Myofascial trigger points (TrPs) are one of the most overlooked and ignored causes of acute and chronic pain (Hendler & Kozikowski 1993), and at the same time, constitute one of the most common musculoskeletal pain conditions (Hidalgo-Lozano et al.2010, Bron et al. 2011a). There is overwhelming evidence that muscle pain is commonly a primary dysfunction (Mense 2010a) and not necessarily secondary to other diagnoses. Muscles feature many types of nociceptors, which can be activated by a variety of mechanical and chemical means (Mense 2009). As a primary problem, TrPs may occur in absence of other medical issues; however, TrPs can also be associated with underlying medical conditions, e.g. systemic diseases, or certain metabolic, parasitic, and nutritional disorders. As a co-morbid condition, TrPs can be associated with other conditions such as osteoarthritis of the shoulder, hip or knee (Bajaj et al. 2001) and also injuries such as whiplash (Freeman et al. 2009). Pain elicited by muscle TrPs constitutes a separate and independent cause of acute and especially chronic pain that may compound the symptoms of other conditions and persist long after the original initiating condition has been resolved. TrPs are also associated with visceral conditions and dysfunctions, including endometriosis, interstitial cystitis, irritable bowel syndrome, dysmenorrhea and prostatitis (Weiss 2001, Anderson 2002, Doggweiler-Wiygul 2004, Jarrell 2004). The presence of abdominal TrPs was 90% predictive of endometriosis (Jarrell 2004).

Throughout history TrPs have been referred to by different names (Simons 1975). The current TrP terminology has evolved during the past several decades (Simons et al. 1999, Dommerholt et al. 2006, Dommerholt & Shah 2010). Although different definitions of TrPs are used among different disciplines, the most commonly accepted definition maintains that ‘a TrP is a hyperirritable spot in a taut band of a skeletal muscle that is painful on compression, stretch, overload or contraction of the tissue which usually responds with a referred pain that is perceived distant from the spot’ (Simons et al. 1999).

From a clinical viewpoint, we can distinguish active and latent TrPs. The local and referred pain from active TrPs reproduces the symptoms reported by patients and is recognized by patients as their usual or familiar pain (Simons et al. 1999). Both active and latent TrPs cause allodynia at the TrP and hyperalgesia away from the TrP following applied pressure. Allodynia is pain due to a stimulus that does not normally provoke pain. In latent muscle TrPs, the local and referred pain do not reproduce any symptoms familiar or usual to the patient (Simons et al. 1999). Active and latent TrPs have similar physical findings. The difference is that latent TrPs do not reproduce any spontaneous symptom. In patients with lateral epicondylalgia, active TrPs commonly reproduce the symptoms within the affected arm (Fernández-Carnero et al. 2007), but patients may also present with latent TrPs on the non-affected side without experiencing any symptoms on that side (Fernández-Carnero et al. 2008).

Although latent TrPs are not spontaneously painful, they provide nociceptive input into the dorsal horn (Ge et al. 2008, 2009, Li et al. 2009, Wang et al. 2010, Xu et al. 2010, Zhang et al. 2010, Ge & Arendt-Nielsen 2011). The underlying mechanism is not clear at this point in time and requires more research. Certain regions within a muscle may only be connected via ineffective synapses to dorsal horn neurons and referred pain may occur when these ineffective synapses are sensitized (Mense 2010b). Latent TrPs can easily turn into active TrPs, which is at least partially dependent on the degree of sensitization and increased synaptic efficacy in the dorsal horn. For example, the pain pressure threshold of latent TrPs in the forearm muscles decreased significantly after only 20 minutes of continuous piano playing (Chen et al. 2000). Active TrPs induce larger referred pain areas and higher pain intensities than latent TrPs (Hong et al. 1997). Active TrPs and their overlying cutaneous and subcutaneous tissues are usually more sensitive to pressure and electrical stimulation than latent TrPs (Vecchiet et al. 1990 1994).

