BRAIN
AND
LANGUAGE
39, 1-13 (I!@@
The Role of the Left and Right Thalamus in Language Asymmetry: Dichotic Listening in Parkinson Patients Undergoing Stereotactic Thalamotomy KENNETH HUCDAHL,* *Depurtment
KNUT WESTER,~ AND ARVE
of Somatic Psychology, Neurosurgery,
ASBJP)RNSEN*
University of Bergen and TDepartment University of Bergen
of
Fourteen patients with Parkinson’s disease (rigidity and unilateral tremor as main symptoms) were treated with stereotactic thalamotomy. The operation was performed in either the left or right ventrolateral thalamus, depending on which hand (or foot) was most affected with tremor. Nine patients were operated on the left side and four on the right side. All patients were tested for asymmetry of language functioning with dichotic presentations of consonant-vowel (CV) syllables. The patients had to indicate which of the two syllables he/she heard on each trial. Dichotic listening was performed before and after the operation, as well as during electrical stimulation of the VL nucleus just before the lesion was carried out. The results revealed essentially three things: an overall reduced right ear advantage (REA) in the patient group compared to what is known from studies of healthy individuals; an increase in REA during left-sided stimulations; and a marked reduction in REA after left-sided lesions. It is concluded that the present data support the notion put forward by Ojemann (e.g., 1975) of a lateralized activating gating mechanism in the left VL nucleus that gates access for language information to the appropriate cortical areas. The gating mechanism seems to be activated by stimulation, and deactivated after lesions. Dichotic listening may thus be a heuristic instrument in assessment of language functions in Parkinson patients. 0 1990 Academic Press, Inc.
In this paper we present data from dichotic listening performance in 14 Parkinson patients undergoing stereotactic thalamotomy that are relevant for elucidation of the role of the left and right thalamus in language asymmetry. The data also elucidate the processes involved in dichotic listening. The study involves comparisons of pre- and postlesion dichotic performance, as well as comparisons of dichotic listening scores during stimulation of the left or right ventro-lateral thalamus. Send reprint requests to Kenneth Hugdahl, Department of Somatic Psychology, University of Bergen, Aarstadveien 21, N-5009 Bergen, Norway. I 0093-934x/90 $3.00 Copyright 0 1990 by Academic Press. Inc. All rights of reproduction in any form reserved.
2
HUGDAHL,
WESTER,
AND ASBJORNSEN
For two decades after Spiegel and Wycis (1952) performed the first stereotactic thalamotomies in 1947, the operation was used frequently in the treatment of Parkinsonian tremor. When levodopa was introduced in the late 1960s interest in thalamotomy declined throughout the world, and very few such operations were performed in the following years (Tasker, Siqueira, Hawrylyshyn, & Organ, 1983). However, it gradually became clear that tremor was often unaffected by this type of medication. Tremor and to some degree rigidity are the main indications for thalamotomy, and interest again focused on this operation (Manen, Speelman, & Tans, 1984; Wester & Hauglie-Hanssen, in press). It is a well-known phenomenon that thalamic lesions have effects on language, and especially lesions in the left ventro-lateral and pulvinar nuclei. Ojemann and his colleagues (e.g., Mateer & Ojemann, 1983; Ojemann, 1975) have over the years reported that lesions in the lateral thalamus produce language disruptions characterized by dysnomia with perserveration, dysarthria, dysphasia, and sometimes intrusions of extraneous words. Others have reported not only that left-sided lesions affect expressive language functions but also that receptive verbal capacities deteriorate, including short-term memory (e.g., Riklan & Cooper, 1977; Crosson, 1984; Fedio & Van Buren, 1976). In addition, aphasia has often occurred after therapeutic lesions of the dominant ventral lateral (VL) thalamus (Crosson, 1988). However, since the aphasia usually disappears after a while, it may be the interaction of the VL nucleus and neighboring nuclei that causes the language dysfunction. One of these neighboring nuclei may be the reticular nucleus (see Fig. 2) believed to be a part of the nonspecific thalamic activation system. It may therefore be that the observed effects are due to shifts in attention/activation especially after left-sided lesions. These findings are underscored by reports of anatomical asymmetries in the cell architecture of the posterior lateral nucleus, reported by Galaburda and associates (e.g. Galaburda, 1986). Taken together, the data suggest: (1) that the human thalamus plays an important role in both language comprehension and speech expression, (2) that the left ventro-lateral and pulvinar nuclei seem more involved than the corresponding right nuclei, and (3) that asymmetry of language functions is not confined to cortical structures but also to subcortical structures which contribute to the overall left-to-right asymmetry (cf. Riklan & Cooper, 1977). A frequently used noninvasive method for the study of language asymmetry is dichotic listening (Kimura, 1967; see also Hugdahl, 1988). Dichotic listening is a relatively reliable technique for the study of language asymmetry (Bryden, 1988). The right ear advantage (indicative of a left hemisphere language specialization) is probably one of the more robust
ROLE OF THALAMUS
IN LANGUAGE
ASYMMETRY
3
empirical phenomen reported in experimental neuropsychology (Bryden, 1982; 1988; Hugdahl & Andersson, 1984; 1986). It could be argued that if the left thalamus is involved in language, then this should be reflected in dichotic listening performance when comparing patients with left vs. right thalamic lesions (and stimulations). The connections between dichotic listening and language asymmetry are perhaps still best explained by Kimura’s structural model from 1967. Kimura’s (1967) model may be described as a set of four assumptions about hemisphere asymmetry and brain functioning: 1. The left hemisphere is specialized for language. 2. Auditory input is more strongly represented in the contralateral hemisphere. 3. Dichotic stimulation initially lateralizes the input, as information in ipsilateral hemisphere is suppressed by the contralateral input. 4. Information from the nondominant hemisphere is transferred across the corpus callosum. Ear differences may thus reflect the relative processing superiority of one hemisphere or information loss due to interhemispheric transmission, or an interaction between these two factors. The only available data on dichotic listening performance in thalamic patients (at least to our knowledge) are the reports by Ojemann and Mateer in 1983 based on eight patients, the study by Ojemann in 1985, and the case study by Bhatnagar et al. in 1987. Mateer and Ojemann (1983) reported that frequency of correct responses increased during brain stimulation, and especially during left-sided stimulation. No data were reported concerning dichotic performance during or after surgery. Ojemann (1985) studied 12 patients (three had Parkinson’s disease) with dichotic presentations of words. He found significantly more words presented during left-sided stimulation correctly reported, with a slight tendency for a contralateral ear advantage. Bhatnagar et al. (1987) using a different approach, found a decrease in REA after a prolonged 20-min stimulation in the left thalamus, probably caused by setting the neurons temporarily out of function. Extrapolating from these findings, it may now be predicted that a relatively short stimulation of the left ventro-lateral (VL) thalamus should yield increased REA, while a lesion on the same side should result in a reduced REA or possibly a left ear advantage (LEA). Stimulation and lesion of the right VL thalamus should not yield any dramatic changes in dichotic listening reports. METHOD Patients The subjects were 14 Parkinson patients with a mean age of 64 (range 54 to 72). They were all referred to surgery from neurological departments because of severe and disabling
4
HUGDAHL,
WESTER, AND ASBJP)RNSEN
drug-resistant tremor and rigidity. Tremor was more profound in the upper limbs, and also more manifest on one side (either right or left). Nine patients were operated on the left side, and five on the right side. All 14 patients were right-handed. Handedness was defined with the help of a Norwegian translation of a modified version of the questionnaire developed by Raczkowski, Kalat, and Nebes (1974). The questionnaire consists of 15 questions related to manual items. To be classified as a right-hander, the patient had to indicate 13 of the 15 items as performed with the right hand (=87% consistency). Hearing acuity was determined by a Tegner screening audiometer. Patients with an imbalance in hearing levels between ears of more than 5 dB (from 20 dB) were not included in the study.
