Changes in maps of language activity activation following melodic intonation therapy using magnetoencephalography: Two case studies

Transcription

1 JOURNAL OF CLINICAL AND EXPERIMENTAL NEUROPSYCHOLOGY 2010, 32 (3), NCEN Changes in maps of language activity activation following melodic intonation therapy using magnetoencephalography: Two case studies MEG LANGUAGE MAPS AFTER MELODIC INTONATION THERAPY Joshua I. Breier, 1 Shannon Randle, 2 Lynn M. Maher, 2 and Andrew C. Papanicolaou 1 1 Center for Clinical Neurosciences in the Department of Pediatrics, The University of Texas Health Science Center at Houston, Houston, Texas, USA 2 Department of Communication Sciences and Disorders, University of Houston, Houston, Texas, USA Two patients with chronic expressive aphasia underwent two blocks of melodic intonation therapy (MIT) each. Maps of language-specific neurophysiological activity were obtained prior to and after each MIT block during a covert action naming task using magnetoencephalography. Both patients exhibited increased left hemisphere activation after MIT. The patient who responded positively to therapy exhibited decreasing activation within areas of the right hemisphere homotopic to left hemisphere language areas compared to baseline after both blocks of MIT. In contrast, the patient who did not show improvement after therapy exhibited increasing activation in these areas of the right hemisphere after therapy. Results are consistent with hypotheses that melodic intonation therapy acts through promotion of left hemisphere activation. Keywords: Aphasia; Magnetoencephalography; Melodic intonation therapy; Stroke; Functional imaging. Chronic aphasia after stroke in the dominant hemisphere is a significant problem, occurring in up to 20% or more of patients (e.g., Kertesz, 1984; Pedersen, Jorgensen, Nakayama, Raaschou, & Olsen, 1995). Data regarding the means by which the brain reorganizes to support recovery of language function in response to therapy for chronic aphasia after stroke may provide a basis for the evaluation and design of rehabilitative strategies. In the current study we report the changes in language-specific neurophysiological activity, as indexed by magnetoencephalography (MEG), in two patients who were administered melodic intonation therapy (MIT) for chronic aphasia after stroke. There are a limited number of functional imaging studies regarding the changes in physiological response in the brain to linguistic stimuli after aphasia therapy. Among these, some report improvement in language function to be correlated with an increase in right hemisphere activation (e.g., Blasi et al., 2002; Musso et al., 1999), while others report improvement in language function to be correlated with either increased activation in perilesional areas in the putative premorbidly dominant hemisphere (e.g., Leger et al., 2002) or increases in activation bilaterally (e.g., Fridriksson, Morrow-Odom, Moser, Fridriksson, & Baylis, 2006; Vitali et al., 2007). Still others report change in language function to be associated with the degree of premorbid language activation in the right hemisphere (e.g., Breier, Maher, Novak, & Papanicolaou, 2006; Richter, Miltner, & Straube, 2008). A number of factors potentially influence findings, including severity and location of stroke lesion, time of administration of therapy after stroke, type of therapy administered, length of therapy, degree of deficit at time of administration of therapy, and imaging modality and activation task used. In the current study we present data regarding changes in language-specific brain activation obtained using MEG during a covert action naming task in 2 participants who This work was supported by National Institutes of Health/National Institute of Neurological Disorders and Stroke (NIH/NINDS) Grant P51-NS to A. C. Papanicolaou. Address correspondence to Joshua I. Breier, Department of Pediatrics, Division of Clinical Neurosciences, University of Texas Houston Health Science Center, 1333 Moursund Street, Suite H114, Houston, Texas 77030, USA ( joshua.i.breier@uth.tmc.edu) Psychology Press, an imprint of the Taylor & Francis Group, an Informa business DOI: /

2 310 BREIER ET AL. underwent two blocks of MIT. MIT is based on the observation that patients with significant expressive language deficits may exhibit an increased ability to repeat verbal stimuli if they attempt to sing rather than speak the stimuli (Albert, Sparks, & Helm, 1973; Sparks, Helm, & Albert, 1974). During MIT patients are asked to repeat stimuli that are presented with exaggerated prosody consisting of a high and low note and short and long duration (Belin et al., 1996). Previous hypotheses suggested that melodic intonation resulted in facilitation of language control in the right hemisphere and perhaps even diminishing the damaged left hemisphere s role in language output (Albert et al., 1973; Sparks et al., 1974; Sparks & Holland, 1976). In a study characterizing the changes