Is immunotherapy the ultimate solution for Alzheimer’s disease?
This essay was written by Marina Lynch and was first published in the 2010 Mill Hill Essays.
Just over 100 years ago Auguste D, a 51 year-old woman, was admitted to the state asylum in Frankfurt under the care of the psychiatrist Alois Alzheimer. She was suffering from loss of memory, progressive cognitive impairment, hallucinations, delusions and psychosocial incompetence. When asked to write her own name, she was unable to and repeated: “I have lost myself” (Ich hab mich verloren). This sentiment was echoed, perhaps more eloquently, by Abraham Schweid, an academic pathologist who was diagnosed with Alzheimer’s disease in 2003. In 2005 he said (see footnote):
I can’t finish my ideas.
My words are upside down.
When I begin an idea,
It’s not there when I go back to it.
According to the World Alzheimer’s Report, one person is diagnosed with Alzheimer’s disease every 7 seconds and an estimated 35.6 million people worldwide are living with dementia in 2010. This number is expected to nearly double every 20 years. Because of progressive decline across a broad range of cognitive functions, daily living tasks become increasingly more difficult and the increasing dependency of the sufferers means that the current and projected cost, both in economic terms and in societal terms, is huge.
Auguste D died five years after she was hospitalized. Alzheimer, now in the medical school in Munich, and his mentor Emil Kraepelin undertook a post-mortem examination of her brain. They linked the deterioration in cognitive function with histological findings, which are the hallmarks of the disease: the presence of abnormal accumulation of fibrous proteins, or plaques, around cells and the presence of microscopic fibrous tangles within cells, coupled with extensive loss of nerve cells (neurons). These findings were published in 1907 and the condition became known as Alzheimer’s disease.
An insoluble protein called amyloid-β is the main component of plaques. In the past two decades or so it has been established that specific forms of amyloid-β are capable of causing neuronal death and also capable of inducing inflammation. These are well-established features of Alzheimer’s disease and therefore it was proposed that amyloid-β was a major contributory factor in the pathogenesis of Alzheimer’s. This idea was formalized with the development of the so-called amyloid hypothesis, which proposes that amyloid-β is the causative factor in Alzheimer’s disease.
Alzheimer’s disease can be divided broadly into two subtypes, early-onset or familial Alzheimer’s, and late-onset Alzheimer’s. There is a strong genetic component to familial Alzheimer’s; persons with inherited mutations in certain genes have a 50:50 chance of developing the disease. Amyloid-β is derived from another protein called amyloid precursor protein. Presenilin is a protein involved in the removal of potentially damaging proteins. Mutations in the genes that code for amyloid precursor protein or presenilin confer a risk of developing familial Alzheimer’s but this form of the disease accounts only for about 5% of sufferers. There is, however, also a genetic risk factor for people who develop late-onset disease. There is a gene that codes for a protein called apolipoprotein E (ApoE), which is involved in the transport of certain fats. The risk of developing Alzheimer’s disease is associated with inheriting one particular form of the ApoE gene; this is identified as the ε4 form (ApoEε4) and the greater the number of copies of this gene that are inherited, the greater is the risk of developing Alzheimer’s disease. However, only about 40% of those with late-onset Alzheimer’s inherit the ApoEε4 gene and so the gene is only one of the risk factors.
Current strategies for the treatment of Alzheimer’s disease are only minimally effective. Part of the reason for this is the absence of methods to diagnose the disease at an early stage. In many cases it is likely that significant neuronal loss has occurred before diagnosis and in this scenario, because restoration of function is unlikely, treatments focus on limiting further neuronal loss. Prevention of the disease is the ultimate goal and it is interesting that the risk of developing Alzheimer’s is reduced in individuals who are treated long-term with non-steroidal anti-inflammatory drugs for conditions such as rheumatoid arthritis. There is some evidence indicating that the cholesterol-lowering drugs, statins, also offer some protection but the evidence is less compelling. However, it is highly unlikely that either strategy will ever be adopted as a means of protection against the disease because neither drug offers complete protection and, like all drugs, however safe, both have side effects.