Apparatus The stereotactic technique and equipment were of the Leksell-type, and the target area was defined from CT-evaluation and located in the ventro-lateral nucleus 3 mm superior, 2 mm posterior, and 14 mm lateral (left or right) to the midpoint of the distance between the anterior (AC) and posterior (PC) commissures. A small hole with about l-cm diameter was drilled just in front of the coronal seam, and two electrodes, having a diameter of 1.5 mm, were inserted into the hole in the skull and placed in the target area (6 mm apart). The lesion was of the thermocoagulative type, and established by emitting radiofrequencies through the electrodes, thereby increasing the temperature between the tips of the electrodes to between 60 and 65°C. Electric stimulations in the target areas through the electrodes were performed before the lesion was carried out. This was done as a control that the electrodes were in the target area. There should usually be an effect on tremor during stimulation if the electrodes are correctly placed. Stimulation was performed with 200 Hz and with an average duration of 5 min. Figures 1 and 2 show the location of the electrodes, and the exact extension of the lesion as seen from the tip of both electrodes.
Stimulus materials The dichotic stimuli consisted of the six stop-consonants b, d, g. p, I, k, all paired with the vowel a to form six Consonant-Vowel (CV) syllables (bo, da, ga, etc.). The syllables were paired with each other for all possible combinations, thus yielding 36 dichotic pairs including the homonymic pairs. The dichotic tape consisted of three lists of 36 dichotic CV-pairs each. The 36 pairs were randomly ordered on each list. Each syllable had a duration of 320 msec, and synchronization of onset between channels was performed for both the consonant and vowel segment onsets. The interstimulus interval was 4 set * 1 sec. The dichotic tape was prepared on a PDP 1l/45 computer with A/D and D/A converters, and with a multiplexer. Due to the resolution of the multiplexer, maximum onset asynchrony was 0.5 msec. The stimuli were copied onto a chrome-dioxid cassette and played to the patient from a SONY WM DD-II minicassette player through SONY plug-in type earphones. The intensity of the output from the earphones were on the average 75 dbA (repeated measurements) when tested with a Bruel and Kjaer 2204 sound level meter.
Procedure The dichotic listening tests were performed the day before (after the patient had arrived to the hospital), just before surgery (inside the operation room), during surgery (stimuFIG. 1. Schematic drawing of electrode locahzations in the lateral ventro-lateral (VL) nucleus. The hatched rectangular area shows the actual lesion area.
ROLE
OF THALAMUS
IN LANGUAGE
ASYMMETRY
5
HUGDAHL,
z P F
WESTER,
AND ASBJORNSEN
ROLE OF THALAMUS
IN LANGUAGE
ASYMMETRY
7
lation), just after surgery (immediately after the surgeon had sutured the opening in the skull), and the day after surgery. The patient had to answer which syllable he/she heard on each trial, and this was marked on a special scoring sheet. In addition, everything that was said during the dichotic tests was taped on another minicassette tape recorder through a “mini” microphone clipped to the patient’s shirt. A short pause was inserted after each list of 36 trials to let the patient rest. During the stimulation and postoperation tests, only 36 or 72 trials could be run in some patients for clinical reasons (signs of fatigue, etc.).
RESULTS Presurgery
Data were collapsed across the two pre- and postsurgery tests taking the mean of the two. Mean percentages of correct recall separated for right and left ear input and for pre- and postsurgery are seen in Fig. 3. There was a slight (4.1%) right ear advantage (REA) presurgery for the patients that were-to-have left-sided lesions. The corresponding ear differences for the group with right-sided lesions were 4.5%. An analysis of variance based on a 2 (pre- vs. postsurgery) x 2 (right vs. left ear input) design did not yield any significant differences (all Fs < 1) (see Figs. 3 and 6). There were however six (of nine) patients with left-sided lesions that showed a REA and three (of five) patients with right-sided lesions. An equal frequency x2 evaluation showed these distributions to be nonsignificant (x2(1) = 1.00 and 0.20, respectively). Postsurgery
Mean percentages of correct recall for both groups of patients are also seen in Fig. 3. During the postsurgery test, the patients with left-sided lesions showed a dramatic reduction in overall performance. A 2 x 2
FIG. 3. Mean percentages of correct recall in the dichotic listening test for the right and left ear, and for the different test conditions. LL = left-sided lesion, RL = rightsided lesion.
