Earlier this year, I wrote about my patient, Andrew, an engineer who developed a heroin habit. An unfortunate series of joint replacements had left Andrew with terrible pain and, when his medication ran out, he turned to heroin.

Months after his surgeries—after his tissue and scars had healed—Andrew remained disabled by a deep, biting pain. I recall puzzling over his pain, how it had spread throughout his body and how previous clinical teams had prescribed progressively higher doses of opioids to tame it. 

Andrew had transitioned from acute pain (i.e., pain from his surgical wounds) to chronic pain (i.e., pain in the absence of an obvious cause), but it was unclear to me whether this reflected a drug tolerance or a different pain process.

Drug Tolerance and Chronic Pain

The difference between drug tolerance and chronic pain is a difficult concept to get hold of. In the hospital workroom one morning, I realized how confused I was by the topic and paged the hospital’s on-call pain specialist. Fortune smiled and Donna-Ann M Thomas, Yale University’s Pain Medicine Division Chief, picked up the phone and patiently explained how tolerance and chronic pain are quite different.

Andrew became tolerant to opioids when his body required progressively larger doses to have the same effect. Opioids activate the Mu opioid receptor, which blocks pain signals in the spinal cord. To find a way around the opioid blockade, Andrew’s body had made more Mu receptors to compensate for the drug, meaning more drug had to be present to stifle the pain signal, hence the escalating doses.

Chronic pain, Thomas explained, differs from tolerance and even acute pain. In Andrew’s case, chronic pain was the agony he felt weeks after one of his (many) surgeries, despite the fact that his body had healed and was no longer sending distress signals from his joint, through his spinal cord.

The Purpose and Problem of Pain

Your body has a system of neurons and neural pathways whose sole purpose is to detect pain and transmit painful signals. Specialized subsets of neurons are devoted to specific forms of pain, such as pain caused by heat, or chemicals, or cuts. Pain is a serious matter. Your brain carefully monitors for threats in the environment and, should you accidently touch a hot stove, for example, your pain system seizes your attention and throws your cognitive machinery into red alert. You yank your hand away Your emotions are riled up—“Who left the damn stove on?” Maybe you cry a little. Maybe you’re in a bad mood for a while.

The goal isn’t to create a moody panic, but rather to mobilize your brain’s learning, memory and decision-making systems. The purpose of pain is to associate an environmental threat with a strongly negative emotion and to demand a course correction away from whatever caused that painful signal. In this sense, pain is a useful alarm—a lighthouse that keeps you from more serious trouble.

The problem is that, like any machine, the brain is vulnerable to errors: if a response to pain is learned, it might be learned incorrectly. Or the alarm might never shut off.

So why does pain transition from acute to chronic? A traditional view is that the more severe an acute injury, the more likely the pain will linger—like the way hitting a cymbal with a hard, heavy mallet creates an enduring crash. And yet it could also be that the cymbal itself, once struck, somehow sustains its own ostinato. Is the problem the mallet or the cymbal?

Chronic Back Pain

The neuroscience question is whether the brains of people who develop chronic pain (known as pain chronification) differ from those who don’t and whether this difference is detectable before or after pain chronification.

To answer the question, a group of neuroscientists led by A. Vania Apkarian at Northwestern University followed a group of thirty-nine patients who had injured their backs.

For three years, Apkarian used M.R.I. to gather incremental measures of the structure and function of patients’ brains. The goal was to see how and whether the brain changes during pain chronification and whether the brains of people who transition from acute to chronic pain differed from those whose pain resolved.

Apkarian’s group had shown that the corticolimbic network was involved. The corticolimbic system includes the medial prefrontal cortex, nucleus accumbens, hippocampus, and amygdala; structures associated with cognitive processes like judgment, motivation, memory, and emotion. Previous research from Apkarian’s lab had used functional M.R.I. to show that strong functional connections between the nucleus accumbens and the medial prefrontal cortex prospectively predict pain chronification.  To follow-up this study, they wanted to know whether the brain’s structure also prospectively predicted patients who would develop chronic pain.

What they discovered was surprising: the presence of pain does not restructure the corticolimbic network. Rather, even before the onset of back pain, the corticolimic network was more densely connected in patients who developed chronic pain than in patients who did not. In addition, patients with smaller hippocampi and amygdalae—structures with strong genetic dependence implicated in emotional learning, anxiety, and stress regulation—were also more likely to develop chronic pain.

Together, these observations suggest that some intrinsic, genetically encoded brain property—and not the pain itself—led to chronic pain. The cymbal, not the mallet, was the culprit.

The implication of Akparian’s work is that a patient’s risk of pain chronification can be measured and therefore clinically managed. For patients like Andrew, it could mean the difference between properly managed pain chronification and a crescendo into opioid tolerance and heroin.