Central sensitization in spinal cord injured humans assessed by reflex receptive fields

Clinical Neurophysiology 125 (2014) 352–362

Contents lists available at SciVerse ScienceDirect

Clinical Neurophysiology journal homepage: www.elsevier.com/locate/clinph

Central sensitization in spinal cord injured humans assessed by reflex receptive fields José Alberto Biurrun Manresa a,⇑, Nanna Susanne Brix Finnerup b, Inger Lauge Johannesen c, Fin Biering-Sørensen d, Troels Staehelin Jensen b, Lars Arendt-Nielsen a, Ole Kæseler Andersen a a

Center for Sensory-Motor Interaction, Aalborg University, Denmark Danish Pain Research Center, Aarhus University Hospital, Denmark The Spinal Cord Unit, Viborg Hospital, Viborg, Denmark d Clinic for Spinal Cord Injuries, The Neuroscience Centre, Rigshospitalet and Faculty of Health Sciences, University of Copenhagen, Denmark b c

a r t i c l e

i n f o

Article history: Accepted 30 June 2013 Available online 22 July 2013 Keywords: Central sensitization Nociceptive withdrawal reflex Reflex receptive fields Descending control Spinal cord injury Capsaicin

h i g h l i g h t s  People with spinal cord injury can develop central sensitization, despite lacking supraspinal input and

altered spinal/supraspinal processing.  Reflex receptive fields are significantly larger and display a distinct different topography in spinal cord

injured volunteers compared to non-injured volunteers.  Protective plastic mechanisms may still be functional in people with spinal cord injury.

a b s t r a c t Objective: To investigate the effects of central sensitization, elicited by intramuscular injection of capsaicin, by comparing the reflex receptive fields (RRF) of spinally-intact volunteers and spinal cord injured volunteers that present presensitized spinal nociceptive mechanisms. Methods: Fifteen volunteers with complete spinal cord injury (SCI) and fourteen non-injured (NI) volunteers participated in the experiment. Repeated electrical stimulation was applied on eight sites on the foot sole to elicit the nociceptive withdrawal reflex (NWR). RRF were assessed before, 1 min after and 60 min after an intramuscular injection of capsaicin in the foot sole in order to induce central sensitization. Results: Both groups presented RRF expansion and lowered NWR thresholds immediately after capsaicin injection, reflected by the enlargement of RRF sensitivity areas and RRF probability areas. Moreover, the topography of the RRF sensitivity and probability areas were significantly different in SCI volunteers compared to NI volunteers in terms of size and shape. Conclusions: SCI volunteers can develop central sensitization, despite adaptive/maladaptive changes in synaptic plasticity and lack of supraspinal control. Significance: Protective plastic mechanisms may still be functional in SCI volunteers. Ó 2013 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction Most of the forms of synaptic plasticity that occur in the spinal cord in response to noxious stimuli, such as post-injury pain hypersensitivity (Woolf, 1983) and spinal long-term potentiation (Woolf and Salter, 2000), can be enclosed into the term central sensitization (Ji et al., 2003). Several types of experimental nociceptive ⇑ Corresponding author. Address: Center for Sensory-Motor Interaction, Department of Health Science & Technology, Aalborg University, Fredrik Bajers vej 7, 9220 Aalborg, Denmark. Tel.: +45 9940 8715; fax: +45 9815 4008. E-mail address: [email protected] (J.A. Biurrun Manresa).

activation can induce central sensitization; commonly used human sensitization models involve chemical irritation (e.g. capsaicin) or electrical stimulation of the skin (Simone et al., 1989; Treede et al., 1992; Magerl et al., 1998; Koppert et al., 2001; Klein et al., 2004; Geber et al., 2007). These surrogate models are commonly used in healthy volunteers to study the underlying mechanisms associated with central sensitization, aiming to extrapolate the findings to those cases where sensitization is present as part of pathophysiological pain disorders (Woolf, 2011). These models, however, have not been extensively studied in cases when there already exists a clinical condition that affects

1388-2457/$36.00 Ó 2013 International Federation of Clinical Neurophysiology. Published by Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.clinph.2013.06.186

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

the central nervous system. In particular, people with complete spinal cord injury (SCI) already present signs of sensitization, such as nociceptive reflex hyperexcitability and enlarged receptive fields (Andersen et al., 2004), so it is crucial to determine if the protective plastic mechanisms triggered after intense nociceptive activation are still functional (Latremoliere and Woolf, 2009). From a clinical perspective, although several maladaptive changes to sensory processing may contribute to the development of neuropathic pain in SCI patients, there is significant evidence indicating that central sensitization plays a prominent role (Tan and Waxman, 2012). Therefore, this reinforces the importance of the investigating the central mechanisms of pain in order to assess their suitability as targets for treatment (Woolf, 2011). Furthermore, a critical issue associated with spinal cord injuries is the partial or total loss of supraspinal control. Most of the early studies on supraspinal control mainly focused on descending inhibition, although it was later established that both inhibitory and facilitatory descending control mechanisms are involved in nociceptive modulation (Fields et al., 2006; Heinricher et al., 2009; Ossipov et al., 2010). In this regard, there is increasing evidence of a significant contribution of supraspinal influences to the development and maintenance of central sensitization (Urban and Gebhart, 1999; Sandkühler, 2009). Thus, it is highly relevant to investigate if central sensitization models can still be established in the presence of adaptive/maladaptive changes in synaptic plasticity and complete lack of supraspinal input. Central sensitization is often assessed using psychophysical measures (Raja et al., 1999; Klein et al., 2005), but these methods are subjective and cannot be applied in people that suffered complete sensory loss after SCI. However, it has been documented that central sensitization can be objectively assessed in humans by the nociceptive withdrawal reflex (NWR) (Biurrun Manresa et al., 2010b; Lim et al., 2011) and the reflex receptive fields (RRF) (Neziri et al., 2010b). Variations in the NWR and the RRF are likely to reflect changes in central processing of nociceptive activity, for instance after repetitive painful stimulation leading to temporal summation (Gozariu et al., 1997; Andersen et al., 2005), or increased excitability in clinical conditions (Banic et al., 2004; Neziri et al., 2010b; Lim et al., 2011). Moreover, descending modulation affects the RRF control following strong nociceptive input, since the responses through the reflex pathways are facilitated by central sensitization and this phenomenon depends on the site of injury and the degree of supraspinal control (Harris and Clarke, 2003). The aim of the present study was to investigate the effects of central sensitization, as elicited by intramuscular injection of capsaicin, in the presence of altered spinal/supraspinal nociceptive processing. Responses of volunteers with complete SCI and spinally-intact volunteers were objectively assessed and compared using the NWR and the RRF. Finally, a model describing the changes of the functional organization of the nociceptive reflex pathways during central sensitization is proposed and discussed.