8
HUGDAHL,
WESTER,
AND ASBJORNSEN
analysis of variance on a similar design as for the presurgery comparisons yielded a significant main effect of lesion side [F( 1, 12) = 5.03, p < .05]. Figure 3 reveals that the effect was due to reduced overall performance in the left-sided group. The interaction of side of lesion by ear of input was not significant [F(l, 12) = 2.37, n.s.). However, a separate t test comparing the right and left ear input for the group with left-sided lesions was significant If(8) = 2.01, p < .05]. This was due to better recall from the left compared to the right ear. Separate t tests were also performed for comparison of left ear presurgery vs. left ear postsurgery, and for right ear presurgery vs. right ear postsurgery, respectively. Both tests for the left lesion group were significant (both ts > 1.92, p < .0.5). No significant difference was observed for the right lesion group in the analysis of variance. All patients underwent a postoperative CT-control on the first or second postoperative day, and 3 to 6 months after the operation. Figure 4 shows a typical lesion from one patient. The lesion seen in Fig. 4 corresponds to the predetermined target coordinates. All patients had similar lesions verified through postoperative CT-scans.
FIG. 4. Horizontal CT-scans through the maximal extent of a thalamotomy lesion in the left thalamus. Left: First postoperative day. Right: Three months postoperatively. Arrow and square, arrow and circle, indicate lesion sites. Note the immediate postoperative edema. Please observe that the left hemisphere is shown to the right and vice versa according to radiological conventions.
ROLE OF THALAMLJS
IN LANGUAGE
ASYMMETRY
9
Brain Stimulation Mean correct recall during brain stimulation separately for each group is seen in Figs. 5 and 6. Note that the pre- and postsurgery data are once again shown in Figs. 5 and 6 to facilitate visual inspection of differences between test conditions. There were no significant effects in the overall analysis of variance. However, there was a significant difference for left and right ear correct recalls during left-sided stimulations when tested with the t statistic [t( 12) = 1.88, p < .05). The means show the difference to be 14.6% in favor of right ear performance. The right-side stimulation group did not reach significance [t(4) < l] (see also Fig. 5). DISCUSSION
To sum up the basic findings: There was generally a reduced REA in the patient group preoperatively compared to what is known from studies of normal healthy individuals. Usually, a 15-20% REA (Bryden, 1988) is obtained in healthy individuals which should be compared with the 4-5% REA seen in the present patient group preoperatively (see Fig. 3). Postoperatively, there was a rather dramatic reduction in recall from both ears in patients with left-sided lesions, while much less dramatic effects were seen in patients with right-sided lesions. However, during stimulation, there was an increase in ear advantage in both patient groups, although in different directions. Thus, while an increased REA was seen in the group with left-sided stimulation, an increased LEA (however non-significant) was seen in the group with right-sided stimulation.
LL E4airrstlm
FIG. 5. Mean percentages of correct and left ear during left-sided stimularion.
LL Fe3t-sug
recall in the dichotic listening test for the right LL = left-sided lesion; RL = right-sided lesion.
10
HLJGDAHL, WESTER, AND ASBJ@RNSEN
60
RL Pre-eug
iX
Brain-stim
FIG. 6. Mean percentagesof correct recall in the dichotic listening test for the right and left ear during right-sided stimulation. LL = left-sided lesion; RL = right-sidedlesion.
The present results are perhaps best explained with reference to the arousal model proposed by Ojemann (e.g., 1975). Mateer and Ojemann (1983) proposed a thalamic attentional, or alerting, mechanism which asymmetrically gates access to the cortex. It has been suggested that there is a gating mechanism in the medial geniculate body, and that this “gate” is more responsive to contralateral than to ipsilateral auditory signals. It has been established with both event-related potentials (ERPs) (e.g., Connolly, 1985) and with recordings of cerebral blood flow (CBF) (e.g., Maximilian, 1982) that the contralateral auditory pathways are both more preponderant and more efficient than the ipsilateral ones in transferring the auditory signal from the cochlea to the primary auditory temporal cortex. The predominance of the contralateral pathways from the cochlea to the temporal cortex may be a consequence of the stronger projection of second-order neurons to the inferior colliculus on the contralateral than on the ipsilateral side (Brodal, 1981). Although the input to, and output from, the inferior colliculi are both ipsi- and contralateral, the projections ascending to the colliculi are greater from the contralateral ear. However, the pathways ascending from the colliculi are greater on the ipsilateral side, thus favoring an ultimate representation of the contralateral ear in the auditory cortex (Brodal, 1981). Applying Ojemann’s model to the present data, we would like to suggest the existence of a language-related activating gating mechanism (AGM) in the ventro-lateral (VL) thalamic nuclei that asymmetrically “gates” information to the left hemisphere. During stimulation of the left VL nucleus, the AGM is set to “n,,” with the result of an increased REA due to facilitation of the contralateral right ear signals to the left hemisphere. The occurrence of electric current across the electrode tips
ROLE OF THALAMUS
IN LANGUAGE
ASYMMETRY
11
