Chronic back pain is associated with marked changes in brain activity compared to acute pain, says a new study from A. Vania Apkarian and colleagues at Northwestern University, Chicago, Illinois, US. In a report published in the September issue of Brain, first author Javeria Hashmi, Apkarian, and colleagues demonstrate a shift over time in brain signatures, or imaged patterns of brain activity, in patients with subacute, short-lasting back pain and those with long-lasting back pain. In general, the representation of pain in the brain shifts over time from classical pain areas to areas implicated in emotion.
This is the first study that examines brain activity in a common pain condition as it transforms over time. The finding of a shift in central nervous system activity in patients moving from acute to chronic pain indicates that there is not just one signature for pain in the brain and could provide new avenues for diagnosis and treatment.
The work is a continuation of an ongoing longitudinal investigation examining changes in brain structure and function in patients with subacute back pain (SBP) as they either recover or develop persistent pain. Previously, Apkarian’s team demonstrated in a yearlong study that differences in functional connectivity between brain areas, and white matter structural differences, can predict the persistence of pain in SBP patients (Baliki et al., 2012, and see PRF related news story on Mansour et al., 2013).
In the current study, the team compared brain signatures between SBP patients, including those who recovered and those whose pain persisted over the course of a year, with chronic back pain (CBP) patients who have experienced back pain for more than a decade. “It’s the first time we’ve done it,” said Apkarian, referring to the study’s longitudinal nature. “It’s been very hard to do.”
First, Apkarian’s team used functional magnetic resonance imaging (fMRI) to compare pain-related brain activity in 94 SBP patients (who had had back pain for less than three months) and 59 CBP patients (who had pain for more than 10 years).
All of the subjects noted and rated instances of spontaneous back pain on an intensity scale of 1 to 100 during their fMRI scans. The images were then analyzed to determine the active regions, distinct from background activity, that were associated with pain perception.
To assign functions to the active regions, the team turned to Neurosynth, a meta-analytic Web-based software that automatically extracts and amasses fMRI data from published research studies (see PRF related news story on Yarkoni et al., 2011). Data from hundreds of brain imaging studies are analyzed to provide a consensus of fMRI data on commonly studied conditions such as “thinking,” “distress,” or “sleep.” “There’s a surprising degree of overlap between individuals,” with regard to brain activity, said Tor Wager, a psychologist and neuroscientist at the University of Colorado, Boulder, US, whose lab developed Neurosynth. Showing conserved areas of brain activity among experimental groups is crucial to any conclusions scientists might draw from their data.
“It gives us a scale of confidence on which we can make our statement,” said Apkarian.
The Apkarian team searched for the terms “pain,” “emotion,” and “reward” following their intuition that subacute pain is more related to classical sensory regions, whereas chronic pain representation might migrate to more affective regions in the brain.
The researchers found that SBP patients demonstrated conserved activity in brain areas including the anterior to mid-insula, thalamus, striatum and lateral parts of the orbitofrontal and inferior cortex, and dorsal anterior cingulate cortex. These areas are known to be involved in acute pain. In contrast, CBP patients exhibited brain activity in areas including the perigenual anterior cingulate cortex, the medial prefrontal cortex, and parts of the amygdala—areas that are considered to mediate emotion.
That finding raised the question of how these distinct brain signatures might arise. To address that question, the researchers analyzed scans of 39 SBP patients followed longitudinally for one year. All of the subjects initially exhibited pain-specific brain activity, including in the insula, anterior cingulate cortex, and thalamus, as expected.
Eventually, 19 of the patients recovered and exhibited no pain-relevant brain activity at one year. But 20 of the patients experienced persisting pain, and that group displayed a transformation in brain activity, with pronounced activity in areas such as the amygdala, medial prefrontal cortex, and basal ganglia, and no sign of previously robust activity in acute pain-related areas. Affirming Apkarian’s ongoing research, SBP patients who had high functional connectivity between the medial prefrontal cortex and nucleus accumbens at the outset of the study were more likely to experience persisting pain.
“[Brain activity] has moved away from sensory representation areas in general to more areas that relate to emotion” in patients with chronic pain, said Apkarian. Brain signatures of pain became more emotional in nature, although, Apkarian stressed, the description by the patients of their pain remained constant.
Brain localization theory, the idea that certain perceptions are only encoded by certain brain regions, has left its mark on psychology and neuroscience. “This is the first unequivocal evidence that that cannot be tenable,” said Apkarian.
“Many people think there’s one system for pain,” said Wager. Now, growing evidence from Apkarian, Wager, and others makes this a difficult proposition.
Besides providing evidence for the idea that pain representation evolves over time, the study provides a possible avenue to explain why treatments that are efficacious for acute pain patients may not work for chronic pain patients with the same injury. “We have to look beyond peripheral sensation to treat pain,” said Apkarian. And he and Wager agreed that brain signatures are a viable future diagnostic tool.
Abdul-Kareem Ahmed is a medical student and freelance science writer in Providence, Rhode Island, US.
Image credit: From Hashmi et al., 2013, courtesy of Oxford University Press
Comments
John Quintner, Arthritis and Osteoporosis WA
Would you please explain what
Would you please explain what you mean by "pain representation in the brain"? What we call "pain" is a lived experience that like all other lived experiences is created by the brain on a moment-to-moment basis. As such, surely pain cannot be a "thing" that is represented by neuro-imaging.
Simon van Rysewyk, University of Tasmania
Hi John,
Hi John,
Abdul-Kareem hasn't yet replied to your question, so allow me to give it a stab. Comments more than welcome.
