Editor’s Note: This article originally appeared on the Pain Research Forum, and has been adapted for RELIEF.
A task force convened by the International Association for the Study of Pain (IASP) and made up of researchers in neuroimaging (i.e., brain imaging), patient care, neuroethics and law has published a Consensus Statement in the journal Nature Reviews Neurology laying out recommendations for the uses of neuroimaging of pain.
The report, aimed at decision makers in health care, government and law, considers the ethical and legal implications of using neuroimaging to detect or diagnose pain. The report strongly cautions against use of functional magnetic resonance imaging (fMRI) in the courtroom, for example to “prove” the existence of chronic pain, which the authors say would be premature and an inappropriate use of the science. The authors also make the case for increased use of fMRI to understand the brain mechanisms underlying chronic pain and to better target specific treatments to individual patients.
Robert Coghill, a pain brain imager at Cincinnati Children’s Hospital, US, who was not an author on the Consensus Statement, applauded the publication.
“There are phenomenal new technologies coming down the pike, and scientists need to be proactive and make sure they’re used to benefit patients rather than to deny them care. This report goes a long way in that direction by pointing out the benefits and the potential pitfalls of using neuroimaging to make decisions about pain,” he says.
However, some researchers not involved in the report are saying it is too stringent, and an overreach of the role of scientists.
For instance, pain brain imager Vania Apkarian, Northwestern University, Chicago, US, called the report a “dark-side view of science” that focuses too heavily on the limitations of neuroimaging without celebrating its recent advances and potential. “They’re missing the positive part of where we are. We are on the threshold of a revolutionary time in the field,” Apkarian says.
The project was spearheaded by Karen Davis, University of Toronto and Krembil Research Institute, Canada, who served as chair of the task force, and by former IASP president Rolf-Detlef Treede, Heidelberg University, Mannheim, Germany.
“There are some fantastic applications that can come down the road,” Davis says, “but we need to keep in mind the potential harm or problems that might arise, and what that might mean for society.” (See previous RELIEF interview with Davis here).
As stated in the report, the task force has come up with a list of seven “standards of evidence that must be satisfied before any brain imaging measure can be considered suitable for clinical or legal purposes.” The criteria include requirements that any biomarker for pain (i.e., an objective, measurable indicator of pain) be both applicable to individuals and generalizable to the patient population being tested.
Is the science ready for prime time?
Chief among the task force’s concerns is the use of neuroimaging in the courtroom as legal evidence that an individual patient has or doesn’t have chronic pain. Although many neuroimaging studies have provided evidence that brain processing differs in patients with chronic pain conditions, those findings have mainly been gleaned from studying large groups of people, which only allow researchers to make generalizations about the conditions, Davis said.
“The problem has always been that MRI is pretty good at finding differences between groups of people, but the ability to find something wrong in an individual—it’s really difficult,” Davis said.
One important issue is the technical challenge of using fMRI. For example, the image quality depends on how still a patient can hold his or her head during the scan. But the biggest barrier to using fMRI to diagnose pain, she added, is “the realization that there is such vast variability across people’s brains,” both between individuals but also within a given person’s brain over time, depending on a range of factors like how much sleep, exercise or caffeine they’ve had. “So to look for an abnormality in one patient—that’s not realistic.”
The authors lay out several obstacles in interpreting brain imaging data as representing a pain state in humans. One is called the “reverse inference” problem, according to which the ability to infer pain perception from a brain activity pattern depends on whether the pattern is present during pain, and also not present during non-painful states. A brain activity pattern present with pain might instead reflect salience, or pain’s attention-grabbing ability, and thus could also be present in other salient situations. No one brain circuit or activity pattern has yet been identified as exclusive to pain, they write in the report. Further, it is unlikely that any given pain biomarker will be present in every type of pain, or in every individual with a given condition.
The nature of chronic pain also makes it difficult to capture with fMRI. Most of the imaging work to date that has identified brain patterns that correlate with painful states have been collected in healthy volunteers by evoking acute pain in the laboratory. Patients with chronic pain, in contrast, by and large suffer from ongoing pain that fluctuates over time. The brain circuitry involved in ongoing chronic pain is different and likely more complex than that activated by acute pain in healthy people.
