Dynamic mechanical allodynia in the secondary hyperalgesic area in the capsaicin model—Perceptually similar phenomena as in painful neuropathy?

2.1 Subjects

2.2 General procedure

All injections (author P.H.) and assessments (author M.S.) were performed by the same investigators. The subjects were carefully familiarized with the different methods to be used before the start of the experiment. Before start of the assessments, the subjects were comfortably seated in a chair or lying on a bed and had time to adjust to the environment and room temperature. The patients were first examined in the area of painful neuropathy and subsequently in the corresponding contralateral control site, i.e., in the potential secondary hyperalgesic area outside the flare after an intradermal injection of capsaicin. The reason for this order of precedence was to avoid any influence on the clinical condition from the acute pain induced by capsaicin. Aged and gender matched healthy subjects were examined in a corresponding control area after an intradermal injection of capsaicin.

2.3 Intradermal injection of capsaicin

A solution of capsaicin (6 mg/ml) and polysorbat 80 diluted in isotonic saline prepared as described previously [19] was injected intradermally. Guided by how well the pain was tolerated by the patients during the first seconds of the injection, either 120 µg (20 µl) or 60 µg (10 µl) capsaicin was injected (Table 1). The matched healthy subjects were given identical amounts of capsaicin as the corresponding patient. The injections were performed manually with the use of a 0.3 ml plastic syringe and a 30-gauge needle.

2.4 Assessment of brush-evoked pain using stimuli with varying temporo-spatial characteristics

A 20 mm long and 20 mm wide test area within the area of maximum brush-evoked pain intensity in patients with painful peripheral neuropathy and, when present, in the secondary hyperalgesic area in capsaicin treated skin in patients and healthy subjects was marked with a pen. Brush-evoked pain was induced by lightly stroking different lengths (10 or 20 mm) in the test area with one of two brushes of different widths (4 or 16 mm) 2 or 4 times. A standardized protocol based on 8 brush stimuli defining the sequence of brushing length, width of the brush and number of strokes was used (i.e., 8 stimuli in the area of painful neuropathy and 8 stimuli in the capsaicin-induced secondary hyperalgesic area, in the patients and in the healthy subjects, respectively). Each three stimulus parameter combination was introduced once. Possible sequence dependency of the stimuli has been investigated previously and was found to be statistically non-significant [17]. The stimuli were delivered manually with an interstimulus interval of 1–2 min depending on differences in duration of aftersensation of brush-evoked allodynia and time consumption for the selection procedure of pain descriptors. The brushes were connected to a pressure sensor of a modified von Frey equipment (Somedic Sales AB, Horby, Sweden). Before each study session the brushing device was calibrated by holding the brush in the air approximating the stroking position. The examiner kept a fairly constant brushing force of 4–25 g and a stroking velocity of approximately 20mm/s. Brushing force was monitored on-line on the computer screen and if exceeding 25 g or if below 4g, the stroke was disregarded and a new attempt was made. The subjects were carefully instructed to rate the intensity of brush-evoked pain separately from spontaneous ongoing pain. Using a computerized VAS, preset to record values exceeding 2 mm and stopped when values were below 2 mm, the subjects continuously rated the intensity of allodynia. Measurements were stored in a database that enabled recordings of seconds to onset of brush-evoked pain (>2 mm VAS), maximum brush-evoked pain intensity, as well as when brush-evoked pain intensity had returned to baseline as a basis for calculating the total brush-evoked pain intensity as the integrated value of the graph over time, i.e., the area under the curve (Fig. 1). The method has been thoroughly described elsewhere [17]. In the present study, aftersensation was defined as the time from cessation of each stimulus plus an additional 5 s to enable the subjects to fully estimate their brush-evoked pain intensity and until the brush-evoked pain intensity had returned to baseline.

Fig. 1

Temporal development profile of brush-evoked pain intensity in one patient with neuropathy and dynamic mechanical allodynia, induced by brushing the skin twice over 10mm or 4 times over 20mm with a 4mm brush.

2.5 Selection of sensory-discriminative and affective pain descriptors

A self-administered pain assessment tool, the Pain-O-Meter^® (POM), has been developed for the purpose of improving assessment of multidimensional aspects of clinical acute and chronic pain [18]. The POM is a hand-held plastic device, comprising two methods for pain assessment, a VAS for assessment of pain intensity and two separate lists of pain descriptors. The Swedish version used in the present study consisted of 12 sensory-discriminative and 11 affective words. Following each stimulus of the protocol, the patients were asked to choose as many sensory-discriminative and affective descriptors as needed to adequately describe the pain experience. The Swedish sensory-discriminative and affective words used in the present study are translated into English according to recent publications of Swedish studies using the POM [20] and a Swedish-English dictionary (Table 2). The descriptors have been demonstrated to discriminate differences in pain intensity but were in the present study used only to describe the experience of pain without a hierarchic intensity rating. Test-retest reliability and concurrent as well as construct validity of the POM has been demonstrated in studies on American patients with acute or chronic pain, i.e., labour pain, rheumatoid arthritis and postoperative pain [18].

