There has been growing interest in the West about the application of acupuncture to control pain since President Nixon’s well-publicized trip to China in 1971. The fascination with this ancient medical modality was heightened when a member of the press corps, James Reston, received acupuncture during an appendectomy. The subsequent publication in 1998 of the National Institutes of Health (NIH) consensus statement on the clinical applications of acupuncture based on over 2000 scientific articles brought a degree of optimism that acupuncture would become a mainstay in the war against pain.1
There has been some dampening of the initial enthusiasm, however, with continued skepticism regarding the efficacy of acupuncture. This skepticism arose from the lack of high-quality randomized controlled clinical trials (RCTs); but the results of a number of recent large RCTs using sham acupuncture controls have nevertheless left many believing that acupuncture is just an elaborate placebo ritual.2 As we shall discuss in more detail, the introduction of sham acupuncture has serious flaws but was based on the desire of researchers to filter the clinical practice of acupuncture through the mesh of standard placebo controlled methodology used in pharmacological research. Although clearly not a universal panacea for all pain syndromes, a more careful reading of the literature in fact does support that acupuncture is a cost-effective method for the treatment of pain.
The goal of this chapter is to lay the basic theoretical and physiological groundwork for understanding the clinical applications of acupuncture for pain. Then the current clinical data regarding the efficacy of acupuncture in various pain syndromes will be discussed. Attention will be given to understanding the pitfalls in devising a true placebo control for acupuncture trials, and our discussion of the literature will focus on the effect this has had on the outcomes of a number of large clinical trials for common pain conditions. Finally, a brief representation of some of the different treatment styles for common pain syndromes is outlined, and the chapter concludes by identifying further educational resources in this field.
BRIEF HISTORY OF ACUPUNCTURE
The term acupuncture is from the Greek acus, “needle,” and punctura, “puncture”; it is the English translation of chan in Mandarin and hari in Japanese.
The clinical practice of inserting needles into the body (initially stone or flint needles) occurred in China by the 5th centuries BC and was followed some time later, between the 2nd and 3rd centuries BC, by the first written medical text on Chinese medicine, the Huang Di Nei Jing, or the Yellow Emperor’s Classic of Internal Medicine. In this text, acupuncture was the most cited treatment method, with Chinese herbal therapies endorsed more cautiously because they were considered dangerous and potentially lethal if used incorrectly. In China, there was a slow evolution of practice, and by the 19th century, acupuncture had lost most of its support in the Imperial Court and herbal therapies were preferred. In the early 20th century, acupuncture had been eliminated from medical training colleges and was practiced mainly by itinerant healers with education provided through family traditions. In the 1930s Cheng Dan’an, a Western-trained Chinese physician, undertook to bring acupuncture out of its superstitious and metaphysical past and ground it on a more secure foundation based on Western concepts of anatomy and thereby bring it into the mainstream. His decision to pursue this effort was based on a personal experience that he had when his father had treated him with acupuncture for back pain. Cheng developed the seminal acupuncture text that is still used in China and is the basis for the acupuncture approach called Traditional Chinese Medicine (TCM) now practiced in the West.3 With this modernized form of acupuncture, point combinations were often written for specific clinical presentations, in direct parallel to how herbal combinations were taught. This led to a proliferation of books giving point combinations that were thought of as canonical in the West but are actually the result of Western scientific and political influences on the evolution of acupuncture in China. In Japan, acupuncture has been practiced for over 500 years and has developed with an emphasis on using classical Chinese texts to guide practice, with less influence by the historical factors that led to the development of TCM in China. The use of TCM formulas has been readily adopted by Western researchers both because of the mistaken viewpoint that these formulas had an uninterrupted and ancient pedigree in China and because of the ease of maintaining scientific reproducibility. As discussed later in the chapter, this has had a significant adverse impact on study design and has led to some of the difficulties interpreting the results of recent large clinical trials in pain.
BASIC ACUPUNCTURE THEORIES
Chinese Taoist theories of yin and yang, or the balance of opposing influences in nature, underlie the theoretical framework used in acupuncture to understand human health. Human beings are seen as an integral part of a larger macrocosm that includes all the elements of the surrounding world. These elements are seen to have varying degrees of influence on the human organism, and factors such as weather, diet, and social environment are all taken to have significant effects on an individual’s health. The dynamic balance of these external factors, together with the internal physical and emotional state of the organism, interacts in a complex way to influence health and disease. As a correlate to this holistic view of human health, Chinese medicine makes no distinction between mental and physical illness and largely bypasses the mind-body dualism that plagues Western medical traditions.
In this integrated framework, the workings and function of the internal organs are believed to have specific, observable effects on the external appearance of the individual. Subtle changes seen on the surface of the body are all seen to reflect accurately on the homeostasis of the internal organ system. For example, alterations in the skin color and skin texture, variations in the suppleness and compliance of underlying muscles, the quality of arterial pulses, and the appearance of the tongue and eyes are important factors that go into making a diagnosis and treatment plan.2
As an outgrowth of Taoist theories of health, forces were postulated to explain how the internal and external systems interrelated. This metaphysical construct led to the concept of qi, or vital energy. As a means of developing organized treatment strategies that could explain empirical observations of human health and disease, qi was postulated to flow in various channels or meridians in the body. There are 12 principal meridians, 8 extra meridians, and a total of 361 classic acupuncture points that are located on these proposed energy channels. Although an anatomical correlation to the meridians has not been located, the concept is useful to understand and treat symptoms seen in various disease and pain states and may provide clues about the deep organization of the nervous system (Fig. 98-1).
