Cure Alzheimer’s Fund’s Dr. Rudy Tanzi’s submitted answers to Senator Hillary Rodham Clinton’s questions from the May 14, 2008, Senate Special Committee on Aging “The Future of Alzheimer’s: Breakthroughs and Challenges”
1. With regard to how we can “best identify those at risk for Alzheimer’s disease”, currently, we can only predict risk with 100% accuracy in patients with early-onset (<60 years), familial Alzheimer’s disease that carry a mutation in one of the three familial genes that we and other’s discovered between 1987-1995. These three genes are the amyloid precursor protein (APP) and presenilin 1 and 2 (PSEN1 and PSEN2) genes. We presently know of >200 different mutations in these three genes, which when inherited, cause early-onset Alzheimer’s with virtual certainty. These mutations are rare, accounting for only 1-2% of all Alzheimer’s and half of the early-onset, familial cases, e.g. the type that afflicts the family of Chuck Jackson who also testified at the hearing on May 14, 2008.
The majority of Alzheimer’s is the sporadic, late-onset (>60 years) form. We know from studies of identical twins, that at least 80% of the common “sporadic” late-onset form of Alzheimer’s also involves inherited genetic risk factors. The only confirmed genetic risk factor in this category is the APOE gene. A risk variant of this gene, called “epsilon-4” occurs in ~25% of the general population and in ~50% of Alzheimer’s population. Unlike the early-onset, familial gene mutations, inheritance of the APOE risk variant only confers increased risk for the disease, and does not guarantee onset. Thus, it is a “susceptibility” gene that requires other genetic and environmental factors to trigger the disease. As such APOE is neither necessary nor sufficient to cause Alzheimer‘s, and is not intended for use as a diagnostic or predictor of the disease. It is only approved for use as a “differential diagnostic”, i.e. for use in a patient presenting with dementia to help determine whether it is due to Alzheimer’s disease. Neither APOE, nor any of several dozen “putative” and unconfirmed Alzheimer’s genetic risk factors are approved for use as sole diagnostics or predictors of the common, late-onset, sporadic form of Alzheimer’s disease. To someday reliably and accurately predict late-onset Alzheimer’s disease, we must first “identify and confirm” the full set (likely dozens) of genetic risk factors that work together with each other (and environmental factors) to trigger this disease.
As an aside, it should be noted that companies like 23andMe, Navigenics, Knome, and DeCode are already charging considerable sums of money for anyone who wishes to pay to be tested for the “unconfirmed” genetic risk factors for Alzheimer’s and other common diseases, e.g. cardiovascular disease, cancer, and stroke. In my view, it is highly premature and both medically and commercially irresponsible to be conducting these tests. To reliably predict disease risk, we will first need to establish the full set of “confirmed” risk factors and then determine how they work together to influence risk in a “multigenic” manner. As these companies become more popular, the public will need to be increasingly informed and educated about the fact these tests are not yet accurate, reliable, or scientifically sound. I am concerned that these tests may increasingly lead to unwarranted anxiety or a false sense of security about one’s genetic destiny as these companies services become more “trendy”.
