Alzheimer’s Disease Research Center Research
Research at NYU Langone’s Alzheimer’s Disease Research Center has led to many discoveries in the field of brain aging and neurocognitive disorders. Through clinical trials and research studies, many of our investigators study areas of health that affect people with Alzheimer’s disease and related dementia.
The Conformational Disorders Lab, led by Thomas M. Wisniewski, MD, focuses on better understanding neurodegenerative diseases such as Alzheimer’s disease and prion-related diseases. In addition, the lab works on other neurological conditions such as autism spectrum disorder and stroke. This work has led to more than 210 peer-reviewed publications. The lab has been continuously funded by the NIH and other groups such as the Alzheimer’s Disease Association for more than 20 years. Discoveries from Dr. Wisniewski’s lab include the following:
- identification of the role of apolipoprotein E4 (apoE4) as a “pathological chaperone,” promoting aggregation of amyloid β (Aβ), in Alzheimer’s disease
- development of therapeutic approaches for Alzheimer’s disease based on the apoE–Aβ interaction, which reduces amyloid plaque burden, cerebral amyloid angiopathy, and Aβ oligomer levels, without toxicity
- production of improved immunization approaches for Alzheimer’s disease using nontoxic, nonfibrillogenic Aβ homologous peptides without Th1 epitopes for greater efficacy and lower toxicity
- development of the first active and passive immunization approaches for prion disease, which are effective in wild-type, nongenetically modified model animals
- reporting the first method to image Aβ deposits in model animals using MRI techniques with ultrasmall superparamagnetic iron oxide (USPIO) amyloid-binding ligands, which cross the blood–brain barrier
- characterization of the neuropathological features associated with different types of autism spectrum disorder
- development of a novel therapeutic approach for Alzheimer’s disease by stimulating innate immunity via Toll-like receptors that is effective in multiple Alzheimer’s disease mouse models and also in nonhuman primates
- characterization of a novel immunization approach with a non-self immunogen (pBri) that produces an immune response targeting the shared pathological conformation of both Aβ and tau oligomers, as well as other amyloid-prone proteins
- development of a novel proteomic methodology that allows use of formalin-fixed, paraffin-embedded tissue and use of this method to help characterize the pathogenesis of rapidly progressive Alzheimer’s disease
- production of novel anti–β-sheet conformational monoclonal antibodies (aβComAb) that have therapeutic potential for multiple neurodegenerative diseases
PET Imaging for Alzheimer’s Disease—From Research to Everyday Practice
Positron emission tomography (PET) has emerged as the key imaging technique for clinical management and research of Alzheimer’s disease. Patients undergoing PET are given a small amount of radioactive tracer that consists of carrier molecules labeled with a positron-emitting radioactive isotope. In the brain, the carrier binds to a specific target. After a while, the radioactive isotope decays and emits a positron, which interacts with a nearby electron resulting in the release of a pair of photons traveling in opposite directions. These photons are detected by the PET system and used to construct three-dimensional images showing the distribution of the tracer in the brain.
FDG PET in Alzheimer’s Disease
The most widely used PET tracer is F-18-labeled fluorodeoxyglucose (FDG), a form of glucose. Glucose is the primary source of energy for adult brain neurons, which consume nearly half of all glucose-derived energy in the body. As FDG is absorbed by the brain, the resulting PET images reflect the metabolic activity in brain tissue.
