Brain imaging of pain: state of the art – J Pain Res. 2016 Sep – free full-text PMC article
Advances made in neuroimaging have bridged the gap between brain activity and the subjective experience of pain and allowed us to better understand the changes in the brain that are associated with both acute and chronic pain.
Additionally, cognitive influences on pain such as attention, anticipation, and fear can now be directly observed, allowing for the interpretation of the neural basis of the psychological modulation of pain.
The use of functional brain imaging to measure changes in endogenous neurochemistry has increased our understanding of how states of increased resilience and vulnerability to pain are maintained.
Introduction: neuroimaging of pain and plasticity and the areas of the brain involved
The poor relationship between regional tissue damage and pain experienced by patients has led from pain being investigated as a localized phenomenon to a more complex process, including the central processing of the brain.
This coincided with new techniques that allowed visual images to be derived from the activity of the brain, known as functional brain imaging (fMRI).
Despite extensive studies, no single area of the brain has been determined as solely responsible for pain processing.
There is no pain center but instead a complex network of brain regions often termed the pain matrix (Figure 1), as originally conceptualized by Melzack and Wall.
This matrix consists of multiple regions that in themselves are not only associated with pain but also involved in other sensory, motor, and cognitive functions where information is often processed in parallel.
The consciousness of pain seems to appear at the later stages of pain processing when the neural information is being integrated across multiple regions of the cortex.
Analysis of experimental pain neuroimaging shows six areas of the brain that consistently respond to acute pain and are believed to play an important role in the sensory-discriminative, cognitive, and affective aspects of pain processing.
These are the thalamus, the insular cortex (IC), the primary and secondary somatosensory cortices (SI and SII), the anterior cingulate cortex (ACC), and the prefrontal cortex (PFC).
These areas differ depending upon factors such as imaging modality, statistical analysis, psychological state, and type of pain elicited.
Functional neuroimaging of pain has been previously reviewed, focusing specifically on positron emission tomography (PET) and functional magnetic resonance imaging (fMRI) studies and neuropathic pain.
This review provides ageneral overview of current neuroimaging techniques and the central processes of pain perception in healthy subjects and patients with chronic pain conditions. How the use of functional imaging can encourage and aid the development of new therapies is also explored.
A schematic representation of pain modularity circuitry and the pain matrix.
Notes on Figure 1:
Nociceptive inputs enter the spinal DH through primary afferent fibers that synapse onto transmission neurons.
The projection fibers ascend through the contralateral spinothalamic tract targeting the thalamus, and collateral projections also target mesencephalic nuclei, including the RVM and the midbrain PAG.
In the pain matrix, there are two complementary pathways through which pain processing takes place.
The medial pathway (dark gray) projects from the medial thalamus to the ACC and IC and processes the affective-motivational component of pain (ie, unpleasantness).
The lateral pathway (light gray) projects from the lateral thalamus to the primary and secondary somatosensory cortices (SI and SII) and IC and processes the sensory-discriminative aspect of pain (ie, location and intensity). Increased activation of the PFC is related to decreased pain affect purportedly by inhibiting the functional connectivity between the medial thalamus and the midbrain.
Descending projections from the hypothalamus (not shown), amygdala, and rACC feed to the midbrain PAG and to the medulla.
Neurons within the RVM project to the spinal or medullary DH to inhibit pain experience.
Imaging techniques used: technology overview
Our understanding of the neural basis of pain has been substantially increased as a result of the development of functional neuroimaging techniques (Table 1).
Functional neuroimaging is based upon being able to measure changes in neuronal activity by measuring alterations in
- either aspects of metabolism (ie, blood flow, volume, oxygen, or glucose metabolism)
- or one aspect of neurochemistry (ie, neurotransmitter precursor uptake or receptor binding).
The most commonly used techniques depend on the premise that increased brain activity leads to increased energy metabolism and a disproportionate increase in regional cerebral blood flow (rCBF).14
fMRI and related modalities
fMRI is one of the most commonly used imaging methods in pain research. fMRI indirectly measures brain activity by detecting associated changes in blood flow (hemodynamic response).