2. Materials and methods 2.1. Volunteers Fifteen volunteers with clinically complete spinal cord injury (15 males, mean age 43 years, range 27–66, see Table 1 for details), classified as grade A according to the American Spinal Injury Association (ASIA) impairment scale (AIS) (Marino et al., 2003) with injuries between T6 and T12 to minimize the risk of autonomic dysreflexia (Helkowski et al., 2003) and fourteen spinally-intact, healthy volunteers (12 males and 2 females, mean age 23 years, range 19–28) participated in the experiment. These groups will

353

be referred to as SCI (spinal cord injured) and NI (non-injured), respectively. Written informed consent was obtained from all volunteers prior to participation and the Declaration of Helsinki was respected. The study was approved by the local ethical committee of the North Denmark Region, Denmark (approval no. VN20060029MCH). 2.2. Setup 2.2.1. Electrical stimulation Eight surface stimulation electrodes (15  15 mm, type Neuroline 700, Ambu A/S, Denmark) were mounted in a non-uniform grid on the plantar side of the right foot. Fig. 1 shows the spatial location of the stimulation electrodes, numbered from 1 to 8. One large common anode (70  100 mm, type Pals Platinum, Axelgaard Ltd., USA) placed on the dorsum of the foot ensured that the stimulus was perceived as coming from the sole of the foot. Thick epidermal layers on the sole of the foot of NI volunteers were ground off in order to reduce the effects of variation in skin thickness. Each train of pulses consisted of a 5 square-wave pulses of 1 ms width delivered at 200 Hz, generated by a computer-controlled constant-current stimulator (Noxitest IES 230, Aalborg, Denmark). The stimulation consisted of a burst of 8 trains of pulses delivered at 3 Hz, in order to elicit temporal summation (ArendtNielsen et al., 2000). In order to set the stimulation intensity, the NWR threshold had to be determined first. The NWR threshold was defined as the stimulation intensity that elicited EMG activity in the tibialis anterior muscle with an amplitude exceeding 20 lV for at least 5 ms in the 60–180 ms post-stimulation interval after a single stimulus in NI volunteers (Neziri et al., 2010a). In SCI volunteers, the quantification interval was extended to 250 ms after stimulation (Andersen et al., 2004). The final stimulation intensity was set as 0.8 times the average NWR threshold over sites 2, 4 and 6. This stimulation intensity was fixed for the rest of the experiment. The stimuli were delivered using a computer-controlled electrical relay, which consists of opto-electrical switches that sustain high voltages and hence provide extra security in the stimulator system. All relays were closed except in the short period when the stimulus was delivered, in which only the selected stimulation channel was opened (Biurrun Manresa et al., 2010a). All sites were stimulated 4 times in a randomized sequence for each condition (see Section 2.2.4 for details on the conditions), and the program delivered the stimuli at random time intervals so that the volunteers were not aware of when or where the stimulus was applied. The inter-stimulus interval ranged from 10 to 15 s for NI volunteers and at least 30 s for SCI volunteers in order to minimize habituation. Each assessment (i.e., 4 stimulations to each of the 10 sites) took between 15 and 25 min. 2.2.2. EMG recordings Activity in the tibialis anterior muscle was measured using surface EMG. Initially the skin was lightly abraded, and then two surface electrodes (30  22 mm, type Neuroline 720, Ambu A/S, Denmark) were placed along the muscle fiber direction over the muscle with an inter-electrode distance of 20 mm. The signal was amplified (up to 20,000 times), filtered (5–500 Hz, 2nd order), sampled (2000 Hz) and stored (3000 ms window including 200 ms of pre-stimulation activity). Fig. 1 shows example traces of EMG from SCI and NI volunteers. 2.2.3. Capsaicin injection A solution of 10 lg of capsaicin in 0.1 ml volume was injected into the flexor digitorum brevis muscle (flexor of the toes) in the central compartment of the foot. The tip of the needle was carefully wiped before the injection, to avoid leakage of capsaicin to the skin. None of the volunteers reported pain in the interval between

354

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

Table 1 Clinical characteristics of spinal cord injured volunteers (i.t.: intrathecal). Patient

Age

Neurological level

Years w/injury

Medication

Chronic pain

1 2 3 4 5

50 33 38 63 35

Th5 Th5 Th5 Th3 Th7

4 11 12 12 4

6

39

Th6

9

Visceral pain None None None Below level neuropathic pain At level neuropathic pain (evoked) Nociceptive shoulder pain

7 8 9 10 11 12

44 44 48 59 30 27

Th7 Th11 Th5 Th10 Th12 Th10

13 14 29 31 8

13

38

Th7

21

14

67

Th7

23

15

39

Th11

1.5

None Baclofen i.t. 8.2 mg/h Tizanidine 6 mg/day Oxybutynin 10 mg Nitrofurantoin 50 mg Baclofen i.t. 11.43 lg/h Baclofen 70 mg Tolterodine 5.6 mg Baclofen i.t. 51.4 lg/h Tolterodine 2.8 mg Tramadol 200 mg + Paracetamol 2000 mg None None Penicillin Diazepam 5 mg None Gabapentin 1200 mg Tolterodine 8.4 mg Diclofenac 100 mg Ketobemidone 25 mg Nitrazepam 10 mg Baclofen 50 mg Paracetamol 1000 mg Gabapentin 900 mg Oxycodone 15 mg Ibuprofen 1200 mg

Below level neuropathic pain None None None None At level neuropathic pain Nociceptive back pain At and below level neuropathic pain Nociceptive back pain Below level neuropathic pain Nociceptive back pain

Below level neuropathic pain Nociceptive back pain

Fig. 1. Nociceptive withdrawal reflex (NWR) setup. Left Position of the stimulation electrodes for reflex receptive fields (RRF) recording. Electrodes 1–3 were placed on the ball of the foot at the 1st, 3rd and 5th metatarsophalangeal joints, respectively. Electrodes 4–5 were placed on the medial and lateral arch of the sole of the foot, respectively. Electrode 6 was placed at the middle of the borderline between heel and arch of the foot. Electrodes 7 and 8 were positioned at the medial and lateral side of a horizontal line drawn in the middle of the heel, respectively. The cross indicates the site of the injection. Right Grand mean EMG traces recorded from the tibialis anterior muscle before the capsaicin injection (mean of all repetitions in all spinal cord injured (SCI) and non-injured (NI) volunteers), depicted for stimulation sites 2, 4 and 6. Signals are shown from 200 ms prior to stimulation until 2500 ms after stimulation. The vertical lines indicate stimuli onset.

the insertion of the needle and the infusion of the capsaicin solution, which indicates that leakage to the skin did not occur. Capsaicin administered to this location will establish central sensitization effects related to the plantar side of the foot. Intramuscular application ensured that no peripheral effects (primary hyper- or hypoalgesia) are evoked at the skin (and effectively none were reported by the healthy volunteers), and electrical stimulation allowed a selective central assessment, since it depolarizes free nerve endings, bypassing the receptor organs. Most NWR pathways will be influenced by sensitization (Harris and Clarke, 2003), affecting the response of most muscles in control of ankle dorsiflexion/plantarflexion and knee and hip flexion. From all these muscles, tibialis anterior was selected because it has a well-delimited RRF that does not cover the entire sole of the foot under normal conditions (Andersen, 2007), thus allowing the assessment of a potential expansion of the RRF during sensitization.

