31 October 2008. Around the time the North American ADNI started, similar efforts sprang up in countries across the globe. They are at various stages of planning or completion in Australia, Japan, several European countries, and most recently, China. The grand idea that there will be one openly accessible, interoperable global database on thousands of volunteers, all studied in the same standardized way, remains out of reach at this point as each national study works to resolve its own hurdles. But according to scientists heading these international studies, they share this goal, and it may be more a question of when than if it will come to pass. Here is a summary of ADNIs from around the globe.
This Sub-Volume Probabilistic Atlas (ASVPA) was obtained by merging the MRI volumes of a large family of AD subjects into a common reference space. White matter, gray matter, and CSF are modeled separately. The green and blue colors indicate the average distribution of the CSF and gray matter, respectively, across the AD population. For a detailed caption, search the LONI image database for ASVPA. Image credit: Laboratory of Neuro Imaging (LONI), UCLA
The Land Down Under: Don’t Call It ADNI!
Mention “Australian ADNI” to scientists from Down Under and they’ll bristle ever so slightly before explaining patiently that the Australian Imaging Biomarkers and Lifestyle Flagship Study of Ageing (AIBL) is anything but a clone of the American initiative. Funded at AUS$10 million by the Australian government, this three-year study started as an imaging and plasma proteomics effort with a dietary and lifestyle component. It only recently added a CSF measurement; whole genome screens and genetic analysis are being planned as well.
AIBL started without a mechanism in place to make its data fully and freely available to the research community at large, but this has since changed. According to Chris Rowe of the University of Melbourne, AIBL’s imaging data with matching neuropsychology and demographic data will be made available through the LONI website after a six-month moratorium. “Baseline data should be available in early 2009,” Rowe wrote to ARF.
AIBL started enrolling in 2007 and by now has completed baseline evaluation of 1,165 participants. They are grouped similarly as in ADNI, though with an additional group of people who complain about memory loss while appearing normal on neuropsychological test batteries. (American ADNI leaders are hoping to fill the gap between normal controls and MCI cases for ADNI2; subjective memory complaint is one attempt to capture that in-between group.) AIBL is at present beginning the 18-month follow-up of its patients, and at that point will request consent for a lumbar puncture. Longitudinal data on some 95 patients who were being studied before AIBL began are being analyzed already. AIBL conducts extensive amyloid imaging. It brings sophisticated software capabilities to image analysis and 3D rendering, for example, of how amyloid deposition spreads from cognitively normal to a full-blown AD load.
As to initial results, at the International Conference on Alzheimer’s Disease (ICAD) last July in Chicago, Kathryn Ellis of the University of Melbourne, with other AIBL colleagues, reported that the groups were well matched. AIBL scientists presented technical results on various ways of analyzing PIB-PET data. In addition, they reported that of 265 people who had undergone amyloid imaging in addition to MRI, 98 percent of AD cases and 60 percent of the MCI cases had an AD-like PIB pattern. Thirty percent of all healthy participants had brain amyloid, though that fraction was 50 percent among people carrying an ApoE4 allele.
One of 25 PIB-negative healthy controls and three of 10 PIB-positive healthy controls had converted to MCI. Five of 13 PIB-negative people with MCI had progressed to non-AD dementia (DLB or FTD), and 12 of 17 PIB-positive people with MCI had since advanced to AD. Kerryn Pike of Melbourne and her AIBL colleagues reported that healthy participants with brain amyloid performed worse on episodic memory tasks even though their overall neuropsychology scores were normal. To the Australian scientists, the data available up to this point would suggest that brain amyloid imaging can help pinpoint early memory impairment, or even latent AD.
In the Land of the Rising Sun, ADNI Is Coming Up
“I believe the ADNIs are the most important trend for bringing the fruits of basic research to the clinic,” Takeshi Iwatsubo of the University of Tokyo, wrote to this reporter. This is perhaps why the Japanese researchers have decided to clone ADNI (sort of). Iwatsubo, who leads a Japanese study that was modeled closely on the North American study, presented a summary with a slide presentation downloadable at adni-info.org.
