Timothy C. Hain, MD. Page last modified: September 1, 2019
|MRI (Magnetic resonance image) of the brain. This is section is in the horizontal plane, just above the ears. Periventricular white matter lesions can be seen (the white spots towards the bottom). This is a moderate case.||MRI showing moderate PVM. Moderate PVM corresponds to roughly a 15% burden of Leukoariaiosis.||Severe PVM. 90 year old woman. There is confluence (merging together) so that there is a cloud surrounding the upper ventricles. (Flair)|
MRI studies of older persons with disequilibrium and gait disturbances of unknown cause often show frontal atrophy and subcortical white matter T2 hyperintense foci. (Kerber et al, 1998). Pathological studies, though scanty, suggest frontal atrophy (shrinkage), ventriculomegaly (i.e. more shrinkage), reactive astrocytes in the frontal periventricular white matter (i.e. scarring), and increased arteriolar wall thickness (Whitman et al, 1999).
There are several locations for white matter lesions. Those around the center black spaces are called "periventricular white matter lesions". Those located between the cortex and ventricles, with some space between, are just called "white matter lesions". There are also subtypes in the "deep white matter", below the ventricles, some in the cerebellum, and sometimes they are seen in the brainstem. This page is focused on the higher lesions around the ventricles.
White matter lesions are best seen on the "T2 FLAIR" MRI sequence of brain imaging. CT scans are not nearly as sensitive. The better (stronger) the MRI magnet, the more lesions are seen. Thus, MRI's done on contemporary 3T units will see more and smaller white matter lesions than scans done on "open" scanners. Pathologically, PWM correspond to areas of myelin thinning and gliosis, and are often accompanied by lacunar (small holes) infarctions and small vessel atherosclerotic disease.
Cerebral white matter lesions are common, alarming, and often called "incidental" by physicians. Perhaps for this reason, the author of this page (Dr. Hain) has been emailed several times with vigorously phrased requests to weaken the language concerning the cognitive consequences of white matter lesions. I just report what the literature has to say, and unfortunately, "it is what it is". Still, in response, I have adjusted the language in some places to use more "academic" terms for reduced mental function.
There is an immense body of literature about white matter lesions, and here we are just discussing a small subset of these thousands of papers. Because there are so many papers, one can generally find a paper supporting nearly any conjecture -- this is good to keep in mind when reading reviews like this one.
The short answer is that nearly anything that damages the brain can cause white matter lesions.
Usually these lesions are blamed on "small vessel disease" (SVD) , with the idea that a small blood vessel closes off. The term is vague, and might be applied both to blood vessels closing off as well as bleeding. In the case of the ischemic (closed off) type of SVD, the brain supplied by this blood vessel is damaged, and a white spot appears on the FLAIR sequence of the MRI.
Hypertension (high blood pressure) is "a major risk factor" for PVM. They are also linked to smoking, type-2 diabetes, and recently obesity. Visceral obesity in particular may increase deep white matter lesions through inflammation (Lampe et al, 2018).
White matter lesions are common in individuals with HIV infections, even in persons with good viral suppression (Su et al, 2016). Haddow et al (2014) reported that WML were found in 161 out of 254 HIV+ patients, of whom 35% had "diffuse WM signal abnormality", and 24% had a pattern suggestive of "small vessel disease".
These two images show the cerebellum. On the left is the Flair image, and on the right, the GRE image. The black spots are "blooming" from iron deriving from blood breakdown products..
Other causes of brain damage that can be spotted on MRI include bleeds, which also cause brain damage. These lesions occur in about 11-23% of the community-dwelling elderly. As they contain iron from blood, they can be distinguished on MRI from other types of damage by viewing them with a "sequence" sensitive to iron, such as gradient echo (GRE) (Greenberg et al, 2009). Cerebral microbleeds can also occur in the white matter, and can be due to trauma, hypertension, or cerebral amyloid. As is shown above, small bleeds can appear very dramatic on GRE (on the right), but not be spotted on FLAIR. Due to "blooming", the size of the black spots on MRI can be larger than the actual area of iron deposition. GRE images with longer echo times (e.g. TE > 50) may increase blooming (i.e. cause false positives). Cerebral microbleeds are associated with accelerated cognitive decline. (Ding et al, 2017)
There are many scales that are mainly used in research studies to quantify white matter lesions. Schelten and associates reviewed 26 different scales in 1998 (Scheltens et al, 1998) . Right now (2019), the terms "mild", "moderate", and "severe" are favored clinically, and thus you can ignore this section if you wish.
