Digital Cognitive Testing Uncovers Hidden Impairment in MS

Caroline Helwick

February 09, 2018

SAN DIEGO — In persons with multiple sclerosis (MS), digital cognitive testing can detect impairment across a range of domains, gauge the severity of impairment, monitor changes with treatment, and track the impact of disease according to economically important milestones that are critical to patients, researchers say.

The computerized neuropsychological screen (NeuroTrax)  can provide what they view as more "patient-centric" cognitive assessment.

Senior investigator, Mark Gudesblatt, MD, medical director of South Shore Neurologic Associates, PC, in Patchogue, New York, spearheaded the research that resulted in several late-breaking abstracts here at Americas Committee for Treatment and Research in Multiple Sclerosis (ACTRIMS) 2018. The studies were independent and not funded by NeuroTrax.  

NeuroTrax produces a standardized computerized neuropsychological battery of tests that measures the seven domains of memory, executive function, attention, visual spatial, verbal function, problem solving, and working memory. The investigators maintain that it provides much more "granular" information than conventional tests, including the Symbol Digit Modalities Test (SDMT), which assesses a single domain of cognitive impairment (processing speed).

"The use of digital analytics takes the guesswork out of documenting disease impact and disease progression, and when to change therapy and which treatment might be appropriate," Dr Gudesblatt told Medscape Medical News.  

The Expanded Disability Status Scale (EDSS), MRI, and relapse rate are used to measure disease progression, but these techniques cannot detect cognitive impairment and decline, according to Dr Gudesblatt. While comprehensive neuropsychological evaluation is the gold standard for tracking cognitive function, it is not available for many patients because of expense and limited resources. SDMT, which screens for cognitive dysfunction, is not commonly used in routine care and provides limited information.

"Traditional instruments are insensitive and nonlinear, and patient perceptions of their disease can only give you their perceptions," he commented. "You can't buy a pair of pants with just one measurement. Why do we think we can track disability with one test, such as the SDMT, EDSS, or 25-foot timed walk?"

"If you ask 300 patients whether they have mild, moderate, severe, or no cognitive impairment, their answers do not track with any cognitive domain score at all, and when you ask their physicians, it's 50%, a coin toss," Dr Gudesblatt said in an interview with Medscape Medical News.

"Computerized screening can provide much more information regarding cognitive deficit and can be easily incorporated into routine care."

He said digital testing adds only about $100 to the patient visit and can be used to screen for patients who might benefit from additional neuropsychological testing or intervention.

Identifying Deficits

One study presented at the ACTRIMS meeting evaluated the ability of the computerized screen to identify deficits across a range of cognitive domains. The study included 80 persons with MS, stratified by severity of cognitive impairment as determined by SDMT (none, mild, moderate, severe), and 15 healthy controls.

The study found statistically significant differences between controls and individuals with MS on almost all traditional measures, with controls outperforming the MS cohort. Across a wide range of NeuroTrax domains, there were robust, statistically significant differences in global cognition (P < .001), memory (P < .001), executive function (P < .002), attention (P < .001), processing speed (P < .016), and visual spatial (P < .001) and motor (P < .007) parameters.

Per logistic regression, the screening tool discriminated between the MS group and controls with a correct classification rate of 81%. Persons with MS who had higher degrees of cognitive impairment demonstrated lower scores on the instrument's global composite measure. The authors concluded that the screening tool provides information "that is unique and not provided by SDMT, EDSS, relapse rate, or MRI findings."

Digital cognitive testing can predict a patient's likely disease trajectory beyond EDSS/MRI or relapse, particularly for "patient-centric" consequences that may be "unseen" and may relate to the loss of important milestones, the researchers said.

They showed this in a retrospective cross-sectional review of a prospective database of patients from the South Shore practice. Patients with EDSS score less than< 6 underwent standardized computerized cognitive testing via NeuroTrax to evaluate the relationship of cognitive impairment, as determined by number of cognitive domains impaired (CDI) more than one standard deviation from age/education norms.