2.2.4. Experimental procedure Volunteers were placed in supine position with a back support, and a pillow was placed under the knee joint, resulting in a knee flexion of approximately 30°. Surface electromyography (EMG) and electrical stimulation electrodes were mounted as described above, and then the volunteers were thoroughly familiarized with electrical stimulation and pain intensity ratings (in the case of NI volunteers) to reduce any effects due to high initial vigilance and/or anxiety (French et al., 2005). Afterwards, the NWR threshold was determined for sites 2, 4 and 6, the stimulation intensity was determined, and baseline reflex responses were assessed. Subsequently, the capsaicin injection was applied, and reflexes were reassessed 1 min and 60 min after the injection. These three conditions are from now on referred to as before capsaicin, 1 min after capsaicin and 60 min after capsaicin. A visual analogue scale (VAS) was used to assess pain intensity ratings after the capsaicin

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

355

injection in the NI volunteers. The scale was anchored at 0 indicating detection threshold, at 3 indicating pain threshold, and at 10 indicating worst imaginable pain. The electrocardiogram and blood pressure were monitored regularly in all volunteers during the study, in order to ensure a normal reaction to the capsaicin injection.

sented as mean ± standard error of the mean (SEM) unless stated otherwise. P values smaller than 0.05 were regarded as significant.

2.3. Data analysis

Fig. 2 shows pain intensity ratings for NI volunteers after the capsaicin injection. Pain intensity ratings peaked between 20 and 40 s after the injection, with an average value of 7.2 ± 2.2 (mean ± SD) in the VAS scale. The ratings declined continuously to an average value of 2.3 ± 2.1 (mean ± SD) after 300 s. Pain intensity ratings were not quantified in SCI volunteers, since they did not feel the capsaicin injection.

2.3.1. RRF sensitivity maps EMG reflex responses were quantified using root–mean–square (RMS) amplitude in the 60–180 ms post-stimulation interval for each stimulus in the train for non-injured volunteers and in the 60–250 ms interval for SCI volunteers. RRF maps were derived for each stimulus in the train by two-dimensional interpolation of the corresponding RMS amplitudes. RRF sensitivity areas were calculated as the fraction of the sole of the foot delimited by a threshold set by the peak RMS amplitude minus two times the standard deviation of the remaining RMS amplitudes (Neziri et al., 2009). Thus, RRF sensitivity areas are a normalized measure (ranging from 0 to 1), that takes into account the spatial relationship between the NWR magnitude and the level of variation (standard deviation) of the NWR at the different sites.

3. Results 3.1. Pain intensity ratings

3.2. NWR thresholds Fig. 3 shows initial NWR thresholds for SCI and NI volunteers at three different sites on the sole of the foot. Group (F(1,27) = 6.247, p = 0.019) was a significant main effect in the RM ANOVA analysis for NWR thresholds, and there was a significant group  site interaction (F(2,26) = 9.123, p < 0.001). The post hoc analysis revealed

2.3.2. RRF probability maps The probability of occurrence of a NWR at each site was calculated as the ratio between the number of reflexes detected and the number of stimulations performed. NWR detection was performed using interval peak z-score with a threshold value of 12 (France et al., 2009; Biurrun Manresa et al., 2011). RRF probability maps were also derived for each stimulus in the train using the same interpolation methodology as for the RRF sensitivity maps. RRF probability areas were calculated as the fraction of the sole of the foot from which a NWR can be elicited by at least 60% of the stimulations (Biurrun Manresa et al., 2011). RRF probability areas are also a normalized measure across subjects (ranging from 0 to 1), and unlike measures of NWR magnitude (such as RMS amplitude), they are unaffected by volitional EMG activity elicited between stimuli during temporal summation stimulation paradigms. 2.3.3. RRF areas as quantitative measures RRF sensitivity and probability areas have two main advantages compared to traditional NWR magnitude analysis: they are intrinsically normalized measures (minimizing large inter-individual differences) and they are more robust against habituation [10]. RRF areas derived from the first stimulus were quantified as responses to single stimulation (SS), whereas RRF areas derived from the 2nd to 8th stimulus were averaged and quantified as responses to temporal summation (TS), since a preliminary analysis showed no differences between the areas elicited from these stimuli.

Fig. 2. Pain ratings during capsaicin injection. Injection time is at 0 s, and the vertical dashed line indicates the beginning of RRF assessment. Mean ± SD values across NI volunteers are shown.

2.4. Statistics NWR thresholds and RRF sensitivity and probability areas were compared using repeated measures analysis of variance (RM ANOVA). Site (2, 4 or 6) was used as within-subject factor for NWR thresholds, and stimulus type (SS or TS) and time (before, 1 min after or 60 min after capsaicin) were regarded as within-subject factors for RRF areas. In both cases, group (SCI or NI) was regarded as a between-subject factor. A multivariate approach was used whenever possible to bypass the compound symmetry and sphericity assumptions required in a classic RM ANOVA approach (Park et al., 2009). The appropriate Student–Newman–Keuls (SNK) statistics were used for post hoc pair-wise comparison when a main effect or an interaction was found significant. All values are pre-

Fig. 3. Nociceptive withdrawal reflex (NWR) thresholds at different sites of spinal cord injured (SCI) and non-injured (NI) volunteers, at the beginning of the experiment. Thresholds at site 4 were larger for SCI volunteers (⁄p < 0.05) and smaller for NI volunteers (⁄p < 0.05), compared to all other thresholds. Mean ± SEM values across all volunteers are shown.

356

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

that NWR threshold in SCI volunteers at site 4 was significantly higher than the thresholds at all other group  site combinations (SNK, p values ranging from 0.001 to 0.036), and that the NWR threshold in NI volunteers at site 4 was significantly lower than the thresholds at all other group  site combinations (SNK, p values ranging from 0.001 to 0.050).

