Vania Apkarian and the holy grail

I love hearing Vania Apkarian speak. He is always so positive about his data and so forthright in his views that you have no doubt what he thinks and no doubt that he loves going to work.  This workshop I went to today was no exception – it was Apkarian at his best. And, quite possibly, he has good reason to be excited. His group at NorthWestern in Chicago has been working for over a decade on functional and structural brain imaging in people with chronic pain. Indeed, he makes a real point of making sure that his team of students and post-docs are acknowledged for their monumental contributions. Over that time, his group has stamped its name on much of what is known about the brain of people in chronic pain. Here is a bit of what his group, and others, have discovered:

  • – That acute evoked pain and spontaneous pain is associated with different ‘signatures’ of activity in the brain. Activation of the insula seems to code best for thermally evoked pain, and medial prefrontal cortex seems to code for spontaneous pain. Indeed, mPFC activation during spontaneous pain reports can differentiate between chronic back pain and other types of chronic pain.
  • – As pain persists, so too does the neural signature shift from a sensory-dominant to an emotional-dominant one, although the same broad areas are involved in both.
  • – Regional gray matter changes in chronic pain depend on the type of the chronic pain – or the location perhaps. The thing that is consistent is that the gray matter changes are not distributed across the brain, as they are for normal, or pain-free, controls.
  • – For evoked pain, Apkarian is most excited about the data relating to the nucleus accumbens (NAc). He reckons that activation of the NAc during evoked pain differentiates between chronic back pain patients and healthy controls pretty much all the time. In fact, when a noxious stimulus is delivered, controls have an increase in NAc when the stimulus stops (which is consistent with reward system activation and NAc is important for reward), but for patients, there is a decrease in NAc activation at the same time. Apkarian reckons this is about value judgements of the pain – more on that in a bit.
  • –  The functional connectivity between NAc and mPFC is only present in cbp and not in healthy controls and the extent of this connectivity correlates with spontaneous pain with a correlation of a whopping r=0.86. Apkarian said a few times that he could determine how bad a chronic pain patient’s pain was on a given day by looking at mPFC activation during spontaneous pain. I could answer the same question without getting out the scanner but I don’t think that was his point.  So, NAc seems important – Nestler et al < > reviews the data on NAc and conclude it does a fair bit (which, it seems is pretty much normal for functional neuroanatomy stuff), including reward expectancy, decision making under uncertainty, habit formation, stress and now, thanks to Apkarian, it is important in pain.
  • – Apkarian’s Piece de resistance was a longitudinal study in which he got 39 acute back pain patients – 4 – 12 weeks in – with pain over 5/10, and followed them with 4 scans over the following year (the amount of grants we have NOT got proposing this very thing makes me groan for a second, but let’s move on).  The short story: nearly all the changes we see in people with chronic pain develop as the pain develops, and there was almost nothing different at entry between those who go onto chronicity and those who do not, with one significant difference – those who still had pain unchanged at 1 year could be identified at entry by the connectivity between NAc and mPFC, with an odds ratio of about 5. If early treatment and the presence of radicular pain was added to the model, the odds ratio was 13, although the 95% CI was pretty big – 2 – 66, so the size of the effect is a bit hard to nail down. Still, it is, as it stands, a huge finding and one that will be out in Nat Neurosci some time soon.

Apkarian is sure this relationship between NAc-mPFC and chronicity is causal (as sure as he is that catastrophising doesn’t exist, but that is for another day). One question I am hoping that Vania responds to before this posts is whether or not the NAc-mPFC effect could have emerged between injury and entry to the study. NEWS FLASH – he has, he said that he presumes it reflects the individual’s response or interpretation of their injury, but that it remains to be shown. There are recent data that connectivity changes can occur as little as 2 weeks in a learning model, which would seem to support this possibility.

A final reflection here is that Apkarian reckons this NAc effect reflects a value judgement on the pain – sounds like something that might help from an intervention that could help the patient reconceptualise their pain…..

On Vania Apkarian – this is what his website at NorthWestern University in Chicago says: Dr. Apkarian has been studying pain for over two decades, both in animal models and fMRI studies in humans. His current interests include cortical dynamics of pain as well as brain plasticity. His overall goal is the uncovering of brain mechanisms underlying PAIN QUALIA. This work aims to alleviate clinical pain conditions and achieve a more profound theoretical and mechanistic understanding of the brain.  It should also say that he is Professor of Physiology, Surgery & Anesthesia at NorthWestern University, has received enough grant money to support a medium sized country, written 130 papers, 20 book chapters and 2 books. He is a real life heavyweight so we should all take notice.

