Clinical evaluation

Individuals with cognitive impairment related to white matter involvement come to clinical evaluation via one of three primary routes: (1) a white matter disorder is known, in most cases supported by a brain neuroimaging study, usually magnetic resonance imaging (MRI); (2) white matter abnormalities of potential significance have been disclosed by MRI or another neuroimaging study; or (3) the medical history suggests the possibility of brain white matter neuropathology. In all of these scenarios, the time-honored methods of careful history taking and thorough physical examination are crucial, with special emphasis on mental status testing.

A diagnosed white matter disorder may pose little or no challenge if the cognitive impairment is obvious and no other disorder is likely to explain the problem. A young adult chronic toluene abuser with no other notable medical history who has dementia associated with severe, diffuse MRI white matter hyperintensity serves an illustrative example. Similarly, a genetically proven case of MLD in an adult who first developed psychosis and then dementia as confluent MRI white matter hyperintensity advanced over several years poses no diagnostic difficulty. Far more often, however, the patient has a complex clinical picture that may include a disease that also affects gray matter, such as MS, or one that prominently affects other organs outside the nervous system, such as systemic lupus erythematosus (SLE). In these cases, clinical judgment must be applied to ascertain the extent to which white matter involvement explains cognitive impairment.

More difficult is the situation in which white matter abnormalities have been identified by neuroimaging and cognitive impairment is an issue, but no clear diagnosis has been made. Many patients seen in MS clinics, for example, have minor degrees of nonspecific white matter hyperintense lesions that are not clearly related to demyelination. In these cases, clinical assessment should first be applied to determine the primary diagnosis so that any cognitive disturbances can be properly interpreted. If a white matter disorder is diagnosed, the appropriateness of diagnosing WMD or MCD can be determined, and if no such disorder is found, the origin of cognitive symptoms can be sought elsewhere.

The clinical history may be highly suggestive or even diagnostic. A patient with a known leukodystrophy, MS, human immunodeficiency virus (HIV) infection, inhalant abuse, or cerebrovascular disease who develops progressive cognitive decline in the absence of other etiologies of dementia may well qualify for the descriptor WMD or MCD. In a patient with unexplained MRI white matter abnormalities and cognitive decline, the medical, psychiatric, family, and social history may all be relevant and can help guide the diagnostic evaluation. White matter involvement may produce a picture of rapidly progressive dementia, prompting an urgent evaluation, as in some cases of infectious disease (Jones et al., 1988), inflammatory disease (Kirk, Kertesz, and Polk, 1991), demyelination (Hardy and Chataway, 2013), and neoplasia (Filley et al., 2003; Rollins et al., 2005; Deutsch and Mendez, 2015).

Also challenging is the patient with symptoms that could suggest the presence of a white matter disorder affecting cognition. These symptoms include a host of complaints, ranging from fatigue, lassitude, somnolence, and apathy to inattention, multitasking difficulty, impaired word finding, and memory loss. As MRI is a rather costly procedure, and these many diverse symptoms can be due to a great variety of medical, neurologic, and psychiatric disorders that do not implicate brain white matter, clinical evaluation is important as the initial step to find alternative explanations. But MRI is at times an appropriate diagnostic test to consider.

