), it is now 4285.

Q: What other surprising results were found during your research?

A: The average baseline age of the patients included in the meta-analysis increased over time. Rates of stroke fell over time despite patients getting older. Also, there was a fall-off in the number of current smokers at baseline. Clearly, some people understood the dangers of smoking and gave it up. In addition, the meta-analysis demonstrated how diagnosing diabetes, high cholesterol, and high blood pressure changed over time. The thresholds lowered and intervention occurred earlier. Medical treatment changed in a number of ways as well including the introduction of new drugs such as statins. The bottom line is that it is all helping to prevent stroke very effectively. It also means the patient, with correct information and support, now has the most power to prevent their own stroke. After all, it is only the patient who can stop smoking, present for risk factor detection, adopt healthy lifestyle habits, and take medication appropriately. This is a substantial swing away from the current mindset that a proceduralist is required to best prevent stroke caused by asymptomatic carotid stenosis. The public needs to know this if we are to move forward and best prevent stroke and, at the same time, make health services affordable.

Q: What is the next step for you in terms of research?

A: I’m a neurologist who finished up doing studies of stroke prevention. These studies have shown me that stroke is not primarily a disorder of the brain (the organ at the center of my neurologist training). Instead, stroke is usually a complication of disease of the heart and blood vessels. To help my patients best prevent stroke, I should have been a cardiologist (to better understand atrial fibrillation and other heart disorders), a kidney specialist (to better understand hypertension), an endocrinologist (to better understand diabetes), a metabolism specialist (to better understand blood lipids), and a psychologist (to better understand smoking addiction), just to name a few! In other words, current knowledge about stroke prevention comes from many different traditional medical specialties. This knowledge has become vast and is ever evolving. Guidelines are many and lengthy, making them poorly accessible to all those who need to help prevent complications of heart and blood vessel disease (all of us). It is now time to summarize and standardize current best knowledge about diagnosis and management of these risk factors in a way all specialists and generalists can utilize in routine practice. Then we need to measure the impact of this simplified, clearly defined current best practice on risk of stroke and other complications in patients identified by various organ/artery-focused manifestations, like asymptomatic and “symptomatic” carotid stenosis, stroke, TIA, or atrial fibrillation. We need to compare this best medical intervention with ‘usual care’ so we can set new standards for routine practice, compare new interventions, measure cost effectiveness, and work out if screening for carotid stenosis (for example) is helpful to better direct medical intervention.

Q: Are you asking for efforts from other physicians? Are you trying to fund a study where you can do this research with others involved?

A: Yes. I would like help in defining which risk factors have ever been shown to be useful in reducing risk of stroke and other complications of heart and blood vessel disease,  how to best diagnose and manage these risk factors to prevent these complications, and understanding the expected magnitude of the absolute risk reduction we can then achieve. As said above,  current best medical intervention needs to be tested in academic and routine practice. Such testing requires multidisciplinary efforts for study design and implementation, not only because of the compartmentalized source of current knowledge but because all the complications we need to measure (like heart attack, stroke, ischemic limbs and ischemic damage of other organs) are seen by different specialists.

Q: Is there anything further that you would like to mention that we haven’t covered so far?

A: We need to make basic research, like accurately measuring the risk of stroke in patients with carotid stenosis, much easier.  We need to integrate research with routine patient care reporting activities. This can be achieved by standardizing the content and terminology used in doctors’ letters/patient reports written in outpatient clinics, hospitals, and private practice. In addition, we need to do away with the paper basis of these letters/reports. Instead, we need direct electronic entry of information already recorded in these letters/reports into a database that would provide us information on what our patients’ problems are, how we intervene, compliance, and outcomes. We clearly need more quality observational data, and not just derived from randomized trials. It is all achievable these days. We just need to get the ball rolling.

Dr. Anne L. Abbott is a neurologist and Senior Research Fellow at the Baker IDI Heart & Diabetes Institute in Australia.