- Email: [email protected]
Differential cerebellar activation related to perceived pain intensity during noxious thermal stimulation in humans: a functional magnetic resonance imaging study
Neuroscience Letters 335 (2003) 202–206 www.elsevier.com/locate/neulet
Differential cerebellar activation related to perceived pain intensity during noxious thermal stimulation in humans: a functional magnetic resonance imaging study C. Helmchen a,*, C. Mohr a, C. Erdmann a, D. Petersen b, M.F. Nitschke a a Department of Neurology, Medical University of Lu¨beck, Ratzeburger Allee 160, D-23538 Lu¨beck, Germany Department of Neuroradiology, Medical University of Lu¨beck, Ratzeburger Allee 160, D-23538 Lu¨beck, Germany
b
Received 24 July 2002; received in revised form 19 September 2002; accepted 3 October 2002
Abstract Little is known about the cerebellar involvement in pain processing in spite of the fact that the cerebellum probably plays a crucial role in pain-related behavior. Using functional magnetic resonance imaging we examined the differential cerebellar activation in 18 healthy subjects in relation to their perceived pain-intensity of noxious and non-noxious thermal stimuli. In contrast to non-noxious (408C) stimuli, noxious (48.58C) stimuli revealed activation in the deep cerebellar nuclei, anterior vermis and bilaterally in the cerebellar hemispheric lobule VI. With the same noxious stimulus (48.58C) there was differential cerebellar activation depending on the perceived pain intensity: high pain intensity ratings were associated with activation in ipsilateral hemispheric lobule III–VI, deep cerebellar nuclei and in the anterior vermis (lobule III). This differential cerebellar activation pattern probably reflects not only somatosensory processing but also perceived pain intensity that may be important for cerebellar modulation of nociceptive circuits. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Pain; Cerebellum; Vermis; Nociception; Functional magnetic resonance imaging
Little is known about the cerebellar involvement in pain processing. Although there is some evidence for nociceptive afferents to the anterior cerebellum [6] several positron emission tomography (PET) studies have shown cerebellar activation during pain as induced by experimental thermal noxious stimuli [2,3] and capsaicin [7,9] but without precise function-related localization yet. It remained unclear: (i) which cerebellar sites are activated by noxious as opposed to non-noxious thermal stimuli; and (ii) whether cerebellar activation is related to somatosensory, nociceptive, pain perceptual or motor signal processing. Using functional magnetic resonance imaging (fMRI) with its better spatial resolution we tried to precisely identify cerebellar anatomical regions which may encode thermal pain intensity as opposed to thermal non-noxious stimuli. Using fMRI, we examined the differential cerebellar activation of 18 right-handed healthy volunteers (nine male, nine female subjects; ages 20–45) during thermal noxious * Corresponding author. Tel.: 149-451-500-2927; fax: 149-451500-2489. E-mail address: [email protected] (C. Helmchen).
(48.58C) and non-noxious (408C) stimulation of the right hand in relation to their perceived pain-intensity. All subjects gave informed consent after the study has been approved by the local ethics committee. In a block design alternating episodes of baseline temperature (for 30 s) and two different thermal stimuli (40 or 48.58C for 24 s) were delivered separately and repetitively (six times each) by a Peltier thermode (3 £ 3 cm thermo-conducting surface; TSA II, Medoc Inc., Israel) tapped at the finger tips of the right hand (digits II and III). Prior and after the imaging session, the subjects rated pain intensity on a visual analog scale (VAS; 1–10, 0 ¼ no heat sensation and no pain, 5 ¼ hot and barely painful, 10 ¼ maximal, unbearable pain). In all subjects 48.58C-stimuli were above and 408C-stimuli below individual thermal hot thresholds. Mean pain ratings for noxious stimuli on a VAS did not differ significantly between females (mean VAS: 5.5) and males (mean VAS: 7.1). Subjects were divided into two groups by means of their subjective pain rating following the noxious 48.58C stimulus, i.e. the pain-group (A) had an VAS .5 (mean 7.8 ^ 1.1, n ¼ 11) whereas the non-pain group (B) rated VAS ,5 (mean 3.5 ^ 1.3, n ¼ 7).
