Irene Tracey, University of Oxford, UK, summarized the contributions of neuroimaging to pain research, and directions for future investigations, during “Translating Neuroimaging Discovery Science for Patient Benefit,” a plenary lecture held at the IASP 16th World Congress on Pain, which took place September 26-30, 2016, in Yokohama, Japan. Her take-home message was that findings from neuroimaging will lead to a brighter outlook for patients suffering from chronic pain. “The aim, ultimately, is to use metrics [discovered by imaging studies] to guide diagnosis and therapies,” she said. Tracey called for neuroimaging work to help bridge scientific knowledge from cells to systems and across species in order to meet the unmet clinical need for new pain treatments.
Neuroimaging of descending pain modulatory systems
While basic science has advanced understanding of the descending pain modulatory system, a similar grasp in humans via neuroimaging has been a challenging feat—but new findings are emerging. Tracey described findings from studies using brainstem neuroimaging in healthy human volunteers that are consistent with animal evidence of central sensitization (Zambreanu et al., 2005; Lee et al., 2008). More recently, evidence of central sensitization has been demonstrated in patients with chronic pain. For instance, increased activity of descending pain modulatory systems, identified by functional magnetic resonance imaging (fMRI), in osteoarthritis patients prior to hip surgery was associated with increased pain ratings to punctate stimuli (Gwilym et al., 2009).
Along these lines, Tracey pointed to unpublished work from her lab, presented during a poster session at the meeting, in patients with knee osteoarthritis awaiting surgery (Soni et al., 2016). Here, patients with neuropathic pain features had higher pain scores in response to cold and punctate stimuli, and greater levels of pain catastrophizing, than patients without neuropathic pain features. The patients with neuropathic pain features also exhibited higher brainstem activity, according to fMRI, suggesting increased descending facilitation of pain in these individuals, who were also in significantly more pain after surgery.
Tracey further said that in addition to translating research from animal models to experimental models in humans and then to clinical research in chronic pain patients, the field is now also translating research to industry and clinical trials. For instance, brainstem biomarkers have been used to inform analgesic drug development (Wanigasekera et al., 2016; Iannetti et al., 2005).
Anti-nociception and placebo analgesia
In the face of severe injury, when threatened by dangerous situations, “something happens in the brain when you’re in these fight-or-flight situations, and you just don’t feel anything,” Tracey said. This paradoxical and protective response results from activation of anti-nociception systems in several regions throughout the cortex, midbrain, and brainstem involving multiple neurotransmitter systems.
Distraction is one aspect of how analgesia is produced during times of danger. The anti-nociceptive systems activated during fight-or-flight situations can be studied in the laboratory setting by using conditions of distraction while assessing pain sensitivity. Citing earlier neuroimaging work from her group (Tracey et al., 2002), Tracey explained that recruitment of descending analgesia during distraction depends on the individual: Some people are good at it, and others are not. Her group is interested in further studying this variation among healthy individuals and patients through neuroimaging studies of brain activity within regions of descending control.
As an interesting parallel to neuroimaging research on distraction, Tracey pointed to studies of placebo analgesia. “Placebo hijacks the anti-nociception brainstem pathway,” she said. Previous investigations showed that the hypothalamus and brainstem are active during the placebo analgesic experience (Eippert et al., 2009). Interestingly, in a subset of healthy individuals, greater “conversation” between these brain regions and the prefrontal cortex appears to drive their activity and produce a greater placebo effect.
Data currently under review from Tracey’s group suggest that in neuropathic pain, patients show evidence that the placebo network in the brain is recruited in the placebo arm of clinical trials. Tracey says studies like these will help researchers and clinicians answer questions such as, Do effects observed in the placebo arm of a clinical trial occur in the drug arm as well? Are drug and placebo mechanisms of pain reduction additive or non-additive? And, importantly, does the drug disrupt placebo mechanisms? Researchers hope that neuroimaging will be able to tease apart the fine details of drug and placebo mechanisms and provide a better understanding of these complex interactions.
