Stahl's Essential Psychopharmacology: Neuroscientific Basis and Practical Applications, 4th Ed.

Chapter 10. Chronic pain and its treatment

   What is pain?

    “Normal” pain and the activation of nociceptive nerve fibers

    Nociceptive pathway to the spinal cord

    Nociceptive pathway from the spinal cord to the brain

   Neuropathic pain

    Peripheral mechanisms in neuropathic pain

    Central mechanisms in neuropathic pain

    The spectrum of mood and anxiety disorders with pain disorders


    Decreased gray matter in chronic pain syndromes?

   Descending spinal synapses in the dorsal horn and the treatment of chronic pain

   Targeting sensitized circuits in chronic pain conditions

   Targeting ancillary symptoms in fibromyalgia


This chapter will provide a brief overview of chronic pain conditions associated with different psychiatric disorders and treated with psychotropic drugs. Included here are discussions of the symptomatic and pathophysiologic overlap between disorders with pain and many other disorders treated in psychopharmacology, especially depression and anxiety. Clinical descriptions and formal criteria for how to diagnose painful conditions are only mentioned here in passing. The reader should consult standard reference sources for this material. The discussion here will emphasize how discoveries about the functioning of various brain circuits and neurotransmitters – especially those acting upon the central processing of pain – have impacted our understanding of the pathophysiology and treatment of many painful conditions that may occur with or without various psychiatric disorders. The goal of this chapter is to acquaint the reader with ideas about the clinical and biological aspects of the symptom of pain, how it can be hypothetically caused by alterations of pain processing within the central nervous system (CNS), how it can be associated with many of the symptoms of depression and anxiety, and finally how it can be treated with several of the same agents that can treat depression and anxiety. The discussion in this chapter is at the conceptual level, not at the pragmatic level. The reader should consult standard drug handbooks (such as Stahl's Essential Psychopharmacology: the Prescriber's Guide) for details of doses, side effects, drug interactions, and other issues relevant to the prescribing of these drugs in clinical practice.

What is pain?

No experience rivals pain for its ability to capture our attention, focus our actions, and cause suffering (see Table 10-1 for some useful definitions regarding pain). The powerful experience of pain, especially acute pain, can serve a vital function – to make us aware of damage to our bodies, and to rest the injured part until it has healed. When acute pain is peripheral in origin (i.e., originating outside of the CNS) but continues as chronic pain, it can cause changes in CNS pain mechanisms that enhance or perpetuate the original peripheral pain. For example, osteoarthritis, low back pain, and diabetic peripheral neuropathic pain begin as peripheral pain, but over time these conditions can trigger central pain mechanisms that amplify peripheral pain and generate additional pain centrally. This may explain why research has recently shown that chronic pain conditions of peripheral origin can be successfully targeted for relief by psychotropic drugs that work on central pain mechanisms.

Table 10-1 Pain: some useful definitions


An unpleasant sensory and emotional experience associated with actual or potential tissue damage, or described in terms of such damage

Acute pain

Pain that is of short duration and resolves; usually directly related to the resolution or healing of tissue damage

Chronic pain

Pain that persists for longer than would be expected; an artificial threshold for chronicity (e.g., 1 month) is not appropriate

Neuropathic pain

Pain that arises from damage to, or dysfunction of, any part of the peripheral or central nervous system


The process by which noxious stimuli produce activity in the sensory pathways that convey “painful” information


Pain caused by a stimulus that does not normally provoke pain


An increased response to a stimulus that is normally painful


Any process that reduces the sensation of pain, while not affecting normal touch

Local anesthesia

Blockade of all sensation (innocuous and painful) from a local area

Noxious stimulus

Stimulus that inflicts damage, or would potentially inflict damage, on tissues of the body

Primary afferent neuron (PAN)

The first neuron in the somatosensory pathway; detects mechanical, thermal, or chemical stimuli at its peripheral terminals and transmits action potentials to its central terminals in the spinal cord; all PANs have a cell body in the dorsal root ganglion


A primary afferent (sensory) neuron that is only activated by a noxious stimulus


The process by which a nociceptor detects a noxious stimulus and generates a signal (action potentials) that is propagated towards higher centers in the nociceptive pathway

Dorsal root ganglion (DRG)

Contains the cell bodies of primary afferent neurons; proteins, including transmitters, receptors, and structural proteins, are synthesized here and transported to peripheral and central terminals


Neuron with its cell body, axon and dendrites within the spinal cord; can be excitatory (e.g., containing glutamate) or inhibitory (e.g., containing GABA)

Projection neurons

Neuron in the dorsal horn that receives input from PANs and/or interneurons, and projects up the spinal cord to higher processing centers

Spinothalamic tract

Tract of neurons that project from the spinal cord to the thalamus

Spinobulbar tracts

Several different tracts of neurons that project from the spinal cord to brainstem nuclei

Somatosensory cortex

Region of the cerebral cortex that receives input mainly from cutaneous sensory nerves; the cortex is topographically arranged, with adjacent areas receiving input from adjacent body areas; stimulation of the somatosensory cortex creates sensations from the body part that projects to it

Many other chronic pain conditions may start centrally and never have a peripheral causation to the pain, especially conditions associated with multiple unexplained painful physical symptoms such as depression, anxiety, and fibromyalgia. Because these centrally mediated pain conditions are associated with emotional symptoms, this type of pain has until recently often been considered not to be “real” but rather a nonspecific outcome of unresolved psychological conflicts that would improve when the associated psychiatric condition improved; therefore, there was not a perceived need to target this type of pain. Today, however, many painful conditions without identifiable peripheral lesions and that were once linked only to psychiatric disorders are now hypothesized to be forms of chronic neuropathic pain syndromes that can be successfully treated with the same agents that treat neuropathic pain syndromes not associated with psychiatric disorders. These treatments include the SNRIs (serotonin–norepinephrine reuptake inhibitors: discussed in Chapter 7 on antidepressants) and the α2δ ligands (anticonvulsants that block voltage-gated calcium channels or VSCCs: discussed in Chapter 8 on mood stabilizers and in Chapter 9 on anxiety disorders). Additional psychotropic agents acting centrally at various other sites are also used to treat a variety of chronic pain conditions and will be mentioned below. Many additional drugs are being tested as potential novel pain treatments as well.