0304-3940/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S0 30 4- 39 40 ( 02) 0 11 64- 3
C. Helmchen et al. / Neuroscience Letters 335 (2003) 202–206
Functional imaging was performed at 1.5 T (Siemens Symphony, Erlangen, Germany; standard head coil) with echo-planar spin-echo imaging (TR/TE ¼ 187/81 ms per slice; matrix size ¼ 128, Flip angle ¼ 908, voxel size 2.0 £ 2.0 £ 5.0 mm, 38 slices). In a block design, for each stimulus (40 or 48.58C) a sequence of measurements consisted of one baseline measurement (5 £ 6 ¼ 30 s), followed by six blocks of 54 s each (24 s stimulus and 30 s baseline). An anatomical data set was obtained using a 3D T1-weighted GE sequence (TR ¼ 40 ms, TE ¼ 5 ms, matrix size ¼ 256 £ 256, 2.5 mm slice thickness) perpendicular to the AC-PC line extending throughout the cerebellum. Image analysis was performed on a PC (LINUX 7.1) using Matlab (Mathworks Inc., Natiek, MA, USA) and the statistical parametric mapping package (SPM 99, Wellcome Department of Cognitive Neurology, London, UK) [17]. Following motion correction and smoothing (5 mm isotropic Gaussian kernel) individual and group data were analyzed. For spatial normalization, fMRI data were aligned to structural MRI data and transformed to Talairach space. In a first step we looked for differences between the active (noxious) condition and the baseline using a contrast matrix design (random effect analysis, P , 0:005) utilizing box-car function. Only pixels passing a threshold of Z ¼ 2:58 (P , 0:005) were clustered. Clusters of at least five voxels were considered significant. In a next step, Talairach coordinates for group-averaged brain activation in response to the two different thermal stimuli were compared. Differences between volumes of activation were determined for each subject and, in a contrast matrix design, for the between-task comparison (noxious versus non-noxious stimulation) on a P , 0:005 level (t-test, SPM) [17]. The SPM random effect analysis at P , 0:005 was used to account for interindividual variance. The anatomical localization was determined by using the appropriate axial and sagittal sections according to the Talairach-transformed coordinates of the MRI atlas of the human cerebellum [16]. All 18 subjects showed task-related signal increases in the cerebellar hemispheres during noxious thermal stimulation (48.58C). Noxious stimulation (n ¼ 18) elicited bilateral activation (P , 0:005) in both cerebellar hemispheres and the anterior vermis which is given for horizontal slices (no figure). In the caudal parts [Z: 250 to 232] of the cerebellar hemispheres there was a bilateral activation in lobuli VIIIA and VIIIB (Z: 250 to 238) and contralateral activation in the left Crus I [242 to 234] and lobule VI [Z: 236 to 232]. In the rostrocaudal middle of the hemispheric lobuli VI [Z: 232 to 222] there was nearly symmetrical bilateral activation which, however, differed depending on the perceived pain intensity (see below). More rostrally, the activation was found ipsilaterally more anterior extending over the primary fissure into lobuli V and IV as compared with the contralateral activation which remained in lobule VI (Z: up to 218). In the rostral hemisphere there was activation in lobuli IV and V, to a smaller extent in lobule III [Z: 220 to 218]. The vermis was activated in its anterior parts involving lobuli
203
V–III [Z: 228 to 226]. Caudally, dorsal and immediately adjacent to lobule III, there was activation corresponding to the location of the midline deep cerebellar nuclei (e.g. nuclei fastigii) [Z: 232 to 228] and further caudally in lobule VIIIB, with only small activation in the posterior vermal lobule VII. All of the subjects (n ¼ 18) tested for thermal nonnoxious (408C) stimulation noticed the 408C-stimulus as warm but none appreciated it as painful or unpleasant. Activation (SPM, P , 0:005) was found in the contralateral left hemisphere in Crus I [Z: 238 to 230] and the dorsomedial aspect of lobule VI [Z: 230 to 234]. In contrast, there was no vermal activation (no figure). When activation was compared intraindividually in subjects during noxious hot (48.58C) versus non-noxious warm (408C) stimuli both stimulations activated the dorsomedial aspects of contralateral hemispheric lobule VI and Crus I. The between-task comparison in all 18 subjects (contrast matrix: noxious minus non-noxious stimulation) on a P , 0:005 level (SPM) revealed bilateral (Z: 222 to 230) but a stronger ipsilateral activation of hemispheric lobule VI, extending into vermal lobule V–III (Z: 222 to 218) (Fig. 1). Midline activation was found in vermal lobule III [Z: 228 to 218] and the deep cerebellar nuclei [Z: 230 to 228] (Table 1). The cerebellar activation pattern differed between group (A) and group (B) during the same noxious (48.58C) stimulation depending on the perceived pain intensity (random effect analysis, P , 0:005): group (A) showed a stronger activation
Table 1 Areas of significant differences (contrast matrix of between-task analysis, random effect, P , 0.005, uncorrected) in cerebellar activation (lobules with X, Y, X-coordinates and Z-score of statistical image analysis) sorted by ascending Z-values [mm] as depicted from transversal slices (SPM) and anatomical reconstruction using 3D-atlas of the human cerebellum [16] during noxious (48.58C) as compared with non-noxious (408C) stimulation for the right (negative) and left (positive) cerebellum a Lobule
X [mm]
Y [mm]
Z [mm]
Z-score
V V, VI III, IV VI VI III V, VI VI Nuclei VI, nuclei Cr I VI Cr I VI IX VIII
24 222 14 24 30 4 218 230 10 28 244 14 12 230 0 24
262 260 240 262 258 256 256 258 250 268 260 264 274 244 256 250
216 220 220 224 228 228 228 228 230 230 232 234 236 236 238 246
3.02 3.53 2.91 3.18 4.33 3.92 3.89 3.87 3.63 3.04 3.92 3.79 3.31 3.11 2.94 3.22
a
Cr I ¼ Crus I; nuclei ¼ cerebellar nuclei.