New roles for important regions of descending control are still being discovered which may help with the discovery of novel treatments for pain. For instance, the midbrain periaqueductal gray (PAG) is a region with a well-established role in descending control of pain. Tracey described a recent study in healthy volunteers that identified a neural region within the PAG that encodes a prediction error (Roy et al., 2014). “When you aren’t expecting something to be painful” and in response to a “mismatch of pain expectation and actual pain,” the “PAG becomes very interested,” she said. This “means that you learn, in effect, from your painful mistakes.” In the future, similar new and complex roles for other regions of descending control, such as the locus coeruleus (LC) and rostroventral medulla (RVM), may be identified using neuroimaging and may aid the development of novel therapies.
New findings on phantom limb pain
There is evidence that phantom limb pain results from preserved representation of the cortical region associated with a missing limb. Tracey explained that many studies have used mirrors to “re-plasticize” the cortex in order to alleviate phantom limb pain. “The phantom cortex is capable of activating” and looks no different than in controls, she said, according to neuroimaging research in patients studied decades after the initial injury (Makin et al., 2013). This is really “good news,” she said, because it means that even a long time after a region of the body is lost, the brain is likely malleable to therapy.
While neuroimaging studies have revealed that more activity in the brain area representing a phantom hand is correlated with more pain (Makin et al., 2013), so, too, is greater disruption of the conversation between the cortical region associated with the phantom limb and the contralateral hemisphere, according to fMRI studies of functional connectivity (Kikkert et al., 2016). Tracey also pointed to unpublished data from studies using transcranial direct current stimulation (tDCS) to correct the conversation between brain regions implicated in phantom limb pain. With anodal stimulation of the region of the primary somatosensory cortex representing the lost limb, patients showed decreased pain up to one week later, compared to baseline.
All eyes on the spinal cord
Preclinical studies of pain processing focused on the spinal cord have produced a wealth of knowledge about the pain system in health and disease. However, it has been difficult to study the human spinal cord. “The spinal cord is a very tough place to get data from,” Tracey said, but this is something that new imaging studies are beginning to address. For example, recent work in healthy human volunteers has identified resting-state networks—regions that show similar slow oscillations of activity when participants are at rest—in the spinal cord (Kong et al., 2014). Tracey acknowledged that “we don’t know what these [findings] mean at this stage.” Specifically, these resting-state networks could reflect descending modulation inputs, interneuron activity, and/or peripheral (sensory, motor, and proprioceptive) nerve input. The “next objective is to determine how these [factors] change in our experimental models of sensitization,” Tracey said.
The future of pain neuroimaging
Tracey also discussed the relevance of pain neuroimaging for law and society. She pointed to a recent court case in Australia ruling that post-traumatic stress disorder (PTSD) is “bodily injury,” based on evidence for representations of pain, stress, and anxiety within the brain. However, the use of pain neuroimaging in the courtroom is far outpacing the science, and researchers strongly caution against its use to “prove” whether someone is or is not in pain. Tracey stressed that it is the pain field’s duty to define and follow guidelines for this area of research. Pain brain imager Karen Davis, University of Toronto, Canada, is chairing an IASP workforce that is creating guidelines for appropriate restraint and ethical use of neuroimaging findings on pain.
“Remember, pain is old and shared,” Tracey concluded her talk, emphasizing the social aspects of the pain experience; it is normal for humans and animals to feel another’s pain and suffering, and to care for others. Therefore, Tracey said, it is important to understand the basis of non-physical (empathic or vicarious) pain. She described recent neuroimaging studies comparing vicarious pain, such as that of individuals shown photographs of painful experiences of others versus pain experienced oneself (Krishnan et al., 2016). Together, these studies show that the vicarious pain experience can be thought of as “a mentalization” of the pain system, where certain patterns of brain activity within this system predict vicarious pain.
Neuroimaging is making great strides toward understanding the human experience of pain, and there is promise to translate new findings to improve clinical care of individuals with chronic pain. While much remains to be explored in animal studies and in human neuroimaging studies, we must continue to “try to understand all [of these findings] as good scientists,” Tracey said.
Katherine Martucci is a pain neurobiologist and postdoctoral scholar at Stanford University in the Department of Anesthesiology, Perioperative and Pain Medicine, Division of Pain Medicine.
Image: International Association for the Study of Pain