[Update: 12 Dec 2012 – the link NorthWestern University is no longer working but this is what it said in May! ]

About Lorimer Moseley

Lorimer is NHMRC Senior Research Fellow with twenty years clinical experience working with people in pain. After spending some time as a Nuffield Medical Research Fellow at Oxford University he returned to Australia in 2009 to take up an NHMRC Senior Research Fellowship at Neuroscience Research Australia (NeuRA). In 2011, he was appointed Professor of Clinical Neurosciences & the Inaugural Chair in Physiotherapy at the University of South Australia, Adelaide. He runs the Body in Mind research groups. He is the only Clinical Scientist to have knocked over a water tank tower in Outback Australia.

Link to Lorimer’s published research hereDownloadable PDFs here.



  1. stuart miller says

    […] In terms of activation of the so-called ‘pain matrix’, I have read there is activation of primary and secondary somatosensory, insular, anterior cingulate, and prefrontal cortices as well as thalamus with novel and uncertain stimuli (salient) just as with nociceptive input – ‘saliency hypothesis’. Would someone care to comment ? Soros et al in 2001 showed that functional reorganization of the primary somatosensory cortex could occur within hours of intense acute nociceptive input but i too would like to know how quickly the connections can be perceived as ‘changed’. When does chronicity become a certainty ? There are definitely instances in which I perceive that patients have central sensitization within a very short period after their injury. What about de-centralization ? English is not necessarily concise – what is the difference between a matrix and a signature ? Thanks […]

  2. stuart miller says

    It was Apkarian that highlighted the connection between nucleus accumbens and medial prefrontal cortex in chronic pain. What else ? In terms of activation of the so-called ‘pain matrix’, my understanding is that with the saliency hypothesis, there is activation of primary and secondary somatosensory, insular, anterior cingulate, and prefrontal cortices as well as thalamus with novel and uncertain stimuli (salient) just as with nociceptive input. Would someone care to comment ? Soros et al in 2001 showed that functional reorganization of the primary somatosensory cortex could occur within hours of intense acute nociceptive input but i too would like to know how quickly the connections can be perceived as ‘changed'(why 2 weeks ?) There are definitely instances in which I perceive that patients have central sensitization within a very short period after their injury. What about de-centralization ? I have seen rapid shifts from agonizing pain to absence with mirror therapy, swelling control and a little education in the acute stage. I appreciate the recent work of Dr. Moseley on the body matrix – again layers upon layers in terms of complexity but frameworks are helpful. How many frameworks (sub frameworks) do we need is the question ?- Melzack’s original proposal was made when there was far less neuroimaging on the main areas activated in the brain with nociceptive input but was possibly more comprehensive. Moseley is brilliant in explaining some fascinating findings (like crossed hands phenomena and rubber hand illusion) as well as providing a framework for working with patients with CRPS. What’s next ? Finally, English is not necessarily concise – what is the difference between a matrix and a signature ?

  3. stuart miller says

    Just a quick point but the concept of implicit memories dominating the person’s thinking (in Gabor Mate’s work) or the lack of differentiation of neural networks in lack of ‘higher’ awareness (modification of Edelman’s work) may have relevance. The coordinated link between nAC and mPFC seems critically important. I seem to remember Damasio’s work on moral judgements and mPFC. I have worked with some brain injured patients with complex pain and seizure disorders and it was interesting although always a bit frightening that post seizure with some patients that there was a perseverative thought with a judgement of others or people having failed them (not always me) that persisted until they became more ‘aware’ of their surroundings. I am looking for insight but it seems that just like the ‘pain orchestra’ that keeps playing there are persistent ‘implicit’ memories that seem to set people up for a hard road. I am always challenged working with brain injured patients with addictions and chronic or complex pain as to the challenges with neuroplasticity. Diffuse damage can sometimes be helped with aerobic exercise (with diffuse effects on brain health) and compassionate, non-threatening environments but I am always challenged. Would appreciate insight and guidance from others. Thanks.

  4. Exciting stuff! Thanks for sharing. So these data suggest that patients have difficulty responding to positive stimuli (i.e. removal of a noxious stimuli)? Or is it that they have difficulty distinguishing whether a stimulus is positive or aversive? If the latter is true then it would have huge implications in deciding what sensory information is dangerous.

  5. Hesitated to post, as perhaps a little off topic in regards to the central theme, however a nagging question arose after a bit of snooping around and reading at Apkarian’s site linked to above.

    In a review he wrote for Pain Management “The brain in chronic pain: clinical implications” Pain Manag. (2011) 1(6), 577-586, Apkarian discusses the neuromatrix model and seems to somewhat dismiss the notion, or at least its usefulness as a concept particularly in chronic pain.

    He writes:
    a consistent and reproducible activation of a set of brain regions in response to acute painful stimuli in healthy subjects has “been dubbed the pain ‘neuromatrix’, with the associated notion that this network is necessary and sufficient for pain perception and that relative changes in extent of activity among these regions will be helpful in uncovering various pathological pain conditions.”
    “the notion of a specific pain matrix remains unconvincing”
    and again;
    “the bulk of the evidence for clinical pain conditions shows unique brain activity patterns with many brains regions identified outside of this neuromatrix”

    Returning to the source material; R Melzack in Journal of Dental Education (2001) (is there a story as to why this paradigm shifting paper was published in what strikes me as a relatively obscure journal) “Pain and the Neuromatrix in the Brain”, Melzack talks of the body-self neuromatrix as a ‘widely distributed neural network” that is capable of producing certain characteristic neurosignature patterns of nerve impulses that result in the output of perceptual, homeostatic and behavioral ‘programs’.