204
C. Helmchen et al. / Neuroscience Letters 335 (2003) 202–206
Fig. 1. Areas of pain-related cerebellar activation in the between-task comparison in all 18 subjects (contrast matrix: noxious minus nonnoxious stimulation; random effect analysis) on a P , 0:005 level. Only areas responding to noxious (48.58C) but not to non-noxious stimuli (408C) are shown. According to the atlas of the human cerebellum [16] pain-related activation sites are shown from caudal to rostral consecutive cerebellar transversal slices. Note the activation in the deep cerebellar nuclei and the anterior vermis. The bilateral frontotemporal cortical activation reflects propagation effects of the strong bilateral insula activation (not shown).
in ipsilateral anterior hemispheric lobules III–VI, deep cerebellar nuclei (Fig. 2) and anterior vermis (lobule III), contralaterally in lobule VIII (Z: 246 to 248) whereas group (B) had a stronger contralateral lobule VI activation. In a direct contrast calculation (P , 0:05) only ipsilateral anterior hemispheric lobules III–VI were activated. Based on previous imaging [1,3,4,7,9,12] and physiological [6] studies we provide evidence for our initial hypothesis that the anterior cerebellum is not only involved in nociceptive processing but also in pain perception. In accordance with a previous PET study [4] using the same non-noxious thermal stimulus (408C) we found activation in the contralateral caudal cerebellar hemisphere, more specifically in the dorsomedial aspect of lobule VI. This was one of the common sites that are activated by noxious and non-noxious thermal stimuli at the fingertips indicating some sensory or motor function not exclusively related to pain. Our comparison of noxious versus non-noxious thermal stimulation clearly showed that the paramedian anterior hemispheric lobuli III–V, the anterior vermis and the medial deep cerebellar nuclei are only activated during noxious
thermal stimuli. Thus, this activation probably does not reflect pure sensory stimulation. In contrast, the contralateral activation of the caudal parts of the hemispheric lobuli VI and Crus I/II of group (A) is not necessarily involved in nociceptive transmission since they were also partly activated during non-noxious thermal stimuli. These areas partly coincide caudally with Larsell’s hemispheric (H) lobuli VII that have been shown to be activated during imagined but not exerted right hand movements [8]. It possibly indicates nociceptive-motor integration, i.e. sensory-guided preparation of motor (nocifensive) behavior during the perception of noxious stimulation. The mid-cerebellar activation (Z ¼ 228 to 2 24) in hemispheric lobuli VI is anatomically compatible with pain-intensity-related bilateral but predominantly ipsilateral cerebellar activation in a previous [1] and recent [3] PET study. This is in accordance with the ipsilateral C-fiber afferents to the cerebellum [6]. Since this ipsilateral activation was not only found in the comparison between noxious versus non-noxious stimuli but in particular between different pain perceptions (group A versus group B) it probably
C. Helmchen et al. / Neuroscience Letters 335 (2003) 202–206
205
Fig. 2. Areas of cerebellar activation (P , 0:005) during noxious stimulation (48.58C) of group (A) (upper panel) and group (B) (lower panel) are shown in mid-cerebellar transversal slices at the level of the deep cerebellar nuclei according to the atlas of the human cerebellum [16]. There is an ipsilateral hemispheric activation in group (A) and a midline activation in the deep cerebellar nuclei and the vermis.
