Effect of local anesthesia on atypical odontalgia – A randomized controlled trial

Pain 122 (2006) 306–314 www.elsevier.com/locate/pain

Effect of local anesthesia on atypical odontalgia – A randomized controlled trial Thomas List

a,*

¨ ster d, Peter Svensson , Go¨ran Leijon b, Martti Helkimo c, Anders O

e,f

a Orofacial Pain Unit, Faculty of Odontology, Malmo¨ University, Malmo¨, Sweden Division of Neurology, Department of Neuroscience and Locomotion, University Hospital, Linko¨ping, Sweden Department of Stomatognathic Physiology, The Institute for Postgraduate Dental Education, Jo¨nko¨ping, Sweden d Department of Stomatognathic Physiology, County Hospital, Kalmar, Sweden e Department of Clinical Oral Physiology, School of Dentistry, University of Aarhus, Denmark f Department of Oral Maxillofacial Surgery, Aarhus University Hospital, Aarhus, Denmark

b c

Received 13 June 2005; received in revised form 13 January 2006; accepted 1 February 2006

Abstract The aim of the study was to evaluate the analgesic effect of lidocaine in a double-blind, controlled multi-center study on patients with atypical odontalgia (AO) – a possible orofacial neuropathic pain condition. Thirty-five consecutive AO patients (range 31–81 years) with a mean pain duration of 7.2 years (range 1–30 years) were recruited from four different orofacial pain clinics in Sweden. In a randomized cross-over design, 1.5 ml local anesthesia (20 mg/ml lidocaine and 12.5 lg/ml adrenaline) or 1.5 ml saline (9 mg/ml NaCl solution) (placebo) was injected to block the painful area. The VAS pain scores showed an overall effect of time (ANOVA: P < 0.001) and treatment (ANOVA: P = 0.018) with a significant interaction between the factors (ANOVA: P < 0.001). Overall, VAS pain relief was significantly greater at 15–120 min following the lidocaine injections compared to the placebo injections (Tukey: P < 0.05). All patients demonstrated significant disturbances in somatosensory function on the painful side compared to the non-painful side as revealed by quantitative sensory tests, however, only one significant inverse correlation was found between percentage pain relief and the magnitude of brush-evoked allodynia (Spearman: P < 0.01). In conclusion, AO patients experienced significant, but not complete, pain relief from administration of local anesthetics compared with placebo. The findings indicate that the spontaneous pain in AO patients only to some extent is dependent on peripheral afferent inputs and that sensitization of higher order neurons may be involved in the pathophysiology of AO.  2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. Keywords: Atypical facial pain; Lidocaine; Neuropathic pain; RCT; Trigeminal pain mechanisms

1. Introduction Atypical odontalgia (AO) is a severe persistent pain condition located in the teeth and the jaws (Melis et al., 2003) and appears to share many of the characteristics of atypical facial pain (Woda et al., 2005).

*

Corresponding author. Tel.: +46 406658424; fax: +46 406658420. E-mail address: [email protected] (T. List).

Although the clinical manifestations of other persistent pain conditions in the face such as temporomandibular disorders (TMD) have been reasonably well described (Svensson and Graven-Nielsen, 2001), relatively few studies have systematically evaluated AO. Patients with AO are often described as having an intense and constant pain in the teeth or in areas that have formerly had teeth, but there are currently no evidence-based diagnostic criteria available for AO (Forsell and Svensson, in press) although AO and atypical facial pain

0304-3959/$32.00  2006 International Association for the Study of Pain. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.pain.2006.02.005

T. List et al. / Pain 122 (2006) 306–314

recently have been proposed to be labeled ‘‘undifferentiated orofacial pain’’ (Woda et al., 2005). Some authors have asserted, but not yet provided supporting evidence for, that AO is a neuropathic pain condition in the oral cavity (Graff-Radford and Solberg, 1992; Vickers and Cousins, 2000; Woda and Pionchon, 2001). Neuropathic pain is defined by IASP as a chronic pain condition caused by a lesion or dysfunction of peripheral or central afferent pathways in the nervous system (Jensen, 2002) although the definition and in particular the term ‘‘dysfunction’’ are currently under intense debate (Hansson, 2002). A frequent characteristic of neuropathic pain conditions is changes in somatosensory sensitivity, for example, hypo- or hyperesthesia, hypo- or hyperalgesia, windup-like pain, or after-sensations, which may be revealed by quantitative sensory testings (QST) (Jensen et al., 2001; Hansson, 2002). Nerve injury, which occurs during endodontic procedures and tooth extractions, may be associated with the development of persistent neuropathic orofacial pain (Lynch and Elgeneidy, 1996). Indeed, animal studies have demonstrated both morphological and structural changes in axonal population and functional changes in second order brainstem neurons following deafferentation of the tooth pulp (Hu et al., 1999; Sessle, 2000). Although, these changes normally are reversible in animal preparations, tooth pulp deafferentation could potentially constitute a risk factor for subsequent development of orofacial neuropathic pain in susceptible individuals, but controlled studies are needed to address this hypothesis. Neuropathic pain conditions are complex and involve peripheral sensitization and neuronal plasticity of the central and peripheral systems (Sessle, 2000). Clinically, these neuronal mechanisms may be difficult to dissect out but QST and simple diagnostic procedures such as the utilization of local anesthetic blocks may assist both researchers and clinicians in getting a better description of the clinical manifestation of a painful condition. Patients with likely neuropathic orofacial pain complaints may demonstrate a wide range of responses to topical anesthesia or local anesthetic (LA) blocks, for example either complete, partial, or no pain reduction to somatosensory blocks (GraffRadford and Solberg, 1992; Vickers et al., 1998). Few studies have, however, been adequately controlled and there is a need for randomized controlled trials (RCT). The overall hypothesis in the present RCT study is that AO is a neuropathic pain condition with signs of disturbances in somatosensory sensitivity (sensory tests) in the painful region. The specific aim was to determine the pain relieving effect of a peripheral LA block in AO patients in order to assess the contribution of peripheral afferent inputs to the AO pain.