Both active and latent TrPs can provoke motor dysfunctions, e.g. muscle weakness, inhibition, increased motor irritability, spasm, muscle imbalance, and altered motor recruitment (Lucas et al.2004, 2010) in either the affected muscle or in functionally related muscles (Simons et al. 1999). Lucas et al. (2010) demonstrated that latent TrPs were associated with impaired motor activation pattern and that the elimination of these latent TrPs induces normalization of the impaired motor activation pattern. In another study, restrictions in ankle range of motion were corrected after manual release of latent TrPs in the soleus muscle (Grieve et al. 2011).

Neurophysiological basis of muscle referred pain

Referred pain is a phenomenon that has been described for more than a century and has been used extensively as a diagnostic tool in the clinical setting. Typically, pain from deep structures such as muscles, joints, ligaments, tendons, and viscera is described as deep, diffuse, and difficult to locate accurately in contrast to superficial types of pain, such as pain originating in the skin (Mense 1994). Pain located at the source of pain is termed local pain or primary pain, whereas pain felt in a different region away from the source of pain is termed referred pain (Ballantyne et al. 2010). Referred pain can be perceived in any region of the body, but the size of the referred pain area is variable and can be influenced by pain-induced changes in central somatosensory maps (Kellgren 1938, Gandevia & Phegan 1999). Referred pain is a very common phenomenon in clinical practice; most patients with chronic pain present with what is commonly described as ‘a summation of referred pain from several different structures.’ Understanding at least the basic neurophysiological mechanisms of muscle referred pain is required to make a proper diagnosis of myofascial pain and to manage patients with TrPs.

Clinical characteristics of muscle referred pain (Arendt-Nielsen & Ge 2009, Fernández-de-las-Peñas et al. 2011)

1. The duration of referred pain could last for as short as a few seconds or as long as a few hours, days, or weeks, or occasionally indefinitely.

2. Muscle referred pain is described as deep, diffuse, burning, tightening, or pressing pain, which is completely different from neuropathic or cutaneous pain.

3. Referred pain from muscle tissues may have a similar topographical distribution as referred pain from joints.

4. The referred pain can spread cranial/caudal or ventral/dorsal.

5. The intensity of muscle referred pain and the size of the referred pain area are positively correlated to the degree of irritability of the central nervous system or sensitization.

6. Referred pain frequently follows the distribution of sclerotomes, but not of dermatomes.

7. Muscle referred pain may be accompanied by other symptoms, such as numbness, coldness, stiffness, weakness, fatigue, or musculoskeletal motor dysfunction. The term referred pain is perhaps not complete and a preferred term can be ‘referred sensation’ as non-painful sensations such as burning or tingling would still be considered referred phenomena from TrPs.

Mechanisms and neurophysiological models of referred pain (Arendt-Nielsen & Ge 2009)

Muscle referred pain is a process of central sensitization which is mediated by a peripheral activity and sensitization, and which can be facilitated by sympathetic activity and dysfunctional descending inhibition

Arendt-Nielsen & Ge 2009

The exact neuropathways mediating referred pain are not completely understood. Several neuro-anatomical and neurophysiological theories regarding the appearance of referred pain have been suggested. All models agree that nociceptive dorsal horn or brainstem neurons receive convergent inputs from different tissues. Consequently, higher brain centers cannot identify the input source properly. Most recent models have included newer theories in which sensitization of dorsal horn and brainstem neurons also play a relevant and central role. We briefly summarize the most common theories below.

Convergent-projection theory

Ruch (1961) proposed that afferent fibers from different tissues, such as skin, viscera, muscles, and joints, converge onto common spinal neurons, which can lead to a misinterpretation of the source of nociceptive activity from the spinal cord. The source of pain of one tissue can be misinterpreted as originating from other structures. The convergent-projection theory would explain the segmental nature of muscle referred pain and the increased referred pain intensity when local pain is intensified. This theory does, however, not explain the delay in the development of referred pain following the onset of local pain (Graven-Nielsen et al. 1997a).