3.3. RRF areas 3.3.1. RRF sensitivity areas Fig. 4a shows the RRF sensitivity areas for both groups at the three assessment times. The RM ANOVA analysis showed that time (F(2,26) = 5.979, p = 0.007), group (F(1,27) = 8.416, p = 0.007), and stimulus type (F(1,27) = 9.701, p = 0.004) were significant main effects, and there were no significant interactions. RRF sensitivity areas were larger 1 min after capsaicin than before and also larger than 60 min after capsaicin (SNK, p < 0.001 and p = 0.022 respectively), whereas the difference before and 60 min after capsaicin

was not significant (SNK, p = 0.149). Specifically for group, SCI volunteers displayed larger RRF sensitivity areas than NI volunteers (SNK, p = 0.007). Regarding stimulus type, temporal summation elicited significantly larger RRF sensitivity areas compared to single stimulation (SNK, p = 0.004). Fig. 5 depicts the grand mean RRF sensitivity maps of SCI and NI volunteers. It can be seen that the topographies of the RRF maps were reversed for SCI volunteers compared to the normal topography observed in NI volunteers. Indeed, the most sensitive area in NI volunteers was the arch of the foot, whereas the sensitivity of the RRF in SCI volunteers reflects a completely opposed behavior, i.e., the arch of the foot became the less sensitive area, and vice versa.

3.3.2. RRF probability areas Fig. 4b shows the RRF probability areas for both groups at the three assessment times. The RM ANOVA analysis showed that group (F(1,27) = 25.210, p < 0.001) and stimulus type (F(1,27) = 22.443, p < 0.001) were significant main effects, and there were

Fig. 4. Reflex receptive field (RRF) areas of spinal cord injured (SCI) and non-injured (NI) volunteers. (a) RRF sensitivity areas were larger 1 min after capsaicin than before (⁄⁄⁄p < 0.001) and 60 min after capsaicin (⁄p < 0.05). RRF sensitivity areas were also larger in SCI compared to NI volunteers (⁄⁄⁄p < 0.001). Finally, RRF sensitivity areas were larger during temporal summation (TS) compared to single stimulation (SS) (⁄⁄p < 0.01). (b) RRF probability areas were larger 1 min after capsaicin than before (⁄⁄p < 0.01) and 60 min after capsaicin (⁄p < 0.05) during SS. RRF probability areas were also larger in SCI compared to NI volunteers (⁄⁄⁄p < 0.001) during TS. Mean ± SEM values across all volunteers are shown.

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

two significant interactions: time  stimulus type (F(2,26) = 3.497, p = 0.045) and group  stimulus type (F(1,27) = 21.970, p < 0.001). The post hoc analysis of the time  stimulus type interaction revealed that RRF probability areas were larger 1 min after capsaicin than before and also larger than 60 min after capsaicin (SNK, p = 0.002 and p = 0.017 respectively) for SS, whereas there were no time differences during TS, albeit all RRF probability areas were larger during TS compared to SS (SNK, p < 0.001 for all comparisons). The post hoc analysis of the group  stimulus type interaction revealed that SCI volunteers displayed significantly larger

357

RRF probability areas during TS compared to all other group  stimulus type combinations (SNK, p < 0.001 for all comparisons). Fig. 6 depicts the grand mean RRF probability maps of SCI and NI volunteers. The same reversed topography can also be observed in this RRF maps. 4. Discussion This study showed for the first time that SCI volunteers can develop central sensitization, despite alterations in spinal

Fig. 5. Grand mean reflex receptive field (RRF) sensitivity maps of spinal cord injured (SCI) and non-injured (NI) volunteers. RRF maps for each of the eight stimuli in the train are shown before, 1 min and 60 min after the capsaicin injection. The black line delimits the reflex receptive field (RRF) sensitivity area. The white dots spot the electrical stimulation sites, and the white cross marks the injection site. Mean RMS amplitudes across all volunteers are shown.

358

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

nociception and lack of supraspinal control. Both SCI and healthy volunteers presented expansion of the RRF and lowered NWR thresholds immediately after capsaicin injection, as reflected in the enlargement of RRF sensitivity areas (p = 0.007) and RRF probability areas (p < 0.017). Moreover, RRF sensitivity and probability areas were significantly larger in SCI volunteers compared to NI volunteers, especially during temporal summation (p = 0.007 and p < 0.001 respectively). An interesting observation is the reversed topography of RRF maps in SCI volunteers compared to RRF maps of healthy volunteers in Figs. 5 and 6, indicating a pivotal role of

the descending control in maintaining the integrity and functional organization of spinal NWR pathways. 4.1. Central sensitization in SCI and healthy volunteers The injection of capsaicin provides a unique model to study the mechanisms of central sensitization in humans via a neurogenic inflammation (Simone et al., 1991), resembling the sensitization symptoms observed for instance in neuropathic pain (e.g., hyperalgesia, allodynia, enlargement of receptive fields) without any evi-

Fig. 6. Grand mean reflex receptive field (RRF) probability maps of spinal cord injured (SCI) and non-injured (NI) volunteers. RRF maps for each of the eight stimuli in the train are shown before, 1 min and 60 min after the capsaicin injection. The black line delimits the reflex receptive field (RRF) probability area. The white dots spot the electrical stimulation sites, and the white cross marks the injection site. Mean NWR probability of occurrence across all volunteers are shown.

