In like a lamb, out like a lion. It’s a full season too late for this variation on the March weather proverb, but hopefully our generous readers will forgive their roving reporter for bending it anyway, because starting meekly and finishing with a roar is exactly what the Alzheimer’s Association’s 12th International Conference on Alzheimer’s Disease (ICAD) did when it unfolded in Vienna 11-16 July. And it’s not about changing weather, either. All week, the Austrian capital bathed meeting attendees in sparkling sunshine (which seemed to seduce many into sneaking off for daytrips into this beautiful city as attendance seemed a bit sparse at times, but no names shall be named). No, the scientific program started with humbling reports of the failure—abject and otherwise—of the latest round of clinical trials. The program continued with confirmatory studies on the ability of exercise, a heart-healthy diet, and modest drinking while one is cognitively healthy, to modify one’s risk of cognitive decline somewhat. This is an important life message people should take to heart, but it is not the kind of news close followers of the field are looking for. But then the last morning—when many had departed, the exhibit hall was emptied out, and the press room had closed up shop—featured within the space of three hours arguably the biggest news in AD genetics research in the past decade. Three independent genomewide association studies, one being the largest performed to date on some 4,000 cases, together reported three new genes for late-onset AD. Hence, overall comments about ICAD in Vienna went from a plaintive “I haven’t heard anything really new” at the beginning to a confident “We can declare three new genes for Alzheimer’s” at the end.
Most remarkable from an observer’s perspective was the consensus among independent genetics groups that the new genes would likely hold up to scrutiny in future studies even if their individual effect was small. When is the last time you remember applause and nodding faces, not skepticism or criticism, greeting the first introduction of a new candidate gene at a major AD conference? The genes reported today are ApoJ (aka clusterin), CR1 (which encodes a complement receptor), and PICALM (which encodes an endocytic gene). Together, they point partly to Aβ generation/clearance playing a role in late-onset AD, said Philippe Amouyel of the Institute Pasteur in Lille, France, but more broadly, they also implicate other upstream processes as leading to AD pathology and symptoms. “The GWAS data suggest that innate immunity and cholesterol are key themes in the disease,” said Julie Williams of Cardiff University, U.K. (For more on these presentations, see upcoming ICAD news story.)
Regarding clinical trials, the news in Vienna was mostly dire. Among the trials that reported results, GSK’s Phase 3 rosiglitazone program failed, the ADCS’s valproate and DHA trials were negative, as were the Phase 2b trials of AstraZeneca’s α4β2 nicotinic acetylcholine receptor agonist AZD3480 and of a monthly implant formulation of donepezil called mimopezil. The large anti-amyloid trials are many months to years away from yielding results, and without substantiation the rumors that inevitably fly in such periods of suspended animation are not worth reporting. The presentation by Bengt Winblad of the Karolinska Institute in Stockholm, Sweden, of Novartis’s CAD106 active vaccine, which is currently in Phase 2, dealt mainly with additional Phase 1b efforts to refine the dosing regimen for obtaining stable and sufficient antibody titers, and with expanding the human safety record for this virus-like particle vaccine to a larger sample of patients. Steven Jacobsen’s presentation of Wyeth’s new γ-secretase inhibitor GSI-953, which according to ClinicalTrials.gov has finished three Phase 1 trials, the latest in the summer of 2008, covered mostly preclinical data.
The trials landscape offered some unexpected highlights, though. A new agonist against the α7 nicotinic acetylcholine receptor showed early hints of possible efficacy on top of stable AChE inhibitor treatment in a Phase 1b/2a study (see ARF related conference story). And John Breitner of the University of Washington, Seattle, reported that the ADAPT primary prevention trial of two NSAID drugs in some 2,500 people age 70 and older provided a pleasant surprise at longer-term follow-up. The ADAPT treatments were halted mid-stream in December 2004 amid concerns about the safety of celecoxib and other NSAIDs for long-term use in the elderly (for prior coverage of this topic, see ARF Safety Data news story, ADAPT Trials and Tribulations news story; see also Safety Concerns Halt ADAPT Trial).