Convergence-facilitation theory

The somatosensory sensitivity changes reported in referred pain areas could, in part, be explained by sensitization mechanisms in the dorsal horn and brainstem neurons, whereas the delay in appearance of referred pain could be explained since the creation of central sensitization needs time (Graven-Nielsen et al. 1997a).

Axon-reflex theory

Bifurcation of afferents from different tissues was suggested as an explanation of referred pain (Sinclair et al. 1948). Although bifurcation of nociceptive afferents from different tissues exits, it is generally agreed that these pathways are unlikely to occur (McMahon 1994). The axon-reflex theory cannot explain the delay in the appearance of the referred pain, the different thresholds required for eliciting local pain vs referred pain, and the somatosensory sensitivity changes within the referred pain areas.

Thalamic-convergence theory

Theobald (1949) suggested that referred pain may appear as a summation of input from the injured area and from the referred pain area within neurons in the brain, but not in the spinal cord. Several decades later, Apkarian et al. (1995) described several pathways converging on different sub-cortical and cortical neurons. There is evidence of pain reduction following anesthetization of the referred pain area, which suggests that peripheral processes contribute to referred pain, although central processes are assumed to be the most predominant.

Central hyper-excitability theory

Recordings from dorsal horn neurons in animal models have revealed that new receptive fields at a distance from the original receptive field emerged within minutes after noxious stimuli (Hoheisel et al. 1993). That is, following nociceptive input, dorsal horn neurons that were previously responsive to only one area within a muscle began to respond to nociception from areas that previously did not trigger a response. The appearance of new receptive fields could indicate that latent convergent afferents on the dorsal horn neuron are opened by noxious stimuli from muscle tissues (Mense 1994), and that this facilitation of latent convergence connections induces the referred pain.

The central hyper-excitability theory is consistent with most of the characteristics of muscle referred pain. There is a dependency on the stimulus and a delay in appearance of referred pain as compared with local pain. The development of referred pain in healthy subjects is generally distal and not proximal to the site of induced pain (Arendt-Nielsen et al. 2000). Nevertheless, several clinical studies have demonstrated proximal and distal referred pain in patients with chronic pain (Graven-Nielsen 2006). The differences between healthy individuals and persons with chronic pain may indicate that the pre-existing pain could induce a state of hyper-excitability in the spinal cord or brainstem resulting in proximal and distal referred pains.

Neurophysiological aspects of muscle/TrPs

The nature of TrPs

Taut bands

TrPs are located within discrete bands of contractured muscle fibers called taut bands. Taut bands can be palpated with a flat or pincer palpation and feel like tense strings within the belly of the muscle. It is important to clarify that contractures are not the same as muscle spasms. Muscle spasms require electrogenic activity, meaning that the α-motor neuron and the neuromuscular endplate are active. A muscle spasm is a pathological involuntary electrogenic contraction (Simons & Mense 1998). In contrast, a taut band signifies a contracture arising endogenously within a certain number of muscle fibers independent of electromyogenic activity, which does not involve the entire muscle (Simons & Mense 1998).

In 1997, Gerwin and Duranleau first described the visualization of taut bands using sonography, but until recently it was not yet possible to visualize the actual TrP (Lewis & Tehan, 1999) mostly due to technological limitations (Park & Kwon 2011). With the advancement of technology, more recent studies have found that TrP taut bands can be visualized using sonographic and magnetic resonance elastography (Chen et al. 2007, 2008, Sikdar et al. 2009, Rha et al. 2011). Chen et al. (2007) demonstrated that the stiffness of the taut bands in patients with TrPs is higher than that of the surrounding muscle tissue in the same subject and in people without TrPs. Sikdar et al. (2009) showed that vibration amplitudes assessed with spectral Doppler were 27% lower on average within the TrP region compared with surrounding tissue. They also found reduced vibration amplitude within the hypoechoeic region identified as a TrP. In summary, TrP taut bands are detectable and quantifiable, providing potentially useful tools for TrP diagnosis and future research.

