A variety of different types of biomarkers are present in individuals over the course of their disease process. The emerging research that informs clinical understanding and application of these biomarkers for diagnosis and prognosis of disease, as well as more personalized treatment regimens, is exciting. However, the connections between currently validated biomarkers and the progression of these diseases and conditions to a level of disability that prevents an individual from engaging in substantial gainful activity as understood under the Social Security law remains tenuous. This chapter summarizes the positive findings and potential uses of biomarkers, as well as their challenges from several perspectives. Finally, it offers considerations to SSA for incorporating biomarkers in its process for determination of disability.
In reaction to the discussions throughout the workshop, a panel made up of several members of the workshop planning committee shared the key points they heard in terms of the state of the science and the possible challenges that lay ahead as biomarkers continue to be developed. The panel included Betty Diamond, director of the Institute of Molecular Medicine at the Feinstein Institutes for Medical Research; Sarah Morris, chief of the Adult Psychopathology and Psychosocial Intervention Research Branch at the National Institutes of Mental Health; Ralph Nitkin, deputy director for the National Center for Medical Rehabilitation Research at NIH’s National Institute of Child Health and Human Development; and Ira Shoulson, professor of neurology, pharmacology, and medicine at the University of Rochester School of Medicine.
Potential Use of Biomarkers
Various types of biomarkers were highlighted throughout the workshop, including molecular, chemical, and systemic—showing the breadth of potential biomarkers that researchers can study when looking for indicators to assist in more accurate diagnostic, prognostic, or treatment determinations. Shoulson commented that the whole field of biomarkers has immediate and long-term application to disease—both related to diagnosis and prognosis. Prognostic biomarkers, especially, could be very valuable in determining the outlook for a person, he said.
Nitkin noted the importance of a functional goal for biomarkers, saying biomarkers need to be useful, whether it is to do things more
quickly or in different environments (e.g., at point-of-care instead of limited to certain lab capacity), or to offer more predictive or precise care for patients. He acknowledged “disability rehabilitation is a tricky field” because it is contextually and environmentally dependent. Morris emphasized the importance of research that capitalizes on big data—such as data already available in electronic medical records that do not require a lot of new expertise or equipment.
Diamond called attention to the potential for being wrong, asking what are the risks to individuals or to society? This comment also echoed Peterson’s previous remarks about the need to grapple with the risk of false positives and false negatives and understand what the risk of being wrong with biomarker assessment will be in a clinical context or in a disability application. To help understand the risks of biomarkers, Diamond agreed with Menetski, saying systemic redefinition of terms may be needed. For example, is SSA trying to identify people who can or cannot work, or just determine what services and supports are needed to make that person a more productive member of society, she asked.
Morris reiterated the disconnect between approaches to research and the challenge of making a disability determination that Rosenbaum raised, which exists because most research studies are focused on comparing a group of patients with a group of comparison subjects. Studies also may sometimes identify subgroups within a sample of participants and discover relationships between brain and behavior and functioning. However, Morris said, the gaps in research become obvious when an agency, like SSA, asks what the cutoff is for certain markers. She said the current research findings
do not necessarily translate nicely into informing a dichotomous disability determination that has to be made on the basis of one individual tested in isolation, compared to the context of a whole group of research participants, which is where the findings originated.
She noted the burden would be on SSA to weigh what acceptable rates of false positives and false negatives would be. To address the research side, Morris called for better standardized tests and measures that can be used and interpreted for one individual in each site. Finally, Shoulson pointed out that most of the research presented was informative, but it currently seems emerging at best, and thus it is a long way from clinical application.
As biomarkers are validated, Nitkin said, they are typically based on retrospective studies and populations that are often representing the majority and are easily accessible. But the results of those studies should not be applied to the generalized population, he cautioned. Lack of diversity in studies could inadvertently cause health disparities by using biomarkers too broadly, he said.
Shoulson noted that what patients report (e.g., patient-reported outcomes) of their symptoms alone accounts for approximately 80 percent of accurate diagnoses—as agreed upon by experienced, expert clinicians. So, this “low-hanging fruit” offers an opportunity, he added. Using tools and technology such as machine learning, he said, and by listening to patients and pairing their subjective reports with objective clinical information can help clinicians further determine a patient’s disease progress. Shoulson also thought it would be useful if biomarkers related highly to function. He suggested exploring and examining this connection more closely—especially in relation to gainful activity—to understand how to make a better determination for people using that lens.