ARWMC - age related white matter changes.
This scale is a 4 point one, based on MRI images with either proton density (PD), T2, or T2-FLAIR. White matter changes were defined as "ill-defined hyperintensities >= 5 mm. Lacunes were defined as well-defined areas > 2 mm, with the same signal characteristics on MRI as spinal fluid. (Wahlund et al, 2001)
ARWMC Rating scale White Matter lesions Frontal, pareto-occipital, temporal, infratentorial/cerebellum, basal ganglia. Left and right hemispheres rated separately 0 No lesions 1 Focal lesions 2 Beginning confluence 3 Diffuse involvement of an entire region Basal Ganglia Lesions 0 No lesions 1 1 focal lesion (>= 5 mm) 2 > 1 focal lesion 3 Confluent lesions
This scale then produces 10 different numbers -- which is 5 regions * 2 sides. This makes the scale impractical for clinicians, who do not have the resources to count 10 different regions of the brain. This is probably why it is not as popular as the two simpler scales below.
Simpler but similar scales include the Fazekas scale, which considers only 2 regions of the brain --(Fazekas et al, 1998). This scale was the most popular one as of 2016.
Fazekas Rating scale White Matter lesions Periventricular white matter (PVWM) -- these are largely from small vessel disease 0 No lesions 1 Caps or pencil-thin lining 2 Smooth halo 3 Irregular periventricular signal extending into the deep white matter Deep white matter These are often not ischemic. 0 No lesions 1 punctate 2 beginning confluence 3 Large confluent areas
Van Swieten scale
Another similar scale for rating white matter lesions is the van Swieten scale (van Swieten et al, 1990). This scale requires grading of two regions -- one involving the anterior white matter in the slice through the choroid plexus, and another for the posterior white matter. There were three values for MRI--0 (nothing or one lesion), 1 -- multiple focal lesions 2, multiple confluent lesions. The two regions and 2 numbers were then added together to provide a score ranging from 0-4. This one is pretty close to the Fazekas scale - it doesn't require making a distinction between two different types of confluence, and it excludes the deep white matter. It seems pretty close to the clinical scale (nothing, mild, moderate, severe) however, and the advantage of the clinical system is that it is pretty obvious even to persons who are not radiologists.
|Lin et al (2017). WML (white matter lesions) in 4693 hospitalized patients. Here the percent positive are all kinds of WML, small or large. Also, these patients are not selected to be well.||Chowdhury et al (2011). WML in patients with NO vascular risk factors. DSWMH refers to some confluence of WML. Note that these are only confluent lesions, and "punctate" lesions are excluded. Thus this graph would make the situation appear better than the study of Lin et al on the left, which includes all types of WML in sicker patients.|
Between a third and 80% of MRI scans done in persons over the age of 65 have changes in their cerebral white matter (Wong et al, 2002). Others quote as high a prevalance as 95% above the age of 60 (e.g. Lampe et al, 2018).
The frequency of white matter lesions depends on your threshold for reporting them and population being studied. Lin et al (2017) recently reported on 4683 hospitalized Chinese patients. They found a rather high prevalence, roughly 40%-80 in persons from 50 onward. They included as positive patients ranging from mild (small lesions) to severe (large confluent lesions). Prevalence was higher in persons with hypertension, diabetes and smokers.
Chowdhury et al (2011), reported on a healthier group -- 1108 subjects without vascular risk factors (i.e. no diabetes, hypertension). These were subjects largely between 50 and 70. Somewhat similar to Lin et al, they found a rather high prevalence -- 39.6% for "grade 1", which are "punctate" meaning small, and only 0.4% for grade 4 (large). They found a much lower prevalence for grade 2-3, which were defined as lesions having some confluence (i.e. blending together). These were subjects with no vascular risk factors, and cannot be directly compared to the subjects of Lin et al, who did have vascular risk factors.
Wen et al (2009) reported on 428 individuals aged 44-48 recruited from a healthy community study. Overall, 50.9% had white matter lesions. They commented that " small punctate or focal WMHs are common in the brains of individuals in their 40s, and may represent an early stage of development of these lesions."
Combining these three studies together, it is clear that small (punctate) white matter lesions are extremely common, they are found in roughly half of the otherwise healthy population in their 40's, and WML increase with age. In addition, as people age, they not only get more white matter lesions, but the WML start to merge together into bigger patches (confluent white matter lesions). People who have vascular risk factors (such as diabetes, smoking, hypertension), have more of these lesions.