Apkarian, however, believes the science is further along—and that scientists should stick to the science. “At the individual level, we can tell with high confidence whether the brain has chronic pain or not—that’s the bottom line.” As for how the technology is used in the courts, he says, “I cannot control how it’s being used. That’s between courts and industry and society. It’s not a scientific issue; it’s a societal issue.”
Order in the court
Joy Hirsch, a brain imager at Yale University, New Haven, US, who was also not an author on the Consensus Statement, said that while she agrees with the scientific content of the report, “it kind of puts a wet blanket” on using fMRI in the clinic and the courtroom. “Basically, the authors say, ‘because we don’t understand everything about pain, we can’t document it for legal purposes.’ It’s an extreme view.”
Can fMRI be used as proof that someone has chronic pain? In some cases, Hirsch said, such as widespread pain, “I would say absolutely not. But in some cases, we can, such as when an individual has an injury on a unilateral side of the body like the hand or foot.” If brain activity representing the injured side is exaggerated compared to the uninjured side in response to the same stimuli, Hirsch contends, “this is consistent with the fact that sensation is extreme on one side relative to the other.”
One such case was that of Carl Koch, a truck driver who sued his employer, Western Emulsions, in 2015 after hot asphalt burned his hand and left him with severe chronic pain. Hirsch, then working at Columbia University, New York, US, provided an fMRI of Koch’s brain, which his lawyer introduced as evidence in the case. Sean Mackey, a pain researcher at Stanford University School of Medicine, Palo Alto, US, and an author on the new report, testified for Western Emulsions against the use of fMRI in the case, arguing that the test was not specific enough to indisputably demonstrate pain.
“There was no doubt in my mind that he had pain,” Mackey says of Koch. Nevertheless, “I thought it was a clear abuse of the science—not by Hirsch, but by the attorneys using the test she did—using it inappropriately.” (Western Emulsions ultimately settled with Koch).
The case, Mackey and Davis say, was an impetus for writing the new report. But the authors’ motivation was not to deny patients a tool that could provide proof of their pain, but rather to protect them.
“There is a fear in the scientific community that lawyers are going to use this technology prematurely, which will lead to misuse in the legal setting, and that patients can be harmed,” Mackey said. For example, a false negative—the absence of a pain marker on the fMRI of someone with chronic pain—might lead to their being denied health care or workers’ compensation.
“When imaging is used as a tool to deny care, that’s a real problem,” said Coghill. “Given that this could be used in a medical-legal context where financial incentives come down to denying patient care rather than providing it, false negatives have to be made very clear to juries,” Coghill said. “Absence of a positive finding is absolutely not the same as a negative result.”
The future of imaging
In the end, Mackey said, the question of whether fMRI can detect the presence or absence of pain is “uninteresting and boring. If I want to know if my patients have pain, I’ll just ask them. I believe them.” That sentiment underscores another tenet of the report: that fMRI or any other biomarker should be used as an adjunct to patients’ self-report of pain.
Much more interesting to researchers, Mackey said, is “where the future lies: in the area of neuroprediction and neuroprognosis.” Neuroprediction is another term for precision medicine. The aim is to use brain scans combined with other biomarkers to predict a patient’s response to therapy, be it a surgical intervention, medication or a psychological treatment, for example. Meanwhile, “in neuroprognosis, we want to use these scans to determine who will develop chronic pain or become addicted to opioids after injury or surgery,” Mackey said. Scans might even be able to predict whether a patient experiences side effects from treatments, or whether certain treatments might worsen chronic pain.
“These are still early days, but there are tremendous possibilities that might help patients avoid unnecessary surgery or drugs,” Davis says.
Apkarian was more enthusiastic about the current state of fMRI. “We are ready to use brain imaging for individualized medicine, and to know who should be treated with what. The whole field has exploded into a fantastic scientific endeavor, and that is exciting.”
Stephani Sutherland, PhD, is a neuroscientist, yogi, and freelance writer in Southern California.
The Neuroscience of Pain (The New Yorker, July 2, 2018).