Introduction

Many physical therapists and other clinicians have adopted a contemporary pain management approach and incorporate graded exercise, restoration of movement, pain science education, and psychosocial perspectives into the examination, assessment, and therapeutic interventions of patients presenting with pain complaints (Gifford & Butler, 1997; George et al., 2010; Nijs et al., 2010; Hodges & Tucker, 2011). The question emerges whether these approaches by themselves are sufficient to address persistent pain states without eliminating peripheral nociceptive input?

Current pain science research supports that pain is an output by the brain, when there is a perception of bodily danger requiring specific action (Moseley, 2003a). In other words, the ‘issues are not just in the peripheral tissues’ (Butler, 1991) and considering the meaning of pain in the context of the patient’s overall situation is critical (Moseley, 2012a). The effects of trigger point (TrP) dry needling (DN) cannot be considered without this broader biopsychosocial model (Dommerholt, 2011). TrP DN must be approached from a pain science perspective, as it is no longer sufficient to consider TrP therapy strictly as a tool to address local muscle pathology.

As Moseley pointed out, nociceptive mechanisms that contribute to threatening information should be treated, where possible (Moseley, 2003a). Especially in persistent pain conditions, TrPs are a constant source of nociceptive input (Melzack, 2001; Giamberardino et al., 2007; Ge et al., 2011), and it follows that removing such peripheral input is indicated and consistent with the concepts of Melzack’s neuromatrix (Melzack, 2001). In Moseley’s words, ‘trigger points are present in all patients with chronic musculoskeletal pain and are thought to reflect sensitisation of nociceptive processing in the central nervous system’ and ‘elimination of trigger points is an important component of the management of chronic musculoskeletal pain’ (Moseley, 2012b). In addition to their contribution to nociception, TrPs can contribute to abnormal movement patterns (Lucas et al., 2004, 2010; Bohlooli et al., 2016). TrP-DN immediately improved these abnormal muscle activation patterns (Lucas et al., 2004, 2010). The combination of tissue-based interventions, such as TrP-DN (bottom–up techniques) should be combined with neuroscience pain education (top–down techniques) (Louw et al., 2017). TrP-DN is effective when combined with neuroscience pain education in patients with low back pain (Téllez-García et al., 2015). Once patients realise that the somewhat uncomfortable stimulus of DN actually has the potential to reduce or even eliminate their pain, an endogenous inhibitory conditioned pain modulation system will be activated that inhibits early nociceptive processing (Bjorkedal & Flaten, 2012). Fear of needles did not seem to influence pain tolerance and sympathetic responses (Joseph et al., 2013). Pain after DN, usually referred to as postneedling soreness, is a common finding (Brady et al., 2014) that must be addressed as well. Patients must be reassured that postneedling soreness is normal and more or less irrelevant regarding the overall therapeutic outcome. Studies show that spray and stretch (Martin-Pintado Zugasti et al., 2014), local pressure (Martin-Pintado Zugasti et al., 2015), and low-load exercise (Salom-Moreno et al., 2017) can reduce the soreness.

By creating a therapeutic environment in which the expectation of pain reduction combined with an increased sense of self-efficacy are the focus, usually the pain intensity of the treatment can be dissociated from the magnitude of responses in the pain matrix (Bandura et al., 1987; Legrain et al., 2011a, 2011c; Mueller et al., 2012). Expectancy can significantly influence the anticipation and experience of pain (Wager et al., 2004). The context in which a painful stimulus (i.e., DN) is delivered affects patients’ experience and anticipation that this noxious stimulus may trigger (Moseley & Amtz, 2007). Legrain and colleagues (2011b) established that patients should be able to focus and maintain their attention on the processing of pain-unrelated information without being too preoccupied by nociceptive stimuli. When applied to using DN in clinical practice, a top–down approach can facilitate an activation of those executive functions that are involved in the control of selective attention and focus on the important overall goal of achieving pain reduction and return to function. In physical therapy, it is often thought that in the presence of central nervous system sensitization, treatments must be pain free to avoid further wind-up of the sensitized central nervous system (Jull 2012). To cause increased wind-up, input from multiple receptors is required over tens of seconds. A single stimulus, such as a pinch or needle stick, is usually insufficient to induce central sensitization.