during the stimulation test may actually drive the neurons in the VL nucleus to open the AGM. However, when a lesion is applied to the same area, the AGM is set to “off’ position. The result should then be a decreased REA (or even a LEA) due to inhibitory blocking of the contralateral right ear signals to the left hemisphere. The thalamic alerting or activating circuitry may thus act as a switching mechanism to activate appropriate focal cortical areas involved in language processing. There is a preprocessing of verbal input that selectively gates input to the cortical hemispheres. Ojemann (1975) has argued that the arousal mechanism in the left thalamus is a “gate” also for access to short-term memory and that brain stimulation should evoke “an intense direction of attention to information [perhaps only verbally coded information” (p. 116)]. Thus, the left thalamus seems to be uniquely related to language functioning (see also Crosson, 1984). It may further be that only portions of the left thalamus, including the anterior superior pulvinar and medial central ventrolateral nuclei are related to language in this respect (cf. Ojemann, 1975). In addition to the presented results on DL performance after and during thalamic stimulation and lesion, several patients showed clinical signs of other language disturbance after left-sided, but not after right-sided lesions. These disturbances involved perseveration, dysarthria, and in one case speech arrest during stimulation. Similar findings were also reported by Ojemann (1975). However, following the findings by Penfield and Roberts (1959) of speech arrest occurring after cortical stimulation, this may be a response related rather to the cortical than to the subcortical level. Mateer and Ojemann (1983) reported that the total number of correct responses in dichotic listening performance made during either left- or right-sided stimulation was higher than was the total number of correct responses made under nonstimulation conditions. Increased accuracy of performance in their DL task was interestingly particularly “dramatic and consistent with left thalamic stimulation” (p. 184). Since we did not find an increase in REA in all nine patients stimulated on the left side, it may seem that our data are less “consistent” than the data presented by Mateer and Ojemann (1983). One explanation for this may be that whereas we used 200 Hz stimulation (over 3-5 min periods) with up to 108 CV trials, Mateer and Ojemann (1983) and Ojemann (1985) used 60 Hz stimulation and only 20 CV trials. Since a CV normally lasts about 350 msec, a 20-trial presentation would require stimulation of less than 1 min (with a reasonable intertrial interval). A 200-Hz stimulation may thus temporarily fatigue neurons in the VL nucleus, thus providing an effect on language more like the one observed after lesions. This seems however to be subject to individual variation. It should also be mentioned that only 3 of the 12 patients studied by Ojemann (1985) were diagnosed
12
HUGDAHL,
WESTER, AND ASBJe)RNSEN
with Parkinson’s disease (the other nine were multiple sclerosis, dystonia, and athetosis). To summarize then, differential DL responding, and particularly the size of the REA, was observed after left- and right-sided stimulations and lesions. It is proposed that this difference in language functioning after left- and right-sided manipulations of the VL nucleus is related to the switching on or off of an AGM that enhances contralateral cortical input through activation of appropriate focal areas in the left cerebral hemisphere when the gate is open. REFERENCES Bhatnagar, S. C., Andy, 0. J., Korabic, E. W., Saxena, V. K., Tikofski, R. S., & Hellman, R. S. 1987. Thalamic functional tuning of the cortex in dichotic listening tasks. Applied Neurophysiology,
50, 457-458.
Brodal, A. 1981. Neurological anatomy (3rd ed.). New York: Oxford University Press. Bryden, M. P. 1982. Laterality. New York: Academic Press. Bryden, M. P. 1988. An overview of the dichotic listening procedure and its relation to cerebral organization. In K. Hugdahl (Ed.), Handbook of dichotic listening: Theory, methods and research. Chichester, U.K.: Wiley & Sons. Connolly, J. F. 1985. Stability of pathway-hemispheric differences in the auditory event22, 87-96. related potential (ERP) to monaural stimulation. Psychophysiology, Crosson, B. 1984.Role of dominant thalamus in language: A review. Psychological Bulletin, 96, 491-517. Crosson, B. 1988. Subcortical language mechanisms: Window on a new frontier. In H. A. Whitaker (Ed.), Phonological processes and brain mechanisms. New York: Springer Verlag. Fedio, P., & Van Buren, J. M. 1975. Memory and perceptual deficits during electrical stimulation in the left and right thalamus and parietal subcortex. Brain and Language, 2, 78-100. Galaburda, A. 1986. Role of the thalamus in auditory lateralization: Histologic data. Revue 142, 441-444. Neurologic, Hugdahl, K. (Ed.) 1988. Handbook ofdichotic listening: Theory, methods and research. Chichester, UK: Wiley & Sons. Hugdahl, K., & Andersson, L. 1984. A dichotic listening study of differences in cerebral organization in dextral and sinistral subjects. Cortex, 20, 135-141. Hugdahl, K., & Andersson, L. 1986. The “forced-attention paradigm” in dichotic listening to CV-syllables: A comparison between adults and children. Cortex, 22, 417-432. Kimura, D. 1967. Functional asymmetry of the brain in dichotic listening. Cortex, 3, 163178. Manen, J. van, Speelman, J. D., & Tans, R. J. J. 1984. Indications for surgical treatment of Parkinson’s disease after levodopa therapy. Clinical Neurology and Neurosurgery, 86, 207-212. Mateer, C., & Ojemann, G. A. 1983. Thalamic mechanisms in language and memory. In S. Segalowitz (Ed.), Languagefunction and brain organization. New York: Academic Press. Maximilian, V. A. 1982.Cortical blood flow asymmetry during monaural verbal stimulation. Brain and Language, 15, l-11. Ojemann, G. A. 1985. Enhancement of memory with human ventrolateral thalamic stimulation. Effect evident on a dichotic listening task. Applied Neuropsychology, 48, 212215.
ROLE OF THALAMUS
IN LANGUAGE
ASYMMETRY
13
Ojemann, G. A. 1983. Brain organization for language from the perspective of electrical stimulation mapping. The Behavioral and Brain Sciences, 6, 189-230. Ojemann, G. A. 1975. Language and the thalamus: Object naming and recall during and after thalamic stimulation. Brain and Language, 2, 101-120. Princeton, N.J.: Princeton Penfield, W., &Roberts, L. 19.59.Speech andbrain mechanisms. University Press. Raczkowski, D., Kalat, J. W., & Nebes, R. 1974. Reliability and validity of some handedness questionnaire items. Riklan, M., & Cooper, I. S. 1977. Thalamic lateralization of psychological functions: Psychometric studies. In S. Harnad et al. (Eds.), Lateralization in the nervous system. New York: Academic Press. Spiegel, E. A., & Wycis, H. T. 1952. Stereo-encephalatomy (Thalamotomy and related procedures). Part I. New York: Grune and Stratton. Tasker, R. R., Siqueira, J., Hawrylyshyn, P., & Organ, L. W. 1983. What happened to 46, 68-83. VIM thalamotomy for Parkinson’s disease? Applied Neurophysiology, Wester, K., & Hauglie-Hanssen, E. in press. Stereotaxic thalamotomy-experiences from the levodopa era. Journal of Neurology, Neurosurgery, and Psychiatry.