ADAPT received funding to follow the trial participants for two more years, and in Vienna Breitner showed results from subgroup analyses. They suggested that people who already had mild cognitive problems at the beginning of the trial indeed worsened if they took NSAIDs, which accelerated their underlying disease. In contrast, people who were cognitively normal at baseline and took naproxen for one to three years appeared to fare better. At the long-term examination, they not only had a lower incidence of AD than did those on placebo, but going along with this apparent clinical protection was a biomarker effect as well. The naproxen group had a favorable, i.e., lower, tau/Abeta42 ratio in their CSF drawn at spinal taps 21-42 months after treatment. “Primary prevention, not disease modification, is the ultimate cure, and timing is everything,” Breitner asserted provocatively. ADAPT has recently received funding to ask the trial cohort back for two further examinations. Breitner suggested that the biomarker results give good hopes to the possibility that one to three years of treatment with naproxen in cognitively intact people may ultimately reduce the incidence of AD for five to 10 years out.
Overall, this ICAD conference was marked by a notable absence of many scientists who tend to be regulars. Citing the growth in the number of conferences in the field, many begged off this time. Some said they are on the road too much to get much else done, and some deplored the cost of trying to bring junior researchers to AD conferences these days. At 3,700, attendance at Vienna’s attractive, light-flooded “Messe” center fell below the 5,400 from Chicago in 2008 and the 5,000 from Madrid in 2006. That said, those who were in Vienna cited a different spirit than in Chicago. Last year, there was palpable tension at ICAD when the Flurizan Phase 3 negative trial and Bapineuzumab’s complicated Phase 2 results failed to meet the expectations that had been building in the months prior. This renewed doubt in the amyloid hypothesis. This time around, researchers expressed a greater sense of calm. In Vienna, Henrik Zetterberg of Sahlgrenska University Hospital in Goteborg, Sweden, spoke for many when he said, “The field has been humbled, and we can now focus on how best to regroup. We can do research to improve the next round of drug testing without the pressure of imminent trial results hanging over our head.”
To an outside meeting observer, two countercurrents seemed to be building simultaneously, though the twain never met for open debate in Vienna. On the one hand, researchers who tend to subscribe to the amyloid hypothesis as the source of the most tangible drug targets at present, drew renewed confidence from results emerging from ADNI and other biomarker studies in Europe and elsewhere. Overall, those results converged around the realization that biomarker signatures of impending AD seen at the MCI and perhaps even earlier stages will make it practical to move the clinical testing of drug candidates to stages of the disease where symptoms are milder—and neurodegeneration less extensive—than at the current “mild to moderate” stage defined by the current NINCDS-ADRDA diagnostic criteria, which are widely seen as being behind the curve of biological research on AD pathogenesis. As a group, these scientists have worried that otherwise respectable drugs may fall by the wayside because they are being tested too late in the disease. “I am coming out of this ICAD meeting with a bump because I think the tools are coming together, from the ADNI study and others, that make it possible to run earlier-stage trials now,” said Reisa Sperling of Brigham and Women’s Hospital in Boston, who oversees AD clinical trials at that institution.
ICAD featured abundant debate about a range of specific biomarker issues, particularly how to diagnose pre-dementia with biomarkers yet avoid false diagnoses. An entire session focused on how the field might be able to implement new diagnostic criteria proposed two years ago by Bruno Dubois at Hôpital de la Salpetrière in Paris and colleagues (see Dubois et al., 2007) and how the stage of “pre-AD” compared with amnestic MCI as defined by Ron Petersen of the Mayo Clinic in Rochester, Minnesota (in short: the two are quite similar; see upcoming story). But all the same, most scientists agreed that at the very least, the more validated ones among the current crop of biomarkers can already serve to ensure that people enrolled in anti-amyloid drug trials indeed have amyloid in their brain, or that people enrolled in AD drugs have hippocampal shrinkage or a pathological Aβ/tau signature in their CSF.
There is no officially anointed “consensus package” of predictive biomarkers yet, if ever there will be one, but some companies are not waiting for it. Word went around in Vienna that Bristol-Myers Squibb, for example, is now recruiting for a 75-center Phase 2 trial of their new γ-secretase inhibitor, BMS-708163, in what that company calls “prodromal AD.” Unusually, that new trial allows study in people as young as 45 and as old as 90, and it uses a combination of a subjective memory complaint plus a defined, low CSF Aβ42 ratio as an inclusion criterion. If people or their partners think their memory is off, and if their Aβ42 is indeed below a set cutoff, then they can ace the MMSE test with a perfect score of 30 and still enter the drug trial. The thinking is that people could be so well educated to begin with that the MMSE would not pick up subtle cognitive changes in them. In addition, a handful of drug companies are already using amyloid imaging either with the original Pittsburgh compound B (PIB) or with Avid Radiopharmaceutical’s AV-45 not as inclusion criteria or primary endpoint, but to gain additional information.