reflects not only nociceptive transmission but also perceived pain, particularly in group (A). The vermal activation is in accordance with PET studies [2,7] and one fMRI investigation [12]. More specifically, several lines of evidence indicate that the anterior vermis (specifically lobule III) and the deep cerebellar nuclei are related to perceived pain of the noxious stimuli. First, they were not activated during non-noxious warm stimuli, i.e. it probably does not encode sensory (thermal) transmission but rather pain. Second, we have previously shown that pure tactile-sensory stimulation elicits activation more laterally in Larsell lobuli IV and V [10,11]. Third, lobule III receives nociceptive spinal afferents [6] and cerebellar activity is modulated by nociceptive stimulation [13]. Fourth, as the vermis, the medial cerebellar nuclei contain opioid receptors [15] and morphine-microinjections into the anterior cerebellum produced analgesia in rats [5]. The deep cerebellar nuclei are thought to be involved in sensory gating of spinal nociceptive responses [13], presumably by inhibiting brainstem antinociceptive neurons that activate antinociceptive descending inhibitory pathways [14]. Fifth, anticipation is unlikely to contribute because anticipation of pain activates more dorsolaterally from our vermal and deep cerebellar nuclei activation [12]. We conclude that vermis and deep cerebellar nuclei show pain-related activation that may be important for cerebellar modulation of nociceptive circuits. [1] Casey, K.L., Minoshima, S., Morrow, T.J. and Koeppe, R.A.,
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
Comparison of human cerebral activation pattern during cutaneous warmth, heat pain, and deep cold pain, J. Neurophysiol., 76 (1996) 571–581. Casey, K.L., Morrow, T.J., Lorenz, J. and Minoshima, S., Temporal and spatial dynamics of human forebrain activity during heat pain: analysis by positron emission tomography, J. Neurophysiol., 85 (2001) 951–959. Coghill, R.C., Gilron, I. and Iadarola, M.J., Hemispheric lateralization of somatosensory processing, J. Neurophysiol., 85 (2001) 2602–2616. Derbyshire, S.W. and Jones, A.K., Cerebral responses to a continual tonic pain stimulus measured using positron emission tomography, Pain, 76 (1998) 127–135. Dey, P.K. and Ray, A.K., Anterior cerebellum as a site for morphine analgesia and post-stimulation analgesia, Indian J. Physiol. Pharmacol., 26 (1982) 3–12. Ekerot, C.F., Garwicz, M. and Schouenborg, J., The postsynaptic dorsal column pathway mediates cutaneous nociceptive information to cerebellar climbing fibers in the cat, J. Physiol., 441 (1991) 257–274. Iadarola, M.J., Berman, K.F., Zeffiro, T.A., Byas Smith, M.G., Gracely, R.H., Max, M.B. and Bennett, G.J., Neural activation during acute capsaicin-evoked pain and allodynia assessed with PET, Brain, 121 (1998) 931–947. Lotze, M., Montoya, P., Erb, M., Hulsmann, E., Flor, H., Klose, U., Birbaumer, N. and Grodd, W., Activation of cortical and cerebellar motor areas during executed and imagined hand movements: an fMRI study, J. Cogn. Neurosci., 11 (1999) 491–501. May, A., Kaube, H., Bu¨ chel, C., Eichten, C., Rijntjes, M., Juptner, M., Weiller, C. and Diener, H.C., Experimental cranial pain elicited by capsaicin: a PET study, Pain, 74 (1998) 61–66. Nitschke, M.F., Hahn, C., Melchert, U.H., Handels, H. and Wessel, K., Activation of the human anterior cerebellum by finger movements and sensory finger stimulation: a
206
[11]
[12]
[13]
[14]
C. Helmchen et al. / Neuroscience Letters 335 (2003) 202–206 functional magnetic resonance imaging study, J. Neuroimaging, 8 (1998) 127–131. Nitschke, M.F., Kleinschmidt, A., Wessel, K. and Frahm, J., Somatotopic motor representation in the human anterior cerebellum detected by functional magnetic resonance neuroimaging, Brain, 119 (1996) 1023–1029. Ploghaus, A., Tracey, I., Gati, J.S., Clare, S., Menon, R.S., Matthews, P.M. and Rawlins, J.N., Dissociating pain from its anticipation in the human brain, Science, 284 (1999) 1979–1981. Saab, C.Y. and Willis, W.D., Nociceptive visceral stimulation modulates the activity of cerebellar Purkinje cells, Exp. Brain Res., 140 (2001) 122–126. Saab, C.Y. and Willis, W.D., Cerebellar stimulation modu-
lates the intensity of a visceral nociceptive reflex in the rat, Exp. Brain Res., 146 (2002) 117–121. [15] Schadrack, J., Willoch, F., Platzer, S., Bartenstein, P., Mahal, B., Dworzak, D., Wester, H.J., Zieglga¨ nsberger, W. and To¨ lle, T.R., Opioid receptors in the human cerebellum: evidence from [11C]diprenorphine PET, mRNA expression and autoradiography, NeuroReport, 10 (1999) 619–624. [16] Schmahmann, J.D., Doyon, J., Toga, A., Petrides, M. and Evans, A., MRI Atlas of the Human Cerebellum, Academic Press, New York, 2000. [17] Turner, R., Howseman, A., Rees, G.E., Josephs, O. and Friston, K., Functional magnetic resonance imaging of the human brain: data acquisition and analysis, Exp. Brain Res., 123 (1998) 5–12.