The sine qua non of acupuncture treatment is for the practitioner to elicit the de qi response when inserting and manipulating the needle with a twisting and thrusting motion. This needle technique has been misinterpreted as a sensation that is as felt by the patient, indicating that the qi has been moved. This misinterpretation has been tacitly accepted in Western clinical research as a key component of ensuring the authenticity of the acupuncture treatment. Current clinical evidence strongly suggests that this phenomenon is not an adequate method of validating the treatment protocol.
Over the last 30 years, a great deal of scientific evidence has accumulated to verify that both acupuncture point (AP) stimulation and electroacupuncture (EA) stimulation have reproducible physiologic effects. Three main lines of evidence are presented in the following discussion. All go to the heart of the neurologic mechanisms that are currently understood to modulate and influence pain.
The evidence for the release of endogenous opioids with AP and EA derives from the seminal work done by Pomeranz and Chiu4 in animals and Mayer et al.5 in humans in the 1970s. Since that time, a large body of evidence has developed to show that both AP and EA lead to the release of endorphins and enkephalins into the cerebrospinal fluid (CSF). Furthermore, the release of these neuropeptides has been demonstrated to play a role in the analgesic effect of acupuncture as evidenced by opioid-receptor antagonists that can abolish the analgesia obtained with acupuncture in both human and animal models of acute pain.
Since the initial studies, both the met-enkephalin-responding neurons in the dorsal column of the spinal cord and the endorphin and enkephalin active sites in the periaquaductal gray zone of the brain have been shown to be involved in acupuncture analgesia. Both the parameters of stimulation and the site of stimulation have significant effects on the type of chemical releases. In particular, antiserum to met-enkephalin abolished acupuncture analgesia but antiserum to dynorphin did not when a true acupuncture point was stimulated, whereas the reverse was true when a nonacupuncture or sham point was stimulated.5 Manual acupuncture (MA) involves the insertion of an acupuncture needle into an acupuncture point followed by application of a stimulation technique, the most classic of which is a Chinese method of twisting the needle while thrusting up and down to elicit the de qi response. This method has been found to activate a broad range of afferent fibers, including Aβ, Aδ, and C.6 In EA, a stimulating, alternating current via the inserted needle is delivered and has been found to activate Aβ and Aδ-fibers.7 Centrally, the sensory information from acupuncture stimulation ascends through the spinal ventrolateral pathway to the brain with multiple effects that involve a network of brain regions, including the nucleus raphe magnus (NRM), periaqueductal gray (PAG), locus coeruleus, arcuate nucleus (Arc) of the hypothalamus, and the accumbens, caudate nuclei, and amygdale. There is a growing list of chemical releases that are felt to mediate acupuncture analgesia, including the opioid peptides (μ-, δ- and κ-receptors), glutamate (NMDA and AMPA/KA receptors), 5-hydroxytryptamine, and cholecystokinin octapeptide (CCK-8). Among these, the opioid peptides and their receptors in the descending pain modulatory pathway play an important role in mediating acupuncture analgesia.
One factor that may influence the response of a particular individual to acupuncture appears to involve CCK-8 receptor density.8 Interestingly the blockade of CCK receptors also potentiates the placebo analgesic response.9 Recently published evidence suggests that a patient’s analgesic response to acupuncture may be related to the patient’s genetic profile. In a Korean study using cDNA microarrays, investigators found that “high responders” had 375 genes that showed significant up- or downregulation compared to “low responders.” Many of the genes that showed upregulation were related to signal related biomolecules and stress and immune function, suggesting that genetics may play a role in the patient’s response to acupuncture.10
Electro-acupuncture stimulation has also been found to elevate levels of 5-hydroxytryptamine (5-HT) in the raphe nucleus, which enhances acupuncture analgesia presumably through descending inhibitory control mechanisms. Destruction of these neurons in the raphe nucleus of the midbrain or injection of paracholorophenylalanine, which lowers cerebral levels of 5-HT, will attenuate acupuncture analgesia, and injection of pargyline, which slows enzymatic degradation of 5-HT, enhances acupuncture analgesia.11 There has been further elucidation of the specific subtypes of 5-HT receptors influenced by EA. In particular, intrathecal injection of antagonists of 5-HT1A and 5-HT3 receptors, but not 5-HT2A antagonists, significantly blocked EA-induced depression of cold allodynia in the neuropathic rat and reduced spontaneous pain behaviors.12
The hypothalamic-pituitary-axis and catecholamines are also influenced by EA and AP and may further influence the analgesic response to pain through both immune modulation and modulation of the sympathetic responses. It has showed that spinal α2-adrenoceptors play a crucial role in inhibitory descending pain control by noradrenergic projections from supraspinal nuclei to the dorsal horn, particularly in modulating neuropathic pain.13,14 Intrathecal injection of α2 receptor antagonist yohimbine, but not α1 receptor antagonist prazosin, significantly blocked EA analgesia in neuropathic rats.12