Since pain is clearly associated with some psychiatric disorders, and psychotropic drugs that treat various psychiatric conditions are also effective for a wide variety of pain conditions, the detection, quantification, and treatment of pain are rapidly becoming standardized parts of a psychiatric evaluation. Modern psychopharmacologists increasingly consider pain to be a psychiatric “vital sign,” thus requiring routine evaluation and symptomatic treatment. In fact, elimination of pain is increasingly recognized as necessary in order to have full symptomatic remission not only of chronic pain conditions, but also of many psychiatric disorders.

“Normal” pain and the activation of nociceptive nerve fibers

The nociceptive pain pathway is the series of neurons that begins with detection of a noxious stimulus and ends with the subjective perception of pain. This so-called nociceptive pathway starts from the periphery, enters the spinal cord, and projects to the brain (Figure 10-1). It is important to understand the processes by which incoming information can be modulated to increase or decrease the perception of pain associated with a given stimulus, because these processes can explain not only why maladaptive pain states arise but also why drugs that work in psychiatric conditions such as depression and anxiety can also be effective in reducing pain.

Figure 10-1. Activation of nociceptive nerve fibers. Detection of a noxious stimulus occurs at the peripheral terminals of primary afferent neurons and leads to generation of action potentials that propagate along the axon to the central terminals. Aβ fibers respond only to non-noxious stimuli, Aδ fibers respond to noxious mechanical stimuli and subnoxious thermal stimuli, and C fibers respond only to noxious mechanical, heat, and chemical stimuli. Primary afferent neurons have their cell bodies in the dorsal root ganglion and send terminals into that spinal cord segment as well as sending less dense collaterals up the spinal cord for a short distance. Primary afferent neurons synapse onto several different classes of dorsal horn projection neurons (PN), which project via different tracts to higher centers.

Nociceptive pathway to the spinal cord

Primary afferent neurons detect sensory inputs including pain (Figure 10-1). They have their cell bodies in the dorsal root ganglia located along the spinal column outside of the CNS and thus are considered peripheral and not central neurons (Figure 10-1). Nociception begins with transduction – the process by which specialized membrane proteins located on the peripheral projections of these neurons detect a stimulus and generate a voltage change at their peripheral neuronal membranes. A sufficiently strong stimulus will lower the voltage at the membrane (i.e., depolarize the membrane) enough to activate voltage-sensitive sodium channels (VSSCs) and trigger an action potential that will be propagated along the length of the axon to the central terminals of the neuron in the spinal cord (Figure 10-1). VSSCs are introduced in Chapter 3 and illustrated in Figures 3-19 and 3-20. Nociceptive impulse flow from primary afferent neurons into the CNS can be reduced or stopped when VSSCs are blocked by peripherally administered local anesthetics such as lidocaine.

The specific response characteristics of primary afferent neurons are determined by the specific receptors and channels expressed by that neuron in the periphery (Figure 10-1). For example, primary afferent neurons that express a stretch-activated ion channel are mechanosensitive; those that express the vanillinoid receptor 1 (VR1) ion channel are activated by capsaicin, the pungent ingredient in chili peppers, and also by noxious heat, leading to the burning sensation both these stimuli evoke. These functional response properties are used to classify primary afferent neurons into three types: Aβ, Aδ, and C-fiber neurons (Figure 10-1). Aβ fibers detect small movements, light touch, hair movement, and vibrations; C-fiber peripheral terminals are bare nerve endings that are only activated by noxious mechanical, thermal, or chemical stimuli; Aδ fibers fall somewhere in between, sensing noxious mechanical stimuli and subnoxious thermal stimuli (Figure 10-1). Nociceptive input and pain can thus be caused by activating primary afferent neurons peripherally, such as from a sprained ankle or a tooth extraction. NSAIDs (nonsteroidal anti-inflammatory drugs) can reduce painful input from these primary afferent neurons, presumably via their peripheral actions. Opioids can also reduce such pain, but from central actions, as explained below.

Nociceptive pathway from the spinal cord to the brain

The central terminals of peripheral nociceptive neurons synapse in the dorsal horn of the spinal cord onto the next cells in the pathway – dorsal horn neurons, which receive input from many primary afferent neurons and then project to higher centers (Figure 10-3). For this reason, they are sometimes also called dorsal horn projection neurons (PN in Figures 10-110-2, and 10-3). Dorsal horn neurons are thus the first neurons of the nociceptive pathway that are located entirely within the CNS, and are therefore a key site for modulation of nociceptive neuronal activity as it comes into the CNS. A vast number of neurotransmitters have been identified in the dorsal horn, some of which are shown in Figure 10-2.

Figure 10-2. Multiple neurotransmitters modulate pain processing in the spinal cord. There are many neurotransmitters and their corresponding receptors in the dorsal horn. Neurotransmitters in the dorsal horn may be released by primary afferent neurons, by descending regulatory neurons, by dorsal horn projection neurons (PN) and by interneurons. Neurotransmitters present in the dorsal horn that have been best studied in terms of pain transmission include substance P (NK1, 2, and 3 receptors), endorphins (µ-opioid receptors), norepinephrine (α2-adrenergic receptors), and serotonin (5HT1B/D and 5HT3receptors). Several other neurotransmitters are also represented, including VIP (vasopressin inhibitory protein and its receptor VIPR); somatostatin and its receptor SR; calcitonin-gene-related peptide (CGRP and its receptor CGRP-R); GABA and its receptors GABAA and GABAB; glutamate and its receptors AMPA-R (α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor) and NMDA-R (N-methyl-D-aspartate receptor); nitric oxide (NO); cholecystokinin (CCK and its receptors CCK-A and CCK-B); and glycine and its receptor NMDA-R.

