Beyond Poppy Juice: The New Science of Pain

He first noticed what looked like a rash on the left side of his chest that wrapped under his arm and spread out across his back and shoulder. Nine months later, the painful blisters of the rash seemed insignificant in comparison with his new symptoms: constant burning pain, punctuated with bouts of shooting and stabbing pain. The slightest brush of his shirt across his back was agonizing. This patient (we will call him Arthur) had not been helped much by his doctors, who had tried to manage his pain with creams and pills that made him feel greasy, drowsy, and light-headed or nauseated (or both), but still in pain.

Normally, only a stimulus that damages tissues or causes other harm to our bodies leads to pain. In Arthur’s case, however, even an innocuous stimulus such as light touch produced pain. This is called “mechanical allodynia.” Injury or disease also can make changes in neurons and neural circuitry that cause a patient to experience greater-than-normal levels of pain from noxious or painful stimuli. This is called “hyperalgesia.” Allodynia and hyperalgesia are key characteristics of persistent pain.

For most of us, if we were to injure a knee or an elbow—say from tennis or jogging—we would expect the pain to persist for a few days or perhaps weeks. Such pain protects us by forcing us to rest the injured limb for a short time, thereby avoiding further damage. As we heal, the pain lessens and then disappears. By contrast, pain that persists long after the injury has apparently healed is not protective. Whether characterized by mechanical allodynia or hyperalgesia, this pain is pathologic. Pain has become a chronic illness.

At the beginning, Arthur was suffering from shingles (herpes zoster), but he became one of a growing number of unlucky patients whose shingles outbreak results in the development of a pain syndrome called post-herpetic neuralgia (PHN). Shingles itself results from a resurgence of the virus that causes chicken pox, but we do not fully understand why some people develop PHN and others do not. What is clear, though, is that the likelihood of developing PHN following a shingles outbreak increases dramatically with the patient’s age.

Each year, some 50 to 100 million Americans suffer either acute or chronic pain such as that from PHN. The costs in lost income and medical expenses run into billions of dollars. As the percentage of Americans who are elderly keeps increasing, the number of sufferers—and the costs—will burgeon as a result of chronic ailments that have pain as their chief complaint. Although this problem affects more and more of us, the medical management of pain still relies chiefly on treatment approaches that date back to the early twentieth century: local anesthetics, aspirinlike drugs, and opioid or morphinelike compounds. These approaches work for acutely painful injuries that resolve in a few days or weeks, but are woefully inadequate for persistent or chronic pain. In short, Arthur’s experience is anything but unique; he is joined by a legion of patients who suffer chronic pain, seeking relief but finding none.

Pain: “A Disease Unto Itself”

The dearth of new pain-treatment approaches is surprising. In the last decade, research has increased our understanding of pain at every level: conversion of the initial stimulus to electrical impulses, how those impulses are conducted in peripheral nerves, and the impact of the stimulus on central nervous system pathways from the spinal cord to the cerebral cortex. In part because it is a prominent component of neuroscience, pain research benefited from attention to neuroscience during the Decade of the Brain. Then, in 2000, Congress declared the first decade of the twenty-first century the Decade of Pain Control and Research.*

Second, and probably most important, is wider awareness that our ability to treat chronic pain is woefully limited and, unless remedied, will have a profound impact on the U.S. health care system in years to come. PHN is just one of the chronic and debilitating pain syndromes that become more frequent in an aging population. Others are chronic inflammatory diseases such as arthritis, pain syndromes in the central nervous system such as those associated with stroke or spinal cord injury, and pain syndromes in the peripheral nervous system (which are called “neuropathic”), including PHN and diabetic neuropathy. Chronic pain may also result from the treatment of other diseases. Several drugs used to treat cancer, for example, may give patients neuropathic pain. As people live longer, thanks to better nutrition and advances in modern medicine, more and more of us will experience one of these pain syndromes. Fortunately, professional societies and public interest groups devoted to the study of pain and its control are taking advantage of the Decade of Pain Control and Research to increase awareness on the part of policymakers and the public.

We see hope for the future. The study of the nervous system has given impetus to new approaches to diagnosing and treating pain, approaches that will emerge from understanding the nervous system’s own mechanisms for amplifying or suppressing pain. These may usher in an era that will enable patients to exercise control over their own pain and thus improve their quality of life even in the throes, or the aftermath, of serious disease or injury.