Increased activity in the anterior cingulate cortex (ACC) is common in chronic pain conditions and is even present when pain is anticipated (Hsieh et al., 1995; Peyron et al., 2000a, 2000b; Sawamoto et al., 2000; Longo et al., 2012), which supports reducing the anticipation of pain during DN. In addition, several studies have shown that TrPs also can activate the ACC and other limbic structures, but suppress hippocampal activity (Svensson et al., 1997; Niddam et al., 2007, 2008). This suggests that TrPs may be linked to stress-related changes and therefore minimising needling-related stress levels and the patients’ anticipation is very important. Of interest in this context is that clinicians may underestimate their patients’ ability to understand basic concepts of pain neurophysiology (Moseley, 2003b), which emphasises the need for better pain science education of healthcare professionals (Hoeger Bement & Sluka, 2015).

When treating patients with DN techniques, it is imperative to avoid creating the impression that local muscle pathology would be solely responsible for the persistent pain (Nijs et al., 2010; Puentedura & Louw, 2012). Rather than explaining TrPs as a local pathological or anatomical problem, it makes more sense to focus on the nociceptive nature of TrPs and their role in perpetuating central sensitisation (Fernández de las Peñas & Dommerholt, 2014). In general, input from muscle nociceptors is more effective at inducing neuroplastic changes in wide dynamic range dorsal horn neurons than input from cutaneous nociceptive receptors (Wall & Woolf, 1984), and persistent peripheral nociceptive input increases the sensitivity of the central nervous system.

Unfortunately, the contributions of TrPs and the potentially detrimental effect of poorly worded diagnostic considerations with a predominant anatomical emphasis often are not considered, and individual patients may have gone through many unsuccessful treatment regimens with multiple diagnostic pathways. For example, telling patients they have ‘a slipped disc’ or degenerative disc disease, or even a tear in the rotator cuff muscles, as the main focus of the discussion can contribute to feelings of iatrogenic hopelessness and reduce their expectations (Jull & Sterling, 2009; Puentedura & Louw, 2012; Louw, 2016). Patients may develop fear avoidance or kinesiophobia, poor coping skills, and an anticipation of pain (Bandura et al., 1987; Vlaeyen & Linton, 2000; Wager et al., 2004; Coppieters et al., 2006). Additionally, patients’ altered homeostatic systems may start contributing to the overall pain experience (Puentedura & Louw, 2012) with decreased blood flow to the muscles (Zhang et al., 2009), abnormal cytokine production (Watkins et al., 2001; Milligan & Watkins, 2009), constrained breathing patterns (Chaitow, 2004), and abnormal muscle activation patterns (Moreside et al., 2007), among others. In some patients the anticipation of pain and the pain associated with DN itself may activate threatening inputs, at which point DN would become counterproductive. Fortunately, fear of needles does not seem to have a major effect (Joseph et al., 2013), and for most patients TrP DN is a viable intervention (Dilorenzo et al., 2004; Affaitati et al., 2011).