The field of amyloid imaging itself advanced in Vienna with the presentation of new data on four different F18-based amyloid tracers developed by GE Healthcare, Bayer Health Care, Avid Radiopharmaceuticals, and AstraZeneca, respectively. The key news there, in a word, was that they all perform quite similarly save for small technical differences, said Murali Doraiswamy of Duke University Medical Center in Durham, North Carolina. Doraiswamy presented the Phase 2 data of a multicenter trial, in 164 people, of AV-45 (which for a while went by the name florpiramine but now is referred to as AV-45 again, in case anyone got temporarily confused). Called florbetaben, the Bayer compound was tested in a Phase 2 study of 150 people in 18 centers, and the 18F-labelled PIB derivative GE067 was tested in a smaller multicenter study involving 70 patients and controls. Scientists from multiple groups agreed that together, the data to date suggest that all four agents likely will do a serviceable job of detecting whether people have a significant amyloid load in their brains. At the same time, they all have a tad more noise than C11 PIB and may be too crude a tool to quantify sub-threshold amyloid loads at very early stages of deposition, according to Chet Mathis of Pittsburgh University Medical Center, who co-developed PIB. ADNI-2, for its part, will likely use AV-45, and a small study performing a side-by-side comparison of PIB and AV-45, is gearing up to be run at the University of Philadelphia Medical School under the guidance of John Trojanowski.
As for ADNI, ICAD featured several dozen presentations on this 59-center observational biomarker study, as scientists from around the world avail themselves of the publicly available data at the LONI website. Moreover, this initiative is sparking new collaborative efforts to help with international standardization of biomarker use, for example, a budding effort to make surplus CSF from routine clinical practice in Sweden available as a standard against which other groups could eventually normalize the results of their observational or drug studies (see upcoming ICAD story).
And yet, in parallel to this building sense of optimism and expanding cooperation, other well-regarded researchers insisted that it is time to make a greater effort to move beyond the amyloid hypothesis, though they preferred to withhold their name from quotation. “It is becoming increasingly hard to take data year after year that are not only not mildly supportive but actually counter to the hypothesis, and to keep the amyloid hypothesis as the framework for understanding Alzheimer disease,” was how one leading investigator put it. Previous results of early stage immunotherapy trials are often cited in this regard, as are recent studies showing that very old people with dementia frequently do not have the signature pathologies of AD (see ARF related news story). At ICAD in Vienna, more ammo for this view came from puzzling data that Dimebon, the Russian drug currently in Phase 3 trials in the U.S., actually raises the release of Aβ from cells and in mouse models. This could mean several things. One interpretation is that if Dimebon manages to repeat in the U.S. the sizeable drug effect reported from a Russian Phase 2 trial, then there is much to be gained in the way of AD treatment without a focus on lowering Aβ, as do most other experimental drugs in large clinical trials these days. Such a result would boost morale among scientists who have felt disadvantaged in their interests outside the amyloid hypothesis. “If people will acknowledge that it is more complicated than the one-agent theory, then that opens up opportunities for scientists to pursue projects that they have hesitated on because of a perception that they are barking up the wrong tree, or that there is only one tree,” said one scientist. Follow-up studies of the new genes announced in the conference’s last session (see above), may well boost such a broadening of AD research, as they suggested that the risk genes for late-onset AD are functionally separate from the three genes that define familial AD. Those genes, APP and presenilin 1 and 2, fell far below genomewide significance, as did the gene for tau, implying that other problems will help explain at least some people’s genetic propensity to come down with Alzheimer disease late in life.
Another notable trend at ICAD suggested that interest in blood-based biomarkers is on the rise again. In the past decade, efforts to measure Aβ and tau in plasma and detect patterns that would predict AD have led to such inconsistent results between studies that many researchers had expressed a general sense that, at this stage of research at least, markers in plasma do not reflect what is going on in the brain. There was no clear, robust correlation that replicated from study to study. Because cerebrospinal markers do—the CSF is made by the brain’s choroid plexus, after all, and receives a constant influx of interstitial fluid from the brain itself—measurements of that body fluid have taken primacy in the past five years. That seems to be changing again. In Vienna, a number of presentations were testament to renewed efforts toward an eventual blood test for AD. To quote but one example, Jean-Charles Lambert of the Institute Pasteur in Lille, France, reported data from the Three Cities (3C) study in France. In this work, the authors constructed a case-control study from those people among a population-based cohort of almost 10,000 people who had entered this prospective study with a baseline blood test and some of whom then developed incident dementia four years later. At ICAD, Lambert reported that of a range of measures, only the ratio of Aβ42/40 and of truncated Aβn-42/n-40 at baseline were consistently associated with a person’s risk of getting a dementia diagnosis in four years, or, in other words, of having presymptomatic (aka prodromal or incipient) disease at baseline. People with a high ratio had a lower risk for dementia. This was consistent with a similar Dutch study reported previously (van Oijen et al., 2006). Lambert’s presentation focused on the most established biomarkers coming out of the amyloid hypothesis, but other studies are using unbiased approaches to trawl for new markers as well.