Although TrPs and taut bands can now be visualized, the mechanism for the formation of muscle taut band is still not fully understood. The probable mechanisms of taut band formation have been summarized by Gerwin (2008). The current thinking is that the development of the taut band and subsequent pain is related to local muscle overload or overuse when the muscle cannot respond adequately, particularly following unusual or excessive eccentric or concentric loading (Gerwin et al. 2004, Gerwin 2008, Mense & Gerwin 2010). Muscle failure and TrP formation are also common with sub-maximal muscle contractions as seen for example in the upper trapezius muscles of computer operators (Treaster et al. 2006, Hoyle et al. 2011) or in the forearm muscles of pianists (Chen 2000). The failure of the muscle to respond to a particular acute or recurrent overload may be the result of a local energy crisis. Muscle activation in response to a demand is always dispersed throughout the muscle among fibers that are the first to be contracted and the last to relax. These fibers are the most vulnerable to muscle overload. Unusual or excessive eccentric loading may cause local muscle injury. In sub-maximal contractions, the Cinderella Hypothesis and Henneman’s size principle apply (Kadefors et al. 1999, Chen et al. 2000, Hägg, 2003, Zennaro et al. 2003, Treaster et al. 2006, Hoyle et al. 2011). Smaller motor units are recruited first and de-recruited last without any substitution of motor units. This would lead to local biochemical changes without muscle breakdown, especially in those parts of the muscle that are not substituted and therefore most heavily worked (Gerwin 2008).

Under normal circumstances, acetylcholine (ACh) is released in a quantal fashion, which is a calcium-dependent process (Wessler 1996). ACh stimulates specific membrane-bound protein molecules, such as nicotinic ACh receptors (nAChR), which leads to miniature endplate potentials (MEPP). A summation of MEPPs causes a depolarization of the muscle membrane, an action potential, stimulation of ryanodine and dihydropyridine receptors in the T-tubuli, activation of the sarcoplasmic reticulum, a release of calcium, and eventually a muscle contraction. The nAChRs are temporarily inhibited following stimulation by ACh (Magleby & Pallotta 1981).

With myofascial pain, an excessive non-quantal release of ACh is released from the motor endplate at the neuromuscular junction. Acetylcholinesterase (ACh-esterase) is inhibited, while nAChRs are up-regulated. These and several other factors are thought to cause the localized muscle fiber contractures in the immediate vicinity of the involved motor endplates. It is conceivable that the limited non-quantal release of ACh is sufficient to trigger contracture without inhibiting the nicotinic ACh, dihydropyridine and ryanodine receptors, which would provide a mechanism for sustained contractures (Dommerholt 2011).

The potential role of calcitonin gene-related peptide (CGRP) in myofascial pain and other pain conditions, such as migraines, cannot be underestimated. CGRP is found in higher concentrations in the immediate environment of active TrP (Shah et al., 2008). It is a potent microvasular vasodilator involved in wound healing, prevention of ischemia, and several autonomic and immune functions (Smillie & Brain 2011). CGRP and its receptors are widely expressed in the central and peripheral nervous system. For example, CGRP Type I is produced in the cell body of motor neurons in the ventral horn of the spinal cord and is excreted via an axoplasmatic transport mechanism. CGRP is also released from the trigeminal ganglion and from trigeminal nerves within the dura and as such contributes to peripheral sensitization (Durham & Vause, 2010). It also stimulates the phosphorylation of ACh receptors, which prolongs their sensitivity to ACh (Hodges-Savola & Fernandez 1995). In addition, CGRP promotes the release of ACh and inhibits ACh-esterase.