2.6 Statistics

Data from assessments of the total pain intensity during brush-evoked pain in patients had a skewed distribution and was transformed to a log scale in order to meet the requirements for an adequate analysis of variance (ANOVA). A four-way ANOVA with repeated measures on four factors were used to analyze the total brush-evoked pain intensity [21]. The factors were, ‘side’ with two levels (area of painful neuropathy and of capsaicin-induced secondary hyperalgesia), ‘brushing length’ with two levels (10 mm and 20 mm), ‘brush width’ with two levels (4 mm and 16 mm) and ‘number of strokes’ with two levels (2 and 4 times). In addition, a five-way ANOVA with repeated measures on the aforementioned factors and a between-groups factor ‘dose of injected capsaicin’ (120 µg or 60 µg) was used. In the statistical analysis a non-painful stimulus was included and calculated as a zero value.

The data from assessments of the duration of aftersensation of brush-evoked pain in the area of painful neuropathy and in the capsaicin-induced secondary hyperalgesic area in patients had a skewed distribution and was therefore analyzed using the Sign test.

Furthermore, the aftersensation was categorized into two categories as the variable had a skewed distribution with many outcomes equal to zero. The categories were ‘aftersensation’ or ‘no aftersensation’. The data were analyzed by a generalized estimating equations (GEE) model with the GENMOD procedure in SAS^® [22]. The GEE strategy is a useful approach for repeated measurements analysis of ordered categorical- and binominal outcomes in a longitudinal study. The model was set up with the within factors ‘side’ (area of painful neuropathy and the capsaicin-induced secondary hyperalgesic area), ‘brushing length’ (10 mm and 20 mm), ‘brush width’ (4 mm and 16 mm) and ‘number of strokes’ (2 and 4 times). The estimates from the models were odds ratios and 95% confidence intervals.

The relationship between the duration of the painful aftersensation and the maximum rated pain intensity during assessments of brush-evoked pain was analyzed with Spearman rank order correlation.

Sign test was used to analyze data from assessments of spontaneous ongoing pain 1 min after the intradermal injection of capsaicin in patients and matched healthy subjects.

Statistical significance was accepted at P ≤ 0.05.

3.1 The influence of stimulus parameters on brush evoked pain intensity in the area of neuropathy and in the capsaicin-induced secondary hyperalgesic area in patients

All patients reported brush-evoked pain in their area of painful neuropathy during all stimuli. Six out of 9 patients reported brush-evoked pain in an area outside the flare in the capsaicin treated skin during all stimuli and 2 patients during 6 and 7 out of 8 stimuli, respectively. The area of secondary hyperalgesia developed in the 8 patients with a latency of 7–22 min. Data from the total brush-evoked pain intensity during different stimulus combinations in the area of painful neuropathy and in the capsaicin-induced secondary hyperalgesic area in patients is presented in Fig. 2.

Fig. 2

Total brush-evoked pain intensity at different stimulus combinations in the area of painful neuropathy (A) and in the capsaicin-induced secondary hyperalgesic area (B) in 9 patients. The different stimulus combinations were: (1) 4mm brush, 2 strokes, 10mm brushing length; (2) 4mm brush, 4 strokes, 10mm brushing length; (3) 4mm brush, 2 strokes, 20mm brushing length; (4) 4mm brush, 4 strokes, 20mm brushing length; (5) 16mm brush, 2 strokes, 10mm brushing length; (6) 16mm brush, 4 strokes, 10mm brushing length; (7) 16mm brush, 2 strokes, 20mm brushing length and (8) 16mm brush, 4 strokes, 20mm brushing length. Data is presented as total brush-evoked pain intensity, AUC (area under the curve).

In the four- and five-way ANOVA there was no significant interaction between the factors side, brushing length, brush width, number of strokes and dose of injected capsaicin.

3.1.2 Number of strokes (2 or 4)

Significantly higher total brush-evoked pain intensity was demonstrated for 4 compared to 2 strokes in both sides (F(1, 8) = 20.61, P < 0.01) (Fig. 3).

Fig. 3

The influence of number of strokes on the total brush-evoked pain intensity following brushing of different lengths using brushes of various widths in neuropathic (n.p.) and capsaicin induced pain (caps) in patients; n = 9. Mean VAS ratings of total brush-evoked pain intensity±SEM (area under the curve (AUC)) is presented. In the four-way ANOVA significant differences are indicated by P-values in the figure (**P < 0.01). b, Brush widths (4 or 16mm); mm, brushing lengths (10 or 20mm); stim, number of strokes (2 or 4).