Mechanisms and effects of trigger point dry needling

There are no studies of the effect of DN on the ACC and other limbic structures, but several papers suggest that needling acupuncture and nonacupuncture points does seem to involve the limbic system and the descending inhibitory system (Takeshige et al., 1992a, 1992b; Wu et al., 1999; Hui et al., 2000; Biella et al., 2001; Hsieh et al., 2001; Wu et al., 2002). DN studies of patients with fibromyalgia, which is a diagnosis of central sensitisation (Dommerholt & Stanborough, 2012; Bennet et al., 2014), demonstrate that DN of a few TrPs does not only reduce the nociceptive input from the treated TrPs, but reduces the overall widespread pain and sensitivity (Ge et al., 2009, 2010, 2011; Affaitati et al., 2011). TrP DN often evokes patients’ referred pain patterns and their primary pain complaint (Hong et al., 1997). Needling of TrPs in the gluteus minimus or teres minor muscles may initiate pain resembling a L5 or C8 radiculopathy, respectively (Escobar & Ballesteros, 1988; Facco & Ceccherelli, 2005). Needling of TrPs in the sternocleidomastoid or upper trapezius muscles may trigger a patient’s migraine or tension-type headache (Calandre et al., 2006). Experimentally induced muscle pain impairs diffuse noxious inhibitory control mechanisms (Arendt-Nielsen et al., 2008), and DN does seem to effect central sensitisation, presumably by altering the nociceptive processing (Kuan et al., 2007a; Mense, 2010; Mense & Masi, 2011). It is known that TrP DN reduces segmental nociceptive input and as such is therapeutically indicated (Srbely et al., 2010).

The exact mechanisms of DN continue to be elusive. Because many studies and case reports have confirmed the clinical efficacy of DN, future research must be directed towards examining the underlying mechanisms (Lewit, 1979; Carlson et al., 1993; Hong, 1994, 1997; Hong & Hsueh, 1996; McMillan et al., 1997; Chen et al., 2001; Cummings, 2003; Mayoral & Torres, 2003; Dilorenzo et al., 2004; Ilbuldu et al., 2004; Itoh et al., 2004, 2007; Lucas et al., 2004; Furlan et al., 2005; Kamanli et al., 2005; Mayoral-del-Moral, 2005; Weiner & Schmader, 2006; Giamberardino et al., 2007; Hsieh et al., 2007, 2012, 2014; Fernandez-Carnero et al., 2010; Lucas et al., 2010; Osborne & Gatt, 2010; Tsai et al., 2010; Srbely et al., 2010; Affaitati et al., 2011; Gonzalez-Perez et al., 2012; Mahmoudzadeh et al., 2016; Gattie et al., 2017). Recent systematic reviews confirm the evidence for DN (Furlan et al., 2005; Cagnie et al., 2015; Liu et al., 2015; Espejo-Antûnez et al., 2017). The quality of myofascial pain treatment studies is also improving (Stoop et al., 2017). TrP-DN was more effective than DN randomly in a muscle (Pecos-Martin et al., 2015). Slowly, bits and pieces of the myofascial pain and DN puzzle are beginning to be explored, even though some studies did not observe any significant advantage to adding DN to the options clinicians may use to treat patients with pain and movement dysfunction (Mason et al., 2016; Espi-Lopez et al., 2017; Perez-Palomares et al., 2017).

Mechanically, deep DN may disrupt contraction knots, stretch contractured sarcomere assemblies, and reduce the overlap between actin and myosin filaments. It may destroy motor endplates and cause distal axon denervation and changes in the endplate cholinesterase and acetylcholine receptors similarly to the normal muscle regeneration process (Gaspersic et al., 2001; Domingo et al., 2013). DN may also change the excitability of spinal motor neurons and improve muscle tone separate from its analgesic effect (Casale et al., 2017).