Immunotherapy is perhaps the strategy likely to offer the greatest hope for the prevention of Alzheimer’s disease and for the treatment of mild or moderate Alzheimer’s, particularly as diagnostic tools to identify early disease become more sophisticated. The primary aim of immunotherapy is to eliminate amyloid-β, accumulation of which is detrimental to the health of neurons, and so it is based on the amyloid hypothesis. There are two possible immunotherapy options. The first is active immunotherapy, where amyloid-β is injected and triggers activation of specific immune cells, T cells. These are lymphocytes, which are a subtype of white blood cells, and their activation plays a pivotal role in immunity. A sequence of events is set in motion resulting in the production of specific antibodies. The second option is passive immunotherapy, where an antibody is injected by-passing the need for activation of the immune cells. In both cases, the objective is to ensure that the antibody will prevent accumulation of amyloid-β and therefore the development of the amyloid-β-containing plaque.
Studies into the possibility of using immunotherapy to treat Alzheimer’s disease began in 1995, after the creation of a mouse model of the disease which developed amyloid-β-containing plaques and showed a deterioration in cognitive function with age. The mouse was engineered to produce increased amounts of a mutant form of a human amyloid precursor protein that was isolated from a Swedish family with the inherited form of the disease. The first trial of active immunotherapy was conducted in 1999 and the authors reported that repeated immunization of these mice with amyloid-β prevented plaque formation in young animals, and reduced the plaque burden and prevented cognitive decline in older animals. It was concluded that this active immunization protocol stimulated the production of anti-amyloid-β antibodies and initiated the removal of amyloid-β by microglia, which are the cells in the brain responsible for removing cell debris by a process called phagocytosis. A short time later, another group undertook a similar experiment in a nonhuman primate, the Caribbean vervet, with a similar outcome. In this case, the clearance of amyloid-β from the brain was linked with a decrease in the amount of amyloid-β in the cerebrospinal fluid, the protective fluid that circulates around the brain and spinal cord of the central nervous system, and an increase in the amount of amyloid-β in the plasma. These studies identified two methods whereby amyloid-β-containing plaques can be cleared from the brain: antibody-mediated removal of amyloid-β by phagocytosis, or transport of amyloid-β from the brain. A third possible method is chemical modification of amyloid-β leading to the dissolution of plaques. Theoretically, approaches to stimulate any of these mechanisms should be beneficial; immunotherapy focuses on the first.
The promising preclinical results led the drug companies Elan and Wyeth to initiate in 1999 the first Phase I clinical trial with their amyloid-β vaccine, called AN-1792. This trial involved 80 patients with mild to moderate Alzheimer’s disease. The vaccine consisted of a synthetic amyloid-β peptide and another substance, called QS-21. This is an “adjuvant”, which is necessary to activate the immune response that leads to antibody production in the immunized individuals. There were no apparent side-effects following the four intramuscular injections over a six-month period but few of the patients developed antibodies. As a result, the vaccine was modified by the addition of a substance called polysorbate 80, which was designed to boost the immune response and therefore increase antibody production. This modified vaccine was used in the Phase IIa clinical trial which began in 2001, with an enrolment of 372 patients, but the trial was aborted when a small number of vaccinated patients developed meningoencephalitis, an inflammatory condition of the meninges and brain. The cause of the excessive inflammatory response leading to the meningoencephalitis has been the subject of intense research and debate. One possibility is that it resulted from the addition of polysorbate 80 to the vaccine, which led to unacceptable activation of immune cells, specifically T cells. In 2001, this solubilizing agent was considered to be inert, but it is now known that it can induce extreme allergic reactions. It was initially suggested that patients who developed meningoencephalitis might be those who exhibited the most profound immune reaction to the injected amyloid-β (and therefore who produced the greatest amount of antibody). However antibody levels did not correlate with the development of meningoencephalitis.
Despite these setbacks, some important findings have been reported and follow-up analysis is still ongoing. Of the 80 patients enrolled in the Phase 1 trial, 42 died before or during follow-up. Post-mortem examination of eight of these individuals who had been vaccinated with AN-1792 indicated that vaccination markedly reduced the number of plaques in the brain. Plaque clearance was correlated with antibody level but not with cognitive performance because all eight individuals showed symptoms of severe dementia before death. The most likely explanation for the persistent dementia is that the neuronal damage was too advanced by the time of vaccination but these individuals also did not complete the trial and received only one or two vaccinations.