Major causes of periventricular white matter (PWM) lesions include normal changes from aging (then they are called UBO's, for "unidentified bright objects), small strokes, and disorders related to multiple sclerosis (MS). PWM are also correlated with vitamin B6 (pyridoxine) deficiency. The phrase "normal changes from aging" is really a synonym for "we don't know".
Getting older: Age is certainly the single most common association of PWM. This is presumably a "wear and tear" phenomenon. You get older, and there is more water under the bridge. More fluctuations in blood pressure, more chance for small blood vessels to close, more chance of head injury, more chance for little emboli.
Nevertheless, while clinicians often suggest that changes in the brain that are similar to others of the same age are not important, and call them "incidental", data suggests that even a few of these PWM reduce cognitive performance (see below).
Small strokes: A period of hypertension is a common cause. In the authors experience, just a few days of extreme hypertension may be enough. This is the "stress is not good for you" connection. Progression of these lesions is associated with variability as well (Liu et al, 2016). This might suggest that small bleeds are the cause in some. There is a related disorder called "superficial siderosis" due to cerebral bleeds as well as cerebral microbleeds briefly discussed above.
Clinical studies of PWM also show association with diabetes, but not consistently with atherosclerosis. PWM are often reported in persons with migraine, and occur especially in women with migraine and aura. PWM are also more common in persons with frequent syncope and orthostatic intolerance (Kruit et al, 2013)
WM lesions are associated with retinal microvascular abnormalities. Persons with both WM lesions and retinopathy have a much higher risk of clinical stroke (20% vs 1.4%), (Wong et al, 2002).
MS and related conditions: Demyelinating disorders such as multiple sclerosis and relatives can cause PWM. These generally have a different look on MRI, as they often resemble "fingers" pointing towards the ventricles.
Rarer causes of white matter disease include "Autoimmune processes include multiple sclerosis and related diseases: tumefactive demyelinating lesions, Balo concentric sclerosis, Marburg and Schilder variants, neuromyelitis optica (Devic disease), acute disseminated encephalomyelitis, and acute hemorrhagic leukoencephalopathy (Hurst disease). Infectious processes include Lyme disease (neuroborreliosis), progressive multifocal leukoencephalopathy, and human immunodeficiency virus (HIV) encephalopathy. Vascular processes include different types of small-vessel disease: arteriolosclerosis, cerebral amyloid angiopathy, cerebral autosomal-dominant arteriopathy with subcortical infarcts and leukoencephalopathy (CADASIL), primary angiitis of the central nervous system, Susac syndrome, and neurolupus. Toxic-metabolic processes include osmotic myelinolysis, methotrexate leukoencephalopathy, and posterior reversible encephalopathy syndrome. (Sarbu et al, 2016). This is basically a laundry list of mostly rare conditions that has little practical consequences, but lets one know that there are many possibilities.
Among these, there is an incredibly rare familial variant of white matter disease, called CADASIL. Testing for the notch-3 mutation that causes CADASIL is done by Athena, among other laboratories. As is the case with most genetic testing, this test is prohibitively expensive and the result generally has no clinical implications.
PRES (reversible posterior leukoencephalopathy syndrome), can cause similar white matter lesions, but this condition is acute, and has some possibility of reversal.
There is strong evidence that cerebral white matter lesions impair brain function, and in particular impair thinking ability and walking.
White matter lesions correlate strongly with reduced gait speed as well as reduced mental ability (Starr et al, 2003; Guttman et al, 2000; Whitman et al, 2001; Bazner et al, 2008). Periventricular location of white matter lesions seem to cause the most serious consequences. Here we are mainly talking about the larger white matter lesions - -the ones that are "confluent", or grades 2-3 on the graph above. Persons with one or two "punctate" white matter lesions are not thought to have these problems. There are also patient groups that don't really fall that well into the grades -- perhaps they have 40 or so punctate lesions. They probably belong with the grade 2 (partially confluent) group. We think there might be room to develop a different criterion - -perhaps the % of white matter that is occupied by lesions.
Individuals with PVM lesions perform nearly 1 standard deviation(SD) below average on tasks involving psychomotor speed. Persons with severe periventricular WVL perform about 1 standard deviation lower than average subjects on tasks involving psychomotor speed, and about 0.5 SD lower on global cognitive function (De Groot et al, 2000). To put this into more familiar terms, on the IQ test, 1 SD is 15 points. Even "silent" white matter lesions in middle-aged hypertensive patients predict reduced attention (Sierra et al, 2014). Severe white matter lesions predict poor activities of daily living (Yamashita et al, 2016).
