(sorry, hacked up version of the cover above)
The National Institute of Aging-Alzheimer Association Research Framework (RF) is a scholarly plan for further research in Alzheimer’s Disease (AD) and dementia, published last April. My letter to the editor is now published, journal cover above. I submitted my comments in May, but they are being published just now, eight months later. The article’s scholarship is massive, and I do not question the laudable idea of proceeding with work in biomarkers. [Because the comments are now behind a paywall, I will add them, as submitted, below]
However, I am first and foremost concerned about how the authors want to change the very nomenclature and nosology (disease classification) of Alzheimer’s Disease without really being explicit about their ideas of basic or sequential pathophysiology, since I understand the ideas are evolving. They would like to change terminology based on their biomarker based “biological construct,” noting that the term “Alzheimer’s Disease” is rather imprecise. I was gratified that the Cochrane Dementia Group at Oxford preceded my letter with a similar concern to mine.
As I said in my letter, the terms “leukemia” and “cancer” are not precision-driven terms, but are useful for lay understanding without requiring molecular or biomarker nuances. I recommend using something like “dementia spectrum disorder” with subtypes as they discover them. One can do precision medicine with jargon while preserving general terms for public understanding.
What I really wanted to say is that the RF is really detailing plans to extend old findings, but certainly worth doing. The biomarker data as I understand it in general does correlate with the irretrievable progression of AD, and might even pick up pre-symptomatic subjects. During recent BACE or monoclonal antibody studies, biomarker targets were hit, as demonstrated with imaging or spinal fluid changes, but no real beneficial effects were seen for patients. There were side effects, including hippocampal shrinkage after one study.
Some are suggesting that finding subjects with a certain thresh hold of biomarkers may make them too late for intervention, so seeking cohorts with even fewer biomarkers might be ideal. My understanding is that these cohorts exist, but not all will develop clinical dementia, and the whole thing needs more longitudinal observation. The FDA implies as much in their Feb 2018 Draft Guidance for AD drug development.
But what if biomarkers are really a dead-end already, like seeing vultures in a wilderness, if not yet tombstones in a graveyard? They denote that something has happened, perhaps irretrievably, perhaps not. Amyloid beta and tau are proteins, why not consider them like scar tissue collagen, a disordered array in a keloid or hypertrophic scar? There’s some understanding of that kind of ectodermal phenomenon, but once the big scar is established, even cosmetic surgery isn’t always satisfactory. Are there parallels in the neuroectodermal origin of the brain?
As mentioned, one of the tantalizing but frustrating features of recent anti-amyloid clinical trials is that target engagement was evident, but the subjects did not improve, and furthermore the progression of disease did not seem to halt (that’s reading a lot into what little data is presented in papers). If that’s true, the progression seems relentless, as seen without intervention.
Progression, based on cellular proliferation, is one of the hallmarks of cancer. I had the privilege of caring for several children with Burkitt’s lymphoma, thought to have the fastest doubling time of any human tumor. The first principle is to halt the progression of a tumor before it does irretrievable damage or impinges on a critical structure like the airway or spinal cord (more common in neuroblastoma). We use urgent radiation, chemo, even surgery in those instances.
Alzheimer’s progression doesn’t seem that dramatic, time wise, but can be just as devastating. One of the agents used in leukemia that helps halt progression is asparaginase, which affects protein production. Given IV or IM, it also gets into the CSF, where asparagine depletion can be measured. Could an agent like that be given to try to halt progression? Something similar has been proposed.
Using asparaginase would be to consider AD pathophysiology a “black box”. One could justify its use in some good animal model against amyloid beta or tau because they are proteins, but the unspoken hope would be that it would halt progression, probably by messing up some as yet unidentified protein factor upstream from amyloid. Yes it has side effects, we even reported one of our patients who had a cerebral venous sinus thrombosis.
I don’t claim to be a great historian of medical science, but I know something of cancer therapy development, and it wasn’t always nice and tidy and rational. I’ve already made my plea to make AD clinical endeavors more like cancer clinical trials. The Alzheimer Clinical Trials Consortium was created in late 2017, but I’m not sure what has been accomplished so far, aside from a subject recruitment statement and an aspirational document.
The other point in my letter about the RF is that it seems to provide an achievable path for faculty development in a very difficult field, perhaps to the detriment of other activities. Faculty development is critical, but I hope the RF doesn’t dominate all grant funding.
Now from STAR WARS: “The Greatest Teacher, Failure Is….” (Yoda)
To the Editors:
The 2018 NIA-AA Research Framework (RF) (1) for Alzheimer’s dementia investigation is an imposing document, with its organizational imprimatur and eminent authorship. Much more than a scholarly review or perspective, it almost creates a “central dogma” to be refuted or confirmed, basically requiring validation before “a precise approach to interventional trials” can proceed. The authors recognize the full force of this document; they include unusual statements about its potential misuse, even in the abstract.