Neuropathological analysis has also been completed on three of the 18 individuals who developed meningoencephalitis in the Phase II trial. This revealed that few plaques were observed in these patients, indicating efficient clearance, but there was evidence that cells which are normally found in the circulation, but not in the brain tissue, had infiltrated the brains of the patients. In contrast, while plaque clearance occurred in the brains of vaccinated patients who did not develop meningoencephalitis, there was no significant infiltration of T cells and this led researchers to conclude that the presence of these cells was likely to be the cause of the excessive inflammation. Studies on another 14 individuals have confirmed that vaccination does indeed cause a marked reduction in the number of plaques, but the persistence of some of the other pathological features of Alzheimer’s disease like the presence of microscopic fibrous tangles within cells and amyloid-β accumulation in cerebral blood vessels (another of the effects of the disease) appears to resolve over time. Initial indications were that plaque clearance was accompanied by a slower rate of cognitive decline but the most recent follow-up data suggest that this improvement is confined to a subgroup of patients where antibodies persisted. At this point, the conclusion is that immunization is likely to be beneficial in a specific cohort of individuals but it should be emphasized that, although other trials have been completed, the only comprehensive set of data published is from the AN-1792 trial.
Efforts are currently focused on engineering vaccines that avoid the adverse side effects and this is being done by immunizing with specific fragments of amyloid-β, rather than using the whole protein or by linking amyloid-β with proteins that control cell activation. In theory, either approach should prevent the unwanted side effects observed to date but the results of the ongoing clinical trials will not be available for at least 12 months.
Passive immunotherapy, i.e. when anti-amyloid-β antibodies are administered directly, is the second major immunotherapeutic approach to Alzheimer’s. In this case, because the patient’s own immune system is not activated, the effects of the therapy are short-lived. However, this approach avoids the risk of T cell-mediated inflammation of the meninges and brain, and it also eliminates the problem of a poor response to antibody (which can be a problem in older individuals because the efficiency of the immune system decreases with age). Animal studies indicated that weekly immunization for six months resulted in a decrease in the number of amyloid-β plaques and an improvement in memory, but microhaemorrhages were observed in some animals. Initial Phase I trials generally showed good safety and tolerability with some evidence of improved cognitive function. An additional encouraging finding was reported at the International Conference on Alzheimer’s Disease in July 2010. The results of one clinical study have found that appropriate screening of patients by magnetic resonance is likely to avoid the side effects of a potentially powerful passive immunotherapy drug, apineuzumab. Meanwhile, another passive immunotherapy drug, bapineuzumab, is being evaluated in two Phase II trials and seven Phase III trials, several of which are actively recruiting participants.
Intravenous immunoglobulin (IVIG) treatment is injection of a cocktail of natural human immunoglobulins that includes amyloid-β antibodies. It is also being considered as a possible treatment for Alzheimer’s. This has been used since 1982 as a treatment for autoimmune neurological diseases like myasthenia gravis. More recently IVIG has been shown to promote amyloid-β clearance from the brain and protect neurons from the neurotoxicity induced by amyloid-β. The reports of a pilot study published in 2004, suggested that amyloid-β was decreased in the fluid which bathes the brain, the cerebral spinal fluid, and increased in blood serum. This indicates that amyloid-β is being removed from the brain and, importantly, this was accompanied by a stabilization of cognitive function in some patients. It has since been established that the changes in amyloid-β distribution were maintained only during treatment, and not after treatment was stopped. Results of a Phase II trial by Baxter International have suggested that there is reduced cell loss and improved cognition, over an 18 month period, in patients given their IVIG therapy, Gammagard, compared with placebo-treated patients. A Phase III trial is currently underway and is due to be completed in July 2011. The potential of IVIG treatment is also being evaluated by other companies.
Immunotherapy is designed to eliminate amyloid-β accumulation and therefore presupposes that the causative agent in Alzheimer’s disease is amyloid-β. However this view continues to be challenged by some researchers, although research in the area continues apace. Perhaps we should not overlook the convincing epidemiological evidence which indicates that preventing inflammation reduces the risk of developing Alzheimer’s and consider the combination of anti-inflammatory therapy and immunotherapy. Huge strides have been taken in the development of immunotherapies for the treatment of Alzheimer’s since the initial Elan clinical trial in 2001 and the optimism that this will offer the first truly beneficial therapeutic after 100 years is palpable. More important, perhaps, is the prospect that further development and research will, at last, result in a vaccine that will eradicate this most devastating of diseases.
Abraham Schweid’s wife recorded his words and later published them:
Dear Alzheimer’s why did you pick our sheltered lives to visit?
Helfgott, EA. (2009)
Journal of Poetry Therapy 22(4):185-217. Abstract.