AND
LANGUAGE
39, 1-13 (I!@@
The Role of the Left and Right Thalamus in Language Asymmetry: Dichotic Listening in Parkinson Patients Undergoing Stereotactic Thalamotomy KENNETH HUCDAHL,* *Depurtment
KNUT WESTER,~ AND ARVE
of Somatic Psychology, Neurosurgery,
ASBJP)RNSEN*
University of Bergen and TDepartment University of Bergen
of
Fourteen patients with Parkinson’s disease (rigidity and unilateral tremor as main symptoms) were treated with stereotactic thalamotomy. The operation was performed in either the left or right ventrolateral thalamus, depending on which hand (or foot) was most affected with tremor. Nine patients were operated on the left side and four on the right side. All patients were tested for asymmetry of language functioning with dichotic presentations of consonant-vowel (CV) syllables. The patients had to indicate which of the two syllables he/she heard on each trial. Dichotic listening was performed before and after the operation, as well as during electrical stimulation of the VL nucleus just before the lesion was carried out. The results revealed essentially three things: an overall reduced right ear advantage (REA) in the patient group compared to what is known from studies of healthy individuals; an increase in REA during left-sided stimulations; and a marked reduction in REA after left-sided lesions. It is concluded that the present data support the notion put forward by Ojemann (e.g., 1975) of a lateralized activating gating mechanism in the left VL nucleus that gates access for language information to the appropriate cortical areas. The gating mechanism seems to be activated by stimulation, and deactivated after lesions. Dichotic listening may thus be a heuristic instrument in assessment of language functions in Parkinson patients. 0 1990 Academic Press, Inc.
In this paper we present data from dichotic listening performance in 14 Parkinson patients undergoing stereotactic thalamotomy that are relevant for elucidation of the role of the left and right thalamus in language asymmetry. The data also elucidate the processes involved in dichotic listening. The study involves comparisons of pre- and postlesion dichotic performance, as well as comparisons of dichotic listening scores during stimulation of the left or right ventro-lateral thalamus. Send reprint requests to Kenneth Hugdahl, Department of Somatic Psychology, University of Bergen, Aarstadveien 21, N-5009 Bergen, Norway. I 0093-934x/90 $3.00 Copyright 0 1990 by Academic Press. Inc. All rights of reproduction in any form reserved.
2
HUGDAHL,
WESTER,
AND ASBJORNSEN
For two decades after Spiegel and Wycis (1952) performed the first stereotactic thalamotomies in 1947, the operation was used frequently in the treatment of Parkinsonian tremor. When levodopa was introduced in the late 1960s interest in thalamotomy declined throughout the world, and very few such operations were performed in the following years (Tasker, Siqueira, Hawrylyshyn, & Organ, 1983). However, it gradually became clear that tremor was often unaffected by this type of medication. Tremor and to some degree rigidity are the main indications for thalamotomy, and interest again focused on this operation (Manen, Speelman, & Tans, 1984; Wester & Hauglie-Hanssen, in press). It is a well-known phenomenon that thalamic lesions have effects on language, and especially lesions in the left ventro-lateral and pulvinar nuclei. Ojemann and his colleagues (e.g., Mateer & Ojemann, 1983; Ojemann, 1975) have over the years reported that lesions in the lateral thalamus produce language disruptions characterized by dysnomia with perserveration, dysarthria, dysphasia, and sometimes intrusions of extraneous words. Others have reported not only that left-sided lesions affect expressive language functions but also that receptive verbal capacities deteriorate, including short-term memory (e.g., Riklan & Cooper, 1977; Crosson, 1984; Fedio & Van Buren, 1976). In addition, aphasia has often occurred after therapeutic lesions of the dominant ventral lateral (VL) thalamus (Crosson, 1988). However, since the aphasia usually disappears after a while, it may be the interaction of the VL nucleus and neighboring nuclei that causes the language dysfunction. One of these neighboring nuclei may be the reticular nucleus (see Fig. 2) believed to be a part of the nonspecific thalamic activation system. It may therefore be that the observed effects are due to shifts in attention/activation especially after left-sided lesions. These findings are underscored by reports of anatomical asymmetries in the cell architecture of the posterior lateral nucleus, reported by Galaburda and associates (e.g. Galaburda, 1986). Taken together, the data suggest: (1) that the human thalamus plays an important role in both language comprehension and speech expression, (2) that the left ventro-lateral and pulvinar nuclei seem more involved than the corresponding right nuclei, and (3) that asymmetry of language functions is not confined to cortical structures but also to subcortical structures which contribute to the overall left-to-right asymmetry (cf. Riklan & Cooper, 1977). A frequently used noninvasive method for the study of language asymmetry is dichotic listening (Kimura, 1967; see also Hugdahl, 1988). Dichotic listening is a relatively reliable technique for the study of language asymmetry (Bryden, 1988). The right ear advantage (indicative of a left hemisphere language specialization) is probably one of the more robust
ROLE OF THALAMUS
IN LANGUAGE
ASYMMETRY
3
empirical phenomen reported in experimental neuropsychology (Bryden, 1982; 1988; Hugdahl & Andersson, 1984; 1986). It could be argued that if the left thalamus is involved in language, then this should be reflected in dichotic listening performance when comparing patients with left vs. right thalamic lesions (and stimulations). The connections between dichotic listening and language asymmetry are perhaps still best explained by Kimura’s structural model from 1967. Kimura’s (1967) model may be described as a set of four assumptions about hemisphere asymmetry and brain functioning: 1. The left hemisphere is specialized for language. 2. Auditory input is more strongly represented in the contralateral hemisphere. 3. Dichotic stimulation initially lateralizes the input, as information in ipsilateral hemisphere is suppressed by the contralateral input. 4. Information from the nondominant hemisphere is transferred across the corpus callosum. Ear differences may thus reflect the relative processing superiority of one hemisphere or information loss due to interhemispheric transmission, or an interaction between these two factors. The only available data on dichotic listening performance in thalamic patients (at least to our knowledge) are the reports by Ojemann and Mateer in 1983 based on eight patients, the study by Ojemann in 1985, and the case study by Bhatnagar et al. in 1987. Mateer and Ojemann (1983) reported that frequency of correct responses increased during brain stimulation, and especially during left-sided stimulation. No data were reported concerning dichotic performance during or after surgery. Ojemann (1985) studied 12 patients (three had Parkinson’s disease) with dichotic presentations of words. He found significantly more words presented during left-sided stimulation correctly reported, with a slight tendency for a contralateral ear advantage. Bhatnagar et al. (1987) using a different approach, found a decrease in REA after a prolonged 20-min stimulation in the left thalamus, probably caused by setting the neurons temporarily out of function. Extrapolating from these findings, it may now be predicted that a relatively short stimulation of the left ventro-lateral (VL) thalamus should yield increased REA, while a lesion on the same side should result in a reduced REA or possibly a left ear advantage (LEA). Stimulation and lesion of the right VL thalamus should not yield any dramatic changes in dichotic listening reports. METHOD Patients The subjects were 14 Parkinson patients with a mean age of 64 (range 54 to 72). They were all referred to surgery from neurological departments because of severe and disabling
4
HUGDAHL,
WESTER, AND ASBJP)RNSEN
drug-resistant tremor and rigidity. Tremor was more profound in the upper limbs, and also more manifest on one side (either right or left). Nine patients were operated on the left side, and five on the right side. All 14 patients were right-handed. Handedness was defined with the help of a Norwegian translation of a modified version of the questionnaire developed by Raczkowski, Kalat, and Nebes (1974). The questionnaire consists of 15 questions related to manual items. To be classified as a right-hander, the patient had to indicate 13 of the 15 items as performed with the right hand (=87% consistency). Hearing acuity was determined by a Tegner screening audiometer. Patients with an imbalance in hearing levels between ears of more than 5 dB (from 20 dB) were not included in the study.