307

2. Materials and methods 2.1. Participants Thirty-five consecutive AO patients, 4 men and 31 women, median age 55.8 years (range 31–81 years) with a mean pain duration of 7.2 years (range 1–30 years) participated in the study. The patients were recruited from four different orofacial pain clinics in Sweden: Jo¨nko¨ping (n = 5), Kalmar (n = 9), Linko¨ping (n = 8), and Malmo¨ (n = 13). Inclusion criteria were pain located in a region where a tooth had been endodontically or surgically treated, chronic pain for more than 6 months, pain that had no pathological cause detectable in clinical or radiological examinations, and pain intensity P3 on a 0–10 numerical rating scale. Exclusion criteria were trigeminal neuralgia, herpes zoster, apical periodontitis, maxillary sinusitis, cluster headache, paroxysmal hemicrania, drug allergy, pregnancy, and breastfeeding. The Regional Ethical Review Board at Linko¨ping University Hospital approved the study, and all patients signed an informed, written consent. The patients received no monetary compensation. 2.2. Clinical examination The assessment comprised an intraoral evaluation of the teeth and oral mucosa that included inspection, palpation, percussion, electric pulp testing, periodontal probing, and translumination. In addition, radiographic examinations of the jaws and teeth (panorama and periapical radiographs), an examination of the masticatory system according to the Research Diagnostic Criteria for TMD (RDC/TMD) (Dworkin and LeResche, 1992), and a cervical spine examination were made by an experienced orofacial pain specialist (TL). The dental examination and the examination of the masticatory system will be the subject of a separate report. 2.3. Neurological and somatosensory examination All patients underwent a neurological examination (GL) of the cranial nerves and a chair-side somatosensory examination of the oral mucosa and the teeth. In all qualitative and quantitative somatosensory examinations, the results from the painful side were compared with those from the contralateral side. The pain-free side was always examined first. 2.3.1. Somatosensory examination Since there are no validated guidelines for examinations of the intraoral somatosensory sensitivity (Svensson et al., 2004), a brief description of the applied qualitative and quantitative procedures will be provided. Patients were asked to report their experience of spontaneous pain, dysesthesia, and paresthesia as well as of stimulus-dependent pain evoked by natural intraoral stimuli. The outcome was either ‘‘absent’’ or ‘‘present’’. A comparison of the sensitivity between the painful and the contralateral sides was performed using the following stimuli applied to the gingiva: a cold steel instrument kept cool in ice water, pain evoked by a toothpick, and touch evoked by a cotton swab. The outcome of the comparison was either:

308

T. List et al. / Pain 122 (2006) 306–314

‘‘the painful side was more, (hypersensitive), less (hyposensitive) or equal sensitive (normosensitive) as the contralateral side’’. After the qualitative examination, a standardized quantitative somatosensory examination was performed: Tactile detection threshold. The least force which the patients could identify correctly in 75% of the applications. Calibrated von Frey nylon filaments with forces from 10,000 to 300,000 mg were used (Somedic Sales). The patients were instructed to close their eyes during the entire test procedure and to report as soon as they felt the von Frey filament on the test site. The filament was applied vertically to the test site and slow pressure was applied until the filament began to bend. The time needed to bend the filament was standardized to approximately 1–2 s and the stimulus was maintained for 1–2 s and then removed (Komiyama and De Laat, 2005). At least three ascending and three descending series of stimuli were applied to determine the threshold. Filament-prick pain detection threshold. In a similar way, the von Frey filaments were used to determine the pain threshold, i.e., the least force which the patients reported as a painful sensation in 75% of the applications. Pressure pain threshold. The pressure (kPa) that the patient first perceived to be painful was determined with a pressure algometer (Somedic Sales, Sweden) (Svensson et al., 1995). A constant pressure of 30 kPa/s with a 4.8 mm diameter probe was applied and the average value of three measurements was calculated. There was about 1 min between repeated pressure stimuli. Brush-evoked (and vibrating) allodynia. The increase in VAS pain scores elicited by a cotton-tipped swab and the vibrating bristles of an electric toothbrush. The swab (diameter 10 mm) was repeatedly brushed over the painful intraoral region and the contralateral side for 10 s at approximately 2 Hz, and the increase in pain was assessed on a 0–10 VAS. The vibrating bristles (electrical toothbrush, Oral-B, Brown) were applied over the painful intraoral region and the contralateral side for 10 s, and the increase in pain was assessed on a 0–10 VAS. Repetitive pin-prick test. The increase in VAS pain elicited by repetitive (2 Hz) mechanical stimulation for 10 s. von Frey filaments were applied to the painful intraoral region and on the contralateral side with a force equal to the individual’s filament-prick pain detection threshold. The increase in pain was assessed on a 0–10 VAS. Thermal thresholds. Thresholds for warmth, cold, and heat pain were measured using a thermotester (modular sensory analyzer [MSA], Somedic). The intraoral thermode was custom-made with a 9 · 9 mm2 surface. The baseline temperature was set at 37 C and the temperature change was set at 1.0 C/s. The patients pushed a stop-button when the threshold was reached. The average of three measurements was used. There was between 4 and 6 s in between the repeated thermal stimuli. 2.4. Study design This was a randomized, multi-center, double-blind, crossover study on patients who complained of pain in the jaw and who had been diagnosed with AO. The study comprised four clinical visits: the first for a dental and a neurological