In specific response to your question about how we can best identify the full set of genetic risk factors for Alzheimer’s disease, we must first “identify” novel Alzheimer’s gene candidates in genetic association studies and then attempt to “confirm” them by testing them for replication in independent Alzheimer’s samples. We are approaching this in two ways. First, we have established a very successful and higlhy accessed website called AlzGene (http://alzgene.org), which is supported by the Cure Alzheimer’s Fund. This site compiles and systematically displays all of the data generated in all available publications (>1600) that have addressed Alzheimer’s genetics. Most importantly, for novel genetic risk factors that have not yet been confirmed but are gradually being tested for replication in multiple independent Alzheimer’s populations, we compile all of the published data for the candidate risk factor and perform genetic analyses on the sum data to determine which novel genetic risk factors for Alzheimer’s have the highest likelihood of being confirmed as bona fide risk factors for Alzheimer’s disease. To date, over 1500 gene variants have been tested as genetic risk factors for Alzheimer’s, of which we (AlzGene) have found that only 29 have yielded statistically significant results toward confirmation. Every week, we update these analyses with the ultimate goal of establishing the complete set of confirmed Alzheimer’s genetic risk factors, which determine one’s predisposition for the common, late-onset form of Alzheimer’s. With the overwhelming success of AlzGene, we have established similar sites for Parkinson’s disease (http://pdgene.org) and schizophrenia (http://szgene.org ). The CDC has recently indicated interest in eventually doing the same for all common human disorders with complex genetics. For the success of all these efforts, Alzheimer geneticists will have to work closely with each other and patients and their families to test candidate risk factors in as many independent Alzheimer’s populations as possible. GINA should go a long way in providing protection to patients and their family members who participate in these studies. However, GINA covers employment and health insurance, but not life insurance or long-term care insurance. Thus, I believe that there is still more work to do on a comprehensive genetic privacy act as we move into the age of personalized medicine.
A second strategy for finding the remaining Alzheimer’s genes is our Alzheimer’s Genome Project, which is described in more detail below in answer #3.
With regard to when we will be able to do routine genetic testing for life-long risk of Alzheimer’s disease and other common age-related disorders, e.g. stroke, diabetes, cancer, currently, we can already reliably predict many of the rare, early-onset, familial forms of these diseases, which generally represent 1-2% of these diseases. But, for the vast majority of cases, which are late-onset, we will first need to identify and confirm dozens of genetic risk factors that work in concert to determine one’s life-long risk for disease. For Alzheimer’s and other common, complex genetic diseases, we have established four “confirmed” genetic risk factors and are still investigating dozens of “putative” risk factors that have yet to be confirmed. To reliably and accurately predict disease risk, we will ultimately need the complete set of “confirmed” risk factors and we will need to understand how they work together in a “multigenic” manner. While great progress is being made, these are still the early and “pioneering” days of this effort. Great progress is being made via AlzGene, the Alzheimer’s Genome Project and other Alzheimer’s genetics efforts. However, given the scientific challenges of identifying and confirming novel genetic risk factors, I believe that it will take 5-10 more years to assemble the first reliable multigenic tests for late-onset Alzheimer’s disease and other common, age-related, complex genetic diseases. Routine testing should be possible in 15-20 years. And once again, the genetic testing currently being sold by companies such as 23andMe, Navigenics, Knome, and DeCode is, in my opinion, entirely premature and scientifically unsound, and it would be prudent to educate the public about this. I and other geneticists are currently doing so through the media, e.g. in an upcoming episode of Nova on PBS.
2. The second question regards the new drugs that target toxic A-beta molecules in the brain. These drugs are aimed at retarding disease progression by curbing the accumulation of the neurotoxic protein, A-beta, in the brain. The four established AD genes (APP, presenilins 1 and 2, and APOE) have taught us that the common pathological feature in the AD brains of patients carrying defects in any of these four genes is the excessive of accumulation of neurotoxic A-beta. There are two basic ways to lower A-beta levels in the brain: Promote the clearance of A-beta from brain, or curb the production of A-beta in the brain. Details on the anti-A-beta drugs currently in development and their prospects for success are attached in a separate word file.
With regard to the predicted timeline, I believe the first anti-A-beta therapies should hit the market in 2-3 years, but as is often the case with the first wave of therapies, these will not necessarily be the best ones. They will, however, open the door for more effective versions of anti-A-beta therapies, which should come on line in 5-7 years.
With regard to safety, generally, I believe this class of drugs will be well tolerated with one exception. We will need to carefully watch for adverse events, e.g. micro-hemorrhages and encephalitis, from immunotherapy approaches involving active vaccination or passive immunization.