Apparatus The stereotactic technique and equipment were of the Leksell-type, and the target area was defined from CT-evaluation and located in the ventro-lateral nucleus 3 mm superior, 2 mm posterior, and 14 mm lateral (left or right) to the midpoint of the distance between the anterior (AC) and posterior (PC) commissures. A small hole with about l-cm diameter was drilled just in front of the coronal seam, and two electrodes, having a diameter of 1.5 mm, were inserted into the hole in the skull and placed in the target area (6 mm apart). The lesion was of the thermocoagulative type, and established by emitting radiofrequencies through the electrodes, thereby increasing the temperature between the tips of the electrodes to between 60 and 65°C. Electric stimulations in the target areas through the electrodes were performed before the lesion was carried out. This was done as a control that the electrodes were in the target area. There should usually be an effect on tremor during stimulation if the electrodes are correctly placed. Stimulation was performed with 200 Hz and with an average duration of 5 min. Figures 1 and 2 show the location of the electrodes, and the exact extension of the lesion as seen from the tip of both electrodes.
Stimulus materials The dichotic stimuli consisted of the six stop-consonants b, d, g. p, I, k, all paired with the vowel a to form six Consonant-Vowel (CV) syllables (bo, da, ga, etc.). The syllables were paired with each other for all possible combinations, thus yielding 36 dichotic pairs including the homonymic pairs. The dichotic tape consisted of three lists of 36 dichotic CV-pairs each. The 36 pairs were randomly ordered on each list. Each syllable had a duration of 320 msec, and synchronization of onset between channels was performed for both the consonant and vowel segment onsets. The interstimulus interval was 4 set * 1 sec. The dichotic tape was prepared on a PDP 1l/45 computer with A/D and D/A converters, and with a multiplexer. Due to the resolution of the multiplexer, maximum onset asynchrony was 0.5 msec. The stimuli were copied onto a chrome-dioxid cassette and played to the patient from a SONY WM DD-II minicassette player through SONY plug-in type earphones. The intensity of the output from the earphones were on the average 75 dbA (repeated measurements) when tested with a Bruel and Kjaer 2204 sound level meter.
Procedure The dichotic listening tests were performed the day before (after the patient had arrived to the hospital), just before surgery (inside the operation room), during surgery (stimuFIG. 1. Schematic drawing of electrode locahzations in the lateral ventro-lateral (VL) nucleus. The hatched rectangular area shows the actual lesion area.
ROLE
OF THALAMUS
IN LANGUAGE
ASYMMETRY
5
HUGDAHL,
z P F
WESTER,
AND ASBJORNSEN
ROLE OF THALAMUS
IN LANGUAGE
ASYMMETRY
7
lation), just after surgery (immediately after the surgeon had sutured the opening in the skull), and the day after surgery. The patient had to answer which syllable he/she heard on each trial, and this was marked on a special scoring sheet. In addition, everything that was said during the dichotic tests was taped on another minicassette tape recorder through a “mini” microphone clipped to the patient’s shirt. A short pause was inserted after each list of 36 trials to let the patient rest. During the stimulation and postoperation tests, only 36 or 72 trials could be run in some patients for clinical reasons (signs of fatigue, etc.).
RESULTS Presurgery
Data were collapsed across the two pre- and postsurgery tests taking the mean of the two. Mean percentages of correct recall separated for right and left ear input and for pre- and postsurgery are seen in Fig. 3. There was a slight (4.1%) right ear advantage (REA) presurgery for the patients that were-to-have left-sided lesions. The corresponding ear differences for the group with right-sided lesions were 4.5%. An analysis of variance based on a 2 (pre- vs. postsurgery) x 2 (right vs. left ear input) design did not yield any significant differences (all Fs < 1) (see Figs. 3 and 6). There were however six (of nine) patients with left-sided lesions that showed a REA and three (of five) patients with right-sided lesions. An equal frequency x2 evaluation showed these distributions to be nonsignificant (x2(1) = 1.00 and 0.20, respectively). Postsurgery
Mean percentages of correct recall for both groups of patients are also seen in Fig. 3. During the postsurgery test, the patients with left-sided lesions showed a dramatic reduction in overall performance. A 2 x 2
FIG. 3. Mean percentages of correct recall in the dichotic listening test for the right and left ear, and for the different test conditions. LL = left-sided lesion, RL = rightsided lesion.