Neurotransmitters in the dorsal horn are synthesized not only by primary afferent neurons, but by the other neurons in the dorsal horn as well, including descending neurons and various interneurons (Figure 10-2). Some neurotransmitter systems in the dorsal horn are successfully targeted by known pain-relieving drugs, especially opioids, serotonin- and norepinephrine-boosting SNRIs (serotonin–norepinephrine reuptake inhibitors), and α2δ ligands acting at voltage-sensitive calcium channels (VSCCs). All of the neurotransmitter systems acting in the dorsal horn are potential targets for novel pain-relieving drugs (Figure 10-2), and a plethora of such novel agents is currently in clinical and preclinical development.

There are several classes of dorsal horn neurons: some receive input directly from primary sensory neurons, some are interneurons, and some project up the spinal cord to higher centers (Figure 10-3). There are several different tracts in which these projection neurons can ascend, which can be crudely divided into two functions: the sensory/discriminatory pathway and the emotional/motivational pathway (Figure 10-3).

Figure 10-3. From nociception to pain. Dorsal horn neurons in the spinothalamic tract project to the thalamus and then to the primary somatosensory cortex. This pathway carries information about the intensity and location of the painful stimuli and is termed the discriminatory pathway. Neurons ascending in the spinobulbar tract project to brainstem nuclei and then to both the thalamus and limbic structures. These pathways convey the emotional and motivational aspects of the pain experience. Only when information from the discriminatory (thalamocortical) and emotional/motivational (limbic) pathways combine is the human subjective experience of pain formed (“ouch”).

In the sensory/discriminatory pathway, dorsal horn neurons ascend in the spinothalamic tract; then, thalamic neurons project to the primary somatosensory cortex (Figure 10-3). This particular pain pathway is thought to convey the precise location of the nociceptive stimulus and its intensity. In the emotional/motivational pathway, other dorsal horn neurons project to brainstem nuclei, and from there to limbic regions (Figure 10-3). This second pain pathway is thought to convey the affective component that nociceptive stimuli evoke. Only when these two aspects of sensory discrimination and emotions come together and the final, subjective perception of pain is created can we use the word pain to describe the modality (“ouch” in Figure 10-3). Before this point, we are simply discussing activity in neural pathways, which should be described as noxious-evoked or nociceptive neuronal activity but not necessarily as pain.

Neuropathic pain

The term neuropathic pain describes pain that arises from damage to, or dysfunction of, any part of the peripheral or central nervous system, whereas “normal” pain (so-called nociceptive pain, just discussed in the section above) is caused by activation of nociceptive nerve fibers.

Peripheral mechanisms in neuropathic pain

Normal transduction and conduction in peripheral afferent neurons can be hijacked in certain neuropathic pain states to maintain nociceptive signaling in the absence of a relevant noxious stimulus. Neuronal damage by disease or trauma can alter electrical activity of neurons, allow cross-talk between neurons, and initiate inflammatory processes to cause peripheral sensitization. In this chapter, we will not emphasize peripheral sensitization disorders and mechanisms, but rather central sensitization disorders and mechanisms.

Central mechanisms in neuropathic pain

At each major relay point in the pain pathway (Figure 10-3), the nociceptive pain signal is susceptible to modulation by endogenous processes to either dampen down the signal or amplify it. This happens not only peripherally at primary afferent neurons, as has just been discussed, but also at central neurons in the dorsal horn of the spinal cord as well as in numerous brain regions. The events in the dorsal horn of the spinal cord are better understood than those in brain regions of nociceptive pathways, but pain processing in the brain may be the key to understanding the generation and amplification of central pain in disorders of chronic peripheral pain such as osteoarthritis, low back pain, and diabetic peripheral neuropathic pain, as well as painful physical symptoms in affective and anxiety disorders and in fibromyalgia.

“Segmental” central sensitization is a process thought to be caused when plastic changes occur in the dorsal horn, classically in conditions such as phantom pain after limb amputation. Specifically, this type of neuronal plasticity in the dorsal horn is called activity-dependent or use-dependent, because it requires constant firing of the pain pathway in the dorsal horn. The consequence of this constant input of pain is eventually to cause exaggerated (hyperalgesic) or prolonged responses to any noxious input – a phenomenon sometimes called “wind-up” – as well as painful responses to normally innocuous inputs (called allodynia). Phosphorylation of key membrane receptors and channels in the dorsal horn appears to increase synaptic efficiency and thus to trip a master switch, opening the gate to the pain pathway and turning on central sensitization that acts to amplify or create the perception of pain even if there is no pain input coming from the periphery. The gate can also close, as conceptualized in the classic “gate theory” of pain, in order to explain how innocuous stimulation (e.g., acupuncture, vibration, rubbing) away from the site of an injury can close the pain gate and reduce the perception of the injury pain.

In segmental central sensitization, a definite peripheral injury (Figure 10-4A) is combined with central sensitization at the spinal cord segment receiving nociceptive input from the damaged area of the body (Figure 10-4B). Segmental central sensitization syndromes are thus “mixed” states where the insult of central segmental changes (Figure 10-4B) is added to peripheral injuries such as low back pain, diabetic peripheral neuropathic pain, and painful cutaneous eruptions of herpes zoster (shingles) (Figure 10-4A).

Figure 10-4. Acute pain and development of segmental central sensitization. (A) When peripheral injury occurs, nociceptive impulse flow from primary afferent neurons is transmitted via dorsal horn neurons to higher brain centers, where it can ultimately be interpreted as pain (represented by the “ouch”). (B) In some cases, injury or disease directly affecting the nervous system may result in plastic changes that lead to sensitization within the central nervous system, such that the experience of pain continues even after tissue damage is resolved. Impulses may be generated at abnormal locations either spontaneously or via mechanical forces. At the level of the spinal cord, this process is termed segmental central sensitization. This mechanism underlies conditions such as diabetic peripheral neuropathic pain and shingles.