Of particular interest are local twitch responses (LTR), which are involuntary spinal cord reflexes of muscle fibres in a taut band after DN, injections, or snapping palpation (Dexter & Simons, 1981; Fricton et al., 1985; Hong, 1994; Hong & Torigoe, 1994; Simons & Dexter, 1995; Wang & Audette, 2000; Ga et al., 2007). Eliciting LTR is important when inactivating TrPs and confirms that the needle was placed accurately into a TrP. Several studies have confirmed that an LTR can reduce or even eliminate the typical endplate noise and endplate spikes associated with TrPs, which suggests that DN inactivates TrPs (Hong, 1994; Hong & Torigoe, 1994; Chen et al., 2001; Hsieh et al., 2011; Liu et al., 2017). There is a positive correlation between the prevalence of endplate noise in a TrP region and the pain intensity of that TrP (Kuan et al., 2007b). Endplate noise and endplate spikes reflect a summation of miniature endplate potentials and are characteristic of TrPs (Simons et al., 1995, 2002; Hong & Simons, 1998; Simons, 2001, 2004). Moreover, eliciting LTRs appears to reduce the concentrations of many chemicals found in the immediate environment of active TrPs, such as calcitonin gene related peptide, substance P, serotonin, interleukins, and epinephrine, among others (Shah et al., 2003, 2005, 2008; Shah & Gilliams, 2008; Hsieh et al., 2014). Shah and colleagues (2008) had speculated that the drop in concentrations may be caused by a local increase in blood flow, by interference with nociceptor membrane channels, or by transport mechanisms associated with a briefly augmented inflammatory response. The decrease of concentrations of substance P and calcitonin gene-related peptide corresponds with the clinical observation of a reduction in pain after deep DN (Shah et al., 2008). Hsieh and colleagues (2012) confirmed that TrP-DN modulated the chemical mediators associated with pain and inflammation, such as substance P, β-endorphin, and tumour necrosis factor-α, among others. They also reported that modulating these chemical concentrations with TrP-DN is dose dependent, which implies that an excessive amount of DN may actually increase the concentrations. Liu and colleagues observed that TrP-DN with LTRs also reduced acetylcholine and acetylcholine receptor levels significantly in a well-executed rabbit study (Liu et al., 2017). The increased concentration of multiple chemicals is associated with a lowered pH as a result of ischaemia and hypoxia (Brückle et al., 1990; Sikdar et al., 2010; Ballyns et al., 2011). DN of the trapezius muscle increased the blood flow and oxygen saturation, which suggests another potential mechanism in support of DN (Cagnie et al., 2012).

LTRs are often visible with the naked eye and can be visualised with sonography (Gerwin & Duranleau, 1997; Lewis & Tehan, 1999; Rha et al., 2011). It is not known how many LTRs are required for a positive outcome or if there even is a correlation between the number and size of LTRs and therapeutic outcomes. A recent study reported no clinical differences on pain depending on the number of LTRs obtained in patients with neck pain (Fernández-Carnero et al., 2017). Nevertheless, individuals exhibited significant clinical improvement when eliciting the highest number (n = 6) of LTRs until exhaustion compared with not eliciting any (Fernández-Carnero et al., 2017). Koppenhaver and colleagues (2017) found no differences after 1 week between patients with low back pain experiencing LTR and those not experiencing LTR, although subjects who experienced a LTR reported a greater short-term improvement in the function of the lumbar multifidus muscle compared with those who did not. Discrepancies expressed in published studies have led some authors to question the need of LTRs during TrP-DN (Perreault et al., 2017). It appears that several DN studies do not resemble clinical practice, which may influence the findings and their utility. In the clinic, it would be highly unusual to stick a patient only once with a needle or treat only one muscle. Yet several studies exploring the efficacy of DN use a model with either one needle stick or a very limited number of muscles.

The efficacy of dry needling can be monitored with ultrasound imaging. Turo and colleagues (2015) visualised tissue changes after DN. Participants with at least one active TrP received a 3-week course of DN at their most active TrP. Grey-scale 2D B-mode and colour Doppler ultrasound images were taken at the TrP. A significant reduction in the heterogeneity of muscle stiffness was observed at those TrPs that responded to treatment even 8 weeks after the DN procedures. In addition, the TrP status changed from active to latent and eventually a complete resolution of pain symptoms in a significant number of cases (Turo et al., 2015). Gerber and colleagues (2015) confirmed that a few sessions of DN can effectively reduce pain for at least 6 weeks. Several clinical investigators are exploring whether DN can be used to decrease spasticity with promising results (see Chapter 4) (Salom-Moreno et al., 2014; Ansari et al., 2015; Calvo et al., 2016; Mendigutia-Gomez et al., 2016). Calvo and colleagues (2017) showed that DN had a positive effect on quantitative electroencephalographic activity, especially in both the frontal and prefrontal regions of two stroke patients.