All Pain Is not the Same

Like the sensing of usually innocuous stimuli such as warmth or touch, the sensing of pain involves the activation of discrete sets of neurons. The first step is the activation of primary sensory (afferent) neurons referred to as nociceptors. Ever since the discovery that local anesthetics block pain (and other sensations) by blocking the electrical activity (called action potentials) in primary sensory neurons, researchers have appreciated that these neurons are a critical target for therapies to treat pain. The pain associated with the vast majority of chronic syndromes reflects, at least in part, activity in nociceptors. Blocking it may achieve maximal pain relief with minimal side effects.

1. the type of injury (for example, bacterial infection versus nerve damage);
2. its location (somatic versus visceral);
3. hormonal status (women and men are different); and
4. timing (in terms of both when the injury occurred during our life and when the initial injury was treated).

Arthur’s experience illustrates the impact of two of these factors: the type of injury and timing. During his initial shingles outbreak, his pain was probably caused by both inflammatory and neuropathic processes. The inflammatory processes likely reflected the re-emergence of the herpes virus and the associated release of inflammatory substances. The neuropathic component, in this case, would be associated with changes in the nervous system that occurred following the virus-induced death of a large number of sensory neurons. In contrast, the pain Arthur experienced from his PHN was strictly neuropathic: It was present after all signs of inflammation had cleared up. The point is that the mechanisms underlying the pain associated with the shingles outbreak (pain that may respond to nonprescription medications such as ibuprofen) are distinct from the mechanisms underlying the pain associated with PHN, which have to do with the long-term changes in the nervous system that occur following the virus-induced nerve injury. This type of pain is not responsive to ibuprofen-like drugs.

Beyond Local Anesthetics

How will the health practitioners of tomorrow treat chronic pain like Arthur’s? There are three broad areas of research on the primary afferent (sensory input) neuron from which novel therapeutic interventions are likely to come.

Better Targets for Traditional Pharmacology

The first area involves a traditional pharmacological approach. For us to detect an aspect of our environment, our nervous system must convert the physical energy of that environment (thermal, mechanical, or chemical stimuli) into an electrical signal. That electrical signal, in the form of an action potential, must then be transmitted to the central nervous system. In the last decade, researchers have identified many of the specific proteins, called transducers, that are involved in the functioning of nociceptors (pain receptors) in the skin and elsewhere. For example, our ability to sense painfully hot stimuli involves a specific molecule originally called vanilloid receptor 1 (VR1), now known as TRPV1.¹ The original name for the molecule, VR1, reflected its activation by the vanilloid compound capsaicin, which gives hot chili peppers their “burn.”

The majority of nociceptors respond to various forms of stimuli (and therefore are called polymodal). Each nociceptor contains a number of different transducers, enabling it to respond to noxious thermal, mechanical, or chemical stimuli (or some combination of them). Thus, the allodynia that Arthur experienced both during his shingles outbreak and again while he suffered from PHN—called mechanical allodynia because it was caused by touch— might be treated with a drug that blocks the molecule that enables us to sense mechanical stimuli. TRPV1 is a promising target for pharmaceutical intervention because it does much more than respond to noxious thermal stimuli and capsaicin: Most likely it is responsible for much of the ongoing pain associated with inflammation. TRPV1 is present in exceptionally high concentrations in nociceptors and its properties are changed by many of the events that occur in the presence of inflammation, including a decrease in tissue pH and the a protein unique to sensory neurons and produced almost exclusively by nociceptors. This protein forms a channel embedded in the cell wall that affects the cell’s electrical charge by admitting or blocking the passage of sodium in and out of the cell. Called a voltage-gated sodium channel (referred to as NaV1.8), this protein is needed for converting the signal from transducers like TRPV1 into the cell’s electrical impulse. In fact, NaV1.8 is a member of a family of related proteins that are found in excitable cells throughout the body and that are the primary target for local anesthetics like lidocaine. The problem is that, because these proteins are distributed throughout the body and are critical for normal function of both brain and heart, there is only a small range of doses for local anesthetic-like compounds to treat chronic pain without producing unacceptable side effects.

Gene Expression and Viral Vectors

Taking advantage of new advances in molecular biology, researchers are pursuing two promising approaches to treating pain that target our primary sensory neurons: the targeted knockdown (decrease) of gene expression and viral vector delivery systems.