examination, the second for randomization and the first injection, the third for the second injection, and the fourth for a follow-up. At the first visit, after diagnosis and a somatosensory examination had been made, the patient signed a written consent. The pain intensity and unpleasantness were scored by the patients on 0–100 mm visual analogue scales (VAS) directly before injection (baseline); 15 and 30 min after injection (while the patient was still at the clinic); 45, 60, 90, 120, 180, 240, 300, and 360 min after injection (after the patient had left the clinic); and 1 day after injection. The patients were randomized to receive either physiological saline or lidocaine as an injection in a cross-over design. Between the two injections was at least a 1-week ‘‘washout period’’. The same assessments were made at both injections. Pain relief was reported by the patients as the percentage change compared with baseline. Adverse events were monitored by the patients 2 h after injection and 1 day following the injection. One week after the second injection the patients were followed up. No treatment other than the use of the patient’s usual analgesics (which was registered separately) was allowed during the study period. One operator made the injection and another trained observer obtained the VAS assessment and handled all administration and contact with the patient at each center during the study period. 2.5. Randomization and injection of test substance Ampoules for injection were blinded, given a patient code number according to a randomization list, and packed at the Linko¨ping University Hospital Pharmacy. The ampoules contained either 2 ml physiological saline (9 mg/ml NaCl solution) or lidocaine (20 mg/ml Xylocain and 12.5 lg/ml adrenaline). The randomization list was kept at the pharmacy until a ‘‘clean file’’ was declared. When the pain was located in the anterior part of the maxilla or the mandible, 1.5 ml of the test substance was injected into the buccal or labial sulcus adjacent to the painful region, i.e., as infiltration anesthesia. When the pain was located in the posterior region around the premolars or molars, an inferior alveolar nerve block or maxillary nerve block (posterior or middle superior alveolar part) was given, i.e., as a nerve block. 2.6. Pain assessment and adverse events Two 0–100 mm VAS with the end definitions ‘‘no pain’’ and ‘‘most pain imaginable’’ and ‘‘no unpleasantness’’ and ‘‘most unpleasantness imaginable’’ were used to assess the pain intensity and unpleasantness (Seymour et al., 1985). In addition, pain relief was estimated by the patients on a scale from 0% to 100% with the end definitions 0% = ‘‘no pain relief at all’’ and 100% = ‘‘complete pain relief’’ (Jensen, 2002). Responders were defined as the patients with more than 50% pain reduction in VAS pain intensity 30 min after injection compared with baseline. Furthermore, the number-needto-treat value (NNT) was calculated, i.e., the number of patients who had to be treated in order to obtain one patient with at least 50% pain reduction (McQuay and Moore, 1998).

T. List et al. / Pain 122 (2006) 306–314

309

The amount and kind of analgesics used were registered at baseline, 2 h after injection, and 1 day after injection. On two occasions – 2 h after injection and 1 day after injection – the patient answered the following question in a questionnaire: ‘‘Have you had any health problems since you received the injection?’’ During the clinical examinations, the investigator monitored clinical signs of adverse reactions. 2.7. Statistics