    Further Melzack proposes that the neuromatrix is genetically determined and modified by sensory experience suggesting significant changeability/adaptability and individual variation.

    It strikes me then that Melzack’s original idea of the neuromatrix was far from a “specific and constant pain matrix” as suggested by Apkarian in his review.

    Perhaps the differences can be accounted for by differences in logical levels, Apkarian approaching from a more structural “brain” based approach and Melzack originally from a mind/”conceptual framework” approach which includes the brain (hence a logical level up) but also a broader body-self perspective. (its important to note that the notion of logical levels does not suggest a value judgement i.e.. a higher logical level being somehow “better” than a lower level; just different)

    Apkarian might be responding to such statements as

    “These responses have been shown to originate from an extensive network of brain regions, which has been christened the Pain Matrix and is often considered to represent a unique cerebral signature for pain perception” Iannetti and Mouraux (2010) “From the neuromatrix to the pain matrix (and back)” Experimental Brain Research Vol 205(1)

    Or even
    “pain is a multisystem output that is produced when an individual-specific cortical pain neuromatrix is activated” (no need to reference that one here)

    My point (I think I have one) is the interchanging use of “neurosignature” and “neuromatrix” and “pain matrix”, “neurotag” etc.

    I guess I’m asking/thinking about – Does a specific “pain neuromatrix” exist, or is there just a (one) neuromatrix that is capable of producing (along with many other outputs) an output/neurosignature that we then call pain?

    With the resultant neurosignature being highly individualistic and potentially involving many brain regions, regions which may vary over time and show certain characteristics in this variability dependent on whether the pain is acute, spontaneous, evoked or chronic etc etc (and also specific relating to the nature of the underlying chronic condition as Apkarian suggests in his work)

    In the latter case, Apkarian’s rejection of the neuromatrix model in chronic pain, (from the same source as above)- “We can now resolutely refute the simplistic notion that brain activity for chronic pain is enhanced activity of the ‘neuromatrix’ as identified for acute pain” (again a bit of a straw man argument for mine as I don’t take away this notion from Melzack’s original proposal, this statement being an over-simplified summary of Melzack’s model from my understanding) would be negated as Apkarian’s proposals of brain changes would fit within a broader neuromatrix framework.

    Further the notion that “When pain becomes chronic, the efficacy of the pain neuromatrix (ed. neurosignature?) is strengthened via nociceptive and non-nociceptive mechanisms, which means that less input, both nociceptive and non-nociceptive, is required to produce pain.” (again, no need to reference) Which Apkarian would seem to refute, could be comfortably extended, rather than discarded, to suggest that with this strengthening comes the “recruitment” of other brain regions; other areas of the neuromatrix, to assist with the overall goal of protection and preservation.

    Is it important? – it is to me, and the words above ask, nay, implore and beseech me to ‘speak my mind’.

    Besides, it was the careless interchanging of the ideas and words “pain” and “nociception” that helped get us in to this bloody mess in the first place.

    Mind spoken…

    (I blame Mick Thacker; he recently demanded at noi2012 that we question, question, question; question him, question others, question everyone)


    Geoff Reply:

    Hi Tim,

    Great observations. Have you seen Lorimer and Mick’s commentary in MJA 196(6) “First-person neuroscience and the understanding of pain”? I think they do a good job in clarifying the oversimplification of pain being in the brain – perhaps this is where some of the confusion lies. I like their statement that pain is an “emergent property of the person who is suffering it … and crosses several domains such as neuroscience, immunology, endocrinology, psychology, sociology and philosophy”. I think this way of thinking takes the ‘neuromatrix’ to another level and gets away from an overly structurocentric view that neuroscientists are likely to disagree on. If you have not read their short commentary, its a nice reflection.

    Tim Reply:

    Thanks Geoff

    Yes, Lorimer’s and Mick’s little piece was a great read. I was very fortunate to meet Mick when he was over for noi2012 and see him speak a number of times. Mick said a number of times that “brains don’t think, people think” and similarly the notion that pain is an emergent property of the whole “person”.

    From my understanding, I think this view sits reasonably comfortably with Melzack’s original neuromatrix theory, but perhaps its become a little too “neuro” (ie “brain”) focussed in some peoples minds, hence the interchanging of neuromatrix, neurosignature and pain matrix etc etc.

    We truly are ‘fearfully and wonderfully’ complex

  6. Evert Jan Das says


    Can you give me the reference on:

    …There are recent data that connectivity changes can occur as little as 2 weeks in a learning model, which would seem to support this possibility…

    Greetings from the Netherlands.