Differential cerebellar activation related to perceived pain intensity during noxious thermal stimulation in humans: a functional magnetic resonance imaging study C. Helmchen a,*, C. Mohr a, C. Erdmann a, D. Petersen b, M.F. Nitschke a a Department of Neurology, Medical University of Lu¨beck, Ratzeburger Allee 160, D-23538 Lu¨beck, Germany Department of Neuroradiology, Medical University of Lu¨beck, Ratzeburger Allee 160, D-23538 Lu¨beck, Germany
b
Received 24 July 2002; received in revised form 19 September 2002; accepted 3 October 2002
Abstract Little is known about the cerebellar involvement in pain processing in spite of the fact that the cerebellum probably plays a crucial role in pain-related behavior. Using functional magnetic resonance imaging we examined the differential cerebellar activation in 18 healthy subjects in relation to their perceived pain-intensity of noxious and non-noxious thermal stimuli. In contrast to non-noxious (408C) stimuli, noxious (48.58C) stimuli revealed activation in the deep cerebellar nuclei, anterior vermis and bilaterally in the cerebellar hemispheric lobule VI. With the same noxious stimulus (48.58C) there was differential cerebellar activation depending on the perceived pain intensity: high pain intensity ratings were associated with activation in ipsilateral hemispheric lobule III–VI, deep cerebellar nuclei and in the anterior vermis (lobule III). This differential cerebellar activation pattern probably reflects not only somatosensory processing but also perceived pain intensity that may be important for cerebellar modulation of nociceptive circuits. q 2002 Elsevier Science Ireland Ltd. All rights reserved. Keywords: Pain; Cerebellum; Vermis; Nociception; Functional magnetic resonance imaging
Little is known about the cerebellar involvement in pain processing. Although there is some evidence for nociceptive afferents to the anterior cerebellum [6] several positron emission tomography (PET) studies have shown cerebellar activation during pain as induced by experimental thermal noxious stimuli [2,3] and capsaicin [7,9] but without precise function-related localization yet. It remained unclear: (i) which cerebellar sites are activated by noxious as opposed to non-noxious thermal stimuli; and (ii) whether cerebellar activation is related to somatosensory, nociceptive, pain perceptual or motor signal processing. Using functional magnetic resonance imaging (fMRI) with its better spatial resolution we tried to precisely identify cerebellar anatomical regions which may encode thermal pain intensity as opposed to thermal non-noxious stimuli. Using fMRI, we examined the differential cerebellar activation of 18 right-handed healthy volunteers (nine male, nine female subjects; ages 20–45) during thermal noxious * Corresponding author. Tel.: 149-451-500-2927; fax: 149-451500-2489. E-mail address: [email protected] (C. Helmchen).
(48.58C) and non-noxious (408C) stimulation of the right hand in relation to their perceived pain-intensity. All subjects gave informed consent after the study has been approved by the local ethics committee. In a block design alternating episodes of baseline temperature (for 30 s) and two different thermal stimuli (40 or 48.58C for 24 s) were delivered separately and repetitively (six times each) by a Peltier thermode (3 £ 3 cm thermo-conducting surface; TSA II, Medoc Inc., Israel) tapped at the finger tips of the right hand (digits II and III). Prior and after the imaging session, the subjects rated pain intensity on a visual analog scale (VAS; 1–10, 0 ¼ no heat sensation and no pain, 5 ¼ hot and barely painful, 10 ¼ maximal, unbearable pain). In all subjects 48.58C-stimuli were above and 408C-stimuli below individual thermal hot thresholds. Mean pain ratings for noxious stimuli on a VAS did not differ significantly between females (mean VAS: 5.5) and males (mean VAS: 7.1). Subjects were divided into two groups by means of their subjective pain rating following the noxious 48.58C stimulus, i.e. the pain-group (A) had an VAS .5 (mean 7.8 ^ 1.1, n ¼ 11) whereas the non-pain group (B) rated VAS ,5 (mean 3.5 ^ 1.3, n ¼ 7).