Before turning to neuroimaging, other information from the history and elemental neurologic examination may prove helpful. Seizures and movement disorders are relatively uncommon in white matter disorders affecting cognition because these phenomena reflect pathology in cortical and deep gray matter, respectively (Filley, 2012). In contrast, other signs can be quite helpful. Corticospinal dysfunction often indicates tract damage and can be useful for distinguishing a problem such as BD, MS, HIV-associated dementia, cobalamin deficiency, and normal pressure hydrocephalus (NPH) from cortical disorders such as Alzheimer’s Disease (AD), in which cognitive impairment is not accompanied by motor dysfunction until late in the course. Gait disorder is common in white matter disorders, and urinary incontinence may also occur; both reflect disruption of frontal lobe tracts. Gait disorder in older people has been repeatedly correlated with frontal lobe white matter dysfunction as measured by MRI and diffusion tensor imaging (DTI) (Srikanth et al., 2010; de Laat et al., 2011; Annweiler and Montero-Odasso, 2012; Callisaya et al., 2013; Smith et al., 2015), and whereas this problem is sometimes referred to as “lower body parkinsonism,” a more appropriate term may be “cerebrovascular gait disorder“ (Rektor, Rektorová, and Kubová, 2006). Urinary incontinence can also occur with frontal white matter dysfunction of various etiologies (Vigliani et al., 1999; Graff-Radford, 2007; Sakakibara et al., 2012). Incontinence is of course a defining clinical feature of NPH, and can also be a result of cerebrovascular white matter lesions interfering with the normal inhibitory function of medial frontal lobe structures controlling micturition (Andrew and Nathan, 1964; Sakakibara et al., 2012). Recent neuro-urological studies have in fact shown that overactivity of the bladder detrusor muscle in older people, popularly known as overactive bladder, is more related to white matter lesions than to the pathology of AD (Sakakibara et al., 2014).

The mental status examination is a key component of the evaluation in patients suspected of WMD or MCD. This examination, always a time-intensive process requiring considerable sensitivity to clinical subtleties, can be particularly challenging in the diagnosis of cognitive decline related to white matter involvement, because language is normal or nearly so in these patients, and deficits are more apparent in cognitive speed, executive function, and attention. Whereas the relative subtlety of these deficits means that the cognitive deficits of many impaired patients may remain undetected as other sensorimotor features of the illness often dominate the clinical encounter, a detailed mental status examination can provide the critical information (Arciniegas, 2013), and neuropsychological testing can be equally helpful (Cullum, 2013). Detailed neurobehavioral or neuropsychological evaluation may not always be feasible, however, so brief clinical measures may be called upon. The popular Mini-Mental State Examination (MMSE; Folstein, Folstein, and McHugh, 1975) is heavily weighted toward language, and relatively insensitive to the executive dysfunction of patients with white matter disorders (Franklin et al., 1988; Swirsky-Sacchetti et al., 1992; Román and Royall, 1999; Xu et al., 2014). Other, more useful tests include the Montreal Cognitive Assessment (MoCA; Nasreddine et al., 2005), the Frontal Assessment Battery (FAB; Dubois et al., 2000), and the Clock Drawing Test (CDT; Cosentino et al., 2004). The MoCA (Griebe et al., 2011; Xu et al., 2014), the FAB (Kanno et al., 2011), and the CDT (Kim et al., 2009) have all been found sensitive to white matter dysfunction in various disorders. The MoCA may be the most convenient of these measures as it incorporates executive function and clock drawing tasks into a 30-point format; another advantage is that it enables the testing of memory retrieval, another clinical feature central to the assessment of WMD and MCD.

Laboratory testing

As in any dementia evaluation, the search for reversible causes includes laboratory testing of blood and sometimes other tissues (Miller and Boeve, 2009). A comprehensive metabolic panel, complete blood count, thyroid-stimulating hormone (TSH), and vitamin B₁₂ level are routinely obtained, and in selected cases, rapid protein reagin (RPR), HIV, Lyme disease serology, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), antinuclear antibody (ANA), urinary toxicology screening, and genetic testing for CADASIL and leukodystrophies can be considered. All of these tests can detect various white matter disorders, including the TSH, elevation of which is diagnostic of hypothyroidism; although not widely appreciated, thyroid hormones contribute to normal brain myelination, and recent DTI studies of hypothyroid patients have found a correlation between impaired memory and microstructural white matter damage (Singh et al., 2014). Lumbar puncture is indicated for dementia that is rapidly progressive (Paterson, Takada, and Geschwind, 2012), and white matter involvement is sometimes found to be responsible, as many infectious, inflammatory, demyelinative, and neoplastic diseases can be detected that feature an abrupt onset and fulminant clinical course. In many cases, problems such as hypothyroidism, B₁₂ deficiency, and HIV infection will have been detected by primary care evaluation, but any potentially reversible cause of white matter dysfunction should always be kept in mind. In contrast, a test helpful for excluding white matter involvement is electroencephalography (EEG), which can support the diagnosis of a seizure disorder or nonconvulsive status epilepticus; in selected cases, the use of continuous EEG in dedicated epilepsy monitoring units can be required (Maganti and Rutecki, 2013). Finally, in recognition of the increasingly apparent role of sleep disorders in producing cognitive complaints, evaluation with polysomnography may be beneficial (Colrain, 2011).