3. Results The VAS pain scores showed an overall effect of time (F = 9.205; P < 0.001) and treatment (F = 6.236; P = 0.018) with a significant interaction between the factors (F = 3.769; P < 0.001). The VAS pain scores at 15, 30, 45, 60, 90, 120, and 180 min were significantly lower than at baseline following lidocaine injection (Tukey: P = 0.038). For placebo injections, the VAS pain scores at 15, 30, 45, 60, 90, and 180 min were also significantly lower than at baseline (Tukey: P < 0.040). Direct comparisons of VAS pain scores at baseline showed no significant difference between lidocaine and placebo injections (Tukey: P = 0.104) but significantly lower VAS pain scores at 15, 30, 45, 60, 90, and 120 min following lidocaine injections compared with placebo injections (Tukey: P < 0.008) (Fig. 1A). The VAS unpleasantness scores were significantly influenced by time (F = 8.289; P < 0.001) and treatment (F = 5.106; P = 0.030) with a significant interaction between the factors (F = 2.469; P = 0.007). Thus, the VAS unpleasantness scores were significantly lower at 15, 30, 45, 60, 90, 120, and 180 min following the lidocaine injection compared to baseline scores (Tukey: P < 0.022). However, the VAS unpleasantness scores were also significantly lower at 30, 45, 60, 90, 180, and 240 min following placebo injections (Tukey: P < 0.045). There were no significant differences at baseline in the VAS unpleasantness scores (Tukey: P = 0.229) but following lidocaine injections, the VAS unpleasant-

Fig. 1. Visual analogue scale (VAS) scores of pain (A) and unpleasantness (B) in 35 patients with atypical odontalgia at baseline and following injections with placebo and lidocaine. Mean values and SEM. * indicates values significantly lower than baseline values (Tukey: P < 0.05) and # indicates significant difference between placebo and lidocaine (Tukey: P < 0.05).

ness scores were significantly lower compared to placebo injections at 15, 30, 45, 60, 90, and 120 min (Tukey: P < 0.028) (Fig. 1B). Finally, the analysis of the VAS pain relief data showed a significant effect of time (F = 4.408; P < 0.001) and treatment (F = 9.048; P = 0.005) and a significant interaction between the factors (F = 4.514; P < 0.001). Overall, VAS pain relief was significantly greater following the lidocaine injections compared to placebo injections (Tukey: P < 0.05). Direct comparisons showed significantly greater VAS pain relief following lidocaine injections at 15, 30, 45, 60, 90, and 120 min compared to placebo injections (Fig. 2).

VAS pain relief (0 -100%)

The VAS pain and unpleasantness scores were analyzed with a two-way analysis of variance (2-ANOVA) with time (11 levels: baseline, 15, 30, 45, 60, 90, 120, 180, 240, 360 min, and 1 day) and treatment (2 levels: active, placebo) as repeated factors. The VAS pain relief data were analyzed in a similar way with time (10 levels: 15, 30, 45, 60, 90, 120, 180, 240, 360 min, and 1 day) and treatment as repeated factors. Tukey post hoc tests were used to adjust for multiple comparisons. The paired t-test was used to analyze variables measured on a ratio scale. When data consisted of frequencies in discrete categories, the v2 test was used to determine the significance between independent groups. Spearman’s rank correlation coefficient (q) was used to measure the association between pain relief (in percentage) and quantitative somatosensory measures, i.e., the difference between the painful and contralateral side. For all tests, a significance level of P < 0.05 was used.

100 90 80 70 60 50 40 30 20 10 0

Placebo Lidocaine

#

#

#

#

#

15

in m

30

in m

45

in m

60

in m

90

in m

#

0 12

in m

0 18

in m

0 24

in m

0 36

in m

1

y da

Fig. 2. Patient’s reports of pain relief in percentage of their baseline pain. Mean values and SEM for 35 patients. # indicates significant difference between placebo and lidocaine (Tukey: P < 0.05).

310

T. List et al. / Pain 122 (2006) 306–314

Table 1 Mean and SD for somatosensory quantitative measures for the painful and non-painful side

Tactile detection threshold (von Frey, log10 mg) Filament-prick pain detection threshold (von Frey, log10 mg) Pressure pain threshold (kPa) Brush-evoked pain (manually, VAS) Brush-evoked pain (toothbrush, VAS) Pin-prick test (von Frey, VAS) Warm threshold (MSA, C) Cold threshold (MSA, C) Heat threshold (MSA, C)

Painful side

Non-painful side

P value

Hypersensitivity (%)

Hyposensitivity (%)

3.7 ± 0.6 4.5 ± 0.6 92.3 ± 45.1 9.2 ± 16.4 21.6 ± 25.4 29.2 ± 24.6 45.8 ± 3.9 22.4 ± 3.7 48.1 ± 2.9

3.7 ± 0.5 4.5 ± 0.7 91.2 ± 44.6 1.9 ± 8.0 4.7 ± 12.4 11.9 ± 15.0 46.6 ± 3.5 21.9 ± 4.3 49.1 ± 1.9

0.30 0.94 0.87 0.007 <0.001 <0.001 0.22 0.59 0.003

28.1 24.2 46.9 45.5 60.6 75.0 54.5 42.4 56.3

40.6 36.4 50.0 3.0 6.1 12.5 33.3 54.5 9.4

P values indicate differences between sides (Paired t-tests). The number of patients with lower threshold values or higher pain scores (hypersensitivity) and with higher threshold values or lower pain scores (hyposensitivity) on the painful side indicated in percentage. MSA, modular sensory analyzer; VAS, visual analogue scale.