“Suprasegmental” central sensitization is hypothesized to be linked to plastic changes that occur in brain sites within the nociceptive pathway, especially the thalamus and cortex, in the presence of known peripheral causes (Figure 10-5A) or even in the absence of identifiable triggering events (Figure 10-5B). In the case of peripherally activated suprasegmental central sensitization, it is as though the brain “learns” from its experience of pain, and decides not only to keep the process going, but also to enhance it and make it permanent. In the case of pain that originates centrally without peripheral input, it is as though the brain has figured out how to spontaneously activate its pain pathways. Interrupting this process of sensitized brain pathways for pain and getting the CNS to “forget” its molecular memories may be one of the greatest therapeutic opportunities in psychopharmacology today, not only because this may be a therapeutic strategy for various chronic neuropathic pain conditions, as discussed here, but also because it may be a viable approach to treating the hypothesized molecular changes that may underlie disease progression in a wide variety of disorders, from schizophrenia to stress-induced anxiety and affective disorders, to addictive disorders. Conditions hypothesized to be caused by suprasegmental central sensitization syndromes of pain originating in the brain without peripheral pain input include fibromyalgia, the syndrome of chronic widespread pain, and painful physical symptoms of depression and anxiety disorders, especially PTSD (Figure 10-5B).

Figure 10-5. Suprasegmental central sensitization. Plastic changes in brain sites within the nociceptive pathway can cause sensitization, for instance at the level of the thalamus or the sensory cortex. This process within the brain is termed suprasegmental central sensitization. This can occur following peripheral injury (A) or even in the absence of identifiable triggering events (B). This mechanism is believed to underlie conditions such as fibromyalgia, chronic widespread pain, and painful symptoms in depression and anxiety disorders.

The spectrum of mood and anxiety disorders with pain disorders

A large group of overlapping disorders can have emotional symptoms, painful physical symptoms, or both (Figure 10-6). Although pain in the absence of emotional symptoms has long been seen as a neurological disorder, and pain in the presence of emotional symptoms as a psychiatric disorder, it is now clear that pain is a symptom that can be mapped onto inefficient information processing within the pain circuit, and is largely considered the same symptom with the same treatments whether occurring by itself or as part of any number of syndromes (Figure 10-6). Thus, pain (Figure 10-6, on the right) can occur not only by itself, but also concomitantly with the emotional symptoms of depressed mood and anxiety (Figure 10-6, on the left), and with the physical symptoms of fatigue, insomnia, and problems concentrating (Figure 10-6, in the middle). No matter whether pain occurs by itself or with additional concomitant emotional or physical symptoms, or in the presence of full syndromal psychiatric disorders such as major depressive disorder, generalized anxiety disorder, or PTSD (Figure 10-6, on the left), it must be treated – and the treatments are the same across the spectrum (Figure 10-6), namely SNRIs and α2δ ligands, as will be explained below.

Figure 10-6. The spectrum from mood and anxiety disorders to painful functional somatic syndromes. Affective spectrum disorders include mood and anxiety disorders, while “functional somatic syndrome” is a term used to describe disorders such as fibromyalgia and chronic widespread pain. Pain, though not a formal diagnostic feature of depression or anxiety disorders, is nonetheless frequently present in patients with these disorders. Similarly, depressed mood, anxiety, and other symptoms identified as part of depression and anxiety disorders are now recognized as being common in functional somatic syndromes. Thus, rather than being discrete groups of illnesses, affective spectrum disorders and functional somatic syndromes may instead exist along the same spectrum.


Fibromyalgia has emerged as a diagnosable (Table 10-2) and treatable pain syndrome with tenderness but no structural pathology in muscles, ligaments, or joints. Fibromyalgia is recognized as a chronic, widespread pain syndrome associated with fatigue, nonrestorative sleep and tenderness at 11 or more of 18 designated “trigger points” where ligaments, tendons, and muscle attach to bone (Figure 10-7). It is the second most common diagnosis in rheumatology clinics, and may affect 2–4% of the population. Although symptoms of fibromyalgia are chronic and debilitating, they are not necessarily progressive. There is no known cause and there is no known pathology identifiable in the muscles or joints. This syndrome can be deconstructed into its component symptoms (Figure 10-8), and then matched with hypothetically malfunctioning brain circuits (Figure 10-9). Some studies suggest that 75–90% of identified patients are women, especially Caucasian women. A related syndrome, called chronic widespread pain, is essentially pain without the tenderness, sometimes called “male fibromyalgia” because many men often do not experience (or at least do not report) tenderness on examination of areas of pain.

Table 10-2 American College of Rheumatology (ACR) 1990 criteria for fibromyalgia (Wolfe F, et alArthritis Rheum 1990; 33: 160–72).

History of widespread pain

Considered widespread when present in all of the following:

Left side of the body

Right side of the body

Above the waist

Below the waist

Axial skeleton (cervical spine, anterior chest, thoracic spine, or low back)

Must be present for at least 3 months

Pain in 11 of 18 tender point sites on digital palpation

Digital palpation should be performed with an approximate force of 4 kg

For a tender point to be considered positive, the subject must state that the palpation was painful

Figure 10-7. Tender points for the diagnosis of fibromyalgia. Fibromyalgia is a chronic widespread pain syndrome formally diagnosed based on tenderness in at least 11 of 18 designated “trigger points” where ligaments, tendons, and muscle attach to bone. Other diagnostic features include fatigue and nonrestorative sleep.

Figure 10-8. Symptoms of fibromyalgia. In addition to pain as a central feature of fibromyalgia, many patients experience fatigue, anxiety, depression, disturbed sleep, and problems concentrating.

Figure 10-9. Symptom-based algorithm for fibromyalgia. A symptom-based approach to treatment selection for fibromyalgia follows the theory that each of a patient's symptoms can be matched with malfunctioning brain circuits and neurotransmitters that hypothetically mediate those symptoms; this information is then used to select a corresponding pharmacological mechanism for treatment. Pain is linked to transmission of information via the thalamus (T), while physical fatigue is linked to the striatum (S) and spinal cord (SC). Problems concentrating and lack of interest (termed “fibro-fog”) as well as mental fatigue are linked to the prefrontal cortex (PFC), specifically the dorsolateral PFC. Fatigue, low energy, and lack of interest may all also be related to the nucleus accumbens (NA). Disturbances in sleep and appetite are associated with the hypothalamus (Hy), depressed mood with the amygdala (A) and orbitofrontal cortex, and anxiety with the amygdala.

