Diagnosing Dementing Disease: A Neurologist’s QI Wish List
Norman FosterImaging can play a valuable role in the early diagnosis of dementing disease. In a special-interest session held on November 26, 2012, at the annual meeting of the RSNA in Chicago, Illinois, neurologist Norman Foster, MD, copresented “What Imaging Measurements Are Needed in Clinical Practice?” He explains how imaging can increase physician confidence and accuracy in the diagnosis and prognosis of dementing disease through identification of biomarkers on imaging studies. According to Foster—professor of neurology at the University of Utah, senior investigator at the university’s Brain Institute, and director of its Center for Alzheimer’s Care, Imaging and Research—early identification of Alzheimer disease and other dementing diseases is critical in preparing patients and their families for dramatic lifestyle changes. Accurate prognosis will enable patients to make the necessary modifications to their lifestyles. The problem with dementing disease, Foster says, is that primary-care physicians are reluctant to address memory issues with their patients and lack confidence in their ability to make accurate diagnoses. Uncertain diagnoses and inconclusive evaluations can lead to frustration, inconsistent treatment, and continued confusion for both patient and physician. Brain imaging has been an accepted part of diagnosis since 2001, and the American Academy of Neurology’s practice parameters include structural neuroimaging as part of the routine for patients with dementing disease, Foster notes. There are now validated, quantitative imaging biomarkers for dementing disease, and it is possible to predict outcomes for patients with mild cognitive impairment. Three Stages The US National Institute on Aging (NIA), in conjunction with the Alzheimer’s Association, released new criteria,1 earlier this year, for diagnosing Alzheimer disease. Diagnosis of Alzheimer disease is broken into three recognizable stages: preclinical disease, mild cognitive impairment, and symptomatic dementia. In mild cognitive impairment, mild changes in memory and thinking are noticeable and can be measured on mental-status tests, but aren’t severe enough to disrupt a person's day-to-day life, according to the NIA guidelines. In symptomatic dementia, impairments in memory, thinking, and behavior significantly affect a person’s ability to function independently in everyday life. Preclinical Alzheimer disease is a newly recognized stage of the disease, Foster explains. In this stage, key biological changes occur in the body, but the disease has not yet caused any noticeable clinical symptoms (such as mild cognitive impairment). According to the NIA guidelines, current scientific evidence suggests that in preclinical Alzheimer disease, brain changes caused by the disease might begin years (or even decades) before symptoms such as memory loss and confusion become evident. These changes can be detected through the presence of biomarkers. Biomarkers are measurable benchmarks in the body that indicate the presence or absence of a disease (or the probability of developing a disease). For Alzheimer disease, Foster says, the strongest biomarker candidates include changes seen in brain-imaging studies using MRI or PET and the presence of specific proteins in the cerebrospinal fluid. Collecting the Evidence Diagnosis with high levels of probability and accuracy depends on quantitative evidence observed in imaging studies, Foster says. This is found by comparing patients’ exams to those with patterns characteristic of Alzheimer disease. On MRI exams, for example, physicians can look for patterns typical of Alzheimer disease, particularly hippocampal atrophy. Hippocampal volume can be measured on MRI, and if it is significantly less than expected for age and there is increased ventricular size, then it becomes a very strong indicator of Alzheimer disease. Foster uses NeuroQuant® software (Cortechs Labs). Molecular imaging can also help identify the preclinical stages of dementing disease. In an example from Foster, a radiologist looking at transaxial images has to decide whether there’s enough hypometabolism in the posterior temporoparietal cortex to consider it abnormal (as opposed to consistent with normal variation). With quantitative techniques such as stereotactic surface-projection maps, these determinations become more concrete and data driven. Stereotactic surface projection allows for comparison of a particular individual’s pattern of metabolism with other datasets, such as those of subjects with normal cognitive function. With these quantitative measures, physicians can determine whether there are statistically significant levels of hypometabolic activity (an indication of developing dementia). Failure to Report Despite these advances, imaging, thus far, has not provided this kind of information to physicians, according to Foster. The problem isn’t one of capability, Foster notes. Imaging can play a pivotal role in the diagnosis of dementing disease, and Foster observes that physicians’ confidence and accuracy levels differ before and after imaging. Before imaging, there is low confidence in diagnosis and the probability of misdiagnosis is high, he says, but after imaging, these results improve significantly, boosting both confidence and accuracy. The newly defined preclinical stage of Alzheimer disease will become more important if researchers can develop treatments that slow or stop the progression of Alzheimer disease—treatments that might be most effective if used as early as possible in the course of the illness. Physicians are looking to radiology and imaging for help in identifying the specific causes of dementia, Foster says. With accurate prognosis, patients can better prepare for long-term–care solutions. Imaging is critical for early diagnosis, but it is failing to meet its potential. Foster hopes that radiologists and other physicians can work together to provide more diagnostic confidence and accuracy using quantitative biomarker clinical analysis. Thanh Le is a contributing writer for Radinformatics.com.