Neuroimaging

MRI is clearly the procedure of choice in this context, and, for white matter disorders, computed tomography (CT) is inadequate. As discussed in Chapter 2, the advent of MRI both revolutionized general neurology and inaugurated the detailed study of the behavioral neurology of white matter without the necessity of neuropathological correlation. The utility of MRI in the diagnosis of dementing disorders has been well documented, and many of the disorders identified feature prominent white matter abnormalities (Pantano, Caramia, and Pierallini, 1999). Although gray matter of the cortex and subcortical regions is often affected, complicating the relationship between clinical phenomenology and neuropathology, MRI allows detailed views of white matter lesions that can be generally correlated with cognitive deficits. As in any cognitive evaluation that includes neuroimaging, careful inspection of the MRI by the behavioral neurologist can be most informative, often adding helpful interpretation to the reading provided by the radiologist (Miller and Boeve, 2009). The primary data used to develop the constructs of WMD and MCD are in fact derived from clinical research showing that brain MRI white matter lesions correlate with cognitive deficits.

Whereas MRI has been essential to progress in the dementias, it is clear that the sensitivity of the technique to brain pathology may not be matched by its specificity. That is, MRI can detect a multitude of lesions, especially in white matter, but it is often unclear exactly what those lesions may be. A white matter lesion in the periventricular region of an adult, for example, may be determined by genetic, demyelinative, infectious, inflammatory, toxic, metabolic, ischemic, traumatic, neoplastic, or hydrocephalic pathology, and the clinician may be hard-pressed to settle on a single etiology. A related point is that whereas MRI can be diagnostic of a white matter disorder such as MS, and can even establish the basis of cognitive impairment in many cases, it can also disclose nonspecific white matter abnormalities that have no impact on the patient’s cognition. Thus MRI can be considered very sensitive to white matter lesions, but the specificity of these findings is a topic better left to the clinician who has a more complete understanding of the clinical picture. This discrepancy between the sensitivity and specificity of MRI is a theme that appears once again in the next section, albeit in a different context.

The newer techniques of DTI, magnetization transfer imaging (MTI), and magnetic resonance spectroscopy (MRS) remain research tools at this point. All require extra time for patients in the scanner, and often significant postacquisition time for data analysis, and the cost of these requirements cannot be justified without clear evidence of patient benefit. It is to be hoped, however, that some clinical utility of these techniques, which after all are all noninvasive and without important risk, will be realized as technology advances and the need for assessing the integrity of white matter tracts for patient care is appreciated. Not only are there likely to be advances in the identification and characterization of tracts seen to be damaged on conventional MRI, but the presumably vast terra incognita of lesions within the normal-appearing white matter (NAWM) will become accessible to the eagerly curious eyes of behavioral neurologists.

Neuropsychology

The field of neuropsychology, a close companion of behavioral neurology, has been essential to the establishment of white matter–cognition relationships (Heaton et al., 1985; Filley and Cullum, 1994; Harris and Filley, 2001; Lafosse et al., 2007; Kozora and Filley, 2011; Kozora et al., 2013; Grigsby, 2014). Combined with MRI and other evolving technologies, neuropsychological assessment has been invaluable in documenting consistent patterns of cognitive impairment in many disorders that inform the concepts of WMD and MCD (Filley, 2012; Cullum, 2013).

In the clinic, individuals with white matter disorders who have had neuropsychological testing can often be readily understood to have WMD or MCD because brain MRI is available. However, more difficulty arises when neuropsychological evaluation suggests a white matter disorder but no such diagnosis has been made and brain MRI has not been obtained. In these cases, it should be remembered that, like MRI, neuropsychological testing is in general more sensitive than it is specific, and that cognitive deficits consistent with white matter neuropathology can be equally indicative of subcortical gray matter involvement or a wide range of psychiatric disorders such as depression and anxiety. A careful clinical approach using the available data will be most useful in interpreting the findings of neuropsychological assessment.