Decreased gray matter in chronic pain syndromes?

Some very troubling preliminary reports suggest that chronic pain may even “shrink the brain” in the DLPFC (dorsolateral prefrontal cortex) (Figure 10-10) and thereby contribute to cognitive dysfunction in certain pain states such as fibromyalgia (Figure 10-8) and low back pain. Brain atrophy is discussed in relation to stress and anxiety disorders in Chapter 6 and illustrated in Figure 6-39. It would not be surprising if stressful conditions that cause pain, as well as pain that causes distress, are all involved in causing brain atrophy and/or cognitive dysfunction in fibromyalgia and other chronic pain states. Chronic back pain, for example, has also been reported to be associated with decreased prefrontal and thalamic gray-matter density (Figure 10-10). Some experts have hypothesized that in fibromyalgia and other chronic neuropathic pain syndromes, the persistent perception of pain could lead to overuse of DLPFC neurons, excitotoxic cell death in this brain region, and reduction of the corticothalamic “brake” on nociceptive pathways. Such an outcome could cause not only increased pain perception, but diminished executive functioning, sometimes called “fibro-fog” in fibromyalgia. In Chapter 6 we discussed how stress-related HPA (hypothalamic–pituitary–adrenal) axis abnormalities in CRH-ACTH-cortisol regulation may be linked to hippocampal atrophy (see Figure 6-39), possibly linked to reduced availability of growth factors (Figures 6-37 and 6-38). Alterations in growth factors may be linked to the reports of reduction in gray-matter volume in chronic pain syndromes (fibromyalgia and low back pain), but in different brain regions (DLPFC, temporal cortex, and thalamus: Figure 10-10) than reported for depression (Figure 6-39B). Gray matter may actually be increased in other brain regions in chronic pain.

Figure 10-10. Gray-matter loss in chronic pain. Research suggests that chronic pain, like anxiety and stress-related disorders, may lead to brain atrophy. Specifically, there are data showing gray-matter loss in the dorsolateral prefrontal cortex (DLPFC), the thalamus, and the temporal cortex in patients with chronic pain conditions.

Although still preliminary, these findings suggest a possibly structural consequence to suprasegmental central sensitization (Figure 10-10) not unlike that suspected for depression and stress (Figure 6-39). Abnormal pain processing, exaggerated pain responses, and perpetual pain could hypothetically be linked to deficiencies in the DLPFC circuit and its regulation by dopamine, and provide a potential explanation for the cognitive difficulties associated with chronic pain, especially so-called “fibro-fog” in fibromyalgia (Figure 10-8). Thalamic abnormalities could hypothetically be linked to problems sleeping and nonrestorative sleep seen in chronic pain syndromes as well (Figure 10-8). Thus, chronic pain syndromes not only cause pain, but also problems with fatigue, mental concentration, and sleep as well as depression and anxiety (Figure 10-8). Structural brain abnormalities associated with inefficient information processing in brain areas that mediate these symptoms (Figure 10-9) may explain why these various symptoms (Figure 10-8) are frequently associated with chronic pain syndromes.

Descending spinal synapses in the dorsal horn and the treatment of chronic pain

The periaqueductal gray (PAG) is the site of origin and regulation of much of the descending inhibition that projects down the spinal cord to the dorsal horn (Figure 10-2). The periaqueductal gray is discussed in relation to its connections with the amygdala and the motor component of the fear response in Chapter 9 and illustrated in Figure 9-9. The periaqueductal gray also integrates inputs from nociceptive pathways and limbic structures such as the amygdala and limbic cortex, and sends outputs to brainstem nuclei and the rostroventromedial medulla to drive descending inhibitory pathways. Some of these descending pathways release endorphins, which act via mostly presynaptic µ-opioid receptors to inhibit neurotransmission from nociceptive primary afferent neurons (Figure 10-2). Spinal µ-opioid receptors are one target of opioid analgesics; so are µ-opioid receptors in the periaqueductal gray itself (Figure 10-11). Interestingly, since Aβ fibers (Figure 10-1) do not express µ-opioid receptors, this may explain why opioid analgesics spare normal sensory input. Enkephalins, which also act via δ-opioid receptors, are also antinociceptive, whereas dynorphins, acting at κ-opioid receptors, can be either anti- or pronociceptive. Interesting also is that opioids in general are not only no more effective for chronic neuropathic pain states than SNRIs or α2δ ligands, but in many cases, such as in fibromyalgia, are not proven to be effective at all.

Figure 10-11. Acute nociceptive pain and opioids. Descending opioid projections are activated by severe injury or “danger” to inhibit nociceptive neurotransmission in the dorsal horn, which allows the individual to escape any immediate danger without being compromised. (A) Shown here is nociceptive input from a peripheral injury being transmitted to the brain and interpreted as pain. The descending opioid projection is not activated and thus is not inhibiting the nociceptive input. (B) Endogenous opioid release in the descending opioid projection, or exogenous administration of an opioid, can cause inhibition of nociceptive neurotransmission in the dorsal horn or in the periaqueductal gray and thus prevent or reduce the experience of pain.

