An article published in the June 2009 issue of MIT's Technology Review titled "Manipulating Memory" by Emily Singer recently caught my attention. It concerns an important little revolution now occurring within the field of neuroscience concerning the nature of how memory works which appears to have important implications for the treatment of clinical disorders, such as Post-traumatic Stress Disorder (PTSD), which are intimately concerned with dysfunctional memories.
The memory revolution in question has to do with a phenomena known as memory reconsolidation. I had not heard of it before. However, I figure that I'm not alone. If I have just now learned about memory reconsolidation, then many people out there, patients and clinicians alike, have not heard of it yet either. I can do a service to all such folks by write an essay educating people about this important thing.
Interestingly, we've written about this stuff before without really understanding the importance of the back story. I'm pleased to be able to present that context now.
I'm going to lay out my discussion in four parts. First we will talk about memory in general - it's architecture and mechanisms and some of the assumptions that we have been guided by about how memory functions that have served us up to this point. This review of memory will give us the foundation we need to explain the phenomena of memory reconsolidation, so that will come next. Having explained memory reconsolidation, we are then in a position to talk about the clinical implications of this process and some of the studies that are currently underway dealing with the treatment of PTSD. Finally, we will conclude by talking about additional issues and concerns that may be on people's minds as they digest this somewhat technical material.
Changing Understandings of How Memory Works
Many people aren't thoughtful enough of what a miracle it is that we can remember things, I think. For them, memory is, more or less, something taken for granted; that aspect of the mind which allows you to remain aware of what you had for breakfast the other day, and which you lose if you develop Alzheimer's Disease. The more you know about how memory works, however, the more in awe you have to become. Memory is a really complicated and subtle thing. Scientists have been studying the brain and nervous system for well over a hundred years now and still don't fully understand how memory works. Many details about how the brain manages to remember things have been painstakingly revealed through careful scientific research, however. We will now review the more important of these findings so as to set ourselves up to appreciate what memory reconsolidation implies.
Memory in a Nutshell
The first thing to think about is that the brain is a modular system, meaning that it is comprised of specialized parts, each of which does a particular job. Some parts of the brain are critical for storage and/or retrieval of memories while others are not. Damage to parts of the brain which are not involved with memory will not result in memory problems (although they may cause other sorts of problems!). However, damage to parts of the brain which are involved with memory may render a person unable to remember previously stored information, or unable to store new memories.
Memory is a physical (spatial, chemical, biological) thing. Memories are stored and retrieved inside the brain and nervous system and nowhere else.
There are at least several distinct memory systems within the brain, and each is specialized to store a particular sort of information. If one system is damaged the other can continue to function independently of the other damaged system. Broadly, the storage of "episodic" or explicit memories - which is the sort of memory most people think of when they think of memory at all - what you had for lunch yesterday - is handled by a different system than the storage of spatial, muscular, or emotional (implicit) memories. You can have a situation where someone gets brain damaged and ends up with a perfectly functioning episodic memory, but no ability to remember how to navigate familiar spaces. Vice versa, you can have a situation where someone can no longer remember their own past but still knows how to navigate a city bus system and retains the social awareness necessary to shake a hand that is offered in greeting.
Neuroscience researchers are very confident that memories of various sorts are stored in the form of the growth of connections (synapses) between brain cells (neurons). Broadly, when a new memory is laid down inside the brain, what this means at the cellular level is that certain neurons have literally grown new branches (dendrites) connecting them to other neurons. Think of neuronal dendrites as river channels through which a current flows, down the branches and across synapses into other neurons where the process repeats itself and you'll have the idea. There are tiny but very physical channels that are being created every time you form a new memory. This process of memory channel building starts before birth and continues endlessly until your death. You are always remembering something.
There is a difference between short term memory and long term memory. Short term memory (STM) is temporary memory - the sort you use when you try to remember a new phone number by repeating it again and again. If you don't repeat the number you will lose it quickly, just as you lose the many thousands of details you encounter every day and do not store. Long term memory (LTM) is more permanent memory - the kind that allows you to remember your name, who you are married to or dating, and a variety of other information that makes up your identity and the knowledge base you take for granted. If you're into computers, think about the difference between RAM and hard disk storage and you'll have a rough analogy to the relationship between short term memory and long term memory. The physical connections between neurons we talked about above are the way that long term memory is laid down. Short term memory is handled in a different manner.