Targeted knockdown of gene expression relies on our knowledge of the molecular identity (the genetic code) of several molecules underlying chronic pain. By selectively blocking the expression of these molecules (that is, blocking their production by our genes) doctors may be able to treat pain. The method used longest and most effectively in preclinical trials, called “antisense knockdown,” utilizes small DNA sequences to initiate a process that causes sensory neurons to decrease their production of a given protein. For example, a sequence of DNA has been used in animals to decrease expression of the protein NaV1.8, resulting in relief from both neuropathic and inflammatory pain. A similar approach could be used to treat Arthur.

The Toxin/Marker Approach

Treating chronic pain will also take advantage of new insights into what makes sensory neurons tick. Studying the form and function of sensory neurons, several research teams observed that these neurons may be categorized based on whether they are associated with a specific marker. For example, in a laboratory experiment a certain group of sensory neurons binds to a unique lectin from a plant, so that lectin is a marker. The functional significance of this remains a mystery, but these neurons do share some unique properties and—here is the payoff— their activity may be necessary for the expression of acute pain. Other subpopulations of neurons are associated with other markers. Researchers have recently employed this knowledge about sensory neurons to devise ways of selectively killing them. One successful approach is to attach a toxin to a neuronal marker to take advantage of the fact that the neuron has the cellular machinery to take up that marker and transport it back to the cell body.⁴ Specific subpopulations of sensory neurons that take up their own specific marker and the toxin attached to it, and then bring the toxin back to the cell body, are killed. This approach would make it possible to target not only the neurons underlying a specific pain syndrome, but a specific class of neurons that may be contributing to it. Here is a third, if not last, resort for Arthur.

Beyond Receptors and Peripheral Nerves

Thus far we have discussed sensory neurons and how they detect and transmit pain signals from the site of an injury. The puzzle of persistent pain becomes more complicated when we move away from this local level and the signals reach the central nervous system (CNS). The barrage of impulses from the injured peripheral (local) nerves initiates long-term changes at sites of pain transmission in the CNS, resulting in both amplification and increased duration of the pain. New targets for treatment emerge as we learn more about those changes, which involve receptors, second messenger modules, and elements that control how genes go about making new proteins. The new therapies will include pharmacological approaches, techniques for altering how genes produce proteins that enhance pain, and the injection of toxins that destroy pain-producing neurons. Some of these interventions may mimic those described already for altering peripheral nervous system mechanisms of pain, but the delivery systems will have to take into account the uniqueness of CNS organization and function.

Information moving along CNS pathways for transmission of pain is modified at many stages: the cellular level, the level of local sensory neural networks, and the level of the brain and entire integrative nervous system. Modification at the cellular level occurs, for example, at the first step of pain transmission in the CNS.⁵ Signals from the peripheral nociceptors are transmitted to neurons in the spinal cord and its brainstem equivalent that receives information from the face and mouth. At these nerve terminals, a number of chemical messengers are released, including glutamate—the major excitatory neurotransmitter. Glutamate binds to various different receptors on the cell, leading to an influx of calcium through ion channels in the cell wall, the release of calcium from intracellular sites, and activation of second messenger systems in the cell’s cytoplasm. The second messenger systems, in turn, control the production of various enzymes called kinases, initiating a process known as phosphorylation that alters how receptors act and their sensitivity to transmitters. This sequence of events is complicated but very important because it initiates what we call central sensitization, which increases the excitability of pain-transmission neurons.

An example of pain modification or modulation that occurs at the local neural network level involves inhibitory interneurons in the dorsal horn of the spinal cord—short nerve cells that connect other nerve cells in a reflex arc. These interneurons form local inhibitory circuits that utilize the major inhibitory neurotransmitter in the nervous system, called gamma aminobutyric acid (GABA). After injury, the barrage of signals from peripheral nerves changes the balance between excitation and inhibition, so that these inhibitory circuits become less active. This form of change goes on at the local network level, but it also contributes to the central sensitization just described.

In the past, patients suffering from pain conditions like PHN who reported pain even after apparent healing were often told it was psychological. Now we can seek to find out whether changes in the central nervous system have occurred and, if so, treat those changes.