0304-3940/02/$ - see front matter q 2002 Elsevier Science Ireland Ltd. All rights reserved. PII: S0 30 4- 39 40 ( 02) 0 11 64- 3
C. Helmchen et al. / Neuroscience Letters 335 (2003) 202–206
Functional imaging was performed at 1.5 T (Siemens Symphony, Erlangen, Germany; standard head coil) with echo-planar spin-echo imaging (TR/TE ¼ 187/81 ms per slice; matrix size ¼ 128, Flip angle ¼ 908, voxel size 2.0 £ 2.0 £ 5.0 mm, 38 slices). In a block design, for each stimulus (40 or 48.58C) a sequence of measurements consisted of one baseline measurement (5 £ 6 ¼ 30 s), followed by six blocks of 54 s each (24 s stimulus and 30 s baseline). An anatomical data set was obtained using a 3D T1-weighted GE sequence (TR ¼ 40 ms, TE ¼ 5 ms, matrix size ¼ 256 £ 256, 2.5 mm slice thickness) perpendicular to the AC-PC line extending throughout the cerebellum. Image analysis was performed on a PC (LINUX 7.1) using Matlab (Mathworks Inc., Natiek, MA, USA) and the statistical parametric mapping package (SPM 99, Wellcome Department of Cognitive Neurology, London, UK) [17]. Following motion correction and smoothing (5 mm isotropic Gaussian kernel) individual and group data were analyzed. For spatial normalization, fMRI data were aligned to structural MRI data and transformed to Talairach space. In a first step we looked for differences between the active (noxious) condition and the baseline using a contrast matrix design (random effect analysis, P , 0:005) utilizing box-car function. Only pixels passing a threshold of Z ¼ 2:58 (P , 0:005) were clustered. Clusters of at least five voxels were considered significant. In a next step, Talairach coordinates for group-averaged brain activation in response to the two different thermal stimuli were compared. Differences between volumes of activation were determined for each subject and, in a contrast matrix design, for the between-task comparison (noxious versus non-noxious stimulation) on a P , 0:005 level (t-test, SPM) [17]. The SPM random effect analysis at P , 0:005 was used to account for interindividual variance. The anatomical localization was determined by using the appropriate axial and sagittal sections according to the Talairach-transformed coordinates of the MRI atlas of the human cerebellum [16]. All 18 subjects showed task-related signal increases in the cerebellar hemispheres during noxious thermal stimulation (48.58C). Noxious stimulation (n ¼ 18) elicited bilateral activation (P , 0:005) in both cerebellar hemispheres and the anterior vermis which is given for horizontal slices (no figure). In the caudal parts [Z: 250 to 232] of the cerebellar hemispheres there was a bilateral activation in lobuli VIIIA and VIIIB (Z: 250 to 238) and contralateral activation in the left Crus I [242 to 234] and lobule VI [Z: 236 to 232]. In the rostrocaudal middle of the hemispheric lobuli VI [Z: 232 to 222] there was nearly symmetrical bilateral activation which, however, differed depending on the perceived pain intensity (see below). More rostrally, the activation was found ipsilaterally more anterior extending over the primary fissure into lobuli V and IV as compared with the contralateral activation which remained in lobule VI (Z: up to 218). In the rostral hemisphere there was activation in lobuli IV and V, to a smaller extent in lobule III [Z: 220 to 218]. The vermis was activated in its anterior parts involving lobuli
203
V–III [Z: 228 to 226]. Caudally, dorsal and immediately adjacent to lobule III, there was activation corresponding to the location of the midline deep cerebellar nuclei (e.g. nuclei fastigii) [Z: 232 to 228] and further caudally in lobule VIIIB, with only small activation in the posterior vermal lobule VII. All of the subjects (n ¼ 18) tested for thermal nonnoxious (408C) stimulation noticed the 408C-stimulus as warm but none appreciated it as painful or unpleasant. Activation (SPM, P , 0:005) was found in the contralateral left hemisphere in Crus I [Z: 238 to 230] and the dorsomedial aspect of lobule VI [Z: 230 to 234]. In contrast, there was no vermal activation (no figure). When activation was compared intraindividually in subjects during noxious hot (48.58C) versus non-noxious warm (408C) stimuli both stimulations activated the dorsomedial aspects of contralateral hemispheric lobule VI and Crus I. The between-task comparison in all 18 subjects (contrast matrix: noxious minus non-noxious stimulation) on a P , 0:005 level (SPM) revealed bilateral (Z: 222 to 230) but a stronger ipsilateral activation of hemispheric lobule VI, extending into vermal lobule V–III (Z: 222 to 218) (Fig. 1). Midline activation was found in vermal lobule III [Z: 228 to 218] and the deep cerebellar nuclei [Z: 230 to 228] (Table 1). The cerebellar activation pattern differed between group (A) and group (B) during the same noxious (48.58C) stimulation depending on the perceived pain intensity (random effect analysis, P , 0:005): group (A) showed a stronger activation
Table 1 Areas of significant differences (contrast matrix of between-task analysis, random effect, P , 0.005, uncorrected) in cerebellar activation (lobules with X, Y, X-coordinates and Z-score of statistical image analysis) sorted by ascending Z-values [mm] as depicted from transversal slices (SPM) and anatomical reconstruction using 3D-atlas of the human cerebellum [16] during noxious (48.58C) as compared with non-noxious (408C) stimulation for the right (negative) and left (positive) cerebellum a Lobule
X [mm]
Y [mm]
Z [mm]
Z-score
V V, VI III, IV VI VI III V, VI VI Nuclei VI, nuclei Cr I VI Cr I VI IX VIII
24 222 14 24 30 4 218 230 10 28 244 14 12 230 0 24
262 260 240 262 258 256 256 258 250 268 260 264 274 244 256 250
216 220 220 224 228 228 228 228 230 230 232 234 236 236 238 246
3.02 3.53 2.91 3.18 4.33 3.92 3.89 3.87 3.63 3.04 3.92 3.79 3.31 3.11 2.94 3.22
a
Cr I ¼ Crus I; nuclei ¼ cerebellar nuclei.