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

dent tissue damage (Treede et al., 1992; Klein et al., 2005; Dombourian et al., 2006). Moreover, such effects are evident just seconds after the administration of the substance and may last up to a couple of hours, depending on the delivery method, the dosage and the site of application (Simone et al., 1989; Liu et al., 1998; Witting et al., 2000). As with all chemical stimulation methods, however, several factors need to be carefully considered in order to ensure proper reproducibility and to reduce inter-individual differences (Mørk et al., 2003). Central sensitization can be defined by an enhanced responsiveness of nociceptive neurons in the central nervous system to their normal and/or sub-threshold afferent input, i.e. increased response magnitudes and/or lower thresholds, as well as an enlargement of receptive fields (Cook et al., 1987; Loeser and Treede, 2008). Both phenomena were readily observed in this study. Firstly, the NWR thresholds were lowered in both groups immediately after the injection, as reflected by the increased number of NWR occurrences in response to the same stimulation intensity that can be observed in the RRF probability maps. Secondly, an expansion of the RRF sensitivity areas was observed in both groups right after the capsaicin injection in comparison to baseline. The topography of the RRF is not dramatically changed after central sensitization in any of the groups; instead, there seems to be a general sensitivity/excitability compared to the initial RRF assessments. This RRF enlargement may be explained by spinal hyperexcitability, likely as consequence of increased number of responsive spinal neurons or an expansion of the receptive fields of spinal neurons as a result of increased synaptic sensitivity (Cook et al., 1987; Dubner and Basbaum, 1994). As expected, this effect fades over time, since 60 min after the injection the RRF areas were significantly smaller compared to the areas assessed right after the injection. In healthy volunteers, this RRF behavior is a generalization of previous experiments showing facilitation of the NWR after topical application of capsaicin (Grönroos and Pertovaara, 1993; Andersen et al., 1996). Previous findings in animals following chemical irritation showed widespread reflex facilitation distal to the knee joint in decerebrated, spinal animals, whereas the facilitation was restricted to specific sites on the sole and ankle in spinally intact animals (Harris and Clarke, 2003). Early research on supraspinal control mainly focused on descending inhibitory control as a potential target for analgesic mechanisms, although later on it became evident that supraspinal control in the sensitization process can be both inhibitory and facilitatory (Fields et al., 2006; Heinricher et al., 2009; Ossipov et al., 2010). Descending facilitation of spinal neuronal responses has mainly been investigated in the rostroventral medulla, although it can be elicited (directly or indirectly onto spinal pathways) from other brain sites as well, such as the anterior cingulate cortex, the reticular nuclei and the periacqueductal grey (Sandkühler, 2009). Under physiological conditions, descending inhibition and facilitation can adapt the responsiveness of the nociceptive system to protect an injured site. Recent studies involving animal models of sensitization suggests that descending facilitation may have a very rapid onset (in the order of a few minutes), whereas descending inhibition shows a late occurrence, probably triggered by sufficient C-fiber afferent activity and temporal summation (You et al., 2010). Furthermore, there is increasing evidence to suggest that supraspinal mechanisms may contribute to central sensitization in the transition from acute to chronic pain (Sandkühler and Liu, 1998; Lemon and Griffiths, 2005; Andersen, 2007; Klein et al., 2007; Klauenberg et al., 2008; Heinricher et al., 2009; Wang et al., 2013). The results of this study show that humans with impaired supraspinal control can develop central sensitization. They also indicate that even in the case where central sensitization is already present, as demonstrated by enlarged reflex responses in SCI volunteers at baseline, there still remains a certain degree of plasticity in the spinal synapses to manifest additional effects of sensitization. This

359

is an indication that some protective mechanisms may still function, in which the threshold for nociceptive activation falls and subsequent responses to nociceptive input are amplified (Latremoliere and Woolf, 2009). These results are therefore highly relevant for further understanding of the mechanisms behind central sensitization, since most of the research performed on these topics comes from animal models, in which it is hard to mimic the pathophysiologic mechanisms and the clinical symptomology of spinal cord injuries in humans (Hulsebosch, 2003). Furthermore, the NWR and RRF provide a valuable tool for the objective assessment of these models, since most common methods for assessing sensitization, such as perceptual correlates after chemical, mechanical or thermal stimuli, rely on intact perception and subjective evaluation of the stimulation from the subjects (Latremoliere and Woolf, 2009), and thus cannot be applied in SCI volunteers suffering from complete sensory loss (Andersen et al., 2004). 4.2. RRF differences between SCI and NI volunteers The NWR has been used to investigate differences in spinal nociception between SCI and NI volunteers, often related to the influence of supraspinal control (Hornby et al., 2003). After spinal cord transection, the NWR becomes larger and changes into a stereotyped flexor pattern with flexion of all joints (Grimby, 1963b; Dimitrijevic´ and Nathan, 1968). Moreover, in accordance with previous evidence gathered from animal and human experiments, the RRF expands when supraspinal control is impaired and/or in the presence of hyperexcited spinal neurons (Grimby, 1963a; Schouenborg et al., 1992; Schmit et al., 2003; Andersen et al., 2004; Dietz and Sinkjaer, 2007). However, abnormal or expanded receptive fields are rather vague descriptions, and a more detailed depiction of the sensitivity variations in receptive fields is required to understand the changes in spinal nociception after the injury. In the present experiment, NWR thresholds were higher in SCI volunteers compared to NI, in agreement with previous studies (Hiersemenzel et al., 2000; Andersen et al., 2004). Moreover, the arch of the foot presented the highest thresholds in SCI volunteers and the lowest thresholds in NI volunteers. RRF assessment also showed striking differences between SCI and NI volunteers. Indeed, SCI volunteers not only presented larger RRF areas, but also a reversed RRF topography compared to RRF of NI volunteers (Fig. 5). The sensitivity of the RRF is shaped by excitatory and inhibitory spinal neuronal circuits under supraspinal influence, among other factors (Schouenborg, 2002). In chronic SCI volunteers, the loss of descending control and appropriate peripheral input causes severe changes on the spinal circuitry (Dietz, 2010), which eventually leads to abnormal RRF configurations. In an attempt to explain the findings of this study, an integrative model of the functional organization of the RRF is proposed, presented in Fig. 7, accounting for the functional organization of the NWR under normal physiological conditions (Fig. 7a) (Schouenborg, 2002; Schouenborg, 2004; Andersen, 2007). The proposed model would also account for changes due to central sensitization, in which increased afferent activity (usually through nociceptive C fibres) causes the neurons in the dorsal horn to become hyperexcitable and as a consequence the RRF would be enlarged, primarily by spinal mechanisms (Fig. 7b) (Sandkühler, 2009; Biurrun Manresa et al., 2010b; Woolf, 2011). However, supraspinal control may additionally increase facilitation and/or decrease inhibition in spinally intact subjects, modulating the effects of sensitization (Sandkühler, 2009; Biurrun Manresa et al., 2010b), although it should not be strictly required (as evidenced by the RRF changes in SCI volunteers observed in this study). The role of descending activity in maintaining the RRF topography and the reversal effect observed in SCI volunteers would also be accounted for, since it is hypothesized that neural circuits that were subjected to tonic

360

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

Fig. 7. Functional organization of the nociceptive withdrawal reflex (NWR) pathways. (a) In healthy volunteers, the reflex receptive field (RRF) of the tibialis anterior (TA) muscle is characterized by a high sensitivity (+Sens) in the medial, distal region, resulting in inversion (Inv) and dorsiflexion (DorFl) of the foot when these sites are activated. Functionally antagonist muscles (not shown) have RRF that evoke plantarflexion (PlanFl) or eversion (Eve). The RRF is likely shaped by excitatory (Excit) and inhibitory (Inhib) descending control input coming from supraspinal structures (SupCtrl). SupCtrl may act presynaptically (not shown) or postsynaptically on one or more interneurons (In) and on reflex encoders (RE) in the NWR pathways modifying their excitability (color-coded similarly to RRF), adjusting the weight of afferent information from nociceptive input. The net output is translated by a-motoneurons (a-Mn) into efferent signals that evoke a proper contraction in the target muscle. (b) During central sensitization, increased afferent activity (usually through nociceptive C fibres) causes the In and/or RE neurons to become hyperexcitable. Additionally, SupCtrl may increase facilitation and/or decrease inhibition, and as a consequence the RRF is enlarged. (c) After an injury to the spinal cord, SupCtrl is partially or totally lost, so in that were subjected to tonic inhibitory descending signals increase their excitability due to disinhibition and vice versa, resulting in a reversed topography of RRF maps. Nevertheless, this abnormal RRF maps can still expand during central sensitization (as can be seen in Fig. 7). Please note that this is a functional model of the RRF; for a full description of all possible physiological configurations of the In, RE and SupCtrl neural networks, please refer to (Sandkühler, 2009; Heinricher et al., 2009(For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)).