Here are more recent updates, based on correspondence with Iwatsubo: the Japanese ADNI study engages 36 centers throughout the country. It is funded at about $3.5 million per year, to run for five years. Some 40 percent of that comes from a consortium of 10 pharma companies, whose representatives form an industry advisory board (ISAB) just as is the case for the North American ADNI. The Japanese ministry of economy, trade, and industry (METI/NEDO) and the Ministry of Health, Labor and Welfare (MHLW) together supply 60 percent.
The Japanese scientists intentionally designed their own ADNI to be as compatible with the U.S. ADNI as possible. This means the protocol employs the same method of MRI volumetry, uses the same phantoms, and the same schedule and type of clinical and neuropsychological batteries. Morihiro Sugishita, the leading neuropsychologist of J-ADNI’s clinical core, amended the existing Japanese versions of these tests to fit them to the English ones. According to Iwatsubo, this was no mean feat, but it’s nearly complete and raters have begun training for ADNI.
The Japanese scientists had the benefit of absorbing the gradual improvements of the N.A. ADNI into its original protocol. For one, the N.A. ADNI added amyloid imaging after it was already well underway, hampering this effort somewhat, whereas the J-ADNI implemented a standardized protocol for amyloid imaging from the get go. Some 13 sites out of the 36 will conduct amyloid imaging, because they have the capability to synthesize 11C-PIB or 11C-BF227 probes (BF227 is an amyloid tracer developed and increasingly widely used in Japan, see Furomoto et al., 2007). Some sites will invite patients from nearby centers, so that 17 sites in total and up to 30 percent of participants can obtain amyloid imaging, Iwatsubo wrote. For another, the N.A. ADNI expanded its initially limited CSF measurements after they turned out to be surprisingly promising. It also added a genetic add-on study later. The J-ADNI is planning on collecting CSF samples at multiple time points, aiming for some 40 percent participation, and Ryozo Kuwano at Niigata University is planning genome studies.
The J-ADNI is planning to analyze data jointly with data from the North American ADNI as soon as their own primary analysis is complete. To this end, the Japanese scientists will work with Laurel Beckett from the University of California, Davis, who heads ADNI’s statistics core. Unrestricted access to data as a public resource is another matter, however. The J-ADNI database will not be open to scientists worldwide, according to Iwatsubo. “However, it will eventually be possible to publicize all the data for wide use,” he added.
Planning for J-ADNI began in earnest in 2006, and the funding arrived in September 2007. Subsequent preparation and various training courses then preceded enrollment, which began last month. The J-ADNI will study 150 normal controls, 300 people with MCI, and 150 with early AD by 2010. It’s starting well, Iwatsubo wrote. “Japanese clinicians/researchers, lay people, government, and pharma, especially global pharmas that are starting trials in Japan, are very interested in J-ADNI. Some sites are flooded by normal volunteers who wish to be enrolled.”
The Old Country: Pilot a Success, But Europe-Wide ADNI Won’t Fly
In Europe, the story of ADNI is less straightforward. Academic centers in seven countries have jointly tested the ADNI platform in a pilot study funded by the U.S. Alzheimer’s Association. They found the platform to be feasible for Europe to adopt, but the follow-up support for a full-fledged European ADNI fell short. Instead there will be national initiatives.
At the International Conference on Alzheimer’s Disease (ICAD) last July in Chicago, Giovanni Frisoni from Fatebenefratelli in Brescia, Italy, presented details from the European pilot ADNI, which also appeared in print the same month (Frisoni et al., 2008). In short, academic centers in Amsterdam, Brescia, Copenhagen, Munich, Rome, Stockholm, and Toulouse had banded together to try out whether the N.A. ADNI platform of collecting clinical, neuropsychological, structural, and functional imaging together with biochemical information in a pooled way would work for them.