204
C. Helmchen et al. / Neuroscience Letters 335 (2003) 202–206
Fig. 1. Areas of pain-related cerebellar activation in the between-task comparison in all 18 subjects (contrast matrix: noxious minus nonnoxious stimulation; random effect analysis) on a P , 0:005 level. Only areas responding to noxious (48.58C) but not to non-noxious stimuli (408C) are shown. According to the atlas of the human cerebellum [16] pain-related activation sites are shown from caudal to rostral consecutive cerebellar transversal slices. Note the activation in the deep cerebellar nuclei and the anterior vermis. The bilateral frontotemporal cortical activation reflects propagation effects of the strong bilateral insula activation (not shown).
in ipsilateral anterior hemispheric lobules III–VI, deep cerebellar nuclei (Fig. 2) and anterior vermis (lobule III), contralaterally in lobule VIII (Z: 246 to 248) whereas group (B) had a stronger contralateral lobule VI activation. In a direct contrast calculation (P , 0:05) only ipsilateral anterior hemispheric lobules III–VI were activated. Based on previous imaging [1,3,4,7,9,12] and physiological [6] studies we provide evidence for our initial hypothesis that the anterior cerebellum is not only involved in nociceptive processing but also in pain perception. In accordance with a previous PET study [4] using the same non-noxious thermal stimulus (408C) we found activation in the contralateral caudal cerebellar hemisphere, more specifically in the dorsomedial aspect of lobule VI. This was one of the common sites that are activated by noxious and non-noxious thermal stimuli at the fingertips indicating some sensory or motor function not exclusively related to pain. Our comparison of noxious versus non-noxious thermal stimulation clearly showed that the paramedian anterior hemispheric lobuli III–V, the anterior vermis and the medial deep cerebellar nuclei are only activated during noxious
thermal stimuli. Thus, this activation probably does not reflect pure sensory stimulation. In contrast, the contralateral activation of the caudal parts of the hemispheric lobuli VI and Crus I/II of group (A) is not necessarily involved in nociceptive transmission since they were also partly activated during non-noxious thermal stimuli. These areas partly coincide caudally with Larsell’s hemispheric (H) lobuli VII that have been shown to be activated during imagined but not exerted right hand movements [8]. It possibly indicates nociceptive-motor integration, i.e. sensory-guided preparation of motor (nocifensive) behavior during the perception of noxious stimulation. The mid-cerebellar activation (Z ¼ 228 to 2 24) in hemispheric lobuli VI is anatomically compatible with pain-intensity-related bilateral but predominantly ipsilateral cerebellar activation in a previous [1] and recent [3] PET study. This is in accordance with the ipsilateral C-fiber afferents to the cerebellum [6]. Since this ipsilateral activation was not only found in the comparison between noxious versus non-noxious stimuli but in particular between different pain perceptions (group A versus group B) it probably
C. Helmchen et al. / Neuroscience Letters 335 (2003) 202–206
205
Fig. 2. Areas of cerebellar activation (P , 0:005) during noxious stimulation (48.58C) of group (A) (upper panel) and group (B) (lower panel) are shown in mid-cerebellar transversal slices at the level of the deep cerebellar nuclei according to the atlas of the human cerebellum [16]. There is an ipsilateral hemispheric activation in group (A) and a midline activation in the deep cerebellar nuclei and the vermis.
