People vary markedly in their ability to tolerate and control pain. This underlying variability, which is reflected in differing behavioral responses to pain, to placebo, and to analgesic drugs, can be quite unpredictable. Some people get substantial pain relief from placebo or from opioid analgesics, while others have little or no response, creating a challenge for effective pain treatment in the clinic. Variability also causes problems in clinical trials, where interindividual variation has the potential to obscure differences between pain treatments, or between active analgesic and placebo.
In three recent imaging studies, researchers have detected differences in brain structure and function that might explain some of the interindividual variability in pain, placebo, and analgesic response. If these results are replicated, and the methods are sufficiently robust, these imaging techniques could be developed into biomarkers that would allow physicians and researchers to identify individuals who are more or less likely to respond to placebo or analgesic drugs in trials or in clinical practice.
In addition, because these studies link higher-order processing centers in the cortex to both descending pain modulation and to reward pathways, the results may also provide a physical basis for the influence of psychological functions, such as emotion, memory, and even personality characteristics, on individuals’ placebo and opioid responses.
“Each study shows, in a slightly different way, that individual variability in response is related to how vertically integrated the brain is,” said Tor Wager of the University of Colorado at Boulder, US, an expert in pain pathways in the brain, who was not involved in any of the studies. “What is happening in the cortex may affect the development and persistence of pain over time.”
Predicting placebo power
In the first study, Vania Apkarian and his colleagues at Northwestern University, Chicago, US, used functional magnetic resonance imaging (fMRI) to study placebo effects in 30 chronic back pain patients. The patients had participated in a clinical trial, also carried out by Apkarian’s group, that compared a 5 percent lidocaine skin patch, frequently used in the clinic, with a placebo patch (Hashmi et al., 2012). The trial found no differences in pain relief between the placebo and analgesia groups, although some patients in each group experienced substantial pain relief in what was essentially a placebo effect. After the trial, the researchers retrospectively analyzed baseline fMRI data to see if there were any differences in brain connectivity between the patients who later showed a placebo response and those who did not. The study, led by first author Javeria Hashmi, was published in the December 2012 issue of Pain.
Apkarian and his group found that placebo non-responders did, in fact, show significantly different baseline brain connectivity patterns compared to responders. For example, non-responders had a stronger connection between the left medial prefrontal cortex (PFC), an area previously implicated in the placebo response that also plays a role in emotional responses, and the bilateral anterior insula, a cortical region linked to self-reflection and self-referential thinking. Combining this observation with multiple other differences in baseline connectivity between the two groups, as measured using two different fMRI techniques, the researchers could predict with 90 percent accuracy which individuals would be placebo responders and which were destined to be non-responders. These findings were replicated in a separate, albeit smaller, cohort of patients with chronic back pain (seven patients) or chronic pain due to osteoarthritis (five patients).
“We think of placebo effects as transient and unpredictable, but this was predictable,” Apkarian said. “These images can predict a priori which patients will respond to the placebo.” In addition, he said, “Some of the brain areas identified [as showing different connectivity patterns] are responsible for rational thinking and planning for the future, and the amount of information that they share with pain areas determines if the placebo works. These results suggest that some individuals have better cognitive control of their pain.”
In the second study, which appeared in the November 2012 issue of Pain, Ulrike Bingel and her colleagues at University Medical Center Hamburg-Eppendorf, Germany, used a complementary technique—diffusion tensor MRI—to examine anatomical connectivity in 24 healthy adults who differed in their responses to an experimental protocol designed to elicit a placebo analgesic response. Diffusion tensor MRI is used to visualize the location and orientation of white matter tracts, which are bundles of neuronal axons and glial cells that connect brain regions with one another. First author Niklas Stein and colleagues reported that people who experienced a placebo response showed higher white matter integrity, thought to be reflective of greater connectivity, in neural tracts linking the dorsolateral PFC and rostral anterior cingulate cortex to the periaqueductal grey (PAG). The PAG is a subcortical region that projects to the brainstem’s raphe area, from there connecting to the dorsal horn of the spinal cord and playing an important role in descending pain modulation.