FDG PET has been used to study brain aging and dementia since the early 1980s.1 In healthy aging, brain metabolism gradually declines with age in a predictable fashion. In patients with memory loss or cognitive impairment, a pattern of low metabolic activity in specific brain areas may signal the presence of Alzheimer’s disease or point to other types of neurodegeneration. PET can often identify the likely underlying cause of cognitive impairment and predict the individual patient’s clinical course. Scientists at NYU Langone’s Alzheimer’s Disease Research Center have been at the forefront of this research. As first shown by a team of NYU Langone researchers, FDG PET can predict cognitive decline in normal elderly subjects and their progression to mild cognitive impairment.2 A subsequent study established standardized methods that enabled FDG PET to differentiate Alzheimer’s disease from other dementias with 96 percent accuracy.3 Tracking patients over time, our researchers mapped out the progression of cognitive decline and dementia, and showed that metabolic changes appear on PET images many years before clinical symptoms do.4
Since the early 2000s, special PET tracers have been developed to detect pathological hallmarks of Alzheimer’s disease—beta-amyloid and tau proteins. The first amyloid PET tracer appeared in 2002. In 2012–2014, three amyloid tracers were approved by the U.S. Food and Drug Administration (FDA). Amyloid PET is highly sensitive to the presence of amyloid plaques, and all Alzheimer’s disease patients have amyloid deposits. Thus, a negative amyloid PET scan can help to rule out Alzheimer’s disease. However, amyloid is also present in 25 percent of cognitively healthy individuals over the age of 75. An amyloid PET study led by Alzheimer’s Disease Research Center scientists showed that cognitively normal individuals with maternal family history of Alzheimer’s disease have increased amyloid buildup in brain areas vulnerable to Alzheimer’s disease.5 Amyloid PET may also be used to much more effectively select appropriate patients for anti-amyloid targeted clinical trials, as well as assessing the effectiveness of such investigation therapeutic interventions.6
Another family of PET tracers targets the accumulation of tau protein found in Alzheimer’s disease and other neurodegenerative conditions. Tau PET is still new and under development, but researchers believe that it will have an important role in early detection of underlying Alzheimer’s disease tissue pathology. Progressive accumulation of tau over time is associated with cognitive decline and can be used to predict disease progression. In this respect, tau PET may offer an advantage over amyloid imaging, because amyloid accumulation tends to level off over the course of Alzheimer’s disease. An array of tau tracers are now being actively investigated with promising results for improved diagnosis, staging, and monitoring of anti-tau treatment.7
The Alzheimer’s Disease Research Center plans to make amyloid PET and tau PET available to all new research participants by 2020. With the support of NYU Langone’s extensive resources—which include in-house facilities that produce these PET tracers, state-of-the-art imaging equipment, and cutting-edge technical expertise—these advanced imaging techniques will become everyday practice.
Recently, integrated imaging systems that combine PET with magnetic resonance imaging (PET/MRI) have been gaining popularity as an alternative to the combination of PET and computed tomography (PET/CT). With PET/MRI, a complete imaging workup of brain anatomy and metabolism can be obtained in a single session, which is easier and more convenient for patients and their caregivers. PET and MRI have complementary strengths, and when combined, they provide more diagnostic insight and confidence than each modality alone. MRI has high spatial resolution, excellent contrast between soft tissues, and can also provide “multiparametric imaging” that includes maps of the blood flow, microscopic tissue organization, and other tissue parameters. PET/MRI also exposes patients to less radiation than does PET/CT, making repeat imaging exams safer. Combining PET and MRI also leads to interesting synergies, currently under investigation at NYU Langone and elsewhere, which may lead to improvements in spatial resolution and further reductions of radiation dose. We plan to perform routine imaging on an integrated PET/MRI system developed by Siemens Healthineers USA.
PET Reconstruction and MR-based Attenuation Correction
When PET images are reconstructed from the raw data, they must be corrected for the attenuation of the photons that travel through the body on the way to the detectors. This correction is particularly important in brain imaging, because the photons may be stopped by the skull bones. The attenuation correction is done with the help of the attenuation images that show the opacity of different tissues for photons of a given energy, or radiodensity. In PET/CT, attenuation images are obtained from CT data, which reflect the radiodensity for X-rays and correlate approximately linearly with attenuation of the higher-energy photons associated with PET imaging. In PET/MRI, a different approach is required, because MRI data are not related to tissue radiodensity. For example, conventional T1-weighted MRI has the same signal intensity for bone, air, and cerebrospinal fluid that surrounds the brain, even though these tissues affect PET photons very differently. Without accurate attenuation correction, PET/MRI may introduce errors into the measurements of PET tracer uptake.
To obtain accurate measures of tracer uptake, NYU Langone researchers, together with their international collaborators and Siemens Healthineers, developed a new hybrid method of PET reconstruction that creates the attenuation map from a combination of an MRI-based segmentation image with a model-based map of the skull bones.8 The new method was tested in 16 patients with suspected cognitive impairment at NYU Langone’s Pearl I. Barlow Center for Memory Evaluation and Treatment. Each patient was imaged back-to-back with PET/CT and PET/MRI. PET images were reconstructed with the new hybrid method and compared with the images reconstructed using only the segmentation image (MRI-based method) and the ground-truth images reconstructed using CT.
The hybrid method dramatically improved the accuracy and precision of the tracer uptake measurements, expressed by the standard uptake value (SUV). With the hybrid method, the whole-brain mean SUV was within 0.3 percent of the ground truth, which translates into a 95 percent lower bias compared to that of the MRI-based method. When SUV was measured in 91 regions defined in each brain, the hybrid method produced only 5 regions in which SUV deviated from the ground truth by 5 percent or more, compared to 38 regions showing such bias with the MRI-based method. The regions with larger bias were located in the frontal poles where it is difficult to account for individual variations in air spaces and skull thickness. The hybrid method compared favorably to other novel approaches, received FDA approval and has now become the standard method for all PET/MRI systems provided by Siemens Healthineers in the U.S. and Europe. PET/MRI assessments are now routinely used at the Pearl I. Barlow Center for Memory Evaluation and Treatment to assess patients with cognitive symptoms.