8
HUGDAHL,
WESTER,
AND ASBJORNSEN
analysis of variance on a similar design as for the presurgery comparisons yielded a significant main effect of lesion side [F( 1, 12) = 5.03, p < .05]. Figure 3 reveals that the effect was due to reduced overall performance in the left-sided group. The interaction of side of lesion by ear of input was not significant [F(l, 12) = 2.37, n.s.). However, a separate t test comparing the right and left ear input for the group with left-sided lesions was significant If(8) = 2.01, p < .05]. This was due to better recall from the left compared to the right ear. Separate t tests were also performed for comparison of left ear presurgery vs. left ear postsurgery, and for right ear presurgery vs. right ear postsurgery, respectively. Both tests for the left lesion group were significant (both ts > 1.92, p < .0.5). No significant difference was observed for the right lesion group in the analysis of variance. All patients underwent a postoperative CT-control on the first or second postoperative day, and 3 to 6 months after the operation. Figure 4 shows a typical lesion from one patient. The lesion seen in Fig. 4 corresponds to the predetermined target coordinates. All patients had similar lesions verified through postoperative CT-scans.
FIG. 4. Horizontal CT-scans through the maximal extent of a thalamotomy lesion in the left thalamus. Left: First postoperative day. Right: Three months postoperatively. Arrow and square, arrow and circle, indicate lesion sites. Note the immediate postoperative edema. Please observe that the left hemisphere is shown to the right and vice versa according to radiological conventions.
ROLE OF THALAMLJS
IN LANGUAGE
ASYMMETRY
9
Brain Stimulation Mean correct recall during brain stimulation separately for each group is seen in Figs. 5 and 6. Note that the pre- and postsurgery data are once again shown in Figs. 5 and 6 to facilitate visual inspection of differences between test conditions. There were no significant effects in the overall analysis of variance. However, there was a significant difference for left and right ear correct recalls during left-sided stimulations when tested with the t statistic [t( 12) = 1.88, p < .05). The means show the difference to be 14.6% in favor of right ear performance. The right-side stimulation group did not reach significance [t(4) < l] (see also Fig. 5). DISCUSSION
To sum up the basic findings: There was generally a reduced REA in the patient group preoperatively compared to what is known from studies of normal healthy individuals. Usually, a 15-20% REA (Bryden, 1988) is obtained in healthy individuals which should be compared with the 4-5% REA seen in the present patient group preoperatively (see Fig. 3). Postoperatively, there was a rather dramatic reduction in recall from both ears in patients with left-sided lesions, while much less dramatic effects were seen in patients with right-sided lesions. However, during stimulation, there was an increase in ear advantage in both patient groups, although in different directions. Thus, while an increased REA was seen in the group with left-sided stimulation, an increased LEA (however non-significant) was seen in the group with right-sided stimulation.
LL E4airrstlm
FIG. 5. Mean percentages of correct and left ear during left-sided stimularion.
LL Fe3t-sug
recall in the dichotic listening test for the right LL = left-sided lesion; RL = right-sided lesion.
10
HLJGDAHL, WESTER, AND ASBJ@RNSEN
60
RL Pre-eug
iX
Brain-stim
FIG. 6. Mean percentagesof correct recall in the dichotic listening test for the right and left ear during right-sided stimulation. LL = left-sided lesion; RL = right-sidedlesion.
The present results are perhaps best explained with reference to the arousal model proposed by Ojemann (e.g., 1975). Mateer and Ojemann (1983) proposed a thalamic attentional, or alerting, mechanism which asymmetrically gates access to the cortex. It has been suggested that there is a gating mechanism in the medial geniculate body, and that this “gate” is more responsive to contralateral than to ipsilateral auditory signals. It has been established with both event-related potentials (ERPs) (e.g., Connolly, 1985) and with recordings of cerebral blood flow (CBF) (e.g., Maximilian, 1982) that the contralateral auditory pathways are both more preponderant and more efficient than the ipsilateral ones in transferring the auditory signal from the cochlea to the primary auditory temporal cortex. The predominance of the contralateral pathways from the cochlea to the temporal cortex may be a consequence of the stronger projection of second-order neurons to the inferior colliculus on the contralateral than on the ipsilateral side (Brodal, 1981). Although the input to, and output from, the inferior colliculi are both ipsi- and contralateral, the projections ascending to the colliculi are greater from the contralateral ear. However, the pathways ascending from the colliculi are greater on the ipsilateral side, thus favoring an ultimate representation of the contralateral ear in the auditory cortex (Brodal, 1981). Applying Ojemann’s model to the present data, we would like to suggest the existence of a language-related activating gating mechanism (AGM) in the ventro-lateral (VL) thalamic nuclei that asymmetrically “gates” information to the left hemisphere. During stimulation of the left VL nucleus, the AGM is set to “n,,” with the result of an increased REA due to facilitation of the contralateral right ear signals to the left hemisphere. The occurrence of electric current across the electrode tips
ROLE OF THALAMUS
IN LANGUAGE
ASYMMETRY
11
