This paper was covered by CNN last week, and I didn't have a chance to talk about it then. It is a case study by Tobinick and Gross in the Journal of Neuroinflammation where a patient with Alzheimer's disease (AD) injected intrathecally (into their spinal fluid) with a TNF-alpha antagonist called etanercept showed rapid and consistent improvement in cognitive function.
I want to talk both about the reasons for skepticism and for optimism in this study.
Tobinick and Gross describe an 81 year old patient with late Alzheimer's disease. The patient was being treated with intrathecal administration of etanercept -- a recombinant antagonist for TNF-alpha -- as part of a larger 6 month pilot study looked at etanercept's effectiveness in treating AD. (Incidentally, it is always good to disclose that Dr. Tobinick has financial connections with Amgen, the maker of etanercept, and he holds patents for its use in treating Alzheimer's.)
TNF-alpha is a pro-inflammatory cytokine, meaning that it is a signaling protein secreted by cells to trigger inflammation in the case of infection or damage. For example, it is highly upregulated during rheumatoid arthritis, one reason why drugs like etanercept were developed. Etanercept is a recombinant protein -- in this case sort of like an antibody -- that binds and neutralizes TNF-alpha. It has FDA approval for treatment of inflammatory conditions like rheumatoid, but the authors were attempting an off-label study to determine whether it would be effective for AD.
The reason that the authors suspected it would be effective for AD is that -- like most cytokines -- TNF-alpha has many functions. In addition to its pro-inflammatory function, evidence suggests that it TNF-alpha may regulate activity in glial cells -- the other cells in the nervous system besides neurons -- which in turn regulate the activity in surrounding neurons. (I posted a while back on neuron-to-glia synapses here and here. The more we know about glia the less we believe that they are just there for structural support. Glia do a lot to regulate the rate of synaptic transmission by taking up excess neurotransmitters and managing neuronal metabolism. In a sense, they appear to regulate the tone of synaptic transmission.)
As I understand it, TNF-alpha is proposed to have two general roles in the brain:
- It has an immune function which excites the cells that cause inflammation and scarring -- astrocytes and microglial among others. This inflammation is bad because it exacerbates neuron death associated with AD.
- It also has a synaptic regulatory function which has to do with the tone of synaptic transmission. With respect to this function, there is a set-point for TNF-alpha levels which is ideal. Too little, and you have too little transmission. Too much, and you have inflammation and neuron death. Here is that role in the author's words:
The authors hypothesize that excess TNF-alpha in Alzheimer's disease[5,7-10,12,14,16,18,23] interferes with the synaptic regulatory functions of TNF-alpha. When TNF-alpha is in a normal physiologic range synaptic scaling is enabled, thereby preserving optimal functioning of the brain's neural network. When TNF-alpha is overexpressed, due to glial activation, it is postulated that the synaptic regulatory activities of TNF-alpha are disturbed. Synaptic dysfunction is hypothesized to result from this dysregulation, which may provide a basis for reduced functional connectivity between brain regions in Alzheimer's disease [57,58]. The rapid effects of perispinal etanercept are hypothesized to be the result of rapid neutralization of excess TNF-alpha, which thereby ameliorates this synaptic dysregulation, allowing normal cross talk between different regions of the brain.
Putting aside the theoretical discussion of TNF-alpha in AD, the question for all potential AD treatments is do they modulate the course of the disease or do they simply hide the symptoms? Many treatments have foundered on the rocks of just helping the symptoms. The best examples of these are nearly every drug on the market for AD today, including donepezil hydrochloride (Aricept), rivastigmine (Exelon), and galantamine (Razadyne -- previously called Reminyl). These drugs are all what are called cholinesterase-inhibitors. They inhibit an enzyme in the brain called acetylcholinesterase which breaks down a neurotransmitter called acetylcholine (ACh).
Let me make this clear: in no way do these drugs modify the disease course. Acetylcholine is released by a brain stem system that is involved in arousal, and it is true that patients with AD have less ACh. However, these drugs do not change the underlying etiology of the disease -- namely, that neurons are still dying. Basically all these drugs do is turn up the volume in the brain, giving the patient a couple more valuable months of intact cognitive function. They do have side-effects, but in many cases the patient's families are willing to make the trade.
It is in this context -- awareness that many drugs for AD do not alter disease course -- that I propose to interpret the case study of Tobinick and Gross.
The authors administered a battery of mental status exams both before and immediately after treatment with etanercept. What they found is that the patient improves nearly immediately -- "within minutes" -- of etanercept treatment. When the patient was tested on the cognitive battery immediately after the treatment, he did much better on a variety of benchmarks.
This would be good news were it not for my cautionary note above. I find it difficult to believe that the drug changed the disease course -- inflammation, neuron death, etc. -- in a manner of minutes, so we can probably attribute the immediate improvement to changes in synaptic transmission and arousal. Changes in inflammation probably happen in hours, and reductions in neuron death would probably be days or weeks. On that score, the etanercept is probably behaving much like existing medications -- i.e. by modifying synaptic transmission. This is the part of the paper that I am skeptical about, namely because the immediate improvement was so hyped by the media.
On the other hand, I am much less skeptical about the long-term evaluation of this and other patients. The authors continued to administer the drug on a weekly basis for six weeks, and then they evaluated the patient 2 weeks later. They found that the cognitive improvements in the patient had persisted during that interval. Further, in their pilot study of more patients, they also found consistent improvement over the period of six months.
This persistent clinical improvement gives me much more hope that etanercept is actually changing the disease course. It is possible that TNF-alpha is having two effects. First, it has an early effect of improving synaptic transmission. Second, it has a long-term effect modifying inflammation or limiting neuron death.
To conclude, in spite of the fact that this is a single case study and that I am skeptical of the immediate improvements, I think this is an excellent line of research to pursue. I am not aware of a single AD treatment that can change the disease course over long periods, so if these authors have succeeded in creating one they deserve a whole heap of praise. I hope to see more research into this area to verify the viability of this treatment and to determine the mechanism of its effects.
Hat-tip: Faculty of 1000
Tobinick, E.L., Gross, H. (2008). Rapid cognitive improvement in Alzheimer's disease following perispinal etanercept administration. Journal of Neuroinflammation, 5(1), 2. DOI: 10.1186/1742-2094-5-2
Interesting commentary. Thank you.
Let's be careful, for Alzheimer's, not to underestimate the value of drugs that simply hide symptoms. If patients can be kept more cognitively functional for a few more months late in their lives, that is a tremendous benefit to them and their families.
I am looking forward trying this procedure on my wife.
She is early onset and at stage 7 AD now at the tender age of 51. When your SO is totally hammered, fully aphasic, incontinent, and unable to dress, bath, or eat without assistance, you realize there is nothing left to lose. Even if she were able to recover enough speak for a few months and share some living with me, I would sell the house to make that happen.
This is the first glimmer of hope I have had in 4 years, and I am almost afraid of feeling optimistic.
Still, compared to the anticipation of watching her lingering death curled up in the fetal position, this seems a worthwhile risk.
It is all relative.
I too read this with great interest, but I have one question in particular: given that the drug is administered intrathecally, doesn't each injection compromise the spinal tissue a little bit as the needle makes its way through to the canal? This is fine for a few injections, but if you give this drug by this method every two weeks for two or three years (say), won't these little damages accumulate to the point where there is significant spinal injury? And at the neck too where there's a lot going on, spinally speaking.
Perhaps a successful result for etanercept will eventually require administration by means of an implanted pump or something similar.