You assert that pain cannot be a thing. Your argument in support of this conclusion seems to be the following: Pain is unpleasant, but nowhere in physical space. However, brain states all occur in physical space (the physical brain), and none of them are unpleasant. So pain cannot be identical to any brain-state. Instead, pain is a 'lived experience'. Two more relevant conclusions follow: (1) fMRI cannot represent pain; (2) Mind is not the same as brain. Pain is mental, brain is physical. Is that faithful to your position?
You are right to imply that what happens in the brain during pain is not itself unpleasant. But, a state of pain – an episode of pain, if you like, which is always a personal experience – is also not itself unpleasant, and based on robust neuroscientific evidence, does in fact occur in the (mammalian) brain, likely in insular and cingulate cortices.
Pain is a certain state of experience, which we call ‘being in pain’, or ‘having a pain’. And when I observe you in pain, I can use the same expressions to characterize your personal experience. So, the word ‘pain’ refers to a type of experience, not a type of object. In other words, you are correct: a pain is not a weird object. It is a sensory, emotional and cognitive experience, which is unpleasant, hurtful, surreal, burning, throbbing, typically accompanied by injury, and so on.
In migraine headache, being in pain is not located in the head, but a state of migraine is identical to a brain state. Pain is neither an object, nor a thing, but a personal event identical to a brain state, and the language of pain may obscure this.
Sincerely,
Simon
John Quintner, Arthritis and Osteoporosis WA
Hi Simon. Thank you for
Hi Simon. Thank you for providing a most welcome and detailed response to my question. In my opinion, the Apkarian group has fallen into the trap of believing that they are imaging a "thing" called pain. Clearly, they have a case to answer.
Simon van Rysewyk, University of Tasmania
Welcome John.
Welcome John.
Andrew McMullan, MCSP
John/Simon
John/Simon,
Would I then be correct to assert that imagery like that cited above, or that by Wager et al., 2013, are revealing a physical correlate of the experience?
Regards,
Andy
John Quintner, Arthritis and Osteoporosis WA
Andrew, the point I am trying
Andrew, the point I am trying to make is that the brain imagery is NOT the lived experience of pain.
By the way, can anyone make sense of this comment?
"Brain signatures of pain became more emotional in nature, although, Apkarian stressed, the description by the patients of their pain remained constant."
Regards
John
Simon van Rysewyk, University of Tasmania
Hi Andy,
Hi Andy,
Short answer: yes.
I think we should be cautious about what such correlational evidence means for pain. Correlational evidence is consistent with any of the following: (1) the measured neural activity is part of the background condition for pain, (2) the measured neural activity is part of the cause of pain, (3) the measured neural activity is part of the sequelae of the awareness of pain, (4) the neural activity parallels, but has no direct causal role in pain, and (5) the measured neural activity is what pain can be identified with.
Ultimately, if we want to be able to explain the nature of pain, what we need is the identification of some type of neural activity with some type of personal pain, i.e., data should justify interpretation (5).
I think the authors in the Apkarian et al. and Wager et al. studies present evidence to mean (2).
That some neural activity x is a correlate of some pain y tells us only a little, such as that we may be on the right path to find data to make interpretation (5). This is obviously not everything, but we have to start somewhere.
Thanks.
Simon van Rysewyk, University of Tasmania
Hi John,
Hi John,
I think it means the patients reported personal pain in terms of both sensory (e.g., burning stabbing, pincing, etc.) and emotional qualities (e.g., unpleasantness, awfulness), but brain activity simultaneous with such reports was more strongly correlated with emotional regions of cortex (e.g., insula, anterior cingulate cortex) than with sensory regions (e.g., SI, SII), which the authors hint could biomark the shift from acute to chronic pain.
Of course, they might be wrong about this, but without an embarrassment of theoretical riches at our fingertips, we'll never come close to the truth of it. I hope they are right.
Sincerely,
Simon
Vania Apkarian, Northwestern University Medical School
The above discussion
The above discussion regarding our study of shape shifting pain with first author Dr. Hashmi harks back to the theroretical writings of late 19th centruy German psychophysicists. Helmholz, Muller, and Weber are part of the German pioneers who established the principles that subjective perceptual states have well organized mathematical rules that can be captured and related across subjects specifically for each sensory modality. Modern human brain imaging studies are an effort of identifying the brain correlates to those psychophysical principles. At least in the work that we have been publishing, when we state "pain perception," we actually imply the intensity of discomfort associated with the perception. The latter is a valid psychophysical concept, with well defined mathematical properties, and brain activity related to this outcome closely tracks subjects' report of how uncomfortable the experience of their pain is.
Is "pain perception" more complex than its intensity? Of course it is. Do we know how to untangle individual subjects' entirety of perception that the subjects associate with their pain state? Certainly not. However, we can still make important advances in our understanding of the links between pain and the brain by associating specific brain properties with unique characteristics of pain, and this line of research does promise to become clinically relevant as it begins to identify macroscopic and molecular mechanisms that point to novel therapy ideas.
John Quintner, Arthritis and Osteoporosis WA
Hi Vania
Vania, thank you for your response. Could you please address the following questions?
1. In relation to the German psychophysicists, are you referring to the Weber-Fechner Law for grading pain intensity by "just notable differences." If so, have you found functional MRI correlates of their experimental work comparing and contrasting changes in "pain sensitivity" and "pain sensibility"?
2. May I quote from the paper by Baliki et al. (2011)? Therein it is stated that there is "... accumulating evidence that chronic pain alters brain dynamics beyond pain perception ..." and that "these results suggest that brain morphological changes may be reversible in nature and are a consequence of pain perception."
In the light of your clarification of what is meant in your publications by "pain perception," would you please comment on whether or not you agree that "pain" is best described as a sensory and emotional experience? If you do agree, then how is it possible for "pain" to be an agent of structural changes within the brain?