A 50% or more reported reduction in VAS pain scores was found in 54% (lidocaine) and in 29% (placebo) of the patients 30 min after injection (v2: P < 0.0001). A 50% or more reported increase in VAS pain scores was found in 2 patients (lidocaine) and none for the placebo 30 min after injection. The NNT value 30 min after injection was 3.3 (CI 2.0–9.8). All patients participating in the study exhibited one or more qualitative or quantitative somatosensory disturbances in the painful area. The qualitative examination found hyper- or hyposensitivity in 50% for touch and in 47% for cold, and either hyper- or hypoalgesia for pin-prick in 57% of the patients. Overall there were significant differences between the painful and the non-painful sides for four of the quantitative tests (Table 1). Taking into consideration whether the sensory changes were hyperesthesias or hyperalgesias or were hypoesthesias or hypoalgesias, a more complex pattern was observed (Figs. 3A–I) (Table 1). However, hypoesthesia/hypoalgesia to at least one sensory test was demonstrated in all patients, and brush-evoked allodynia was observed in the majority of patients (Fig. 3E). A significant inverse correlation between percentage pain relief and brush-evoked allodynia (Spearman: P < 0.01) (Fig. 3E) was observed. No significant association was found between pain relief and the other quantitative somatosensory measures (Fig. 3). Two hours following the injection, 29% of the patients reported adverse events after lidocaine and 26% after placebo (v2: P > 0.79). The most common reports for lidocaine were headache, increased pain, and heart palpitation and for placebo headache, increased pain, dizziness, tiredness, and paresthesia. Analgesic consumption was reported at baseline by five individuals (2 took paracetamol, 2 ibuprofen, and 1 ibuprofen + paracetamol) and at the follow-up by three patients (2 took paracetamol and 1 traumadol). No statistical difference in analgesic consumption was seen during the lidocaine or placebo period.

4. Discussion The main finding in this RCT study was that a majority of the AO patients experienced a significant, but not complete, pain relief following local administration of lidocaine in the painful region. The reduction in pain was inversely correlated to the magnitude of brushevoked allodynia in the painful area. Since AO is a rare condition, consecutive patients from four orofacial pain clinics were recruited. Each clinic is responsible for the population in a large district in Sweden, and together the centers represent a population of approximately 2 million people. In our study, we used inclusion criteria, which agreed with the major current classification system for AO (Melis et al., 2003). The patients are representative of a clinical AO population since the age and gender distribution is similar to that in populations reported by others (Schnurr and Brooke, 1992; Vickers et al., 1998). The term AO has been used for a pain in a tooth or the mucosa at the site of a tooth extraction, with no physical or radiographic finding to explain the pain, and often with a history of frequent restorative, endodontic procedures or where extractions have been made (Vickers et al., 1998). During dental procedures, trauma to dental tissues, that is, pulpal tissue, is common and some researchers have suggested that AO is related to deafferentation (Graff-Radford and Solberg, 1992; Marbach, 1993). Experimental studies in animals with deafferentation procedures and clinical observations in humans exposed to nerve lesions have found that these injuries are most often reversible (Hu et al., 1999). It is still uncertain why only some patients develop chronic pain following tooth pulp deafferentation, but genetic factors as well as inflammation, mechanical loading, and psychological factors may all play a role (Woda and Pionchon, 2000; Seltzer and Dorfman, 2004).

A

Spearman’s rho=0.15

C

PPT (kPa)

von Frei (10 Iog mg)

von Frei (10 Iog mg)

B

P=0.29

Pain reduction (%)

Spearman’s rho=0.19

Spearman’s rho=0.33

P=0.29

P=0.88

Pain reduction (%)

Pain reduction (%)

D

VAS

VAS

F

Spearman’s rho=0.31

Spearman’s rho=-0.43

P=0.08

P=0.01

Spearman’s rho=-0.21 P=0.24

Pain reduction (%)

Pain reduction (%)

Pain reduction (%)

G

H

C˚

I C˚

C˚

Pain reduction (%)

T. List et al. / Pain 122 (2006) 306–314

VAS

E

Spearman’s rho=0.18

Spearman’s rho=0.13

Spearman’s rho=0.11

P=0.31

P=0.46

P=0.56

Pain reduction (%)

Pain reduction (%)

311

Fig. 3. Associations between quantitative measures of intraoral somatosensory sensitivity measured on the painful and non-painful sides in AO patients (n = 35) expressed as delta values (painful side minus non-painful contralateral sides) and percentage pain reduction (%) 30 min after lidocain injection. A, tactile detection threshold; B, filament-prick pain detection threshold; C, pressure pain threshold; D, brush-evoked pain (manually); E, brush-evoked pain (tooth-brush); F, pin-prick test; G, warm threshold; H, cold threshold; I, heat threshold. Inserted full line = regression line. Spearman’s q = Spearman’s rank correlation coefficient (q).