The effects of superficial DN are often attributed to stimulation of Aδ sensory afferent fibres, which may outlast the stimulus for up to 72 hours (Baldry, 2005). It is true that stimulation of Aδ nerve fibres may activate enkephalinergic, serotonergic, and noradrenergic inhibitory systems (Bowsher, 1998); however, type I high-threshold Aδ nerve fibres are only activated by nociceptive mechanical stimulation, and type II Aδ fibres require cold stimuli (Millan, 1999). Because superficial DN is neither a painful mechanical nor a cold stimulus, it is unlikely that Aδ fibres would get activated (Dommerholt et al., 2006). When superficial DN is combined with rotation of the needle, the stimulus may activate the pain inhibitory system associated with stimulation of Aδ fibres through segmental spinal and propriospinal hetero- segmental inhibition (Sandkühler, 1996). Deep DN can be also combined with rotation of the needle, after which the needle is left in place until relaxation of the muscle fibres has occurred (Dommerholt et al., 2006). The mechanical pressure exerted with the needle may electrically polarise muscle and connective tissue and transform mechanical stress into electrical activity, which is required for tissue remodelling (Liboff, 1997). It is also possible that superficial DN may activate mechanoreceptors coupled to slow conducting unmyelinated C fibre afferents. This could trigger a reduction of pain and a sense of progress and well-being through activation of the insular region and anterior cingulate cortex (Olausson et al., 2002; Mohr et al., 2005; Lund & Lundeberg, 2006).

Many clinicians combine superficial and deep DN with electrical stimulation through the needles (Mayoral & Torres, 2003; Mayoral-del-Moral, 2005; Dommerholt et al., 2006). Electrical stimulation can activate the endogenous opioid system, which likely has a facilitating effect of pain reduction (Han et al., 1984). In addition, it may activate the peri-aqueductal grey in some patients (Niddam et al., 2007). Modulating high- and low-frequency TENS produced better analgesia than either frequency by itself (DeSantana et al., 2008); however each application has specific effects. Only low-frequency TENS increased spinal concentrations of serotonin during and immediate after treatment (Sluka et al., 2006), which can reduce pain. Several rodent studies have shown that electrical acupuncture can modulate the expression of N-methyl-d-aspartate in primary sensory neurons (Choi et al., 2005; Wang et al., 2006). Only high-frequency TENS reduced spinal concentrations of glutamate and aspartate in rats with joint inflammation and activated delta-opioid receptors, possibly through spinal dorsal horn glial cells, as blocking the delta-opioid receptors with naltrindole caused a reduction in spinal concentrations (Gopalkrishnan & Sluka, 2000; Sluka et al., 2005). High-frequency TENS also reduced primary hyperalgesia to mechanical stimulation and heat (Gopalkrishnan & Sluka, 2000). It appears that combining high- and low-frequency TENS may give the best results, even though there is conflicting research (Lin et al., 2002; Barlas et al., 2006; Leon-Hernandez et al., 2016).

Unfortunately, there are no evidence-based guidelines of the optimal treatment parameters such as optimal amplitude, frequency, and duration. Stimulation frequencies between 2 and 4 Hz are thought to trigger the release of endorphins and enkephalin, whereas frequencies between 80 and 100 Hz may release gamma-aminobutyric acid, galanin, and dynorphin (Lundeberg & Stener-Victorin, 2002). The ideal needle placement for e-stim with DN has not been determined either. White and collagues (2000) recommended placing the needle-electrodes within the same dermatomes as the location of the lesion, whereas other clinicians recommended inserting two converging electrodes directly into a TrP at both sides of a TrP inside the taut band (Elorriaga 2000; Mayoral et al., 2004).

Summary