The profile of deficits in WMD has been discussed in Chapter 7, and remains the basis for suspecting the development of this syndrome from a neuropsychological perspective. Accordingly, MCD would manifest as a similar profile but with less severe impairments, prominently featuring deficits in processing speed and working memory, and the important clinical datum that usual social and occupational function is not compromised. This assessment is crucial in many cases, and can help distinguish a white matter disorder from others impacting cognition.

Neuropsychological testing can take many forms depending on the professional orientation of the examiner and the specific measures selected, and the intent here is not to propose an invariant assessment method or battery of tests. One perspective that is particularly useful in white matter disorders, however, is the Boston Process Approach, a strong academic tradition in neuropsychology championed by Edith Kaplan during her many years at the Boston Veterans Administration Medical Center in the latter half of the twentieth century (Libon et al., 2013). This approach emphasizes the process by which a patient carries out a given task rather than considering only the test score obtained. Test scores in comparison to normative data are considered, but close attention is also paid to the analysis of errors that emerge as a patient undertakes a given task (Libon et al., 2013). The philosophy motivating this approach is one that considers the reason for test failure as much as the failure itself. Given the subtlety of cognitive impairment that can occur in WMD, which is still more evident in MCD, the Boston Process Approach readily lends itself to the assessment of these problems.

An illustrative example of the advantage of the Boston Process Approach can be found in the California Verbal Learning Test (CVLT), a widely used neuropsychological assessment instrument introduced by Kaplan and her colleagues (Delis et al., 1987). The CVLT provides a detailed analysis of memory performance by using various measures that permit the disambiguation of various factors that underlie memory failure. By focusing on the process by which a patient takes on a memory task, the reason for an impaired performance becomes clearer. Of particular importance for the evaluation of patients with suspected WMD and MCD, the CVLT specifically tests for recognition memory, which, if preserved, supports the criterion of retrieval deficit that is included in the clinical profile of white matter disorders affecting cognition (Chapter 7). The Boston Process Approach can also be applied to other domains implicated in WMD and MCD, such as processing speed, working memory, and executive function.

The centrality of slowed processing speed in the disorders of white matter deserves special emphasis. Many neurobehavioral disorders can manifest cognitive slowing, but a specific relationship of this deficit with white matter damage is being increasingly well established (Chapter 7). One of the most useful measures for quantitating this problem is the Paced Auditory Serial Addition Test (PASAT), first developed for use in traumatic brain injury (Gronwall and Wrightson, 1981) and now widely applicable to white matter disorders (Benedict et al., 2008; Kozora et al., 2013). The PASAT can be quite frustrating for patients at times, suggesting the need for alternative means of quantitating slowed cognition, but assessment of this aspect of white matter dysfunction is critical.

One of the areas in which neuropsychological testing for WMD and MCD could be expanded is in the evaluation of procedural learning and memory. As reviewed in Chapter 7, this domain is one of the key areas in which the distinction between WMD and subcortical gray matter dementia can be clarified. Acquisition of motor skills is typically only mildly impaired or even unaffected in patients with white matter disorders (Lafosse et al., 2007), in contrast to those with lesions of the basal ganglia (Martone et al., 1984) and the cerebellum (Sanes, Dimitrov, and Hallett, 1990). Thus the preservation of procedural learning and memory would be useful to document, since this testing can help differentiate patients with white matter disorders from other patients expected to manifest deficits, such as those with Parkinson’s Disease and cerebellar disorders (Yamadori et al., 1996). Yet the acquisition of motor skills is not typically tested in clinical evaluations, and this aspect of the cognitive profile is usually not available. Procedural learning and memory are more readily tested in experimental settings with measures such as rotary pursuit and mirror drawing (Gabrieli et al., 1997; Lafosse et al., 2007), but these methods, however useful, are difficult to apply in the clinic and may yield equivocal results. The development of other, more convenient measures for the assessment of procedural learning and memory would be welcome.