Pain in the Brain

Information coming from our injured skin, muscles, internal organs, or nerves is modified by control systems in our brain. For example, pain information coming from peripheral tissues is influenced by parts of the brain involved in the cognitive, attentional, and motivational aspects of our experience of the pain. Studies in the early 1970s first showed that stimulation of a part of the brain called the midbrain periaqueductal gray could yield pain relief, and the link was made between this “stimulation-produced” analgesia (pain relief) and the brain’s own pain inhibitory systems involving morphine- or opiate-like compounds and other neurotransmitters.⁶

We now know that after injury the modulation of pain signals in the brain increases, and this can either heighten or lessen our experience of pain. For example, a major pathway from the periaqueductal gray (involving a part of the brain called the rostral ventromedial medulla or RVM) affects central sensitization at the level of the spinal cord. From animal experiments, we know that neurons in the RVM can modulate pain.⁶ The activity of one type of neuron (“on” cells) appears to facilitate withdrawal from painful stimuli, whereas the activity of another type (“off” cells) has been linked with inhibiting withdrawal. After injury, this enhanced excitability and lessened inhibition at the spinal cord level tend to amplify signals relayed to the brain; they also produce changes in how returning signals from the brain are modified.

How will these insights into what is called integrative modulatory circuitry influence our diagnosis and treatment of persistent pain disorders? Pain signals relayed from Arthur’s brain may be strengthened at the spinal cord level. Gene therapy might be used to restore his loss of central pain inhibition at the spinal cord level; alternatively, chemical agents might moderate the tendency of the RVM to enhance pain signals relayed from the brain.

Persistent pain in deep tissues such as muscle and internal organs also can be related to changes in the balance of excitatory and inhibitory nerve pathways, changes that alter pain signals coming (descending) from the brain. Incoming signals from deep tissues produce more neural plasticity after injury than those from surface or cutaneous tissues. Conditions such as fibromyalgia and irritable bowel syndrome are often characterized by diffuse pain at multiple sites. The severity and location of the persistent pain vary with previous history of painful episodes, hormonal changes related to age and sex, emotional issues related to work, family, and the meaning of ongoing pain, and other factors. But underlying all this variability in pain sensation may be a pain modulatory site like the RVM. In other words, if pain signals coming from the brain are enhanced, the result can be diffuse—and sometimes more severe—pain that spreads to multiple sites. Treatments of the future aimed at restoring the balance between excitatory and inhibitory circuits coming from the brain could use cognitive, pharmacological, and gene-therapy approaches.

Finding Pain in Higher Brain Circuits

Our understanding of what underlies pain in the cerebrum is far less detailed than what we know about how pain signals are processed at earlier stages—in the peripheral nervous system, spinal cord, and brain stem. The novel approaches to pain relief that we have described are not applicable to the higher brain centers, at least not with our current knowledge, but other promising approaches to treating chronic pain at the brain’s highest levels are on the horizon.

The brain’s role in pain is being revealed by functional neuroimaging studies that use Positron Emission Tomography (PET), Magnetoencephalography (MEG), and functional Magnetic Resonance Imaging (fMRI).⁷ It came as no surprise that brain areas such as the primary and secondary somatosensory cortices would be activated by painful stimulation; pain is, after all, a sensory experience. But several other cortical areas were found to be activated as well, including the insular cortex, part of the anterior cingulate cortex, and multiple areas of the frontal lobe. One interpretation is that these multiple brain areas are collectively responsible for our perception of pain—that there is no single brain center for pain perception. It may be possible, though, to identify specific roles that these different brain regions play in the multidimensional experience that we call pain.

One example of fMRI’s capabilities comes from the case of a patient we will call Margaret, who suffered from what we earlier termed central pain—continuous pain that starts after damage to the CNS (in her case a stroke). Fortunately, central pain is relatively infrequent, estimated to occur in, for example, 8 percent of stroke cases and 28 percent of multiple sclerosis cases; but it is devastating to the affected individual, not least because it responds poorly to available pain treatments. Ironically, the brain itself has no nociceptors, so the central pain resulting from injury to the brain does not arise from the same mechanisms involved in other body tissue injury. Instead, central pain arises because some aspect of CNS processing that normally provides pain-related information is disrupted. For that reason, approaches to pain management that target the nociceptors or the spinal cord are not appropriate in cases of central pain.

Keep in mind some caveats about functional neuroimaging studies of pain. Both healthy people who are tested with transient painful stimulation and patients with chronic pain exhibit great variability in their pain-related brain-imaging patterns. Some of this variability probably reflects the way individuals vary in their sensitivity to painful events. Functional neuroimaging enables us to see whether these individual differences are related to differences in how an individual’s brain reacts.