In its primary form, fMRI uses the blood-oxygen-level-dependent (BOLD) contrast imaging, which is indicative of synaptic activity. The BOLD technique evaluates the difference in magnetic susceptibility between the oxygenated blood (oxyhemoglobin) required by active neurons and the resultant deoxygenated blood (deoxyhemoglobin) and creates the fMRI signal from this difference
The fMRI BOLD technique is an extremely useful measure in acute and experimental pain where there are short periods of pain followed by short periods that are pain free, causing a rapidly changing hemodynamic response. This allows the study of acute pain response in pain-free volunteers and pain patients with chronic pain but is not well suited to the monitoring of responses to changes in chronic pain.
For persistent pain or chronic pain conditions, an alternative fMRI technique utilizing arterial spin labeling (ASL) is more appropriate. ASL provides a direct measure of cerebral blood flow using magnetically labeled endogenous water in the blood to act as a diffusible tracer.
In contrast with BOLD, ASL has reduced temporal resolution but allows for improved quantification of regional blood flow, resulting in better estimation of ongoing blood flow.19 fMRI ASL techniques have consequently been used to assess the central processing of pain in patients with migraine and chronic lower back pain in different pain states.
Positron emission tomography
Electrophysiological studies: electroencephalography and magnetoencephalography
electroencephalography (EEG) and magnetoencephalography (MEG) studies are often used to evaluate separate temporal components of the cerebral pain response
The EEG signal represents voltage fluctuations from neurons firing in the brain and is recorded from multiple electrodes placed on the scalp.
In experimental pain research, a brief noxious stimulus can be time locked to give an event-related potential that provides millisecond accuracy in the timing of pain-related neuronal events. However, EEG has worse spatial resolution than fMRI.
One of the fundamental problems that applies to all functional imaging data is that it is dependent on averaging signals over time. This is dependent on assumptions about the stability of the methods over this time
Magnetoencephalography (MEG) maps brain activity by recording magnetic fields produced by electrical currents that occur when neurons fire.
In the same way as EEG, the signals derive from the net effects of currents flowing through neuronal dendrites. Signals are mainly derived from the cortex.
MEG provides a better spatial resolution in comparison to EEG; however, it is important to note that source analyses to assess deeper regions are unreliable in both modalities
Mechanisms of pain perception and analgesia
When compared to healthy controls, chronic pain patients process acute pain in a different way, with what would normally be nonpainful stimuli, often being reported as painful and activating pain-related brain areas
As such, chronic pain patients appear to have altered brain function and structure.
During periods of OA pain, there is
- extensive activation of the cingulate cortex and
- greater activity in the amygdala,
- orbitofrontal cortex, and
when compared to that during periods of experimental pain
The amygdala, orbitofrontal cortex, and putamen are areas previously associated with aversive conditioning, reward, and fear, and their activation in OA implies that time activation of this fear-related circuitry is perhaps related to fear of additional injury and disability.
These regions are associated with the supervision of attention, and it is hypothesized that descending fibers from the PFC inhibit neuronal coupling along the ascending midbrain–thalamic–cingulate pathway, modulating pain in a “top-down” fashion
This is why we can’t focus when we’re in pain.
Expectations of pain, and the anxiety caused by these expectations, are suggested as a possible source of the increased pain perception seen in chronic pain patients.
Chronic pain often causes changes to the brain that result in pain at sites of the body without tissue damage.
This is because prolonged pain can lead to neuroplastic changes at the cortical level, which induce central sensitization.
Chronic pain is often associated with spontaneous pain that has no physical origin and exhibits moment-to-moment variation in pain intensity
Neuroimaging has facilitated the observation of volume-based changes that occur in the gray matter of chronic pain patients
Interestingly, the decrease seen in the gray matter of chronic pain patients can be reversible.
gray matter volume in the affected cortical areas increases after surgery to alleviate the pain in the damaged joint.