MR-based Joint Reconstruction Improves PET Image Quality
In a 2019 study, Timothy M. Shepherd, MD, PhD, and his colleagues showed that an MR-guided reconstruction technique developed at NYU Langone improved the quality of PET images and helped to detect brain lesions that cause epileptic seizures (seizure foci) in 26 patients.9 Three neuroradiologists rated the quality of the MR-based PET images significantly higher compared to that of the images reconstructed with the conventional method. Using MR-reconstructed PET images, 2 of the 3 readers demonstrated a 40–50 percent improvement in detecting the most difficult to locate seizure foci outside the temporal lobe. All three readers reported the images were higher quality and reported improved confidence in the diagnosis made using MR-based PET images. These methods should apply equally well to other PET tracers, including amyloid or tau.
New Methods Contribute to Making Advanced Imaging Routine
The novel approaches to MR-based PET reconstruction produce higher quality images and improve the accuracy of tracer uptake measurements. These improvements are essential for the clinical use of non-FDG PET, detection of subtle abnormalities, and combination of PET with advanced MRI. The current effort of NYU Langone’s Alzheimer’s Disease Research Center researchers is focused on making these state-of-the-art imaging methods available to all Alzheimer’s Disease Research Center participants.
Sleep and Brain Health
Sleep is a critical factor of health and healthy aging. It has long been known that good sleep is essential for human health. Poor sleep is linked to a variety of health problems, such as high blood pressure, heart disease, weakened immunity, diabetes, and obesity. Recent research increasingly suggests that sleep may also be critical for brain health and healthy aging. As people age, their sleep often gets worse: almost 40 percent of adults over the age of 65 complain of chronic sleep problems1,2. Poor sleep in older adults is associated with worse memory, higher risk of falling, depression, and cognitive decline3. Sleep disturbances are also a common early symptom of Alzheimer’s disease4. But does poor sleep lead to dementia or do dementia-related brain changes cause sleep problems?
Sleep and Brain Health Are Closely Related
“The relationship between sleep and brain health appears to be two-way,” says Ricardo M. Osorio Suarez, MD, assistant professor in the Department of Psychiatry and director of the Center for Sleep and Brain Health at NYU Grossman School of Medicine, and who studies sleep and aging. “How exactly sleep contributes to brain health is yet unknown. But understanding the mechanism of sleep may help to develop approaches to lower the risk of developing Alzheimer’s disease or slowing its progression. Sleep may end up being an important factor in secondary prevention of dementia, along with physical activity, social engagement, and education.”
Poor Sleep Is Linked to Cognitive Decline
There is ample evidence that poor sleep is linked to cognitive decline and increased risk of Alzheimer’s disease. A meta-analysis of 27 studies has found that people with sleep problems had a 1.68 times higher risk of developing cognitive impairment than those with normal sleep5. Poor sleep at midlife is associated with worse cognitive function later in life, as shown by a 20-year follow-up study of over 2,300 twins6. In a three-year study of sleep quality in older men assessed through self-reports and a wrist-worn device tracking movements during sleep, participants who slept poorly were more likely to suffer cognitive decline than those who slept well7.
Sleep Apnea Accelerates Cognitive Decline, While Restored Sleep Slows It Down
Sleep apnea and other sleep-disordered breathing, which are characterized by pauses in breathing, disrupt sleep and cause intermittent hypoxia (lack of oxygen). Among older adults, these disorders affect almost 50 percent of men and 25 percent of women and are associated with worse cognitive outcomes. A five-year study of dementia-free older women showed that those who had 15 or more apnea events per hour of sleep had a higher risk of mild cognitive impairment or dementia8. Dr. Osorio examined multiple imaging, physiological, and cognitive characteristics in over 2,000 older adults. His study demonstrated that sleep apnea was associated with an earlier age of onset of mild cognitive impairment and dementia9. “We also showed that treating sleep apnea slows down cognitive decline. This is good news for many patients who can use continuous positive airway pressure (CPAP) devices to restore correct breathing and improve their sleep,” says Dr. Osorio. Treatment of sleep apnea is one of the few therapies that has shown to slow down cognitive decline among Alzheimer’s disease patients10.