Lacunes (i.e. holes in the brain) are also associated with cognitive disturbances (Jokinen et al, 2011).
Deep white matter lesions are reportedly even more burdensome than periventricular white matter lesions to cognition (Soriano-Raya et al, 2012). Fortunately these are relatively uncommon. These are more correlated with visceral obesity than with hypertension (Lampe et al, 2018).
According to Degroot et al (2002), "After adjusting for age, gender, educational level, measures of depression, and brain atrophy and infarcts, subjects with severe periventricular white matter lesions experienced cognitive decline nearly three times as fast" as the average." This is not surprising as persons with severe PVM have experienced a lot of brain damage.
Acceleration in white matter hyperintensity burden, is a pathologic change that occurs early in the presymptomatic phase leading to mild cognitive impairment. In fact, on average, acceleration occurs 10 years prior to onset of mild cognitive impairment. (Silbert et al, 2012). Only a 3% burden of leukoariosis (white matter lesion percent of brain) is enough to reduce working memory scores 2 standard deviations (Price et al, 2012). A 3% burden on MRI, is the amount of PVM usually called "mild" on MRI reports.
If we consider gait and balance, more white matter lesions predicts decreased mobility (Onen et al, 204) . According to a recent study of Bazner, the slow-down in walking is not extreme -- older people are about 20% slower with a large # of PVM than a small #. It would seem to us that people may simply be more cautious and walk more slowly when they are more prone to fall. (2008)
Practically, PVM seem to be associated with severe consequences. As once you have them, they are there for life, prevention is the main goal of treatment.
Major vascular risk factors for white matter disease include:
- type-2 diabetes
- obesity, particularly visceral.
We advocate attention to reducing vascular risk factors, and especially controlling both static (constant) as well as labile (i.e. rapidly changing) hypertension. "White coat" hypertension is a type of labile hypertension. Thus if your blood pressure is "only up at the doctor's office", this does not mean that you are safe from PVM. A beta blocker might help. Intensive blood pressure control (defined as "targeting" a BP < 120) was associated with a smaller increase in white matter lesions in 449 hypertensive patients over the age of 50 (Sprint group, 2019). This seems like a generally good idea, although not applicable to everyone.
Strict control of diabetes is probably helpful. This means getting one's Hemoglobin A1C to normal -- lets say, about 6.5 or below.
Reducing calorie intake until the BMI becomes normal is prudent. Physical fitness is weakly associated with better white matter (Sexton et al, 2016).
Reducing elevated cholesterol is probably helpful too.
Other things that you can do, without as much evidence for them include:
Small amounts of vitamin b6 (pyridoxine) supplementation are reasonable (i.e. 2 mg/day).
Cognitively complex leisure activity (i.e. doing something difficult in your spare time) is associated with better mental performance in persons with white matter lesions (Saccynski et al, 2008). It is certainly better for your brain to be physically active, and probably better to also keep your brain active.
It is not clear if daily aspirin intake is useful or harmful in persons with PVM, and in our opinion, the decision should be made on an individual basis. Persons with bleeds, probably should not be on aspirin. In other words, an MRI is needed to make this decision. Beta blockers such as propranolol and related medications may be especially suitable to prevention of spikes in blood pressure. These drugs are also cardioprotective. In other words, if someone has "white coat hypertension", a beta blocker might be worth considering. More data would be helpful here, as this is simply speculation.
In persons with migraine and a large load of white matter lesions, we generally suggest a prophylactic regimen such as a combination of low-dose aspirin and a migraine prevention agent such as verapamil. It is not known whether this treatment regimen is effective. Beta blockers would also seem very reasonable for this situation.
Treatment of the demyelinating diseases such as multiple sclerosis may be suitable, but is outside the scope of this brief review.
With respect to the common symptom of unsteadiness, empirical treatment including physical therapy. Physical therapy however is not going to make these things go away.
While generally medications are not helpful in situations where fiber tracts or neurons have died (i.e. PVM), in some occasional cases, trials of medications such as antidepressants or anti-parkinsonian drugs are helpful. Again, medications are not going to improve your MRI or get rid of damage already done.
We think it is unlikely that any treatment will be successful in reversing the mental slowing associated with PVM. However, there may be some room to readjust medications that are slowing down thinking independently -- such as antivertigo medications (e.g. meclizine or clonazepam), and replacing them with medications that do not impair thinking (such as betahistine).