in language activation after training with MIT, however, Belin et al. (1996), using positron emission tomography (PET), found that patients who experienced improvement in word repetition after administration of MIT exhibited increased hemodynamic response in Broca s area in the left hemisphere during a word repetition task when stimuli were similar to those used during MIT. Given the findings of Belin et al. (1996) we hypothesized that improvement in language function would be associated with increased activation within putative premorbid language areas within the left hemisphere. Patients METHOD Patient 1 was a 55-year-old right-handed Caucasian male with a history of stroke five years prior to administration of MIT. Initial symptoms prior to hospital admission at the time of the stroke included right-sided weakness, dragging of the right foot, and mixed expressive/receptive aphasia. A computed tomography (CT) scan done at the time reportedly indicated an acute left middle cerebral artery (MCA) infarct involving the left basal ganglia and hemorrhage change as well as an area of infarction involving the left posterior frontal region. Angiogram performed at the time indicated high-grade critical stenosis of the left internal carotid artery, and a stent was deployed to reduce stenosis to less than 50%. There was a history of myocardial infarction five years prior to stroke. The patient had undergone traditional therapy immediately after the stroke for a period of four months. In addition, constraint induced language therapy (CIT; Meinzer et al., 2004) was administered three months before MIT for a duration of three weeks with a significant positive response. Patient 2 was a 49-year-old right-handed Caucasian male with a history of stroke two years prior to the administration of MIT. The patient had a history of hypertension prior to stroke. Symptoms prior to initial hospital admission included right-sided hemiparesis and mixed expressive/receptive aphasia. Magnetic resonance imaging (MRI) done at the time reportedly indicated a large left MCA territory ischemic stroke. Angiogram performed at the time indicated almost full left internal carotid artery occlusion that was felt to be chronic and significant crossflow from patent left posterior (PCA) and anterior (ACA) cerebral arteries into the distribution of the left MCA. Also noted was 50% stenosis of both external carotid arteries. There was no intervention for these conditions performed. The patient had undergone traditional therapy intermittently after the stroke as well as constraint induced language therapy eight months before MIT for a duration of three weeks. Response to the latter therapy was not significant. Sagittal slices at two levels of the left hemisphere are shown in Figure 1 for each patient. Both patients incurred significant damage to frontoparietal areas as well as some damage to temporal lobe areas within the left hemisphere. Damage to these latter areas was somewhat more extensive in Patient 2 than in Patient 1. Both patients were administered the Western Aphasia Battery (Kertesz, 1982) prior to MIT in order to characterize aphasia severity, but not after any of the MIT blocks. These data are presented in Table 1 for each patient. Both patients exhibited significant aphasia prior to treatment, with deficits in expressive language. Receptive language was significantly more compromised in Patient 2 than in Patient 1. Magnetoencephalography Patients were presented with a simple line drawing taken from the Action Naming Test (Obler & Albert, 1979) and were told to silently name the action it depicted as quickly as possible. Stimuli consisted of 160 drawings. A short practice period during which the participants were introduced to the task preceded data collection. Two separate blocks of trials were obtained at different time points. Stimuli were presented via an arrangement of mirrors with a variable interstimulus interval ( s) for 1 s. The patient was asked to keep his/her eyes open during stimulus presentation, fixating on a dark dot placed in the center of the mirror, in order to reduce eye movements or blinks and to prevent event-related field (ERF) contamination by rhythmic activity (typically in the alpha band), which can seriously interfere with the accurate detection of task-related brain activity. Event related fields (ERFs) time locked to the words were recorded in a magnetically shielded room, using a whole-head neuromagnetometer (4D 3600, 4D Neuro- Imaging, San Diego) equipped with an array of 248 gradiometer sensors and housed in a magnetically shielded room designed to reduce environmental magnetic noise that might interfere with biological signals. The signal was recorded with a band pass filter set at 0.1 and 50 Hz and digitized for 1,000 ms (254-Hz sampling rate) including a 150-ms prestimulus period. Single-trial ERF segments that were identified as contaminated by eye- or head-movement-related magnetic artifacts were removed, and the recordings were filtered with a band pass between 0.1 and 20 Hz and subjected to an adaptive filtering procedure that is part of the 4-D Neuroimaging signal analysis package. Artifact-free epochs in each channel were subsequently averaged, and MEG activity sources of the averaged waveforms were modeled as single equivalent current dipoles (ECDs) and fitted at