reflects not only nociceptive transmission but also perceived pain, particularly in group (A). The vermal activation is in accordance with PET studies [2,7] and one fMRI investigation [12]. More specifically, several lines of evidence indicate that the anterior vermis (specifically lobule III) and the deep cerebellar nuclei are related to perceived pain of the noxious stimuli. First, they were not activated during non-noxious warm stimuli, i.e. it probably does not encode sensory (thermal) transmission but rather pain. Second, we have previously shown that pure tactile-sensory stimulation elicits activation more laterally in Larsell lobuli IV and V [10,11]. Third, lobule III receives nociceptive spinal afferents [6] and cerebellar activity is modulated by nociceptive stimulation [13]. Fourth, as the vermis, the medial cerebellar nuclei contain opioid receptors [15] and morphine-microinjections into the anterior cerebellum produced analgesia in rats [5]. The deep cerebellar nuclei are thought to be involved in sensory gating of spinal nociceptive responses [13], presumably by inhibiting brainstem antinociceptive neurons that activate antinociceptive descending inhibitory pathways [14]. Fifth, anticipation is unlikely to contribute because anticipation of pain activates more dorsolaterally from our vermal and deep cerebellar nuclei activation [12]. We conclude that vermis and deep cerebellar nuclei show pain-related activation that may be important for cerebellar modulation of nociceptive circuits. [1] Casey, K.L., Minoshima, S., Morrow, T.J. and Koeppe, R.A.,
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
Comparison of human cerebral activation pattern during cutaneous warmth, heat pain, and deep cold pain, J. Neurophysiol., 76 (1996) 571–581. Casey, K.L., Morrow, T.J., Lorenz, J. and Minoshima, S., Temporal and spatial dynamics of human forebrain activity during heat pain: analysis by positron emission tomography, J. Neurophysiol., 85 (2001) 951–959. Coghill, R.C., Gilron, I. and Iadarola, M.J., Hemispheric lateralization of somatosensory processing, J. Neurophysiol., 85 (2001) 2602–2616. Derbyshire, S.W. and Jones, A.K., Cerebral responses to a continual tonic pain stimulus measured using positron emission tomography, Pain, 76 (1998) 127–135. Dey, P.K. and Ray, A.K., Anterior cerebellum as a site for morphine analgesia and post-stimulation analgesia, Indian J. Physiol. Pharmacol., 26 (1982) 3–12. Ekerot, C.F., Garwicz, M. and Schouenborg, J., The postsynaptic dorsal column pathway mediates cutaneous nociceptive information to cerebellar climbing fibers in the cat, J. Physiol., 441 (1991) 257–274. Iadarola, M.J., Berman, K.F., Zeffiro, T.A., Byas Smith, M.G., Gracely, R.H., Max, M.B. and Bennett, G.J., Neural activation during acute capsaicin-evoked pain and allodynia assessed with PET, Brain, 121 (1998) 931–947. Lotze, M., Montoya, P., Erb, M., Hulsmann, E., Flor, H., Klose, U., Birbaumer, N. and Grodd, W., Activation of cortical and cerebellar motor areas during executed and imagined hand movements: an fMRI study, J. Cogn. Neurosci., 11 (1999) 491–501. May, A., Kaube, H., Bu¨ chel, C., Eichten, C., Rijntjes, M., Juptner, M., Weiller, C. and Diener, H.C., Experimental cranial pain elicited by capsaicin: a PET study, Pain, 74 (1998) 61–66. Nitschke, M.F., Hahn, C., Melchert, U.H., Handels, H. and Wessel, K., Activation of the human anterior cerebellum by finger movements and sensory finger stimulation: a
206
[11]
[12]
[13]
[14]
C. Helmchen et al. / Neuroscience Letters 335 (2003) 202–206 functional magnetic resonance imaging study, J. Neuroimaging, 8 (1998) 127–131. Nitschke, M.F., Kleinschmidt, A., Wessel, K. and Frahm, J., Somatotopic motor representation in the human anterior cerebellum detected by functional magnetic resonance neuroimaging, Brain, 119 (1996) 1023–1029. Ploghaus, A., Tracey, I., Gati, J.S., Clare, S., Menon, R.S., Matthews, P.M. and Rawlins, J.N., Dissociating pain from its anticipation in the human brain, Science, 284 (1999) 1979–1981. Saab, C.Y. and Willis, W.D., Nociceptive visceral stimulation modulates the activity of cerebellar Purkinje cells, Exp. Brain Res., 140 (2001) 122–126. Saab, C.Y. and Willis, W.D., Cerebellar stimulation modu-
lates the intensity of a visceral nociceptive reflex in the rat, Exp. Brain Res., 146 (2002) 117–121. [15] Schadrack, J., Willoch, F., Platzer, S., Bartenstein, P., Mahal, B., Dworzak, D., Wester, H.J., Zieglga¨ nsberger, W. and To¨ lle, T.R., Opioid receptors in the human cerebellum: evidence from [11C]diprenorphine PET, mRNA expression and autoradiography, NeuroReport, 10 (1999) 619–624. [16] Schmahmann, J.D., Doyon, J., Toga, A., Petrides, M. and Evans, A., MRI Atlas of the Human Cerebellum, Academic Press, New York, 2000. [17] Turner, R., Howseman, A., Rees, G.E., Josephs, O. and Friston, K., Functional magnetic resonance imaging of the human brain: data acquisition and analysis, Exp. Brain Res., 123 (1998) 5–12.