Two other important descending inhibitory pathways are also shown in Figure 10-2. One is the descending spinal norepinephrine (NE) pathway (Figure 10-12), which originates in the locus coeruleus (LC), and especially from noradrenergic cell bodies in the lower (caudal) parts of the brainstem neurotransmitter center (lateral tegmental norepinephrine cell system) (Figure 6-32). The other important descending pathway is the descending spinal serotonergic (5HT) pathway (Figure 10-13), which originates in the nucleus raphe magnus of the rostroventromedial medulla and especially the lower (caudal) serotonin nuclei (raphe magnus, raphe pallidus, and raphe obscuris) (Figure 6-33). Descending noradrenergic neurons inhibit neurotransmitter release from primary afferents directly via inhibitory α2-adrenergic receptors (Figure 10-2), explaining why direct-acting α2 agonists such as clonidine can be useful in relieving pain in some patients. Serotonin inhibits primary afferent terminals via postsynaptic 5HT1B/Dreceptors (Figure 10-2). These inhibitory receptors are G-protein-coupled, and indirectly influence ion channels to hyperpolarize the nerve terminal and inhibit nociceptive neurotransmitter release. However, serotonin is also a major transmitter in descending facilitation pathways to the spinal cord. Serotonin released onto some primary afferent neuron terminals in certain areas of the dorsal horn acts predominantly via excitatory 5HT3 receptors to enhance neurotransmitter release from these primary afferent neurons (Figure 10-2). The combination of both inhibitory and facilitatory actions of serotonin may explain why SSRIs, with actions that increase only serotonin levels, are not consistently useful in the treatment of pain, whereas SNRIs, with actions on both serotonin and norepinephrine, are now proven to be effective in various neuropathic pain states, including diabetic peripheral neuropathic pain and fibromyalgia.

Figure 10-12. Descending noradrenergic neurons and pain. (A) Descending spinal noradrenergic (NE) neurons inhibit neurotransmitter release from primary afferent neurons via presynaptic α2-adrenergic receptors, and inhibit activity of dorsal horn neurons via postsynaptic α2-adrenergic receptors. This suppresses bodily input (e.g., regarding muscles/joints or digestion) from reaching the brain and thus prevents it from being interpreted as painful. (B) If descending NE inhibition is deficient, then it may not be sufficient to mask irrelevant nociceptive input, potentially leading to perception of pain from input that is normally ignored. This may be a contributing factor for painful somatic symptoms in fibromyalgia, depression, irritable bowel syndrome, and anxiety disorders. (C) A serotonin–norepinephrine reuptake inhibitor (SNRI) can increase noradrenergic neurotransmission in the descending spinal pathway to the dorsal horn, and thus may enhance inhibition of bodily input so that it does not reach the brain and get interpreted as pain.

Figure 10-13. Descending serotonergic neurons and pain. (A) Descending serotonergic (5HT) neurons directly inhibit activity of dorsal horn neurons, predominantly via 5HT1B/D receptors. This suppresses bodily input (e.g., regarding muscles/joints or digestion) from reaching the brain and thus prevents it from being interpreted as painful. (B) If descending 5HT inhibition is deficient, it may not be sufficient to mask irrelevant nociceptive input, potentially leading to perception of pain from input that is normally ignored. This may be a contributing factor for painful somatic symptoms in fibromyalgia, depression, irritable bowel syndrome, and anxiety disorders. (C) A serotonin–norepinephrine reuptake inhibitor (SNRI) can increase serotonergic neurotransmission in the descending spinal pathway to the dorsal horn, and thus may enhance inhibition of bodily input so that it does not reach the brain and get interpreted as pain. However, the noradrenergic effects of SNRIs may be more relevant to suppression of nociceptive input.

Descending inhibition, mostly via serotonergic and noradrenergic pathways, is normally active at rest and is thought to act physiologically to mask perception of irrelevant nociceptive input (e.g., from digestion, joint movement, etc.: Figures 10-12A and 10-13A). One hypothesis for why patients with depression or fibromyalgia or related chronic pain disorders perceive pain when there is no obvious sign of peripheral trauma is that descending inhibition may not be acting adequately to mask irrelevant nociceptive input. This leads to the perception of pain from what is actually normal input that is ordinarily ignored (Figures 10-12B and 10-13B). If this descending monoaminergic inhibition is enhanced with an SNRI, irrelevant nociceptive inputs from joints, muscles, and the back in fibromyalgia and depression, and from digestion and the gastrointestinal tract in irritable bowel syndrome, are hypothetically once again ignored and thus are no longer perceived as painful (Figures 10-12C and 10-13C). SNRIs include duloxetine, milnacipran, venlafaxine, desvenlafaxine, and some tricyclic antidepressants (TCAs). SNRIs and TCAs are discussed extensively in Chapter 7.

Descending inhibition is also activated during severe injury by incoming nociceptive input, and in dangerous “conflict” situations via limbic structures, causing the release of endogenous opioid peptides (Figure 10-11B), serotonin (Figure 10-13A), and norepinephrine (Figure 10-12A). When this happens, this reduces not only the release of nociceptive neurotransmitters in the dorsal horn (Figure 10-2) but also the transmission of nociceptive impulses up the spinal cord into the brain (Figure 10-3), thereby reducing the perception of pain, dulling it to allow escape from the situation without the injury compromising physical performance in the short run (reduction of “ouch” in Figure 10-3). On return to safety, descending facilitation replaces the inhibition to redress the balance, increases awareness of the injury, and forces rest of the injured part (lots of “ouch” in Figure 10-3).

The power of this system can be seen in humans persevering through severe injury on the sports field and on the battlefield. The placebo effect may also involve endogenous opioid release from these descending inhibitory neurons (Figure 10-11B), since activation of a placebo response to pain is reversible by the µ-opioid antagonist naloxone. These are adaptive changes within the pain pathways that facilitate survival and enhance function for the individual. However, maladaptive changes can also hijack these same mechanisms to inappropriately maintain pain without relevant tissue injury, as may occur in various forms of neuropathic pain ranging from diabetes to fibromyalgia and beyond.