Interestingly, myofascial tension, as seen with TrPs, may also stimulate an excessive release of ACh, which suggests the presence of a self-sustaining vicious cycle (Chen & Grinnell 1997, Grinnel et al. 2003). Experimental research of rodents demonstrated that excessive ACh in the synaptic cleft leads to morphological changes resembling TrP contractures (Mense et al. 2003). Consuming excessive amounts of coffee in combination with alcohol triggers a similar response pattern, which has been attributed to the ability of caffeine to release calcium ions from the sarcoplasmic reticulum (Oba et al. 1997a, 1997b, Shabala et al. 2008).

There is some evidence that TrPs are also associated with increased autonomic activity (Ge et al., 2006), which is likely to be due to activation of adrenergic receptors at the motor endplate (Gerwin et al. 2004). Stimulation of these adrenergic receptors triggered an increased release of ACh in mice (Bowman et al. 1988). Nociceptive stimuli of latent TrPs can lead to autonomic changes such as vasoconstriction (Kimura et al. 2009). The local or systemic administration of the alpha-adrenergic antagonist phentolamine to TrPs caused an immediate reduction in electrical activity, suggesting that TrPs indeed have an autonomic component (Hubbard & Berkoff 1993, Lewis et al. 1994, McNulty et al. 1994, Banks et al. 1998). Such increased autonomic activity may facilitate an increased concentration of intracellular ionized calcium and be responsible again for a vicious cycle maintaining TrPs (Gerwin et al. 2004, Gerwin 2008). Muscle spindle afferents may also contribute to the formation of TrP taut bands via afferent signals to extrafusal motor units through H-reflex pathways (Ge et al. 2009), but are not considered to be the primary cause of TrP formation.

Local twitch response

Manual strumming or needling of a TrP usually result in a so-called local twitch response (LTR), which is a sudden contraction of muscle fibers in a taut band (Hong & Simons 1998). LTRs can be observed visually, can be recorded electromyographically, or can be visualized with diagnostic ultrasound (Gerwin & Duranleau 1997, Rha et al. 2011). The number of LTRs may be related to the irritability of the muscle TrP (Hong et al. 1997), which in turn appears to be correlated with the degree of sensitization of muscle nociceptors by bradykinin, serotonin, and prostaglandin, among others. Hong and Torigoe (1994) reported that, in an animal model, LTRs could be elicited by needling of hypersensitive trigger spots, which are the equivalent of TrPs in humans. LTRs were observed in only a few of the control sites. In addition, LTRs could not be elicited after transection of the innervating nerve. LTRs are spinal cord reflexes elicited by stimulating the sensitive site in the TrP region (Hong et al. 1995). Audette et al. (2004) reported that dry needling of active TrPs in the trapezius and levator scapulae muscles elicited bilateral LTRs in 61.5% of the muscles, whereas dry needling of latent TrPs resulted only in unilateral LTRs.

Eliciting LTRs has been advocated for effective TrP dry needling (Hong 1994). Following LTRs, Shah et al. (2005) observed an immediate drop in concentrations of several neurotransmitters, including CGRP and substance P, and several cytokines and interleukins, in the extracellular fluid of the local TrP milieu. The concentrations did not quite reach the levels of normal muscle tissue. The reductions were observed during approximately 10 minutes, before they appeared to slowly rise again. Due to the short observation times, it is not known whether the concentrations stabilized or increased again over time.

Muscle pain

Muscle pain follows noxious stimuli, which activate specific peripheral nociceptors. Nociceptive impulses are transmitted through second order neurons in the dorsal horn, through the spinal cord, and to primary and secondary somatosensory areas in the brain, including the amygdala, anterior cingulated gyrus, and the primary sensory cortex. Locally, activation of receptors leads to the release of neuropeptides, which also causes vasodilatation and increases the permeability of the microvasculature (Snijdelaar et al. 2000, Ambalavanar et al. 2006). When neuropeptides are released in sufficient quantity they trigger the release of histamine from mast cells, BK from kallidin, serotonin (5-HT) from platelets, and PGs from endothelial cells (Massaad et al. 2004), which leads to a vicious cycle as these chemicals also activate peripheral nociceptive receptors and potentiate dorsal horn neuron sensitization. As such, muscle nociceptors play an active role in muscle pain and in the maintenance of normal tissue homeostasis by assessing the peripheral biochemical milieu and by mediating the vascular supply to peripheral tissue.