Each center recruited three patients each in the control, MCI, and AD groups for one baseline assessment. No follow-up was planned. For the clinical assessments, the European study translated the forms of the U.S. ADNI into six different languages. On imaging, the volunteers submitted to a 1.5 Tesla volumetric scan and a diffusion tensor scan. Unlike the N.A. ADNI, the European ADNI also took a resting fMRI measurement, but no amyloid scan. For calibration, they used three human “phantoms” who traveled from center to center to be scanned in each of the participating machines. Three quarters of the participants agreed to a lumbar puncture.
These are the results: the European patients matched the North American ones well in their clinical measures and cognitive performance. A few differences showed up (e.g., the Europeans were a tad more depressed), but they were small. On the imaging results a manuscript is in preparation, and on the biomarker results a paper is submitted (Buerger et al.). In a nutshell, the results were comparable to the American ADNI. Hippocampal atrophy increased from normal to MCI to AD; white matter abnormalities did not. CSF total tau increased and Aβ42 decreased as expected from normal to MCI to AD. “We came out with study groups that from a clinical, functional, and neuropsychological standpoint are remarkably similar to those of the American ADNI,” Frisoni later wrote to this reporter.
The results of this pilot study support the validity of the American ADNI platform, and a larger European study could be set up fairly quickly based on this model, Frisoni said at ICAD. But it most likely will not. Due to political and structural reasons, a large European ADNI with pooled, openly accessible data is unlikely. Instead, there will be national efforts. “National efforts are more of a necessity than a choice, arising from the much more fragmented way research is funded at the European Union level,” Frisoni wrote.
In Germany, Italy, and France, scientists are setting up their respective networks that Frisoni hopes will choose the ADNI platform to collect multicenter prospective data. In addition, some ADNI-related initiatives already exist in Europe. For example, an EU-funded multinational, public-private initiative already uses an ADNI-compatible MRI protocol. Called AddNeuroMed, it is coordinated by Simon Lovestone at King’s College, London (Lovestone et al., 2007). At ICAD, Christian Spenger of Stockholm’s Karolinska Institute reported that the AddNeuroMed network had enrolled 383 subjects in two years of operation thus far, and he emphasized the study’s focus on quality control. Furthermore, academic centers in Brescia, Amsterdam, and Stockholm are presently building an electronic infrastructure that aims to support advanced computational analysis of ADNI-compatible images. Called neuGRID, it receives European Union funding of €2.8 million.
Whether these separate efforts will consolidate to become compatible with the American and Japanese ADNI, or whether they will remain different is an important but unresolved issue, Frisoni noted. Ditto for the question of whether their data will be openly available. “Appropriate coordination of all these efforts would enhance the impact of EU research,” Frisoni wrote, adding that the political hurdles on the way are considerable.
New Kid on the Block: ADNI-like Studies in China
Chinese researchers have not stood idly by while interest in comprehensive natural history studies of AD has grown elsewhere. In fact, they have started several similar studies. These initiatives come against a backdrop of estimates that the percentage of people older than 60 in China’s population will triple by the year 2050, from about 10 percent now to 30 percent then. Of two multicenter initiatives described here, one is a biomarker study similar to the North American ADNI in design; the other has a more epidemiological bent and focuses on MCI, but its overall idea is the same as ADNI’s.
The principal investigator of one study is Jun Xu, who was a postdoctoral fellow at the University of Pittsburgh before joining the faculty at the Nanjing Brain Hospital and Nanjing Medical University, China, where he is vice dean of the Department of Neurology. His study is a government-funded effort in Eastern China that at present includes 10 centers located in Nanjing, Shanghai, Beijing, Guangzhou, Chengdu, Xi'an, Hangzhou, Wuhan, and Fuzhou. Co-principal investigators are Wu-hua Xu (no relation) at Jinan University 4th Hospital in Guangzhou, and Yong-an Sun at Peking University 1st Hospital in Beijing. According to Xu, this study started up in May 2008 and has to date enrolled 100 of a targeted 500 patients. Of those, 200 will be normal controls, 200 have amnestic MCI, and 100 probable AD. According to Xu, this study uses much the same protocol as the American ADNI, with the express intention of gathering data using the ADNI platform of standardized assessments. The study includes clinical and neuropsychological assessments, MRI, FDG-PET, CSF, and plasma biochemistry. Due in large part to its expense, amyloid imaging is not part of the current protocol, nor are whole genome screens at this point in time, Xu wrote to ARF.