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

inhibitory descending signals (low sensitivity RRF regions, illustrated in blue in Fig. 7a) increase their excitability due to disinhibition (and turn into high sensitivity RRF regions, illustrated in red in Fig. 7c) and vice versa, i.e., the tonic descending facilitation on the arch region of the RRF is lost, and henceforth the arch exhibits lower excitability in SCI volunteers. 4.3. Temporal summation and central sensitization Temporal summation can be described as successive increases in perceived pain intensity to repeated stimuli. As a physiological correlate, the NWR has proven to be a particularly useful tool (Arendt-Nielsen et al., 1994; Arendt-Nielsen et al., 2000; Guirimand et al., 2000; Serrao et al., 2004). In healthy volunteers, it is characterized by a gradual increase in size and duration of the NWR for a few seconds following repetitive stimulation depending on stimulus intensity and frequency (Arendt-Nielsen et al., 2000), after which the NWR size reaches a plateau or even decreases, probably due to descending inhibitory control (Bajaj et al., 2005). Temporal summation bears resemblance to the early part of the wind-up process that involves summation of excitatory post-synaptic potentials (Randic, 1996). The term wind-up is mostly used for C fiber evoked responses, although there is evidence that it can also be evoked from A-delta fibers (Thompson et al., 1990; Sivilotti et al., 1993; Kimura and Kontani, 2008). Furthermore, there is an ongoing debate on whether temporal summation should be encompassed within the term central sensitization (Ji et al., 2003) or not (Sandkühler, 2009); in any case, a lowered temporal summation threshold frequency or an enhanced temporal summation response could be an indicator of central sensitization (Arendt-Nielsen et al., 1996). The present findings show that RRF sensitivity and probability areas were larger in SCI than in NI volunteers during central sensitization. This effect was clearly more pronounced for RRF sensitivity areas during temporal summation, as descending inhibition seems to be one of the factors regulating this phenomenon. Indeed, when inhibition is lost (as with SCI volunteers), the NWR can be elicited during temporal summation from virtually every site with a very high probability of occurrence (Fig. 6). These results expand the knowledge of previous findings showing enlarged RRF in SCI compared to NI volunteers at different stimulation intensities but using a single electrical stimulus (Andersen et al., 2004), as well as gradual enlargement of the RRF in response to repetitive stimulation in NI volunteers (Andersen et al., 2005). Indeed, spinal facilitatory effects (including temporal summation and central sensitization) can be masked or even completely overridden by supraspinal inhibitory processes triggered after high intensity stimulation (Gozariu et al., 1997; Bajaj et al., 2005; Biurrun Manresa et al., 2010b), as also demonstrated by this differential effect between NI and SCI volunteers. 4.4. Limitations A potential limitation in the study is the presence of chronic pain in some of the SCI volunteers. Within this group, eight volunteers presented different forms of chronic pain whereas the remaining volunteers did not. A preliminary analysis revealed no differences in terms of the outcome variables between these subgroups when chronic pain was used as a predictor. Indeed, to date there is no consensus regarding which are the mechanisms underlying the development of chronic pain in SCI volunteers or its effect on the NWR (Finnerup et al., 2003; Finnerup and Jensen, 2004; Finnerup et al., 2007; Wasner et al., 2008). In the light of these results, all SCI volunteers were pooled into the same group for the subsequent analysis. Another important factor in RRF assessment is the skin thickness on the sole of the foot. In previous studies on healthy volunteers, it has been shown that sensitivity at each

361

electrode site is correlated to skin thickness at that site; hence the stimulation intensities at each site were titrated to pain thresholds in order to attempt to provide equal input to the spinal cord (Andersen, 2007). However, SCI volunteers cannot report pain ratings to electrical stimulation; furthermore, they display very little variation in skin thickness because of the low shear/load on the sole in absence of gait (Andersen et al., 2004). Consequently, the same stimulation intensity was used for all sites. Other factors were also analyzed (e.g. the mean age between groups), but the quantitative impact of this factor was negligible for the NWR, especially after repeated stimulation (Neziri et al., 2010a). 5. Conclusion The results of this study demonstrate that people with spinal cord injury can develop central sensitization despite lacking supraspinal input and altered spinal/supraspinal processing, indicating that some protective plastic mechanisms may still be functional. Funding sources The Danish Research Council for Technology and Production Sciences (FTP). Conflict of interest The authors declare that there is no conflict of interest related to this article. References Andersen OK. Studies of the organization of the human nociceptive withdrawal reflex. Focus on sensory convergence and stimulation site dependency. Acta Physiol 2007;189:1–35. Andersen OK, Felsby S, Nicolaisen L, Bjerring P, Jensen TS, Arendt-Nielsen L. The effect of Ketamine on stimulation of primary and secondary hyperalgesic areas induced by capsaicin – a double-blind, placebo-controlled, human experimental study. Pain 1996;66:51–62. Andersen OK, Finnerup NB, Spaich EG, Jensen TS, Arendt-Nielsen L. Expansion of nociceptive withdrawal reflex receptive fields in spinal cord injured humans. Clin Neurophysiol 2004;115:2798–810. Andersen OK, Spaich EG, Madeleine P, Arendt-Nielsen L. Gradual enlargement of human withdrawal reflex receptive fields following repetitive painful stimulation. Brain Res 2005;1042:194–204. Arendt-Nielsen L, Andersen OK, Jensen TS. Brief, prolonged and repeated stimuli applied to hyperalgesic skin areas: a psychophysical study. Brain Res 1996;712:165–7. Arendt-Nielsen L, Brennum J, Sindrup S, Bak P. Electrophysiological and psychophysical quantification of temporal summation in the human nociceptive system. Eur J Appl Physiol O 1994;68:266–73. Arendt-Nielsen L, Sonnenborg FA, Andersen OK. Facilitation of the withdrawal reflex by repeated transcutaneous electrical stimulation: An experimental study on central integration in humans. Eur J Appl Physiol O 2000;81:165–73. Bajaj P, Arendt-Nielsen L, Andersen OK. Facilitation and inhibition of withdrawal reflexes following repetitive stimulation: electro- and psychophysiological evidence for activation of noxious inhibitory controls in humans. Eur J Pain 2005;9:25–31. Banic B, Petersen-Felix S, Andersen OK, Radanov BP, Villiger PM, Arendt-Nielsen L, et al. Evidence for spinal cord hypersensitivity in chronic pain after whiplash injury and in fibromyalgia. Pain 2004;107:7–15. Biurrun Manresa JA, Hansen J, Andersen OK. Development of a data acquisition and analysis system for nociceptive withdrawal reflex and reflex receptive fields in humans. Conf Proc IEEE Eng Med Biol Soc 2010a;2010:6619–24. Biurrun Manresa JA, Jensen MB, Andersen OK. Introducing the reflex probability maps in the quantification of nociceptive withdrawal reflex receptive fields in humans. J Electromyogr Kinesiol 2011;21:67–76. Biurrun Manresa JA, Mørch CD, Andersen OK. Long-term facilitation of nociceptive withdrawal reflexes following low-frequency conditioning electrical stimulation: a new model for central sensitization in humans. Eur J Pain 2010b;14:822–31. Cook AJ, Woolf CJ, Wall PD, McMahon SB. Dynamic receptive field plasticity in rat spinal cord dorsal horn following C-primary afferent input. Nature 1987;325:151–3. Dietz V. Behavior of spinal neurons deprived of supraspinal input. Nat Rev Neurol 2010;6:167–74.