One major concern is the way the authors might create confusion by redefining the term “Alzheimer’s Disease (AD)” itself, to fit their RF “biological construct” concept. AD would then be used by a select few for a select few, in the context of biomarkers. Everyone else would be expected to adopt “Alzheimer’s Clinical Syndrome,” especially for patients with symptoms. I would recommend that they drop this proposed change in nomenclature.
The authors state that the pre-mortem use of the AD term is usually incorrect, if one adheres to a certain strict definition (from 1984). But why change its common usage to imply a level of precision apparently yet to be defined, and to satisfy the precise needs of so few? “Cancer,” even “leukemia,” are imprecise terms; they do not require definition by biomarkers, but are vastly useful for the public and for medicine.
The authors acknowledge the public and clinical use of the term “Alzheimer’s Disease.” It is already “common language” and commonly understood, mainly as a behavioral, cognitive and functional construct, without the research subtlety of biomarkers or pathologic confirmation. ICD-10 uses the term.
This is more than a scientific semantic issue: public disease awareness, along with economic, social and caregiving issues are important arenas for discussion, and require understandable general concepts, not nuances.
There is some historical irony here. The term “Alzheimer’s Disease” was designated over a hundred years ago, after Alzheimer described a case which today might be called “early onset AD (EOAD).” His subject was a 51 year old woman with dementia who evidently had brain pathology that resembled that found with typical “late onset” symptoms and pathology (2).
Recent studies using neuroimaging biomarkers suggest differences between EOAD and “late onset” AD (3,4). Should EOAD prove to have a distinct biomarker pattern, perhaps different from the proposed “Alzheimer’s continuum,” wouldn’t one expect that the term “Alzheimer’s Disease” be reserved for that specific entity?
Clinical medicine will avoid misusing RF concepts if biomarker researchers can create a wholly different nomenclature, so there is no confusion. Other terms, such as “dementia spectrum disorder,” or “progressive cognitive impairment disorder,” are less specific but might be more accurate in the RF biomarker context.
Perhaps using the suggestion in Text Box 5, which describes avoiding the term AD altogether, might be the best approach. By the way, shouldn’t the authors use the term “biomarker change” instead of “pathologic change,” if tissue is not directly examined? Those trained in pathology might be confused by the ambiguity of that term.
My gravest concern is that the RF, even with its own statements about not meaning to restrict clinical research, will nevertheless occupy the careers of several successions of young investigators, with proportional funding. It also creates a platform that will dominate the literature, as it implies a vast set of program projects, despite “reassurances” from three companion articles in the same issue.
Investigators must have academic production to succeed, and one can foresee that biomarker hypotheses will be testable and reportable. The linear and sequential hypotheses proposed in Figure 6 will be the guide for many grants, I suspect. In contrast, clinical trials have a frustrating history, especially in cognitively impaired subjects. Clinical investigations that don’t dovetail with the RF would seem to have few incentives for young investigators.
Defining a Cognitively Unimpaired A+T+ (N)— group might be ideal to test an intervention, if one reads the FDA 2018 draft guidance for early stage AD (5). That level of cohort selection already seems technically available, but the group’s natural history evidently needs further observation. The RF mentions longitudinal studies, but what seems missing in the document is a practical clinical strategy to accomplish that goal.
The goal of eventually understanding enough pathophysiology to enable successful disease-modifying therapy is laudable, and I appreciate the massive scholarship that went into this document. Despite my concerns, I sincerely hope the RF somehow accelerates the discovery of disease-modifying therapy.
Respectfully,
Ron Louie, MD
References:
- Jack CR, Bennett DA, Blennow K, Carrillo MC, Dunn B, Haeberlein SB, et al. NIA-AA Research Framework: Toward a biological definition of Alzheimer’s disease. Alzheimers Dement 2018;14:535 – 562.
- Ryan NS, Rossor MN, Fox NC. Alzheimer’s disease in the 100 years since Alzheimer’s death. Brain 2015; 138:3816–3821.
- Kaiser NC, Melrose RJ, Liu C, Sultzer DL, Jimenez E, Su M, et al. Neuropsychological and Neuroimaging Markers in Early Versus Late-Onset Alzheimer’s Disease. Am J Alzheimers Dis Other Demen 2012; 27: 520 – 529.
- Koychev I, Gunn RN, Firouzian A, Lawson J, Zamboni G, Ridha B, et al. PET Tau and Amyloid-β Burden in Mild Alzheimer’s Disease: Divergent Relationship with Age, Cognition, and Cerebrospinal Fluid Biomarkers. J Alz Disease. 2017; 6:283-293.
- US Food and Drug Administration, Early Alzheimer Disease: Developing Drugs for Treatment, Draft Guidance, February 2018. https://www.fda.gov/downloads/Drugs/GuidanceComplianceRegulatoryInformation/Guidances/UCM596728.pdf
A thoughtful piece. However, though you note Burkitt’s lymphoma with great interest, it is strange you do not remind readers it is associated with Epstein Barr virus, and malaria–two infections. And thereby miss the chance to point out many clues are emerging that Alzheimer’s can likely be triggered by infectious agents. Tragically, major grantmakers have so far provided only paltry funds to deeply investigate these leads.