3 MEG LANGUAGE MAPS AFTER MELODIC INTONATION THERAPY 311 Figure 1. Sagittal slices through the left hemisphere for (a, b) Patient 1 and (c, d) Patient 2. TABLE 1 Pre-MIT language data (Western Aphasia Battery, Boston Naming Test) Patient 1 Patient 2 Information content (10) 3 3 Fluency (10) 2 2 Yes/no questions (60) Auditory word recognition (60) Sequential commands (80) Repetition (100) Object naming (60) Word fluency (20) 3 0 Sentence completion (10) 7 8 Responsive speech (10) 4 4 Aphasia quotient (100) Boston Naming Test (Pre) (60) 6 1 Boston Naming Test (Post 1) 5 2 Boston Naming Test (Post 2) 6 1 Note. Total possible correct in parentheses. MIT = melodic intonation therapy. successive 4-ms intervals (Sarvas, 1987). Up to three sources could be identified at any time point. Two runs were obtained for each patient at each scanning session. Clusters of activity sources were derived from the two runs using an algorithm that ranked sources according their spatial and temporal proximity across the runs. In this manner sources that did not replicate across the runs were removed. Early (before the resolution of the N1m at approximately 200 ms post stimulus onset) dipole activity was generally observed within primary visual cortex and was not considered in this study. Source locations, which were initially computed in reference to the MEG Cartesian coordinate system, were coregistered on the patient s MRI. Transformation of the MEG coordinate system into MRI-defined space was achieved with the aid of three lipid capsules inserted into the ear canals and attached to the nasion, which were easily visualized on the MRIs. Location of individual dipoles was determined with the use of a standard MRI atlas of the human brain (Damasio, 1995). Melodic intonation therapy (MIT) MIT is a rehabilitation program that aims to recover the formulation of propositional language in individuals with nonfluent aphasia. It consists of repetition of linguistic phrases embedded in tonal patterns derived from possible speech prosody patterns (Helm-Estabrooks, Nicholas, & Morgan, 1989). The exaggerated intonation patterns are gradually reduced to yield a more natural prosodic structure (Sparks & Holland, 1976).

4 312 BREIER ET AL. The patients each received individual melodic intonation therapy in two blocks. Each block consisted of two 30-minute treatment sessions per day, two days a week for three weeks, for a total of 12 hours of treatment over the two blocks (6 hours per block). There was an intervening break between the first and second block of three weeks. Independent production on a set of phrases from the MIT program was used before and after each treatment block as the treatment response measure. Independent production on these phrases was also assessed throughout each treatment block (at every second session). Correct information Units (CIUs; Nicholas & Brookshire, 1993) were obtained for all responses as the measure of change. Behavioral changes RESULTS Patient 1 exhibited a significant increase in CIUs (>35%) after the first block of treatment. This improvement was maintained after the break, but little additional improvement after the second block was apparent. In contrast, Patient 2 did not demonstrate substantial change on the treatment response measures after either block of MIT. Neurophysiological changes Early (before the resolution of the N1m) dipole activity was generally observed within primary visual cortex. This activity was not used in the study as it represents primary sensory processing. Language-specific activation maps for each patient are presented in Figure 2. This activation, occurring after the resolution of the N1m, was generally observed in the superior, middle, and occasionally inferior temporal gyri, angular gyrus, temporal pole, and inferior frontal gyrus in either hemisphere. The number of late, language-specific dipoles occurring after the resolution of the N1m in putative premorbid language areas and homotopic areas of the contralateral hemisphere, prior to MIT and after each MIT session, is presented for each patient in Figure 3. Both patients exhibited more left than right hemisphere activity at the baseline scan, prior to MIT. Both Figure 2. Magnetoencephalography/magnetic resonance imaging (MEG MRI) scans for Patient 1 (responder on left side of figure) and Patient 2 (nonresponder on right side of figure) at each scanning session. Green dots represent individual dipoles thresholded after the resolution of the N1m in the left hemisphere, red dots in the right hemisphere. Sources are projected up to 1.5 cm in the coronal plane.

