The Pain Chronicles (35 page)

A UNIVERSE OF HURT

I hope functional imaging will progress in my lifetime enough to have clinical input,” comments John Keltner, who spent several years working with Irene Tracey in her center before deciding to begin new training as a psychiatrist. He points out that CT scanning and MRI technology were revolutionary technologies, with huge immediate clinical impact, because they created the first anatomically accurate pictures of the inside of a body. How much more revolutionary it seemed when, in the late 1980s, functional imaging produced the first 3-D movies of the working brain. But the films were and remain largely indecipherable. Researchers puzzle over the images like Columbus staring at the gray shoreline, thinking,
India?

“We don’t understand virtually anything about the human brain,” Dr. Keltner says soberly. “Pain, sleep, memory, thinking, adding two and two—we don’t understand any of that stuff. When I started doing functional imaging research on pain twenty years ago, I thought it would soon lead to a meaningful diagnostic tool. Now I hope that in the next forty years I will help come up with a test that will be able to answer a simple clinical question about a patient’s pain, such as, Should we focus on treating your toe or your emotional state? That’s such a basic question, and right now there isn’t a single diagnostic test that can answer it.

“Brain function turns out to be so complicated. It would be a lot easier if there were a part of the brain associated with pain and only pain, but so far we haven’t been able to find a single unique marker that would allow us to definitively identify a pain state. If you show me a brain scan and say,
Is
this person in pain or thinking about running from a tiger?
I wouldn’t be able to tell you.” A brain scan of a person in a state of repose would look different, of course, but pain and fear are both salient experiences with strong activation in common brain regions.

“We had to start recognizing that the fundamental pillars of human experience—pain, fear, anxiety, sadness, joy—involve the whole brain, with dozens of areas switching on and off. And many of those parts also light up in scans that have nothing to do with pain. Of the countless possible brain states, perhaps only ten thousand happen if a person is in pain. But nobody has come up with a rich and complicated enough model to analyze the complexity of the distributed patterns of neural networks and deduce any underlying rules. The daunting aspect is that it’s a bit like chess. Chess is eight spaces by eight spaces and you have thirty-two pieces, yet by the third move of any game, there are a thousand possibilities.” But instead of thirty-two pieces, the brain has a hundred billion neurons that can form an unknown number of neural networks.

“The pictures are so complicated,” he says for the fourth time. “If we change one parameter in an experiment—say, change a visual cue from blue to red or change from color to sound—we’d expect to see corresponding changes in the auditory and visual sections of the brain. But instead we see changes in a dozen areas.

“I’m not discouraged,” he adds, sounding like a hiker who has realized that he can’t figure out where he is on a map, but is reminding himself that he likes hiking and should trudge on. “We’re literally grappling with the fundamental aspects of human beings. We naively believed that pain is simple—it hurts or it doesn’t hurt—so there should be a single brain state we could see every time someone is in pain. But what we’ve stumbled into is the discovery that there’s a relative
universe of hurt
—that hurting is an immense, rich, and varied human experience, associated with an unknown number of possible brain states. From a scientific position, we’re overwhelmed at how large that universe is. We’re still at the stage where each step forward makes us realize how far we have to go . . .”

“We’re getting there faster than we thought possible,” Sean Mackey responds. He and his Stanford colleagues recently made a significant encroachment in an experiment in which they were able to distinguish, with approximately 85 percent accuracy, brain scans of volunteers given a painfully hot stimulus from those given a non-painfully hot stimulus or no stimulus. A further step, he points out, would be to ask volunteers to simply
imagine
being given a thermal stimulus and to see if he could distinguish those scans from the scans of volunteers who actually had the heat stimulus (in other experiments, imagined pain has been shown to engage similar brain regions as physical pain).

Although Dr. Mackey believes that this kind of simple, acute pain probably has very little to do with the experience of chronic pain, he also feels that “we are rapidly moving to a point where we may be able to detect the subjective experience of pain and to find a separate signature for it that allows us to distinguish it from other affective states, such as depression or anxiety.” He points to the amazing acceleration of certain kinds of technology, such as machine learning techniques and pattern classifiers—complex software algorithms that can be fed a set of known examples (scans of people experiencing thermal pain, or not) and then used to classify new, uncategorized scans.