Targeting sensitized circuits in chronic pain conditions

Chronic pain perpetuated as a marker of an irreversible sensitization process within the central nervous system has already been discussed as a disorder triggered by progressive molecular changes due to abnormal neuronal activity within the pain pathway, sometimes called central sensitization. When this occurs at the spinal or segmental level, it is likely linked to the multiple different neurotransmitters released there, with each neurotransmitter's release mechanism requiring presynaptic depolarization and activation of N-type and P/Q-type voltage-sensitive calcium channels (VSCCs: Figures 10-14 and 10-15), which is often coupled to the release of glutamate but also to aspartate, substance P (SP), calcitonin-gene-related peptide (CGRP), and other neurotransmitters (Figure 10-2). When this occurs at suprasegmental levels in the thalamus and cortex, it is likely linked to release mostly of glutamate via the same N-type and P/Q-type VSCCs (Figures 10-14 and 10-15). The idea is that low release of neurotransmitter creates no pain response because there is insufficient neurotransmitter release to stimulate the postsynaptic receptors (Figure 10-14A). However, normal amounts of neurotransmitter release cause a full nociceptive pain response and acute pain (Figure 10-14B). Hypothetically, in states of central sensitization there is excessive and unnecessary ongoing nociceptive activity causing neuropathic pain (Figure 10-15A). Blocking VSCCs with the α2δ ligands gabapentin or pregabalin (Figure 10-16) inhibits release of various neurotransmitters in the dorsal horn (Figures 10-210-15B10-17A) or in the thalamus and cortex (Figures 10-15B and 10-17B) and has indeed proven to be an effective treatment for various disorders causing neuropathic pain (Figure 10-15B). Gabapentin and pregabalin (Figure 10-16), may more selectively bind the “open channel” conformation of VSCCs (Figure 10-18), and thus be particularly effective in blocking those channels that are the most active, with a “use-dependent” form of inhibition (Figures 10-15B10-1710-18B). This molecular action predicts more affinity for centrally sensitized VSCCs that are actively conducting neuronal impulses within the pain pathway, and thus having a selective action on those VSCCs causing neuropathic pain, ignoring other VSCCs that are not open, and thus not interfering with normal neurotransmission in central neurons uninvolved in mediating the pathological pain state.

Figure 10-14. Activity-dependent nociception in pain pathways, part 1: acute pain. The degree of nociceptive neuronal activity in pain pathways determines whether one experiences acute pain. An action potential on a presynaptic neuron triggers sodium influx, which in turn leads to calcium influx and ultimately release of neurotransmitter. (A) In some cases, the action potential generated at the presynaptic neuron causes minimal neurotransmitter release; thus the postsynaptic neuron is not notably stimulated and the nociceptive input does not reach the brain (in other words, there is no pain). (B) In other cases, a stronger action potential at the presynaptic neuron may cause voltage-sensitive calcium channels to remain open longer, allowing more neurotransmitter release and more stimulation of the postsynaptic neuron. Thus, the nociceptive input is transmitted to the brain and acute pain occurs.

Figure 10-15. Activity-dependent nociception in pain pathways, part 2: neuropathic pain. (A) Strong or repetitive action potentials can cause prolonged opening of calcium channels, which may lead to excessive release of neurotransmitter into the synaptic cleft, and consequently to excessive stimulation of postsynaptic neurons. Ultimately this may induce molecular, synaptic, and structural changes, including sprouting, which are the theoretical substrates for central sensitization syndromes. In other words, this can lead to neuropathic pain. (B) α2δ ligands such as gabapentin or pregabalin bind to the α2δ subunit of voltage-sensitive calcium channels, changing their conformation to reduce calcium influx and therefore reduce excessive stimulation of postsynaptic receptors.

Figure 10-16. Gabapentin and pregabalin. Shown here are icons of the pharmacological actions of gabapentin and pregabalin, two anticonvulsants that also have efficacy in chronic pain. These agents bind to the α2δ subunit of voltage-sensitive calcium channels (VSCCs).

Figure 10-17. Anatomic actions of α2δ ligands. (A) α2δ ligands may alleviate chronic pain associated with sensitization at the level of the dorsal horn. As illustrated here, α2δ ligands may bind to voltage-sensitive calcium channels (VSCCs) in the dorsal horn to reduce excitatory neurotransmission and thus alleviate pain. (B) α2δ ligands may also alleviate chronic pain associated with sensitization at the level of the thalamus or cortex. As illustrated here, α2δ ligands may bind to VSCCs in the thalamus and cortex to reduce excitatory neurotransmission and thus alleviate pain.

Figure 10-18. Binding of α2δ ligands. (A) Calcium influx occurs when voltage-sensitive calcium channels (VSCCs) are in the open-channel conformation. (B) α2δ ligands such as gabapentin and pregabalin have greatest affinity for the open-channel conformation and thus block those channels that are most active. (C) When VSCCs are in the closed conformation α2δ ligands do not bind and thus do not disrupt normal neurotransmission.

Treatment of pain, including neuropathic pain conditions, may be less costly when you “pay” for it in advance, or at least early in the game. The hope is that early treatment of pain could interfere with the development of chronic persistent painful conditions by blocking the ability of painful experiences to imprint themselves upon the central nervous system by not allowing triggering of central sensitization. Thus, the mechanisms whereby symptomatic suffering of chronic neuropathic pain is relieved – such as with SNRIs or α2δ ligands – may also be the same mechanisms that could prevent disease progression to chronic persistent pain states. This notion calls for aggressive treatment of painful symptoms in these conditions that theoretically have their origin within the CNS, thus “intercepting” the central sensitization process before it is durably imprinted into angry circuits. Thus, major depression and anxiety disorders and fibromyalgia can all be treated with SNRIs and/or α2δ ligands to eliminate painful physical symptoms and thereby improve the chances of reaching full symptomatic remission. The opportunity to prevent permanent pain syndromes or progressive worsening of pain is one reason why pain is increasingly being considered a psychiatric “vital sign” that must be assessed routinely in the evaluation and treatment of psychiatric disorders by psychopharmacologists. Future testing of agents capable of reducing pain should be done to determine whether eliminating painful symptoms early in the course of psychiatric and functional somatic illnesses will improve outcomes, including preventing symptomatic relapses, the development of treatment resistance, or even brain atrophy from stress in pain states (Figure 10-10) and hippocampal atrophy from stress in anxiety and affective disorders (Figure 6-39). Pre-emptively treating pain before it occurs, or at least rescuing centrally mediated and sensitizing pain by intercepting such pain before it becomes permanent, may represent some of the most promising therapeutic applications of dual reuptake inhibitors and α2δ ligands, and it deserves careful clinical evaluation.