Beyond simmering debate about the amyloid hypothesis, the meeting featured several presentations that focused on dementia rates in the oldest old, the fastest growing population segment in some Western societies. One study in particular, by Ugo Lucca of the Istituto Ricerche Farmacologiche Mario Negri in Milan, Italy, punctured the myth that if people manage to dodge cancer or other major diseases in their sixties and seventies, they tend to stay mentally sharp and live surprisingly healthy lives through their eighties, nineties, and even into the 100s. This notion was introduced into the popular imagination by accounts of the centenarian studies in the 1990s, and also by reports of rugged Georgian centenarians who may not have always known their precise birth year but swore by a diet of yoghurt. But this happy notion does not hold up in comprehensive population studies of dementia in the oldest old. Lucca’s study took place in the Varese region north of Milan, Italy, where his colleagues went door to door in eight municipalities, assessed all adults older than 80 and reassessed them after three years. At ICAD, Lucca confirmed previous reports that the prevalence and incidence of dementia doubles relentlessly every five years in old age, to the point where all people in their mid-nineties meet NINDS-ADRDA criteria for dementia. The annual incidence of dementia was 20 percent by this age. Several researchers from France, the U.S., and other countries urged the field at large to pay more attention to AD in the oldest old and to make sure study participants represent the general population. For example, Jean François Dartigues at Bordeaux University, France, emphasized in his talk on the PAQUID prospective study of 3,777 elderly people in the Bordeaux region, that because many more younger AD patients (i.e., in their late sixties and seventies) tend to see doctors and participate in research than do people in their eighties and nineties, much of the available research creates a skewed picture of AD.
In other news, Dessa Sadovnik of the University of British Columbia presented the first report of a presenilin 1 mutation in a large North American aboriginal kindred with early onset familial AD who are living in Canada. Sadovnik emphasized the dilemma of trying to provide counseling, genetic testing, and care services to a population whom Western research understands as being afflicted with a severe form of a genetic disease, but who by tradition view their family’s condition as a spiritual rite of passage into old age that should be respected. Does that mean AD research physicians should leave them be? What kind of contact, support, and information is appropriate and helpful? Other reports at ICAD focused more broadly on the increasing recognition of sporadic AD in indigenous peoples in remote areas in India, Australia, and other countries. A plenary lecture by Gladys Maestre at Columbia University, New York, warned that the dementia prevalence in various Latin American countries is actually higher than in the U.S. for relatively young people, i.e., those in their sixties. This is particularly true among groups with little schooling, who are presumably less able to tap their cognitive reserve and compensate for underlying pathology than are highly literate groups. All these reports highlighted the challenges of furthering dementia awareness and providing care in developing countries. Because life expectancy is rising in many indigenous and developing populations, the number of dementia cases in these groups will inevitably rise, Sadovnik said. For more ICAD stories on specific topics, check this space.—Gabrielle Strobel.
Upper Belvedere Palace, Vienna
Image credit: Marc Aert, Creative Commons Attribution
- Vienna: New Shoot Among Ashes of Drug Trials
- ADAPT Safety Data Released—Controversy Persists
- Trials and Tribulations: Does ADAPT Have to Adapt?
- Safety Concerns Halt ADAPT Trial
- Oldest Old—The New Face of Dementia?
- Dubois B, Feldman HH, Jacova C, Dekosky ST, Barberger-Gateau P, Cummings J, Delacourte A, Galasko D, Gauthier S, Jicha G, Meguro K, O'brien J, Pasquier F, Robert P, Rossor M, Salloway S, Stern Y, Visser PJ, Scheltens P. Research criteria for the diagnosis of Alzheimer's disease: revising the NINCDS-ADRDA criteria. Lancet Neurol. 2007 Aug;6(8):734-46. PubMed.
- van Oijen M, Hofman A, Soares HD, Koudstaal PJ, Breteler MM. Plasma Abeta(1-40) and Abeta(1-42) and the risk of dementia: a prospective case-cohort study. Lancet Neurol. 2006 Aug;5(8):655-60. PubMed.