5 MEG LANGUAGE MAPS AFTER MELODIC INTONATION THERAPY 313 Figure 3. Total number of late, language-specific dipoles thresholded after the resolution of the N1m during each scanning session for the left (white bars) and right (grey bars) hemispheres. Responder refers to Patient 1, nonresponder to Patient 2. patients also exhibited an increase in left hemisphere activation after the first block of therapy. Patient 1, who responded with improvement in language function to MIT, exhibited a steady reduction in activation within the right hemisphere across the two therapy blocks, resulting in a strong left hemisphere lateralization of MEG activity. Patient 2, who did not respond positively to MIT, exhibited increased right hemisphere activation after both blocks of therapy compared to baseline, resulting in a right hemisphere lateralization of MEG activity. DISCUSSION Contrary to our hypothesis both the responding and nonresponding patients in this study exhibited an increase in left hemisphere activity during the MEG scanning task after MIT. However, while the responding patient exhibited a decrease in the amount of activation in the right hemisphere after therapy the nonresponding patient exhibited an increase in right hemisphere activation. Thus, while both patients exhibited increased activation during the MEG task after MIT, the responder was lateralized to the left (presumably dominant) hemisphere for language function, and the nonresponder was lateralized to the right hemisphere after therapy. There have been few functional or structural imaging studies regarding the effects of MIT. Belin et al. (1996), using PET, found that patients who responded positively to MIT exhibited an increase in right hemisphere blood flow while repeating spoken words. In contrast, in the same patients, repetition of MIT-loaded words was associated with an increase in left hemisphere activation in Broca s area, accompanied by a decrease in activation in an area within the right hemisphere homotopic to Wernicke s area. The authors interpreted these findings as indicating that recovery of language function was associated with reactivation of left prefrontal areas while abnormal activation within the right hemisphere may be associated with the persistence of aphasia. In a structural imaging study Naeser and Helm-Estabrooks (1985) found that good response to MIT was associated with lesions in Broca s area and/or white matter deep but no significant lesion in Wernicke s area and no lesion in the temporal isthmus or the right hemisphere. In contrast, poor response to MIT was associated with bilateral lesions or a lesion including Wernicke s area or the temporal isthmus. The current findings, similar to those of Belin et al. (1996), support the hypothesis that MIT acts through promoting left hemisphere activation and that increased activation in the right hemisphere after therapy may have limited functionality or even act in a manner detrimental to behavioral response. However, the nonresponding patient in the current study had a history of chronic occlusion of the left MCA, stenosis of the external carotid artery bilaterally, and significant damage to the left temporal lobe. Behaviorally this patient presented with relatively greater deficits in auditory comprehension than the responding patient and, as with MIT, had experienced limited response to CIT, suggesting that damage to areas that potentially support recovery of language function may have been more extensive in this patient. Therefore, while current data suggest that right hemisphere activation after MIT may be either limited in functionality or even inhibitory in nature, this patient, like the poor responders in the Naeser and Helm-Estabrooks (1985) study, may not have been able to activate areas able to subserve language function to any degree in either hemisphere. Improvement in language function in the responding patient in the current study was limited and task