In one marked difference to the other ADNIs, this Chinese study is planned to run for 20 years. That means the study places a particular emphasis on observing the long-term natural history of disease in its participants, and hopes to learn a great deal from the 200 volunteers who were cognitively normal at study entry.
Xu wrote to ARF that the Chinese investigators aim to make their data publicly accessible, much like the North American ADNI data are on LONI, but he added that the requisite software remains under security consideration and still awaits approval. Likewise, Xu noted that the Chinese investigators are interested in connecting their data with the American data pools to enable joint analyses and boost power. On this goal, as well, structural, legal, and privacy issues still need to be sorted out.
How did Xu develop his interest in ADNI? The Internet made it easy. After his return to China, Xu simply kept up to date through the websites of ADRCs, the NIA and ADNI-info.org, LONI, and took it from there. The Chinese investigators are holding a symposium to discuss enrollment and screening data this December. According to Wu-hua Xu, this study is looking to engage more specialists in China and open to interacting with North American ADNI colleagues.
A separate, five-year study that focuses on mild cognitive impairment is gearing up under the direction of Jianping Jia. Jia directs the Department of Neurology of Xuan Wu Hospital of Capital Medical University in Beijing. This study is part of China’s effort at early detection of dementia and MCI. Called the Chinese Dementia and Mild Cognitive Impairment Longitudinal Study, it received funding in 2007 from the Ministry of Science and Technology and the Ministry of Health of the People’s Republic of China. Covering 12 provinces in five geographic regions of the country, this is a community-based epidemiology and natural history study of MCI in China. Its goals are to determine fundamental datasets on prevalence and subtypes of dementia in China, to find risk and protective factors, and to identify predictors for MCI conversion. It also aims to establish normative data for cognitive measures and find cognitive tests that are sensitive for MCI, according to Maria Carrillo of the Alzheimer’s Association, who is working to coordinate worldwide ADNI efforts. To this end, the study will gather data on risk factors, medication, and cognitive impairment history. Using annual follow-up visits, it will include a range of assessments, such as neurological, neuropsychological, as well as some biological data in blood and neuroimaging. Genetic tests for ApoE and FAD genes are also planned.
The study this year began screening a sample of 30,000 elderly people in communities and hospitals, but how many will be enrolled has not been calculated yet, Jia wrote to ARF. “We are hoping they will join the WW-ADNI effort. We are in the planning process of having a WW-ADNI meeting in Japan in November 2009, and these Chinese investigators will of course be invited,” Carrillo wrote.—Gabrielle Strobel.
- As ADNI Turns Four, $64 Million Data Start Rolling In
- ADNI Results: A Story of Standardization and Science
- More ADNI Results 2: It’s All About Power
- Add to ADNI: Imaging Amyloid and Analyzing the Genome
- Bumps Along the Way—ADNI’s Snafus
- Furumoto S, Okamura N, Iwata R, Yanai K, Arai H, Kudo Y. Recent advances in the development of amyloid imaging agents. Curr Top Med Chem. 2007;7(18):1773-89. PubMed.
- Frisoni GB, Henneman WJ, Weiner MW, Scheltens P, Vellas B, Reynish E, Hudecova J, Hampel H, Burger K, Blennow K, Waldemar G, Johannsen P, Wahlund LO, Zito G, Rossini PM, Winblad B, Barkhof F, . The pilot European Alzheimer's Disease Neuroimaging Initiative of the European Alzheimer's Disease Consortium. Alzheimers Dement. 2008 Jul;4(4):255-64. PubMed.
- Lovestone S, Francis P, Strandgaard K. Biomarkers for disease modification trials--the innovative medicines initiative and AddNeuroMed. J Nutr Health Aging. 2007 Jul-Aug;11(4):359-61. PubMed.