312

T. List et al. / Pain 122 (2006) 306–314

Neuropathic pain conditions often include spontaneous pain, paresthesia, and dysesthesia as well as stimulus-provoked sensory disturbances such as hyperalgesia, touch- and cold-evoked allodynia, or non-painful sensory disturbances (Jensen et al., 2001). This somatosensory dysfunction is characterized by altered threshold values for touch, temperature, and pain (Hansson et al., 2001; Essick, 2004; Svensson et al., 2004). Since the diagnosis of neuropathic pain involves the detection of sensory dysfunction, a thorough examination of the sensory functions is important (Hansson and Kinnman, 1996). A few studies in AO patients have indicated sensory abnormalities (GraffRadford and Solberg, 1992; Essick, 2004). In the present study, somatosensory abnormalities were determined by comparing the painful side with the healthy side according to clinical recommendations (Jensen, 2002). A limitation with this approach is that there may also be contralateral changes in sensitivity in neuropathic pain conditions (Eliav et al., 2004). Nevertheless, all the participants in this study exhibited at least one sensory dysfunction in the painful area (Table 1), which in addition to the history of nerve damage (tooth extraction or endodontic treatment) is compatible with the suggestion that AO could represent a neuropathic pain condition. Diagnostic blocks and topical anesthesia have been suggested to be used to gather data for better identifying the source of pain and the underlying mechanism (Rees and Harris, 1979; Graff-Radford and Solberg, 1992). In our study, a double-blind approach was used to evaluate the pain reduction following administration of local anesthetic. A significant difference was found between administration of lidocaine and placebo with respect to pain intensity and unpleasantness. This finding differs with results of another randomized double-blind, placebo-controlled trial where a topical anesthetic and local anesthetic blocks were administered. No significant difference was seen in AO patients, but patients with nociceptive pain significantly improved (Graff-Radford and Solberg, 1992). Two uncontrolled studies using local anesthetic blocks have reported similar findings (Rees and Harris, 1979; Bates and Stewart, 1991); however, corroboratory results were also found in another study using topical anesthetics where a significant reduction in pain in a group of AO patients was reported (Vickers et al., 1998). In a double-blind placebo-controlled study, lidocaine patches (5%) were found to be clearly effective in reducing ongoing pain and allodynia in peripheral neuropathic pain syndromes (Meier et al., 2003). Differences in definition and diagnostic criteria of neuropathic pain, methods of application, and assessment techniques may explain the variation in outcome in the AO studies, but local anesthetics seem to be able to relieve, but not completely abolish, neuropathic pain. To complement the pain intensity measurements, pain relief was estimated by the patients according to

the recommendation by McQuay and Moore (McQuay and Moore, 1998). Lidocaine causes a reversible block to the conduction of impulses along nerve fibers and it has an onset of a few minutes and a duration of 1–2 h, and therefore the time point for estimation of the pain reduction was set at 30 min. A significant difference was seen between saline and lidocaine 15–180 min following the injection. A similar pattern was seen for the reported pain relief in our study. It has been pointed out that the highest level of patient impression of improvement corresponds to a 50% reduction in pain intensity (Farrar et al., 2000). In our study, a 50% or more reduction in VAS pain scores was found in 54% (lidocaine) and in 29% (placebo) of the patients 30 min after injection. Although the present sample size was fairly small, the magnitude of the response to placebo in our study and the NNT value are similar to those reported by other studies where placebo was compared with various active substances such as analgesics given orally or intramuscularly (McQuay and Moore, 1998). The psychobiological mechanisms underlying the placebo response are still under examination (Colloca and Benedetti, 2005). High analgesic success rates following injection of local anesthetics have been reported in experimental studies (Ernberg and Kopp, 2002; Axelsson and Isacsson, 2004) as well as during dental procedures (Moore and Dunsky, 1983; Dunsky and Moore, 1984). This indicates that for a large proportion of the patients with AO, the peripheral input from the painful area may to some extent be involved in the spontaneous pain. However, the frequency of patients reporting effective pain reduction was considerably lower in our study compared with studies on injections with local anesthetics during dental procedures. This indicates that other mechanisms are also involved in the pain of AO. Neuromas can form after peripheral nerve injury. Both neuromas and dorsal root ganglion exhibit spontaneous activity and increased sensitivity to mechanical and chemical stimuli (Wall and Devor, 1983). It has been suggested that the development of spontaneous activity after nerve injury on the molecular level is related to sodium channel accumulation along the entire damaged primary sensory neuron (Devor et al., 1993). In support of this finding, low-dose sodium channel blockers have been found to reduce spontaneous activity in experimental animal neuroma models, although no conduction block of the nerve fibers occurs (Chabal et al., 1989). In several patients with phantom limb pain, a peripheral nerve block failed to block spontaneous pain or spontaneous activity, although it is known that stimulus-evoked abnormal sensations can be blocked (Nystrom and Hagbarth, 1981). In our study, most patients reported only partial pain relief. One possible explanation for this finding may be that dorsal root ganglion cells (trigeminal ganglion) belonging to damaged fibers can