during the stimulation test may actually drive the neurons in the VL nucleus to open the AGM. However, when a lesion is applied to the same area, the AGM is set to “off’ position. The result should then be a decreased REA (or even a LEA) due to inhibitory blocking of the contralateral right ear signals to the left hemisphere. The thalamic alerting or activating circuitry may thus act as a switching mechanism to activate appropriate focal cortical areas involved in language processing. There is a preprocessing of verbal input that selectively gates input to the cortical hemispheres. Ojemann (1975) has argued that the arousal mechanism in the left thalamus is a “gate” also for access to short-term memory and that brain stimulation should evoke “an intense direction of attention to information [perhaps only verbally coded information” (p. 116)]. Thus, the left thalamus seems to be uniquely related to language functioning (see also Crosson, 1984). It may further be that only portions of the left thalamus, including the anterior superior pulvinar and medial central ventrolateral nuclei are related to language in this respect (cf. Ojemann, 1975). In addition to the presented results on DL performance after and during thalamic stimulation and lesion, several patients showed clinical signs of other language disturbance after left-sided, but not after right-sided lesions. These disturbances involved perseveration, dysarthria, and in one case speech arrest during stimulation. Similar findings were also reported by Ojemann (1975). However, following the findings by Penfield and Roberts (1959) of speech arrest occurring after cortical stimulation, this may be a response related rather to the cortical than to the subcortical level. Mateer and Ojemann (1983) reported that the total number of correct responses in dichotic listening performance made during either left- or right-sided stimulation was higher than was the total number of correct responses made under nonstimulation conditions. Increased accuracy of performance in their DL task was interestingly particularly “dramatic and consistent with left thalamic stimulation” (p. 184). Since we did not find an increase in REA in all nine patients stimulated on the left side, it may seem that our data are less “consistent” than the data presented by Mateer and Ojemann (1983). One explanation for this may be that whereas we used 200 Hz stimulation (over 3-5 min periods) with up to 108 CV trials, Mateer and Ojemann (1983) and Ojemann (1985) used 60 Hz stimulation and only 20 CV trials. Since a CV normally lasts about 350 msec, a 20-trial presentation would require stimulation of less than 1 min (with a reasonable intertrial interval). A 200-Hz stimulation may thus temporarily fatigue neurons in the VL nucleus, thus providing an effect on language more like the one observed after lesions. This seems however to be subject to individual variation. It should also be mentioned that only 3 of the 12 patients studied by Ojemann (1985) were diagnosed
12
HUGDAHL,
WESTER, AND ASBJe)RNSEN
with Parkinson’s disease (the other nine were multiple sclerosis, dystonia, and athetosis). To summarize then, differential DL responding, and particularly the size of the REA, was observed after left- and right-sided stimulations and lesions. It is proposed that this difference in language functioning after left- and right-sided manipulations of the VL nucleus is related to the switching on or off of an AGM that enhances contralateral cortical input through activation of appropriate focal areas in the left cerebral hemisphere when the gate is open. REFERENCES Bhatnagar, S. C., Andy, 0. J., Korabic, E. W., Saxena, V. K., Tikofski, R. S., & Hellman, R. S. 1987. Thalamic functional tuning of the cortex in dichotic listening tasks. Applied Neurophysiology,
50, 457-458.
Brodal, A. 1981. Neurological anatomy (3rd ed.). New York: Oxford University Press. Bryden, M. P. 1982. Laterality. New York: Academic Press. Bryden, M. P. 1988. An overview of the dichotic listening procedure and its relation to cerebral organization. In K. Hugdahl (Ed.), Handbook of dichotic listening: Theory, methods and research. Chichester, U.K.: Wiley & Sons. Connolly, J. F. 1985. Stability of pathway-hemispheric differences in the auditory event22, 87-96. related potential (ERP) to monaural stimulation. Psychophysiology, Crosson, B. 1984.Role of dominant thalamus in language: A review. Psychological Bulletin, 96, 491-517. Crosson, B. 1988. Subcortical language mechanisms: Window on a new frontier. In H. A. Whitaker (Ed.), Phonological processes and brain mechanisms. New York: Springer Verlag. Fedio, P., & Van Buren, J. M. 1975. Memory and perceptual deficits during electrical stimulation in the left and right thalamus and parietal subcortex. Brain and Language, 2, 78-100. Galaburda, A. 1986. Role of the thalamus in auditory lateralization: Histologic data. Revue 142, 441-444. Neurologic, Hugdahl, K. (Ed.) 1988. Handbook ofdichotic listening: Theory, methods and research. Chichester, UK: Wiley & Sons. Hugdahl, K., & Andersson, L. 1984. A dichotic listening study of differences in cerebral organization in dextral and sinistral subjects. Cortex, 20, 135-141. Hugdahl, K., & Andersson, L. 1986. The “forced-attention paradigm” in dichotic listening to CV-syllables: A comparison between adults and children. Cortex, 22, 417-432. Kimura, D. 1967. Functional asymmetry of the brain in dichotic listening. Cortex, 3, 163178. Manen, J. van, Speelman, J. D., & Tans, R. J. J. 1984. Indications for surgical treatment of Parkinson’s disease after levodopa therapy. Clinical Neurology and Neurosurgery, 86, 207-212. Mateer, C., & Ojemann, G. A. 1983. Thalamic mechanisms in language and memory. In S. Segalowitz (Ed.), Languagefunction and brain organization. New York: Academic Press. Maximilian, V. A. 1982.Cortical blood flow asymmetry during monaural verbal stimulation. Brain and Language, 15, l-11. Ojemann, G. A. 1985. Enhancement of memory with human ventrolateral thalamic stimulation. Effect evident on a dichotic listening task. Applied Neuropsychology, 48, 212215.
ROLE OF THALAMUS
IN LANGUAGE
ASYMMETRY
13
Ojemann, G. A. 1983. Brain organization for language from the perspective of electrical stimulation mapping. The Behavioral and Brain Sciences, 6, 189-230. Ojemann, G. A. 1975. Language and the thalamus: Object naming and recall during and after thalamic stimulation. Brain and Language, 2, 101-120. Princeton, N.J.: Princeton Penfield, W., &Roberts, L. 19.59.Speech andbrain mechanisms. University Press. Raczkowski, D., Kalat, J. W., & Nebes, R. 1974. Reliability and validity of some handedness questionnaire items. Riklan, M., & Cooper, I. S. 1977. Thalamic lateralization of psychological functions: Psychometric studies. In S. Harnad et al. (Eds.), Lateralization in the nervous system. New York: Academic Press. Spiegel, E. A., & Wycis, H. T. 1952. Stereo-encephalatomy (Thalamotomy and related procedures). Part I. New York: Grune and Stratton. Tasker, R. R., Siqueira, J., Hawrylyshyn, P., & Organ, L. W. 1983. What happened to 46, 68-83. VIM thalamotomy for Parkinson’s disease? Applied Neurophysiology, Wester, K., & Hauglie-Hanssen, E. in press. Stereotaxic thalamotomy-experiences from the levodopa era. Journal of Neurology, Neurosurgery, and Psychiatry.
