Latest News

Contributed Commentary By Mylea Charvat

October 18, 2017 | As a clinical psychologist who specialized in neuro assessment, I’ve seen many patients with dementia. Dementia is a disease of progressive global cognitive impairment, and is characterized by progressive loss of both cognitive function and the ability to perform activities of daily living. It is a cruel and unrelenting illness, and its underlying pathology is usually degenerative.

Given the lack of success in both disease-modifying and symptomatic agents for Alzheimer’s disease (AD), some experts have noted the study endpoints need to be more sensitive. Axovant’s recent failure in the space is another reminder of how difficult it has been for drug development in this sector. Many large pharma companies have also failed, including Merck’s verubecestat, Pfizer/Johnson & Johnson’s bapineuzumab, and Myriad Genetics’ tarenflurbil, among others.

Mental status tests are given by healthcare professionals to assess their patients' cognitive function. Although most frequently given by physicians, testing can also be done by nurse practitioners, physician assistants, and nurses. The tests are designed to be brief and paint a general picture of the patient's mental status, or brain health. Assessing specific neurocognitive disorders requires more complex testing and is usually done by a psychologist.

In Alzheimer’s clinical trials, one of the best known and widely-used tests for patient study inclusion is the Mini-Mental State Exam or MMSE. It was created by Johns Hopkins psychiatrists Marshal F. Folstein and Susan E. Folstein, MDs in 1975. The journal article that first described this mental status test (written with Paul McHugh) has been cited in over 25,000 published articles, double the next most cited paper in psychiatry and neurology. The test quickly caught on as a tool to screen for dementia and to track its course over time.

The maximum MMSE score is 30 points. A score of 20 to 24 suggests mild dementia, 13 to 20 suggests moderate dementia, and less than 12 indicates severe dementia. On average, the MMSE score of a person with Alzheimer's declines about two to four points each year.

There is no question that the MMSE was a breakthrough in mental status testing during its time. It is easy to administer and score, takes about 10 minutes to administer, and is available in multiple languages. Its appeal to healthcare professionals since its inception, and especially in today's era of managed care, is understandable.

My main problem with the MMSE is that it is not a valid test of cognitive function.

The MMSE was developed to distinguish cognitive from psychiatric problems and to document improvement or decline over time by trained professionals. Cognition is the ability of a person to assimilate and process the information that he/she receives from different sources and convert them into knowledge. The MMSE was not designed to test how a person processes information. And it is not sensitive to small cognitive changes.

Another problem with the MMSE is that over half of the points scored are devoted to memory and recall. Retrieval from memory is important, but assessing cognitive abilities is more than just recall. Additionally, the MMSE gives no information on performance in specific cognitive domains like working memory, executive function, and impulse control. And you can't give the test to people with poor language skills or education.

I am not alone in my concern. In a study published in 1989 in the Cleveland Clinic Journal of Medicine the authors said that the MMSE overestimated impairments in persons over age 60 and in persons with less than nine years of education. They also noted that "The MMS scale has been reported to be insensitive to cognitive impairments resulting from right hemisphere dysfunction as well as milder forms of cognitive dysfunction irrespective of cortical origin."

In an article in the Textbook of Clinical Neurology in 2007, the authors wrote:

"Although the MMSE is the most commonly used mental status test, it has several problems. First, one must accept the basic limitation of any screening examination. Since the test has a limited number of questions, adequate testing of cognitive function is not possible. Practically, however, the MMSE score is used as an indicator of intact or impaired performance. The cutoff score of 24 is associated with relatively high false-negative rates (test indicates absence of impairment when impairment is present) and relatively high false-positive rates (test indicates impairment when no impairment is present). Additionally, MMSE score is affected by education, race, and gender. Despite these limitations, the MMSE provides a “quick and dirty” assessment of overall cognitive function and in most contexts a score of 22 or lower is considered an accurate mark of clinically significant cognitive impairment."

Another review of the test noted that "structured, screening questionnaires are still subject to interpretive bias and depend on the skill and experience of the interviewer. All screening questionnaires have a fairly significant false-negative rate, especially in patients with focal lesions of the right hemisphere. Age (especially > 60 years), education (< 9th grade), cultural experience, and low socioeconomic standing limit the usefulness of MSE screening questionnaires. Unlike a detailed mental status exam, screening questionnaires are less too sensitive to subtle cognitive impairment."

Cognition should be a key vital sign -- and is important to measure on a regular basis as blood pressure, temperature, respiration, and pulse. One important aspect to assess in future studies is whether conversation from MCI to dementia could be better predicted by MMSE changes over time using multiple cognitive tests and screenings during the actual clinical trial instead of single measurements.

We need to think about novel endpoints for cognition - going outside of just capturing traditional measures such as accuracy and reaction times, but using digital tools that may be more sensitive by capturing movement and psychomotor changes earlier on before and during disease progression.

Digital tools can be used to identify at risk patients for Alzheimer’s disease earlier, instead of using current pen and paper neuropsychologist testing - but even that has massive limitations as we are still capturing the same old datasets - accuracy and reaction times. Maybe all these clinical trial failures in Alzheimer’s disease is telling us is that we are not sensitive enough in measuring cognition – or broadly enough. We need to do better with our current testing if we want to help improve the lives of patients with Alzheimer’s one day.