3.2 The influence of stimulus parameters on brush evoked pain intensity in the capsaicin-induced secondary hyperalgesic area in patients and healthy subjects

As described above 8 out of 9 patients reported brush-evoked pain in an area outside the flare during all stimuli. Three out of 9 healthy subjects reported brush-evoked pain in an area outside the flare during 4, 5 and 6 out of 8 stimuli, respectively. The area of secondary hyperalgesia developed in the 3 healthy subjects with a latency of 7–11 min. Data from the total brush-evoked pain intensity during different stimulus combinations in the capsaicin-induced secondary hyperalgesic area in patients and healthy subjects is presented in Fig. 4. No statistical analysis was done due to the few healthy subjects reporting an area of secondary hyperalgesia. The patient, and with one exception, healthy subjects who did not develop a secondary hyperalgesic area also reported the lowest spontaneous ongoing pain intensity after the intradermal capsaicin injection.

Fig. 4

Total brush-evoked pain intensity at different stimulus combinations in the capsaicin-induced secondary hyperalgesic area in 9 patients (B) and their matched healthy subjects (C). The different stimulus combinations were: (1) 4mm brush, 2 strokes, 10mm brushing length; (2) 4mm brush, 4 strokes, 10mm brushing length; (3) 4mm brush, 2 strokes, 20mm brushing length; (4) 4mm brush, 4 strokes, 20mm brushing length; (5) 16mm brush, 2 strokes, 10mm brushing length; (6) 16mm brush, 4 strokes; 10mm brushing length; (7) 16mm brush, 2 strokes, 20mm brushing length and (8) 16mm brush, 4 strokes, 20mm brushing length. Data is presented as total brush-evoked pain intensity, AUC (area under the curve).

3.3 The influence of stimulus parameters on frequency and duration of painful aftersensation (s) in the area of painful neuropathy and in the capsaicin-induced secondary hyperalgesic area in patients

Aftersensation was reported during 5–8 stimuli in the area of painful neuropathy by 6 patients and in the capsaicin-induced secondary hyperalgesic area by 2 patients. During 1–4 stimuli 2 patients reported aftersensation in the area of painful neuropathy and 6 patients in the capsaicin-induced secondary hyperalgesic area. Lack of aftersensation was reported by 1 patient in the area of painful neuropathy.

The odds ratio for any aftersensation following a brushing length of 10 mm was 5.4 times higher in the area of painful neuropathy compared to the capsaicin-induced secondary hyperalgesic area (P < 0.05). The odds ratio for any aftersensation in the area of painful neuropathy was 3 times higher following a brushing length of 10 mm compared to 20 mm (P < 0.01).

Significantly longer duration of aftersensation was demonstrated in the area of painful neuropathy compared to the capsaicin-induced secondary hyperalgesic area when brushing 2 or 4 strokes with a 4 mm brush for 10 mm, respectively (P < 0.05) (Fig. 5).

Fig. 5

Significantly longer duration of aftersensation (s) was demonstrated in the area of painful neuropathy (n.p.) compared to the capsaicin-induced secondary hyperalgesic area (caps) in patients; n = 9 when brushing with a 4mm brush, 10mm brushing length using 2 or 4 strokes, respectively (*P < 0.05). Mean values of aftersensation (s)±SEM is presented. b, brush widths (4 or 16mm); stim, number of strokes (2 or 4); mm, brushing lengths (10 or 20 mm).

3.4 The relationship between duration of painful aftersensation (s) and maximum brush-evoked pain intensity (mm) in the area of painful neuropathy and in the capsaicin-induced secondary hyperalgesic area in patients

A non-significant correlation was demonstrated between the duration of the painful aftersensation and the maximum brush-evoked pain intensity for the various temporo-spatial stimulus parameters in the area of painful neuropathy (r_s = 0.02–0.56) and in the capsaicin-induced secondary hyperalgesic area (r_s = 0.18–0.65).

3.5 The intensity of spontaneous ongoing pain after capsaicin injection in patients and healthy subjects

All patients and healthy subjects reported spontaneous ongoing pain at the site of the intradermal injection of capsaicin. Significantly higher spontaneous pain intensity was demonstrated in patients (VAS median 94 mm; VAS range 48–100 mm) compared to healthy subjects (VAS median 30 mm; VAS range 16–51 mm) (P <0.01) (Table 3).

3.6 Choice of pain descriptors

The most commonly used sensory-discriminative descriptors during brush-evoked pain in the area of painful neuropathy were smarting (15%) and burning (14%) and for the affective descriptors troublesome (36%) and annoying (28%). The most commonly used sensory-discriminative descriptors during brush-evoked pain in the capsaicin-induced secondary hyperalgesic area in patients were smarting (30%) and burning (23%) and for the affective descriptors troublesome (48%) and annoying (41%) The most commonly used sensory-discriminative descriptors (n = 3) during brush-evoked pain in the capsaicin-induced secondary hyperalgesic area in healthy subjects were smarting (56%) and burning (13%) and for the affective descriptors (n = 2) annoying (90%) (Table 2).