T. List et al. / Pain 122 (2006) 306–314

become a source of ectopic nerve impulses (Wall and Devor, 1983). That central changes may cause hyperalgesia/allodynia has been shown in human models of patients with peripheral neuropathy (Koltzenburg et al., 1994). Ongoing C-fiber input leads to central changes in animals, which cause nociceptive-specific neuronal sensitization as well as wide dynamic range (WDR) neuronal sensitization to nociceptive as well as non-nociceptive stimuli. Allodynic symptoms in post-herpetic neuralgia were observed to be reduced following treatment with lidocaine cream (Attal et al., 1999). Our design of the study did not permit evaluation of somatosensory changes following injection, so we cannot know if there was a differential effect on spontaneous and stimulus-evoked pain. Inverse correlations were found between measures of brush-evoked allodynia and pain relief (in percentage), i.e., the greater disturbance in somatosensory function – the smaller effect of the lidocaine injection. This finding is in accordance with the notion that the possible neural mechanisms underlying brush-evoked allodynia are related to central sensitization and plasticity (Koltzenburg et al., 1994). Thus, hyperexcitability of higher order trigeminal neurons is likely to contribute to the spontaneous pain in AO patients. Abnormal temporal summation (wind-up) is the clinical equivalent to increased neural activity following repetitive C-fiber stimulation. Wind-up is a characteristic clinical finding in chronic pain conditions such as neuropathic pain (Jensen et al., 2001). It has been suggested that cross-excitation between low-threshold mechanoreceptors and nociceptors in the periphery and/or in the dorsal root ganglion leads to the summation of a response to mechanical stimulus and might be a mechanism involved in the summation of repetitive stimuli (Devor, 1994). Another possible cause of summation is that of central disinhibition. Due to temporal summation mechanisms, dorsal horn neurons will also gradually increase their response to repetitive stimulus (Price et al., 1978). In our study, we found no support of a correlation between pain relief due to administration of lidocaine and increased pain to repetitive pinprick (Fig. 3). The reason for this is not known and further studies need to monitor changes in somatosensory sensitivity following the administration of lidocaine.

5. Conclusion The study found that AO patients experienced significant, but not complete, pain relief from administration of local anesthetics compared with placebo. This indicates that persistent peripheral afferent inputs alone cannot explain the spontaneous pain in AO patients and that sensitization and plasticity of higher order trigeminal neurons may contribute.

313

Acknowledgments This study was supported by the Health Research Council in the South-East of Sweden (FORSS) and The Swedish Research Foundation. References Attal N, Brasseur L, Chauvin M, Bouhassira D. Effects of single and repeated applications of a eutectic mixture of local anaesthetics (EMLA) cream on spontaneous and evoked pain in post-herpetic neuralgia. Pain 1999;81:203–9. Axelsson S, Isacsson G. The efficacy of ropivacaine as a dental local anaesthetic. Swed Dent J 2004;28:85–91. Bates Jr RE, Stewart CM. Atypical odontalgia: phantom tooth pain. Oral Surg Oral Med Oral Pathol 1991;72:479–83. Chabal C, Russell LC, Burchiel KJ. The effect of intravenous lidocaine, tocainide, and mexiletine on spontaneously active fibers originating in rat sciatic neuromas. Pain 1989;38:333–8. Colloca L, Benedetti F. Placebos and painkillers: is mind as real as matter. Nat Rev Neurosci 2005;6:545–52. Devor M. The pathophysiology of damaged peripheral nerves. In: Wall PD, Melzack R, editors. Textbook of pain. Edingburgh: Churchill Livingstone; 1994. p. 79–100. Devor M, Govrin-Lippmann R, Angelides K. Na+ channel immunolocalization in peripheral mammalian axons and changes following nerve injury and neuroma formation. J Neurosci 1993;13:1976–92. Dunsky JL, Moore PA. Long-acting local anesthetics: a comparison of bupivacaine and etidocaine in endodontics. J Endod 1984;10: 457–60. Dworkin SF, LeResche L. Research diagnostic criteria for temporomandibular disorders: review, criteria, examinations and specifications, critique. J Craniomandib Disord 1992;6:301–55. Eliav E, Gracely RH, Nahlieli O, Benoliel R. Quantitative sensory testing in trigeminal nerve damage assessment. J Orofac Pain 2004;18:339–44. Ernberg M, Kopp S. Ropivacaine for dental anesthesia: a dose-finding study. J Oral Maxillofac Surg 2002;60:1004–10. [discussion 1010– 11]. Essick GK. Psychophysical assessment of patients with posttraumatic neuropathic trigeminal pain. J Orofac Pain 2004;18:345–54. Farrar JT, Portenoy RK, Berlin JA, Kinman JL, Strom BL. Defining the clinically important difference in pain outcome measures. Pain 2000;88:287–94. Forsell H, Svensson P. Facial pain – atypical facial pain and burning mouth syndrome. In: Cervero FJT, editor. Handbook of neurology. Elsevier [in press]. Graff-Radford SB, Solberg WK. Atypical odontalgia. J Craniomandib Disord 1992;6:260–5. Hansson P. Neuropathic pain: clinical characteristics and diagnostic workup. Eur J Pain 2002;6(Suppl. A):47–50. Hansson P, Kinnman E. Unmasking mechanisms of peripheral neuropathic pain in a clinical perspective. Pain Rev 1996;3:272–92. Hansson P, Lacerenza M, Marchettini P. Aspects of clinical and experimental neuropathic pain: the clinical perspective. In: Hansson P, Fields H, Hill R, Marchettini P, editors. Neuropathic pain: pathophysiology and treatment, vol. 21. Seattle: IASP Press; 2001. p. 1–19. Hu JW, Woda A, Sessle BJ. Effects of pre-emptive local anesthesia on tooth pulp deafferentation-induced neuroplastic changes in cat trigeminal brainstem neurones. Arch Oral Biol 1999;44:287–93. Jensen TS. An improved understanding of neuropathic pain. Eur J Pain 2002;6:3–11. Jensen TS, Gottrup H, Sindrup SH, Bach FW. The clinical picture of neuropathic pain. Eur J Pharmacol 2001;429:1–11.