Certain domains are known to be related to loss of employment, loss of driving, and increased fall risk, and these were correlated with number of CDIs per patient. "These, not EDSS step, are the disabilities that matter to patients," he commented. 

The study assessed 543 patients for employment-related domains, 159 for fall risk, and 115 for driving status. The analysis showed that increasing accumulation of CDIs was associated with a progressive loss of employment, loss of driving, and fear of falling. 

Table. Proportion of Patients With Patients Who Have "Milestones" by CDI Domains

Milestone CDI-0 CDI-1 CDI-2 CDI-3
Likely still employed 61 50 43 33
Likely still driving 100 66 53 21
Likely free of falls with simple daily activities 77 66 36 39
Likely free of falls with complex daily activities 72 58 36 33

 Describing how the test captures "hidden" disabilities, he noted that "Within homologous groups of physical disability, fall risk should be the same, but it is not. When cognitive impact is tied in to physical disability, those with the same physical disability but increased cognitive impairment have greater risks of falling."

Correlation With Changes in Brain Atrophy

The researchers were also able to link brain structure and cognitive function with this screening tool in another study of 59 patients, showing that cognitive subdomains and number of CDIs were both strongly correlated with degree of brain atrophy seen in specific brain regions.

However, among the multiple significant correlations that were found between the NeuroTrax scores and regional atrophy measurements, the only relationship that remained consistently significant in the longitudinal analysis was the correlation between the change in number of CDIs and change in right thalamic atrophy volume (P = .009).

Describing the thalamus as the "network router hub," Dr Gudesblatt said this finding makes sense. "Change in structure has to do more with change or progressive loss of functional network connectivity, or network efficiency, which translates to loss of function," he explained.

With digital cognitive testing, he said, "hidden burden of disease can now be identified…and this allows us to understand the 'disconnect' when patients say they are worse, but the MRI indicates no 'apparent or significant visible change.'"

Their next step is to use digital brain atrophy software to track "granular" change within each cognitive domain over time. "Maybe additional relationships will arise," he said.

Finally, digital cognitive testing documented significant improvements in the global cognitive score (GCS) after 2 years of treatment with natalizumab, in a study of 52 patients adjusted for baseline EDSS score and age. GCS and cognitive domain scores were numerically higher at 1 year than at baseline. Changes from baseline to 2 years included a significant improvement in GCS (P = .003), memory (P = .006), visual-spatial processing (P = .025), attention (P = .017), and information processing speed (P = .002). 

The percentage of patients exhibiting a clinically significant improvement in cognition score from baseline (ie, more than one standard deviation) was 21.6% at 1 year and 32.7% at 2 years, Dr Gudesblatt reported.

The test offers clinicians a more reliable view of treatment efficacy and can be informative in deciding whether to switch drugs, he said.

Utility Lies With the User

"Computerized neuropsychological measures are not inherently better or worse than traditional cognitive tests that are administered by a trained examiner or neuropsychologist. While some aspects of cognition are harder to assess via a computerized test (eg, expressive language), others are easier (eg, measures of reaction time)," according to Brian K. Lebowitz, PhD, director of neuropsychology training and associate professor of clinical neurology at Stony Brook University Medical Center in New York.

Dr Lebowitz commented that as long as computerized measures are developed with the same scientific rigor as traditional neuropsychological tests (ie, using well-established normative comparative groups and demonstrating good reliability and validity), "these tools will continue to grow in popularity because of costs savings, ease of use, and reduced training requirements for administrators."

 "As with any test, the strength of the instrument ultimately lies in the expertise of the healthcare professional who interprets the clinical meaningfulness of the test scores," he said.

Certain validation studies were supported by Biogen and Sanofi. The natalizumab study was supported by Biogen. Dr Gudesblatt has disclosed no relevant financial relationships

Americas Committee for Treatment and Research in Multiple Sclerosis (ACTRIMS) 2018. Abstracts LB275, LB247, LB263, and P162. Presented February 2, 2018.

For more Medscape Neurology news, join us on Facebook and Twitter


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.