PET imaging with the radionucleotide 11C-carfentanil demonstrated that sustained acute pain triggers the release of endogenous opioids in a region-specific manner and also that a reduction in the severity of pain correlated with increased occupation of μ-opioid receptors by endogenous opioids
Reductions in μ-opioid receptor binding in the amygdala, cingulate, and nucleus accumbens have been demonstrated in patients with FM
- In patients with peripheral neuropathic pain, reduced opioid binding is observed bilaterally across brain hemispheres,
- in those with central neuropathic pain, these reductions in opioid binding are observed in one hemisphere contralateral to pain
The differences seen in opioid receptor binding between patient groups may reflect different underlying mechanisms of pain and explain the variability seen in response to treatment with opioids
PET studies have also demonstrated changes in opioid receptor binding in patients in and out of pain, induced by rheumatoid arthritis and trigeminal neuralgia. These results were consistent with increased competition of opioid receptors by endogenous opioids during ongoing pain
This suggests that the endogenous opioid system is activated by chronic pain
Neuroimaging can also be a useful tool to study the effects of analgesics on the brain
After a month of daily opioid administration, chronic back pain patients show decreases in gray matter in the right amygdala and increases in the cingulate (middle, dorsal posterior, and ventral posterior), which are areas known to have high μ-opioid receptor density and binding capacity and strong neural response to opioid administration
Using structural MRI, DTI, and resting-state fMRI, comparisons between prescription opioid-dependent individuals and controls revealed bilateral volumetric loss in the amygdala, decreased anisotropy in axonal pathways specific to the amygdala, and significant decreases in functional connectivity between seed regions that included the anterior insula, nucleus accumbens, and amygdala in opioid-dependent patients. The longer the duration of exposure to prescription opioid, the greater the observed changes in functional connectivity
Informing pain relief and clinical treatment – future perspectives
Modern noninvasive human brain imaging has revolutionized the study and treatment of pain. It allows scientists to understand individual differences in response to pain treatments, how endogenous pain relief such as that seen in the placebo response works, and to identify targets for future drug development
Real-time fMRI is a technique showing potential as a treatment for chronic pain.
Patients can be trained to use neurofeedback to up- or downregulate the BOLD response to influence the activation of a target brain area, in this case, the ACC or anterior IC.
Behavioral pain ratings decreased during feedback, and anterior IC regulation and ACC regulation led to a significant downregulation of parts of the pain network with practice, notably the caudate nucleus.
As pain is subjective and pain self-reports can be unreliable in determining pain experience, researchers have been looking for a physiologically based pain assessment that shows a significant correlation with self-reported pain.
However, in the end, it is the patient’s report of pain relief that will determine efficacy of any new therapy
Science hates vagueness, so self-reports are always viewed with suspicion.
While these methods will provide a more precise understanding of the dynamics of brain connectivity that underpins pain experience, we should be cautious of the concept of a brain scan telling a patient how much pain he or she should be feeling.
The biggest risk of this technology is that our pain will be judged by others looking at scans and our own reports will be progressively less valued.
Philosophically and scientifically, this is probably not tenable and the concept has potential moral side effects in relation to health insurance and other social issues.
So much that’s going on with pain relief is “philosophically and scientifically not tenable”, like forcing millions of people to endure pain so that others don’t overdose on heroin, but it is happening anyway.
Functional imaging techniques can be broadly split into electrophysiological methods and hemodynamic methods.
Electrophysiological methods, EEG and MEG, have unparalleled temporal resolution and so are often used to separate the pain response into temporal components, such as anticipation and early and late response.
In comparison, hemodynamic methods, PET and fMRI, have better spatial resolution and therefore are utilized to ascertain specific points of cerebral activation. These techniques have defined key brain structures that comprise the pain matrix consisting of a medial and lateral pain system.
The division of function of these two systems may be broadly defined as related to affective-motivational and sensory-discriminative processing, respectively
The pattern and distribution of activity within the pain matrix is crucially affected by the emotional and cognitive context of experimental pain stimuli, including expectation.
More recently, potential mechanisms of chronic pain have been identified in relation to the processing of expectation of pain within the medial pain system, suggesting problems with top-down regulation, particularly involving the interactions between the DLPFC and limbic components of the medial pain system
Neurochemical deficits within the dopaminergic and opioid systems may contribute to some of these candidate mechanisms.
Functional brain imaging has allowed the identification of new candidate brain mechanisms of chronic pain that provide both physiological and pharmacological therapeutic targets for us to collectively develop