314

T. List et al. / Pain 122 (2006) 306–314

Koltzenburg M, Torebjork HE, Wahren LK. Nociceptor modulated central sensitization causes mechanical hyperalgesia in acute chemogenic and chronic neuropathic pain. Brain 1994;117(Pt 3):579–91. Komiyama O, De Laat A. Tactile and pain thresholds in the intra- and extra-oral regions of symptom-free subjects. Pain 2005;115:308–15. Lynch ME, Elgeneidy AK. The role of sympathetic activity in neuropathic orofacial pain. J Orofac Pain 1996;10:297–305. Marbach JJ. Is phantom tooth pain a deafferentation (neuropathic) syndrome? Part II: psychosocial considerations. Oral Surg Oral Med Oral Pathol 1993;75:225–32. McQuay H, Moore A. An evidence-based resource for pain relief. Oxford: Oxford University Press; 1998. p. 14–31. Meier T, Wasner G, Faust M, Kuntzer T, Ochsner F, Hueppe M, et al. Efficacy of lidocaine patch 5% in the treatment of focal peripheral neuropathic pain syndromes: a randomized, doubleblind, placebo-controlled study. Pain 2003;106:151–8. Melis M, Lobo SL, Ceneviz C, Zawawi K, Al-Badawi E, Maloney G, et al. Atypical odontalgia: a review of the literature. Headache 2003;43:1060–74. Moore PA, Dunsky JL. Bupivacaine anesthesia – a clinical trial for endodontic therapy. Oral Surg Oral Med Oral Pathol 1983;55:176–9. Nystrom B, Hagbarth KE. Microelectrode recordings from transected nerves in amputees with phantom limb pain. Neurosci Lett 1981;27:211–6. Price DD, Hayes RL, Ruda M, Dubner R. Spatial and temporal transformations of input to spinothalamic tract neurons and their relation to somatic sensations. J Neurophysiol 1978;41: 933–47. Rees RT, Harris M. Atypical odontalgia. Br J Oral Surg 1979;16:212–8. Schnurr RF, Brooke RI. Atypical odontalgia. Update and comment on long-term follow-up. Oral Surg Oral Med Oral Pathol 1992;73:445–8.

Seltzer Z, Dorfman R. Identifying genetic and environmental risk factors for chronic orofacial pain syndromes: human models. J Orofac Pain 2004;18:311–7. Sessle BJ. Acute and chronic craniofacial pain: brainstem mechanisms of nociceptive transmission and neuroplasticity, and their clinical correlates. Crit Rev Oral Biol Med 2000;11:57–91. Seymour RA, Meechan JG, Blair GS. An investigation into post-operative pain after third molar surgery under local analgesia. Br J Oral Maxillofac Surg 1985;23:410–8. Svensson P, Graven-Nielsen T. Craniofacial muscle pain: review of mechanisms and clinical manifestations. J Orofac Pain 2001;15:117–45. Svensson P, Arendt-Nielsen L, Nielsen H, Larsen JK. Effect of chronic and experimental jaw muscle pain on pain-pressure thresholds and stimulus-response curves. J Orofac Pain 1995;9:347–56. Svensson P, Baad-Hansen L, Thygesen T, Juhl GI, Jensen TS. Overview on tools and methods to assess neuropathic trigeminal pain. J Orofac Pain 2004;18:332–8. Vickers ER, Cousins MJ. Neuropathic orofacial pain part 1–prevalence and pathophysiology. Aust Endod J 2000;26:19–26. Vickers ER, Cousins MJ, Walker S, Chisholm K. Analysis of 50 patients with atypical odontalgia. A preliminary report on pharmacological procedures for diagnosis and treatment. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 1998;85: 24–32. Wall PD, Devor M. Sensory afferent impulses originate from dorsal root ganglia as well as from the periphery in normal and nerve injured rats. Pain 1983;17:321–39. Woda A, Pionchon P. A unified concept of idiopathic orofacial pain: pathophysiologic features. J Orofac Pain 2000;14:196–212. Woda A, Pionchon P. Orofacial idiopathic pain: clinical signs, causes and mechanisms. Rev Neurol (Paris) 2001;157:265–83. Woda A, Tubert-Jeannin S, Bouhassira D, Attal N, Fleiter B, Goulet JP, et al. Towards a new taxonomy of idiopathic orofacial pain. Pain 2005;116:396–406.