362

J.A. Biurrun Manresa et al. / Clinical Neurophysiology 125 (2014) 352–362

Dietz V, Sinkjaer T. Spastic movement disorder: impaired reflex function and altered muscle mechanics. Lancet Neurol 2007;6:725–33. Dimitrijevic´ MR, Nathan PW. Studies of spasticity in man: 3. Analysis of reflex activity evoked by noxious cutaneous stimulation. Brain 1968;91:349–68. Dombourian MG, Turner NA, Gerovac TA, Vemuganti R, Miranpuri GS, Türeyen K, et al. B1 and TRPV-1 receptor genes and their relationship to hyperalgesia following spinal cord injury. Spine 2006;31:2778–82. Dubner R, Basbaum AI. Spinal dorsal horn plasticity following tissue or nerve injury. In: Wall PD, Melzack R, editors. Textbook of pain. 3rd ed. Edinburgh: Churchill Livingstone; 1994. p. 225–41. Fields HL, Basbaum AI, Heinricher MM. Central nervous system mechanisms of pain modulation. In: McMahon S, Koltzenburg M, editors. Wall and melzack’s textbook of pain. 5th ed. London: Elsevier; 2006. p. 125–42. Finnerup NB, Gyldensted C, Nielsen E, Kristensen AD, Bach FW, Jensen TS. MRI in chronic spinal cord injury patients with and without central pain. Neurology 2003;61:1569–75. Finnerup NB, Jensen TS. Spinal cord injury pain – mechanisms and treatment. Eur J Neurol 2004;11:73–82. Finnerup NB, Sørensen L, Biering-Sørensen F, Johannesen IL, Jensen TS. Segmental hypersensitivity and spinothalamic function in spinal cord injury pain. Exp Neurol 2007;207:139–49. France CR, Rhudy JL, McGlone S. Using normalized EMG to define the nociceptive flexion reflex (NFR) threshold: further evaluation of standardized NFR scoring criteria. Pain 2009;145:211–8. French DJ, France CR, France JL, Arnott LF. The influence of acute anxiety on assessment of nociceptive flexion reflex thresholds in healthy young adults. Pain 2005;114:358–63. Geber C, Fondel R, Krämer HH, Rolke R, Treede RD, Sommer C, et al. Psychophysics, flare, and neurosecretory function in human pain models: capsaicin versus electrically evoked pain. J. Pain 2007;8:503–14. Gozariu M, Bragard D, Willer JC, Le Bars D. Temporal summation of C-fiber afferent inputs: competition between facilitatory and inhibitory effects on C-fiber reflex in the rat. J Neurophysiol 1997;78:3165–79. Grimby L. Normal plantar response: integration of flexor and extensor reflex components. J Neurol Neurosurg Psychiatr 1963a;26:39–50. Grimby L. Pathological plantar response: disturbances of the normal integration of flexor and extensor reflex components. J Neurol Neurosurg Psychiatr 1963b;26:314–21. Grönroos M, Pertovaara A. Capsaicin-induced central facilitation of a nociceptive flexion reflex in humans. Neurosci Lett 1993;159:215–8. Guirimand F, Dupont X, Brasseur L, Chauvin M, Bouhassira D. The effects of ketamine on the temporal summation (wind-up) of the R(III) nociceptive flexion reflex and pain in humans. Anesth Analg 2000;90:408–14. Harris J, Clarke RW. Organisation of sensitisation of hind limb withdrawal reflexes from acute noxious stimuli in the rabbit. J Physiol (Lond) 2003;546:251–65. Heinricher MM, Tavares I, Leith JL, Lumb BM. Descending control of nociception: Specificity, recruitment and plasticity. Brain Res Rev 2009;60:214–25. Helkowski WM, Ditunno Jr JF, Boninger M. Autonomic dysreflexia: incidence in persons with neurologically complete and incomplete tetraplegia. J Spinal Cord Med 2003;26:244–7. Hiersemenzel LP, Curt A, Dietz V. From spinal shock to spasticity: neuronal adaptations to a spinal cord injury. Neurology 2000;54:1574–82. Hornby TG, Rymer WZ, Benz EN, Schmit BD. Windup of flexion reflexes in chronic human spinal cord injury: a marker for neuronal plateau potentials? J Neurophysiol 2003;89:416–26. Hulsebosch CE. Central sensitization and pain after spinal cord injury. Semin Pain Med 2003;1:159–70. Ji RR, Kohno T, Moore KA, Woolf CJ. Central sensitization and LTP: do pain and memory share similar mechanisms? Trends Neurosci 2003;26:696–705. Kimura S, Kontani H. Separate recording of A-delta and C fiber-mediated nociceptive flexor reflex responses of mouse hindlimb using electromyography and the characteristics of wind-up appearing in the responses. J Pharmacol Sci 2008;108:172–8. Klauenberg S, Maier C, Assion HJ, Hoffmann A, Krumova EK, Magerl W, et al. Depression and changed pain perception: hints for a central disinhibition mechanism. Pain 2008;140:332–43. Klein T, Magerl W, Hopf HC, Sandkühler J, Treede RD. Perceptual correlates of nociceptive long-term potentiation and long-term depression in humans. J Neurosci 2004;24:964–71. Klein T, Magerl W, Rolke R, Treede RD. Human surrogate models of neuropathic pain. Pain 2005;115:227–33. Klein T, Magerl W, Treede RD. Forget about your chronic pain. Pain 2007;132:16–7. Koppert W, Dern SK, Sittl R, Albrecht S, Schüttler J, Schmelz M. A new model of electrically evoked pain and hyperalgesia in human skin: the effects of intravenous alfentanil, S(+)-ketamine, and lidocaine. Anesthesiol 2001;95:395–402. Latremoliere A, Woolf CJ. Central sensitization: a generator of pain hypersensitivity by central neural plasticity. J Pain 2009;10:895–926. Lemon RN, Griffiths J. Comparing the function of the corticospinal system in different species: organizational differences for motor specialization? Muscle Nerve 2005;32:261–79. Lim ECW, Sterling M, Stone A, Vicenzino B. Central hyperexcitability as measured with nociceptive flexor reflex threshold in chronic musculoskeletal pain: a systematic review. Pain 2011;152:1811–20.