Targeting ancillary symptoms in fibromyalgia

We have repeatedly mentioned the proven usefulness of the α2δ ligands gabapentin and pregabalin and the SNRIs duloxetine, milnacipran, venlafaxine, and desvenlafaxine for treating the painful symptoms of fibromyalgia, yet these two classes have not been studied extensively in combination. Nevertheless, they are frequently used together in clinical practice on an empiric basic and anecdotally have been shown to give additive improvement in relieving pain. Each class of drug may also help different ancillary symptoms in fibromyalgia, so the combination of α2δ ligands with SNRIs may lead to broader symptom relief than using either alone, although both are effective for pain in fibromyalgia. That is, α2δ ligands may reduce symptoms of anxiety in fibromyalgia (see discussion of α2δ ligands in anxiety in Chapter 9, illustrated in Figures 9-25C and 9-26C) and for improving the slow-wave sleep disorder of fibromyalgia (sleep disorders and their treatment are discussed in further detail in Chapter 11). SNRIs can be useful in reducing symptoms of depression and anxiety in fibromyalgia (see Chapter 7 on antidepressants and Figures 7-30through 7-34; see also Chapter 9 on anxiety and Figures 9-25D and 9-26D) and for treating fatigue as well as the cognitive symptoms associated with fibromyalgia, sometimes also called “fibro-fog.”

Problems with executive functioning in a wide variety of clinical conditions are generally linked to inefficient information processing in the dorsolateral prefrontal cortex (DLPFC), where dopamine neurotransmission is important in regulating brain circuits (see Chapter 4 on schizophrenia and Figure 4-41; see Chapter 6 on depression and Figures 6-456-48, and 6-49; see Chapter 9 on anxiety disorders and Figure 9-17). This concept of dopaminergic regulation of cognition in DLPFC and the role of boosting dopamine neurotransmission to improve executive dysfunction is also discussed in Chapter 12 on attention deficit hyperactivity disorder. Since SNRIs increase dopamine concentrations in DLPFC (see Figure 7-34C), they can also potentially improve symptoms of “fibro-fog” in fibromyalgia patients. This may be particularly so for the SNRI milnacipran, which has potent norepinephrine reuptake binding properties at all clinically effective doses (Figure 7-32), or for higher doses of the SNRIs duloxetine (Figure 7-31), venlafaxine, and desvenlafaxine (Figure 7-30), which have increased norepinephrine reuptake blocking properties and thus act to increase concentrations of dopamine in the DLPFC (Figure 7-34C). Other strategies for improving fibro-fog in fibromyalgia patients include the same ones used to treat cognitive dysfunction in depression, including modafinil, armodafinil, selective norepinephrine reuptake inhibitors (NRIs) such as atomoxetine, norepinephrine–dopamine reuptake inhibitors (NDRIs) such as bupropion, and, with caution, stimulants. SNRIs, sometimes augmented with modafinil, stimulants, or bupropion, can also be useful for symptoms of physical fatigue as well as mental fatigue in fibromyalgia patients.

Second-line treatments for pain in fibromyalgia can include mirtazapine and tricyclic antidepressants, as well as the tricyclic muscle relaxant cyclobenzaprine. Sleep aids such as benzodiazepines, hypnotics, and trazodone can be helpful in relieving sleep disturbance in fibromyalgia. Evidence is also accumulating for the efficacy of γ-hydroxybutyrate (GHB or sodium oxybate) in fibromyalgia (use with extreme caution because of diversion and abuse potential). GHB is approved for narcolepsy, enhances slow-wave sleep, and is discussed in Chapter 11 on sleep. In heroic cases the use of GHB by experts for the treatment of severe and treatment-resistant cases of fibromyalgia may be justified. A number of anticonvulsants other than the α2δ ligands (Figure 10-19) are also used second line for chronic neuropathic pain states, including fibromyalgia. These agents are thought to target voltage-gated sodium channels rather than voltage-gated calcium channels (Figure 10-19) and thus seem to have a different mechanism of action than α2δ ligands and may be effective in patients with inadequate response to α2δ ligands. Other adjunctive or experiemental treatments for various chronic pain syndromes include botulinum toxin injections, cannabinoids, NMDA antagonists, and various new anticonvulsants.

Figure 10-19. Anticonvulsants in chronic pain. A number of anticonvulsants other than the α2δ ligands are also used second line for chronic neuropathic pain states, including fibromyalgia. These agents are thought to target voltage-sensitive sodium channels (VSSCs) rather than voltage-sensitive calcium channels (VSCCs) and thus seem to have a different mechanism of action than α2δ ligands.


This chapter has defined pain, and has explained the processing of nociceptive neuronal activity into the perception of pain by pathways that lead to the spinal cord, and then up the spinal cord to the brain. Neuropathic pain is discussed extensively, including both peripheral and central mechanisms, and the concept of central sensitization. The key role of descending inhibitory pathways that reduce the activity of nociceptive pain neurons with the release of serotonin and norepinephrine is explained, and shown to be the basis for the actions of serotonin–norepinephrine reuptake inhibitors (SNRIs) as agents that reduce the perception of pain in conditions ranging from major depression to fibromyalgia to diabetic peripheral neuropathic pain, low back pain, osteoarthritis, and related conditions. The critical role of voltage-sensitive calcium channels (VSCCs) is also explained, providing the basis for the actions of α2δ ligands as agents that also reduce the perception of pain in diabetic peripheral neuropathic pain, fibromyalgia, painful physical symptoms of depression and anxiety disorders, shingles, and other neuropathic pain conditions. Finally, the spectrum of conditions from affective disorders to chronic neuropathic pain disorders is introduced, with emphasis on the condition of fibromyalgia and its newly evolving psychopharmacological treatments.