Short term memory is transformed into long term memory through an active process known as Consolidation. Consolidation is not an instant process; it takes a few hours to occur. If you mess with the brain before consolidation has had a chance to complete, for instance, by providing an electric shock such as occurs in an ECT treatment session, no long term storage of working memory held in the short term will occur. That memory will have just disappeared without a trace just as if you turned a computer off before saving your document.
There are several ways to interfere with the process of consolidation. You can use electric current as described above. This works most probably because the current disrupts and scrambles the flow of electrical signals within the brain (neurons conduct electrical signals), effectively erasing short term memory before it has a chance to consolidate. You can also use various drugs that interfere with the process of consolidation at a more chemical level, by interfering with the construction of new dendrites by inhibiting the construction of new proteins needed for that project, for instance.
Long Term Memory is not as permanent as we once thought
The question of how permanent a long term memory pattern of synaptic connections is once it has been consolidated is critical to our present discussion.
It used to be thought that once memories had been converted from short term form into long term form that they were permanently stored and would remain more or less stable and resistant to decay or alteration regardless of how many times they were remembered. This is a vision of long term memory as functioning like a CD-ROM which will play back a perfect rendition of its digital contents each time it is read, more or less, no matter how many times it is read. ROM means "read only memory"; The memory is not altered by the process of playback.
Newer evidence suggests that in many cases, memories are not "read only", but rather are rendered fragile and changeable by the process of remembering. What most probably happens is that memories in long term storage stay fairly stable until they are remembered. The process of remembering alters stored memories in some still unknown but physical fashion so that they become temporarily unstable and need to be consolidated (made stable) once again. The term used to described this new understanding is, consequently "reconsolidation".
Just as is the case with the initial consolidation process, reconsolidation takes time to occur (another several hour period is required), and can be interfered with by various means so as to prevent reconsolidation from occurring If you mess with the brain in just the right way during the period of memory reconsolidation you can keep reconsolidation from happening, which effectively results in a weakening of that memory. The memory may not be erased outright, but it loses some of its strength.
Clinical Research based on Memory Reconsolidation
Therapists, like all other human beings, are guided and limited by their assumptions of what is possible. The new memory reconsolidation literature is important because it changes fundamental assumptions of what is possible to achieve. If you believe that memory is a fixed thing once it has been laid down, then you are going to work around that memory as you plan treatment options for patients who are suffering from memories. You will encourage new learning (the creation of new memories) to counter-balance and eventually outweigh the effects of older memories, for example. When you start to understand that memory is not really a fixed thing, then the possibility that you can simply go in and directly edit a troubling memory comes into reach. This is a radical possibility, fraught with potential. "A power so great it can only be used for good or evil", if I may borrow an appropriate quote from the Firesign Theater.
At the time the Technology Review article was written Alain Brunet, a clinical psychologist at McGill University (also home to Karim Nader, the neuroscientist who has championed the reconsolidation paradigm shift), and Roger Pitman, a psychiatrist at Harvard. Pitman and Brunet were researching the effectiveness of Propranolol, a widely available beta blocker medication used in the treatment of high blood pressure, for the treatment of pre-existing PTSD.
Post-traumatic Stress Disorder or PTSD is, of course, an anxiety disorder that occurs sometimes in the aftermath of exposure to a horrendous traumatic experience involving death or the threat of death. It essentially involves a person ingesting a trauma memory that is so "hot" and emotionally overwhelming that it cannot be processed and grieved normally and instead is just avoided as best as can be managed. The trauma memory does not degrade as normal memories do, but instead stays fresh and intrudes into the traumatized person's awareness in an unwanted, unbidden and very frightening manner. Three classes of classic PTSD symptoms come out of this situation: 1) hypervigilance for threats, 2) attempts to avoid threats (anything which will trigger the trauma memories), and 3) chronic intrusive remembering of the trauma memory despite attempts to avoid it.