Although these differences in structural connectivity could reflect changes due to increased use of these pathways during past pain experiences, the authors argue that this is unlikely because the study subjects were healthy individuals without a previous history of extended pain. Instead, the researchers suggest that the differences reflect an innate variability in the subjects’ abilities to engage in descending pain control. Because it takes time for these tracts to develop, this work provides evidence for relatively permanent differences among individuals, although whether these structural properties reflect accumulated changes created over a lifetime or underlying genetic or developmental differences is still a matter of speculation. In any case, “these findings support the idea that being a placebo responder is a stable trait,” commented Wager.
A marker for opioid response
The third study, performed by Vishvarani Wanigasekera and her coworkers in Irene Tracey’s group at the University of Oxford, UK, examined interindividual variability in the responses of 25 healthy individuals to opioid analgesics. Multiple studies have linked the wide variability in opioid response to genetic differences in the endogenous opioid system, but these genetic differences are too complex and multifactorial to serve as a basis for distinguishing among patients, at least for now.
In addition to its functions in reducing pain, the endogenous opioid system enhances the pleasantness of rewarding stimuli and influences individual responsiveness to rewards. Pain relief is, in a sense, a reward, so Wanigasekera and colleagues asked whether individual variability in the psychological trait reward responsiveness, as measured by a questionnaire, might be predictive of opioid response and thus act as an indirect measure of underlying variation in the endogenous opioid system. They used Carver’s BIS/BAS scale of behavioral avoidance or activation, which measures an individual’s sensitivity to reward signals and ability to respond positively when exposed to such signals, to measure reward responsiveness as a personality trait. In addition, they used fMRI to measure activity in brain reward circuitry in response to noxious stimuli and to opioids in the same individuals.
The 25 healthy volunteers were subjected to a moderately painful heat stimulus applied to the right forearm before, during, and after a 40-minute infusion of the opioid remifentanil or a control saline solution. Brain activity was measured using fMRI, and subjects were asked to rate their pain levels at intervals during the protocol.
The investigators found that individuals with high reward responsiveness, as measured by the questionnaire, reported higher levels of analgesia in response to remifentanil. The magnitude of pain relief was predicted by the baseline level of neural activity in the endogenous, opioid-rich regions of the reward circuitry in response to painful heat, as measured by fMRI. The results were published online October 8 in the Proceedings of the National Academy of Sciences USA.
While these results are preliminary, Wanigasekera and her colleagues argue that these two measures (the questionnaire combined with brain imaging of the heat response) could be used together to reliably identify individuals most likely to benefit from opioid therapy. They are planning a larger study in the setting of acute pain with a uniform cohort of surgical patients to further validate these measures. “It is complex, but it can be done,” said Wanigasekera.
Objective measures of interindividual variability such as the imaging techniques used in these three studies represent an “important clinical opportunity,” said Apkarian. His group intends to test the robustness and replicability of their findings in a larger population of chronic pain patients. If placebo responders can be identified reliably, he envisions a time in which “the placebo can become a predictable tool for use in medicine.”
Fabrizio Benedetti of the University of Turin, Italy, an expert on the placebo response who was not involved in the current studies, commented in an e-mail that the possibility of using imaging biomarkers to guide treatment choices or stratify patients for clinical trials is “highly speculative right now, but surely promising and interesting.” On a more basic science level, he noted that “these studies shed light on some of the physical bases of cognitive and emotional processing. For example, stress and anxiety do play a role in placebo responses, and neuroimaging gives us information on the brain regions that are involved.”
Although preliminary, these studies represent early steps towards linking psychological modulators of placebo analgesia with their underlying neurobiological pathways in the brain, and should promote research in this area by allowing investigators to select and characterize study subjects more carefully.
Megan Stephan, PhD, is a science and medical writer based in Cambridge, Massachusetts, US.
Image Credit: Borsook et al. Neuroimaging revolutionizes therapeutic approaches to chronic pain. Mol Pain. 2007 Sep 11;3:25.