There is evidence to suggest that acupuncture analgesia may also work through blockade of NMDA and AMPA/KA receptors. In the rat spinal nerve ligation model, EA decreased nerve injury-induced mechanical allodynia.15 Immunochemical studies revealed that nerve ligation increased the expression of NMDA receptor subtype NR1 immunoreactivity, which could be reduced by low-frequency EA.16
Spinal cord glia (microglia and astrocytes) make important contributions to the development and maintenance of inflammatory and neuropathic pain.17 It appears that EA can also act to inhibit microglial activation in mice and rat models of chronic pain.18,19
Recent technological advances in mapping brain activity using functional magnetic resonance scanning (fMRI) have begun to be applied to acupuncture. Comparison has been made between tactile sensation (tapping the skin with a wire at 2 Hz) and AP using a manual stimulation technique. The acupuncture stimulation used in this study involved twisting the needle at 2 Hz in L I4 (a point in the first dorsal interosseous muscle of the hand). Stimulation of an acupuncture point in this manner produces a deqi sensation, which is a full, aching feeling at the point of the needle and is believed to be important in obtaining the clinical effect with AP. The results of unilateral AP showed bilateral neural modulation of cortical and subcortical structures. The primary action was to decrease signal intensity in the limbic region and other subcortical areas. Tactile stimulation did not produce these changes in fMRI. In addition, if the needle was placed in the point and left at rest or placed subcutaneously and not in the muscle, fMRI signal decrease in these deep subcortical structures was not seen. This suggests that the response of the organism to AP depends on activation of the muscle sensory afferents and not the superficial afferents in the skin.20
Although most of the focus on the physiology of acupuncture has been on the release of endogenous opioids, the use of positron emission tomography (PET) has allowed the study of acupuncture effects on the opioid receptors themselves. In a study using 11C-carfentanil (an analog of the μ-opioid receptor) as the tracer, both short- and long-term increases in μ-opioid receptor binding potential in the cingulate, caudate, and amygdala were observed in patients receiving acupuncture therapy. In the comparison group that received sham acupuncture, the effects on μ-opioid receptor binding potential were not present, suggesting that acupuncture and sham acupuncture function by different mechanisms.21
PET has also allowed us to study the effects of acupuncture on brain metabolism. In a recent study by Park et al. using fluorodeoxyglucose (an analog of glucose) as the tracer, acupuncture stimulation was shown to increase glucose brain metabolism in the left insula, bilateral thalami, the superior frontal region of the right frontal lobe, and the inferior frontal region of left frontal lobe. At the same time, glucose metabolism was decreased in the cingulate and parahippocampal regions of the left limbic lobe. This study suggests that acupuncture changes regional brain glucose metabolism patterns in a very specific and specialized way based on the acupoints stimulated.22
These studies suggest that the grid of acupuncture points may be focused regions in the peripheral nervous system that represent a network of nodes that have profound and specific effects on modulating and regulating the activity of the central nervous system.
It is still an open question whether acupuncture has an influence on pain and other disease states that goes beyond the direct effect of the chemical releases previously mentioned. The early data using fMRI suggest that the sensory stimulation provided by acupuncture may have direct and selective effects on CNS function. Although the demonstration that endogenous opioids can be consistently released in both animal and human experimental models has been an important step in verifying that acupuncture analgesia has a physiologic basis, there continues to be debate about whether this effect is sufficient to explain the observed clinical benefits. Humoral effects are nonspecific and short-lived and cannot explain why a certain treatment method for a particular condition would have a sustained or permanent disease-modifying result. The chemical releases observed with EA and AP may just be an epiphenomenon, indicating that there is an influence on the CNS without yet comprehending what the actual changes are. Table 98-1 lists the problems with our current understanding of acupuncture analgesia.
Humoral Theories and Problems Associated with Acupuncture Analgesia
Humoral Theory of Acupuncture Analgesia
Endogenous opioid effect
Humoral effect short-lived.
Fails to explain importance of point selection and meridians.
Fails to explain disease modification and sustained analgesia obtained with acupuncture.
Difficult to implement theory to explain effects of nonpain-related conditions such as stroke.
Fails to capture neuromodulating effects of acupuncture.
An example of how the neurohumoral model fails to comprehend the clinical effect of AP is a recently published study using heat stimulation (moxabustion) of an acupuncture point on the fifth toe (B67) to turn breech babies after the 33rd week of pregnancy. The results of the study were profound, showing a significantly improved turning of the infants to the cephalic position at delivery compared with the control group. Of 130 fetuses in the intervention group, 98 (75.4%) were cephalic, compared with 62 (47.7%) of 130 fetuses in the control group (P < .001; relative risk = 1.58; 95% CI, 1.29–1.94).23