3. Regarding the third question of how critical is “understanding the genetic causes of Alzheimer’s in developing a treatment and eventual cure”, the vast majority of researchers and clinicians in the Alzheimer’s field would agree that the contribution of genetics to solving the mystery of Alzheimer’s disease has been unmatched and unprecedented. The genetic component of Alzheimer’s disease is very strong with at least 80% of cases involving inheritance, according to large twin studies. Most of what we now know about the etiology and pathogenesis of Alzheimer’s disease has come from the discovery and characterization of the four known Alzheimer’s genes (APP, PSEN1, PSEN2, and APOE). Moreover, most Alzheimer’s therapies currently in development, e.g. anti-A-beta therapies have been made possible from studies of the four known Alzheimer’s genes, particularly, three early-onset genes.
In 1987, we, and others, reported the isolation of the first AD gene (APP) then went on to co-discover two more early-onset genes (presenilin 1 and 2) in 1995. I wrote about these discoveries and their impact on Alzheimer’s research in my book “Decoding Darkness: The Search for the Genetic Causes of Alzheimer’s Disease” and would be more than happy to send the Senator and her staff a copy. In addition to these early onset genes, a late-onset genetic risk factor gene, APOE, was discovered in 1992. These four genes account for only 30% of the inheritance of AD with 70% still remaining a mystery. If one considers what the field has accomplished with the known genes, imagine what we can achieve with the remaining 70%. Every new gene we identify and confirm as a bona fide genetic risk factor in AD provides a new biological target for drug discovery while also enhancing our ability to predict and diagnose the disease. While, I am generally optimistic about the ongoing clinical trials of anti-A-beta therapies, we must ready ourselves for the possibility that they may not be sufficient to fully treat or prevent the disease, or may even fail. This is why we must identify the genes underlying the remaining 70% of the inheritance of Alzheimer’s. As history as shown us, every new gene we can identify will provide another shot on goal to effectively treat, prevent, or even cure this disease.
To elucidate the complete set of Alzheimer’s genes, labs all around the world are trying to identify the remaining AD genes. Toward this end, we are carrying out the AlzGene project (described above in answer 1 above) and the Alzheimer’s Genome Project (AGP). The AGP, which is based in my laboratory at Massachusetts General Hospital, is a three-year, approximately $3 million effort mainly being funded by the Cure Alzheimer’s Fund with additional support from the NIMH and NIA. We are scheduled to publish the first set of results by the summer of 2008. The AGP is the first family-based whole genome association study for Alzheimer’s disease, being carried out in over 1300 AD families. This study requires the newest technology, e.g. microarray genotyping “chips”, sophisticated statistical analyses, large family samples for DNA, and especially, the many databases made possible by the NIH-funded human genome project. The databases have been absolutely essential, if not indispensable, to the success of the AGP and other efforts like it. They provide details about individual genes as well as the structure and organization of the human genome. We need this information to interpret or genetic findings from studies of patients and their family members.
4. The fourth question regards the relationship between Alzheimer’s and traumatic brain injury (TBI). After age, family history, and gender, the greatest risk factors for Alzheimer’s disease are head injury and stroke. Over the past several years, we and others have discovered both stroke and TBI significantly increase production of the neurotoxic A-beta protein in the brain. This, in turn, leads to increased risk for Alzheimer’s disease over the ensuing years following injury. In 2007, we published the molecular mechanism by which stroke leads to increased generation of A-beta in the brain. Over the past year, we have found that TBI increases cerebral A-beta levels in the same manner. Consequently, we believe that those who suffer from a stroke or undergo head trauma, e.g. soldiers in Iraq and Afghanistan, also incur increased risk for Alzheimer’s disease. We are currently studying the molecular mechanism underlying the stroke/TBI-induced increase in A-beta in order to develop strategies to reduce A-beta generation immediately following brain injury. Such therapies could include the anti-Abeta drugs currently in clinical trials for the treatment of Alzheimer’s. If successful, one could envisage a medical protocol in which patients entering the emergency room or soldiers undergoing TBI in the battlefield would immediately be given such drugs to help ward off downstream risk for Alzheimer’s disease later in life.