He feels it is important, however, to understand that this technology is a long way from being able to recognize—let alone provide insight into—the state of chronic pain. He has great concerns about how this kind of technology will be used, because he thinks it is “ripe for abuse by the legal community and insurance companies to try to disprove that somebody is having chronic pain and deny care.” He has already seen a lawsuit that relied on claims about using scanning to detect pain. The suit involved a worker who developed chronic pain after his arm was injured by molten tar. The worker’s lawyer claimed that a cognitive neuropsychologist had validated the man’s chronic pain by scanning his brain. The expert had, in fact, scanned the man doing various activities, such as squeezing a ball, with both his injured and his uninjured arm. Because the two brain scans were different and the scan of the injured side showed more brain activity, the expert inferred that the scans proved the patient had more pain on the injured side.

Yet, Dr. Mackey points out, there is no evidence that greater brain activity necessarily indicates greater pain. The difference could reflect the man squeezing the ball harder with one hand (or feeling more anxious while using the injured hand, or any number of other factors). As a physician, Dr. Mackey believed in the worker’s pain, but he felt it was important to refute what he believed was specious methodology, which could as easily be used in other cases to falsely discredit patients’ pain.

“What was remarkable is that the expert who was arguing for it is an internationally acclaimed researcher in cognitive systems,” he recalls. “It showed me that even a really smart person can be naive in regard to pain and approach it as if it were this Cartesian experience”—as if increased brain activity was the bell in the brain that simply rang louder when there was more pain.

“The holy grail” of functional imaging, he says, “would be to be able to distinguish different patterns for different types of pain and to use that information to tailor specific treatments for a particular person. As you know, when I treat a person with pain, I go through a process of trial and error of different medications that is very laborious and frustrating for both the patient and me. I hope that one day we could use scanning, along with other information, as a predictive tool in which we could tell someone,
You have this genetic profile, this type of injury, and based on this imaging information, we believe you are going to respond to this particular therapy.
”

When asked if he thinks there will be a breakthrough in the treatment of chronic pain that is similar to that of anesthesia, he points out that a century and a half later, we don’t know how anesthesia works, and although anesthesia allows us “to conquer pain, the person can’t be conscious!” Moreover, turning off an entire system, the way anesthesia does, is going to be much simpler than trying to turn off only the pain system, because pain is so intricately woven throughout the brain, with so many networks available to reroute it if one is disabled.

At the moment, the pictures offered by the current technology are too primitive—the resolution of the images too coarse—to fathom pain. But “ultimately,” he says, “if we believe that our experiences and beliefs and perceptions are made up of firing of neurons and flow of information in neural circuits, then each experience has to be characterized by a different pattern. The limitation in understanding those patterns is only technological, and technology continuously improves.” By contrast, he says, “some people believe that consciousness is uniquely human and God-given: that it is inherently nondeterministic and cannot be defined or distilled down to the firing of neurons. I believe that it can, and I’m of the opinion that we’re inching closer and closer.”

It’s easy to think of obstacles. Perhaps we need to see how the neurons are wired together to recognize pain, which the scans cannot show. Perhaps—even though the technology is improving—the gap between the coarse current technology and what we would really need will always be too great. Or perhaps each individual’s pain is simply too individual. Even if perceptions are all composed of neurons firing in particular patterns in particular networks, it may be that these patterns differ sufficiently from person to person in ways that will make it perpetually difficult for us to interpret them, even when we have functional imaging scans with far higher resolutions than today’s.

The neural networking patterns of pain might turn out to be like fingerprints—something that everyone has and that, in essence, serve the same purpose for everyone, but differ too much in random detail from person to person to be able to meaningfully categorize them. After all, one could say that just as thoughts all spring from neuronal patterns, all the wondrous qualities of a painting depend on arrangements of paint on canvas. But we cannot now and probably never will be able to teach a computer to be able to analyze a new painting and say whether it has any merit—whether it is interesting, pleasing, or stirring—by showing it examples of thousands of famous paintings and hoping it can discern an underlying, predictive pattern.

“I’m not saying that we’re going to be able to see the ghost in the machine—the experience of pain itself,” Dr. Mackey responds. “My hope is that we can get to the point where imaging can become a clinically useful tool, in the same way that we can use a cholesterol test as a biomarker for heart disease to guide us to choose an effective therapy. I don’t think functional imaging will allow us to ‘see’ pain or suffering or love in the foreseeable future. But in the same way that pattern-classifying software might enable us to identify a painting as being from the Impressionist period, or possibly even as a Monet, I think it will be able to identify different types of pain. In regard to using functional imaging as a diagnostic tool, I believe the question is now a
when
, not an
if.
”

“The idea of functional imaging as a pain meter is unrealistic,” counters Scott Fishman, the head of pain services at UC Davis. “Humans are variable enough in our physiologies that doctors can’t even agree on reading an EKG or on what a stroke means, which are much more clear-cut things. How are scans going to prove or disprove someone else’s pain and suffering—or even illuminate its nature?”