Liu M, Max MB, Robinovitz E, Gracely RH, Bennett GJ. The human capsaicin model of allodynia and hyperalgesia: sources of variability and methods for reduction. J Pain Symptom Manag 1998;16:10–20. Loeser JD, Treede RD. The kyoto protocol of IASP basic pain terminology. Pain 2008;137:473–7. Magerl W, Wilk SH, Treede RD. Secondary hyperalgesia and perceptual wind-up following intradermal injection of capsaicin in humans. Pain 1998;74:257–68. Marino RJ, Barros T, Biering-Sorensen F, Burns SP, Donovan WH, Graves DE, et al. International standards for neurological classification of spinal cord injury. J Spinal Cord Med 2003;26(Suppl. 1). Mørk H, Ashina M, Bendtsen L, Olesen J, Jensen R. Experimental muscle pain and tenderness following infusion of endogenous substances in humans. Eur J Pain 2003;7:145–53. Neziri AY, Andersen OK, Petersen-Felix S, Radanov B, Dickenson AH, Scaramozzino P, et al. The nociceptive withdrawal reflex: normative values of thresholds and reflex receptive fields. Eur J Pain 2010a;14:134–41. Neziri AY, Curatolo M, Bergadano A, Petersen-Felix S, Dickenson A, Arendt-Nielsen L, et al. New method for quantification and statistical analysis of nociceptive reflex receptive fields in humans. J Neurosci Methods 2009;178:24–30. Neziri AY, Haesler S, Petersen-Felix S, Müller M, Arendt-Nielsen L, Manresa JB, et al. Generalized expansion of nociceptive reflex receptive fields in chronic pain patients. Pain 2010b;151:798–805. Ossipov MH, Dussor GO, Porreca F. Central modulation of pain. J Clin Invest 2010;120:3779–87. Park E, Cho M, Ki C-S. Correct use of repeated measures analysis of variance. Kor J Lab Med 2009;29:1–9. Raja SN, Meyer A, Ringkamp M, Campbell JN. Peripheral neural mechanisms of nociception. In: Wall PD, Melzack R, editors. Textbook of pain. Churchill Livingstone; 1999. Randic M. Plasticity of excitatory synaptic transmission in the spinal cord dorsal horn. Prog Brain Res 1996;113:463–506. Sandkühler J. Models and mechanisms of hyperalgesia and allodynia. Physiol Rev 2009;89:707–58. Sandkühler J, Liu X. Induction of long-term potentiation at spinal synapses by noxious stimulation or nerve injury. Eur J Neurosci 1998;10:2476–80. Schmit BD, Hornby TG, Tysseling-Mattiace VM, Benz EN. Absence of local sign withdrawal in chronic human spinal cord injury. J Neurophysiol 2003;90:3232–41. Schouenborg J. Learning in sensorimotor circuits. Curr Opin Neurobiol 2004;14:693–7. Schouenborg J. Modular organisation and spinal somatosensory imprinting. Brain Res Rev 2002;40:80–91. Schouenborg J, Holmberg H, Weng HR. Functional organization of the nociceptive withdrawal reflexes. Exp Brain Res 1992;90:469–78. Serrao M, Rossi P, Sandrini G, Parisi L, Amabile GA, Nappi G, et al. Effects of diffuse noxious inhibitory controls on temporal summation of the RIII reflex in humans. Pain 2004;112:353–60. Simone DA, Baumann TK, LaMotte RH. Dose-dependent pain and mechanical hyperalgesia in humans after intradermal injection of capsaicin. Pain 1989;38:99–107. Simone DA, Sorkin LS, Oh U, Chung JM, Owens C, LaMotte RH, et al. Neurogenic hyperalgesia: central neural correlates in responses of spinothalamic tract neurons. J Neurophysiol 1991;66:228–46. Sivilotti LG, Thompson SWN, Woolf CJ. Rate of rise of the cumulative depolarization evoked by repetitive stimulation of small-caliber afferents is a predictor of action potential windup in rat spinal neurons in vitro. J Neurophysiol 1993;69:1621–31. Tan AM, Waxman SG. Spinal cord injury, dendritic spine remodeling, and spinal memory mechanisms. Exp Neurol 2012;235:142–51. Thompson SWN, King AE, Woolf CJ. Activity-dependent changes in rat ventral horn neurons in vitro; Summation of prolonged afferent evoked postsynaptic depolarizations produce a D-2-amino-5-phosphonovaleric acid sensitive windup. Eur J Neurosci 1990;2:638–49. Treede RD, Meyer RA, Raja SN, Campbell JN. Peripheral and central mechanisms of cutaneous hyperalgesia. Prog Neurobiol 1992;38:397–421. Urban MO, Gebhart GF. Supraspinal contributions to hyperalgesia. Proc Natl Acad Sci USA 1999;96:7687–92. Wang R, King T, De Felice M, Guo W, Ossipov MH, Porreca F. Descending facilitation maintains long-term spontaneous neuropathic pain. J Pain 2013. Wasner G, Lee BB, Engel S, McLachlan E. Residual spinothalamic tract pathways predict development of central pain after spinal cord injury. Brain 2008;131:2387–400. Witting N, Svensson P, Gottrup H, Arendt-Nielsen L, Jensen TS. Intramuscular and intradermal injection of capsaicin: a comparison of local and referred pain. Pain 2000;84:407–12. Woolf CJ. Central sensitization: implications for the diagnosis and treatment of pain. Pain 2011;152:S2–S15. Woolf CJ. Evidence for a central component of post-injury pain hypersensitivity. Nature 1983;306:686–8. Woolf CJ, Salter MW. Neuronal plasticity: increasing the gain in pain. Science 2000;288:1765–8. You HJ, Lei J, Sui MY, Huang L, Tan YX, Tjølsen A, et al. Endogenous descending modulation: spatiotemporal effect of dynamic imbalance between descending facilitation and inhibition of nociception. J Physiol 2010;588:4177–88.