specific. In addition, as MEG is sensitive to muscle artifact produced during overt speech, we acquired neurophysiological data during a covert naming task. These are both potential sources of variance in comparing this to other studies reviewed above that relate to a wide range of language tasks in terms of determining behavioral and neurophysiological response to therapy. The observations made in the case studies presented here, however, are consistent with suggestions that positive response to MIT, at least in some cases, is associated with increased activation in left hemisphere areas that remain intact after stroke. Original manuscript received 20 June 2008 Revised manuscript accepted 4 May 2009 First published online 5 August 2009 REFERENCES Albert, M. L., Sparks, R. W., & Helm, N. A. (1973). Melodic intonation therapy for aphasia. Archives of Neurology, 29, Belin, P., Van Eeckhout, P., Zilbovicius, M., Remy, P., Francois, C., Guillaume, S., et al. (1996). Recovery from nonfluent aphasia after melodic intonation therapy: A PET study. Neurology, 47, Blasi, V., Young, A. C., Tansy, A. P., Petersen, S. E., Snyder, A. Z., & Corbetta, M. (2002). Word retrieval learning

6 314 BREIER ET AL. modulates right frontal cortex in patients with left frontal damage. Neuron, 36, Breier, J. I., Maher, L. M., Novak, B., & Papanicolaou, A. C. (2006). Functional imaging before and after constraintinduced language therapy for aphasia using magnetoencephalography. Neurocase, 12, Damasio, H. (1995). Human brain anatomy in computerized images. New York: Oxford University Press. Fridriksson, J., Morrow-Odom, L., Moser, D., Fridriksson, A., & Baylis, G. (2006). Neural recruitment associated with anomia treatment in aphasia. NeuroImage, 32, Helm-Estabrooks, N., Nicholas, M., & Morgan, A. (1989). Melodic intonation therapy program. San Antonio, TX: Special Press. Kertesz, A. (1982). Western Aphasia Battery. San Antonio, TX: The Psychological Corp. Kertesz, A. (1984). Neurobiological aspects of recovery from aphasia in stroke. International Rehabilitation Medicine, 6, Leger, A., Demonet, J. F., Ruff, S., Aithamon, B., Touyeras, B., Puel, M., et al. (2002). Neural substrates of spoken language rehabilitation in an aphasic patient: An fmri study. NeuroImage, 17, Meinzer, M., Elbert, T., Wienbruch, C., Djundja, D., Barthel, G., & Rockstroh, B. (2004). Intensive language training enhances brain plasticity in chronic aphasia. BMC Biologyl, 2, 20. Musso, M., Weiller, C., Kiebel, S., Muller, S. P., Bulau, P., & Rijntjes, M. (1999). Training-induced brain plasticity in aphasia. Brain, 122(9), Naeser, M. A., & Helm-Estabrooks, N. (1985). CT scan lesion localization and response to melodic intonation therapy with nonfluent aphasia cases. Cortex, 21(2), Nicholas, L. E., & Brookshire, R. H. (1993). A system for quantifying the informativeness and efficiency of the connected speech of adults with aphasia. Journal of Speech and Hearing Research, 36, Obler, L. K., & Albert, M. L. (1979). The Action Naming Test. Boston: VA Medical Center. Pedersen, P. M., Jorgensen, H. S., Nakayama, H., Raaschou, H. O., & Olsen, T. S. (1995). Aphasia in acute stroke: Incidence, determinants, and recovery. Annals of Neurology, 38, Richter, M., Miltner, W. H., & Straube, T. (2008). Association between therapy outcome and right-hemispheric activation in chronic aphasia. Brain, 131, Sarvas, J. (1987). Basic mathematical and electromagnetic concepts of the biomagnetic inverse problem. Physics in Medicine and Biology, 32, Sparks, R., Helm, N., & Albert, M. (1974). Aphasia rehabilitation resulting from melodic intonation therapy. Cortex, 10, Sparks, R. W., & Holland, A. L. (1976). Method: Melodic intonation therapy for aphasia. Journal of Speech and Hearing Disorders, 41(3), Vitali, P., Abutalebi, J., Tettamanti, M., Danna, M., Ansaldo, A. I., Perani, D., et al. (2007). Training-induced brain remapping in chronic aphasia: A pilot study. Neurorehab. and Neural Repair, 21,