Medications are used to treat PTSD but largely for the reduction of anxiety symptoms associated with the condition. None in wide use are curative. The best current "curative" treatment for PTSD is behavioral in nature, known as Prolonged Exposure, and consists of having the PTSD patient repeat their story in excruciating detail again and again. Exposure therapy is a classic and generally effective strategy for treating all anxiety-based disorders. Its effectiveness springs from the way it systematically interrupts anxiety patient's strong tendency to avoid what they are anxious about and instead force them (with their consent!) to confront that frightened thing. Avoidance has the effect of strengthening anxiety emotions and thoughts. When avoidance is interrupted, people feel the anxiety feelings and think the anxiety thoughts they have been avoiding, and this creates the opportunity for what is known as "reappraisal"; an opportunity to realize that while the anxiety is terribly uncomfortable, it is not really dangerous in of itself. Reappraisal and associated learning phenomena such as habituation (e.g., the anxiety feelings lessen in intensity as people realize in their guts that they are not really dying) lead to new learning with the result that the patient's anxiety tends to become reduced in intensity over time and repeated exposure. There isn't anything wrong with exposure therapy per se, but it is not a terribly efficient or inexpensive process either. There is always room for improvement in this type of treatment protocol.
Pitman, aware of the emerging memory reconsolidation literature, realized that it might be possible to treat PTSD patients with Propranolol, which would have the effect of disrupting memory reconsolidation in the emotional areas of memory but not in the episodic areas of memory (remember - memory for these things is modular and separately stored) - There's a whole biochemical explanation for why this should be so involving inhibition of the neurotransmitter Norepinephrine but this is well beyond the scope of the present essay. A treatment of this sort would need to combine elements of exposure therapy - so as to reactivate the trauma memories to put them into an unstable state and require them to reconsolidate - and Propranolol therapy - so as to interfere selectively with the reconsolidation process before it can occur. As previously reported, Pitman had already showed that Propranolol treatment administered to people immediately after their trauma (prior to the time when normal consolidation would occur) resulted in fewer PTSD like symptoms developing later on. Working together and separately over the last several years, Pitman and Brunet have been testing exactly this new "reconsolidation" therapy. According to the Technology Review article, though the research is not yet complete, the early data strongly suggest that the technique works fairly well: "Preliminary findings show a 40 to 50 percent improvement in self-reported symptoms among those taking the drug". Some people did not take the drug in the study, of course, so as to serve as blind controls (to eliminate the potential for the placebo effect from contaminating the results). If this finding is replicable and stable, these guys are really on to something important.
Reconsolidation interference therapies are bound to be exciting news to anyone who has worked with PTSD patients, or who themselves must live with PTSD. This is an exciting shift in our understanding of the neuroscience of memory, and an exciting and innovative use of this new knowledge to help treat what can be a very painful and entirely disabling condition. Given the unending warfare of the last decade, we're going to be dealing with a metric ton of PTSD cases, and any new hope we have of doing that more efficiently and effectively is welcome news.
It's going to take some time and a bunch more research before this approach will become widely available, however. Multiple questions need to be answered first. Not only will this therapy have to turn out to work well, it will also need to be proven to function better than existing therapies or to offer advantages such as speed of treatment effect, or lowered expense over existing therapies. The mechanism through which reconsolidation is interfered with will surely need to be tweaked as well. Might there be other substances, or other methods which would work more efficiently than Propranolol for the purpose of reducing the strength of emotional memory?
Comparison of this technique for hacking memories to that depicted in the 2004 movie "Eternal Sunshine of the Spotless Mind" are bound to come up for some people. In that excellent movie, a physician had a machine that was capable of entirely erasing people's memories of painful relationships. Much of the plot of that movie revolved around the desire the protagonists had to be rid of the pain of their connection, and then ultimately, their need to retain those memories even though they were painful, because without them, they were not themselves anymore. The point to make here is that thought this technique might be extremely useful for helping people to suffer less, it could also be used to harm people - to steal parts of their identities. And there is surely no lack of sociopathic neo-Joseph-Mengele types in the world who would use it for just that purpose if they could. Technology Review article author Emily Singer suggests that present day techniques for memory editing are nowhere near as powerful as people might think - that memories can be tweaked only - edited in minor ways but not really deleted - and that there is no cause for alarm. Singer also quotes Brunet as rejecting any concerns over the possibility that his techniques could be used to harm people; he does not see potential for abuse. And I agree that given where things are at today, there probably is not anything to be alarmed about. But should this approach pan out and become refined, I would not rule out the possibility of abusive uses of the technology in the future. I'm not overly concerned about that possibility, simply because I doubt there is any stopping it from happening. Once the genie is out of the bottle, as it surely is now, people will work on it for good and for evil, and we have to focus on the good.