In the same vein, patients with similar types of chronic pain may differ in their functional neuroimaging profiles. Arthur’s allodynia may be similar to Margaret’s, but Arthur may not show the heightened activation in the primary somatosensory cortex that Margaret exhibits. It is increasingly evident that patients with similar types of chronic pain, even with similar causes, can reveal distinctly different changes in neural plasticity. This undoubtedly contributes to individual differences in response to pain treatments. Knowing this, physicians can customize pain therapies based on the differences in cerebral functions related to pain. As one concrete example, PET can measure the distribution and density of opiate receptors in the brain, a key to how well an individual will respond to opiate therapy.

Treating Pain in the Brain

Direct manipulation of the brain to alleviate debilitating pain has been practiced by neurosurgeons for decades. For most of this time, the only recourse was destruction: attempts to eliminate the part of the brain deemed responsible for generating unwanted pain—even at the cost of losing other sensory function (including the protective role of pain). Reviewing decades of reports on this “brain lesion” approach to alleviating pain, one is struck by its very limited success.⁸ Destroying portions of the pain pathway, including portions of the thalamus and S1 cortex, has typically produced only transient relief, at best, and occasionally actually resulted in central pain.

Another tool for manipulating specific brain regions to treat intractable pain is Transcranial Magnetic Stimulation (TMS),⁹ which employs a pulsing magnetic field strong enough to modify the electrical properties of nearby neurons. TMS, long used to study cortical (especially motor) function in human volunteers, recently has been explored for treating psychiatric disorders such as depression. A clear advantage is that TMS is completely noninvasive, but for now it has limitations. As typically administered, its effect lasts only as long as the stimulation is applied. Although repetitive TMS exerts effects outlasting the stimulation period, ultimately a practical means of applying stimulation frequently, if not continuously, would need to be developed. Also, this form of stimulation affects only the part of the brain close to the skull, leaving inaccessible many regions potentially relevant to pain. Nevertheless, the potential is compelling enough to command attention over the coming decade.

Manipulating the Body’s Own Pain-Control Systems

Consider now an entirely different approach to pain management: manipulating the body’s own (endogenous) pain-control systems by cognitive or behavioral means. As noted, we have learned much about brain stem and spinal cord mechanisms that can lessen signals from pain receptors, and we have some ability to manipulate this system.

A prime example is hypnosis, which has been studied and used for many decades to reduce pain. Although the exact nature of hypnosis is still debated, inducing relaxation and suggestions of less pain can reduce a person’s reported pain.

This brings us to the recent resurgence of interest in placebo pain relief, or analgesia. For decades, we have known that people’s pain can be reduced if they take a sugar pill but believe it is a pain reliever. Paradoxically, the placebo effect has long been viewed as simply a nuisance in tests to establish the pharmacological effectiveness of painkillers; now it is recognized as another form of cognitively mediated endogenous pain reduction. As in the case of hypnotic suggestion, functional neuroimaging studies have begun to identify specific brain regions where pain response is modified by the placebo effect.¹¹

Taking these imaging capabilities one step further, one can envision using functional neuroimaging as part of biofeedback techniques that enable people to control their pain by training themselves to exercise their inherent analgesic capacity. Researchers have already demonstrated people’s ability to modify their electroencephalographic (EEG) signals with biofeedback training. From there, it is a short conceptual leap—although a larger technological one—to consider more specific control of pain through similar approaches.

The Final Frontier

We may well discover the final frontier in pain management where the nervous system is closest to consciousness, rather than to our injury or illness. Functional neuroimaging is a window on both abnormalities in the brain associated with pathological pain and individual differences in our neural responses to pain, and so helps us better define targets for therapeutic intervention. In time, neuroimaging might become an objective measure of the effectiveness of both traditional and novel treatments.

Many processes involved in chronic pain—peripheral sensitization, central sensitization, and descending modulation—are the normal ones by which pain helps to protect us. We guard the injured site, heal the injury, and recuperate. As the experiences of Arthur and Margaret illustrate, though, pain may exist in the absence of any need for protection and healing. This is the pathology that confronts us with increasing frequency. The tools are at hand to meet the challenge of chronic pain by employing new strategies that take advantage of our knowledge of pain mechanisms at the levels of cells, neural networks, and integrative systems. We can discern already the dawn of a day when “learn to live with the pain” has ceased to be an acceptable recommendation to any patient.

*The Decade of Pain Control and Research began in January 2001 and was authorized by the 106th Congress (HR 3244) and signed into law by President Clinton in November 2000 as Title VI, Section 1603.