Poor Sleep Leads to Buildup of Toxic Brain Proteins
The disruption of sleep commonly seen in Alzheimer’s disease was often assumed to be the result of accumulation of beta-amyloid protein, a hallmark of Alzheimer’s disease. Recent research shows, however, that the reverse may also occur: disturbed sleep may increase the buildup of beta-amyloid and lead to neurodegeneration and cognitive decline11. A brain imaging study of older adults using PET scans with a beta-amyloid-sensitive tracer (Pittsburgh compound B, PiB) showed that participants who slept less or less well had more beta-amyloid deposits in their brains than those with longer and better sleep12. Another PiB PET study showed that dementia-free older adult subjects who suffered from sleep apnea accumulated more beta-amyloid over a two-year period than those who had normal sleep13. Biomarker studies of cognitively healthy older adults at risk for Alzheimer’s disease demonstrated that those subjects who slept poorly had lower levels of beta-amyloid in their cerebrospinal fluid (CSF), an early sign of Alzheimer’s disease14.
Sleep as the Brain’s “Wash Cycle”
These observations have prompted interest in the role played by sleep in clearing away of toxic waste that accumulates in the brain as a result of its activity. A sleep-dependent clearing mechanism has been proposed and demonstrated in animal studies by Maiken Nedergaard of University of Rochester, who coined the term “the glymphatic system” to describe this pathway15. Experiments in mice revealed that the levels of beta-amyloid are dynamic and oscillate with the sleep–wake cycle: they rise during wakefulness and decline during sleep; they also increase in response to chronic sleep deprivation and decrease when sleep is restored16. Subsequent PET imaging in healthy subjects has shown that beta-amyloid level in the human brain increases after just one sleepless night17.
During sleep, humans cycle through predictable sleep stages of non-REM sleep, followed by slow wave sleep (deep sleep), and by REM sleep often associated with dreaming. Recent studies uncovered the importance of slow wave sleep for the removal of toxic products. A study of mostly cognitively normal older adults, which involved specialized PET imaging, CSF biomarkers, sleep monitoring, and cognitive testing, demonstrated that patients who had less slow wave sleep had higher deposits of tau protein, another pathological signature of Alzheimer’s disease18. A closer look at the glymphatic system in humans has been provided by an experiment that combined MRI with electroencephalography (EEG)19. In 11 healthy subjects, the slow brain waves appear just before a drop in blood flow followed by a pulse of CSF that washes over the brain and probably removes toxic waste products20. If slow wave sleep is disturbed, for example by sleep apnea or insomnia, or diminished with advanced age21,22 the brain’s ability to remove waste may be compromised leading to buildup of toxic proteins, such as beta-amyloid and tau. The disruption of slow wave sleep may therefore become a new biomarker and therapeutic target in aging and prevention of neurodegeneration23.
New Methods for Analyzing Sleep Studies
Studies of the mechanisms of sleep and its effect on health are becoming increasingly complex and often combine several different techniques, including imaging, actigraphy (measurements of movements during sleep), and polysomnography. Polysomnography, or sleep study, records the electrical activity of various parts of the brain with EEG during each of the sleep stages, as well as blood oxygen level, blood pressure, heart rate, and breathing. Sleep studies are widely used to monitor sleep cycles and diagnose sleep disorders, such as sleep apnea, chronic insomnia, REM sleep behavior disorder (RBD), and other conditions.
Currently, sleep studies are analyzed manually by trained readers. This method is subjective, and the agreement between readers is often poor. Efforts to develop faster and more objective automatic analysis methods based on machine learning are underway24. NYU Grossman School of Medicine researchers have developed an algorithm to detect spindles—short bursts of brain activity during non-REM sleep—and showed that it provides fast and accurate analysis of publicly available expert-validated EEG datasets25. “The machine learning algorithms may take advantage of the large databases of sleep studies and really boost our ability to analyze sleep data in both humans and animals, usually a slow and laborious process,” says Dr. Osorio. In addition to detecting spindles, Dr. Osorio is interested in studying REM sleep, or “paradoxical sleep,” which is plays an important, but still mysterious role in the formation of certain types of memory, especially navigation memory. “We know that this type of memory is often affected in patients with Alzheimer’s disease, who might wander away and get lost on the way home, so studying this mechanism may be important for our understanding of dementia.”
Although sleep is fundamental for human health, our understanding of it is still limited. Much remains to be elucidated, but the recognition of the importance of sleep for brain health and the advances in experimental techniques that enable to probe the brain activity during sleep promise new insights into the role of sleep in the development of Alzheimer’s disease and other neurodegenerative disorders as well as methods of their prevention.