The responsiveness of receptors is indeed a dynamic process and can change depending upon the concentrations of the sensitizing agents. As an example, under normal circumstances, the BK receptor, knows as a B2 receptor, triggers only a temporary increase of intracellular calcium and does not play a significant role in sensitization. When the BK concentration increases, a B1 receptor is synthesized, which facilitates a long-lasting increase of intracellular calcium and stimulates the release of tumor necrosing factor and other interleukins, which in turn lead to increased concentrations of BK and peripheral sensitization (Calixto et al. 2000, Marceau et al. 2002). There are many interactions between these chemicals making muscle pain a very complicated phenomenon. Babenko et al. (1999) found that the combination of BK and 5-HT induced higher sensitization of nociceptors than each substance in isolation.

Referred pain occurs at the dorsal horn level and is the result of activation of otherwise quiescent axonal connections between affective nerve fibers dorsal horn neurons, which are activated by mechanisms of central sensitization (Mense & Gerwin 2010). Referred pain is not unique to myofascial TrPs but, nevertheless, it is highly characteristic of myofascial pain syndrome. Usually referred pain happens within seconds following mechanical stimulation of active TrPs suggesting that the induction of neuroplastic changes related to referred pain is a rapid process. Kuan et al. (2007a) demonstrated that TrPs are more effective in inducing neuroplastic changes in the dorsal horn neurons than non-TrPs regions.

The multiple contractures found in muscles of patients with myofascial pain are likely to compress regional capillaries, resulting in ischemia and hypoxia. Recent Doppler ultrasound studies confirmed a higher outflow resistance or vascular restriction at active TrPs and an increased vascular bed outside the immediate environment of TrPs (Sikdar et al. 2010), which is consistent with the measurement of decreased oxygen saturation levels within TrPs and increased levels outside the core of TrPs (Brückle et al. 1990). Hypoxia may trigger an immediate increased release of ACh at the motor endplate (Bukharaeva et al. 2005). Hypoxia also leads to a decrease of the local pH, which will activate transient receptor potential vanilloid (TRPV) receptors and acid sensing ion channels (ASIC) via hydrogen ions or protons. Because these channels are nociceptive, they initiate pain, hyperalgesia and central sensitization without inflammation or any damage or trauma to the muscle (Sluka et al. 2001, 2002, 2003, 2009, Deval et al. 2010). Research at the US National Institutes of Health has confirmed that the pH in the direct vicinity of active TrPs is well below 5, which is sufficient to activate muscle nociceptors (Sahlin et al. 1976, Gautam et al. 2010). Different kinds of ASICs play specific roles (Walder et al. 2010) and it is not known which ASICs are activated in myofascial pain. It is likely that multiple types of ASICs are involved in the sensory aspects of TrPs (Dommerholt, 2011), such as the ASIC1a, which processes noxious stimuli, and the ASIC3, which is involved in inflammatory pain (Shah et al. 2005, Deval et al. 2010). A low pH down-regulates ACh-esterase at the neuromuscular junction and can trigger the release of several neurotransmitters and inflammatory mediators, such as CGRP, substance P, BK, interleukins, adenosine triphosphate (ATP), 5-HT, prostaglandins (PG), potassium and protons, which would result in a decrease in the mechanical threshold and activation of peripheral nociceptive receptors. A sensitized muscle nociceptor has a lowered stimulation threshold into the innocuous range and will respond to harmless stimuli like light pressure and muscle movement. When nociceptive input to the spinal cord is intense or occurs repeatedly, peripheral and central sensitization mechanisms occur and spread of nociception at the spinal cord level results in referred pain (Hoheisel et al. 1993).