IN RECENT YEARS, researchers have explored the effectiveness

CLINICAL NOTE Effects of Bromocriptine in a Patient With Crossed Nonfluent Aphasia: A Case Report Anastasia M. Raymer, PhD, David Bandy, MA, John C. Adair, MD, Ronald L. Schwartz, MD, David J.G. Williamson, PhD, Leslie J. Gonzalez Rothi, PhD, Kenneth M. Heilman, MD ABSTRACT. Raymer AM, Bandy D, Adair JC, Schwartz RL, Williamson DJG, Gonzalez Rothi LJ, Heilman KM. Effects of bromocriptine in a patient with crossed nonfluent aphasia: A case report. Arch Phys Med Rehabil 2001;82:139-44. Objective: Because studies have shown some positive effects of the dopaminergic agent bromocriptine for improving verbal production in patients with nonfluent aphasia, we examined its effect in a patient with an atypical form of crossed nonfluent aphasia from a right hemisphere lesion. Design: Open-label single-subject experimental ABAB withdrawal design. Patient: A right-handed man who, after a right frontal stroke, developed nonfluent aphasia, emotional aprosodia, and limb apraxia. Intervention: Escalating doses up to 20mg of bromocriptine in 2 separate phases. Main Outcome Measures: We measured verbal fluency (words/min in discourse, Thurstone letter fluency), expression of emotional prosody, and gesture production. Results: The patient showed substantial improvement in both verbal fluency measures and no significant improvement in gesture or emotional prosody. Verbal fluency improvements continued in withdrawal phases. Conclusions: Our results are less likely caused by practice or spontaneous recovery because we observed little improvement in emotional prosody and gesture tasks. Verbal fluency improvements during treatment and withdrawal phases suggest that the effects of bromocriptine may be long-lasting in its influence on the neural networks subserving verbal initiation. Key Words: Aphasia, Broca; Bromocriptine; Rehabilitation; Verbal behavior; Case report. 2001 by the American Congress of Rehabilitation Medicine and the American Academy of Physical Medicine and Rehabilitation From Old Dominion University, Norfolk, VA (Raymer); VA Medical Center, Gainesville, FL (Bandy, Heilman); University of New Mexico, Albuquerque, NM (Adair); St. Barnabas Institute of Neurology & Neurosurgery, West Orange, NJ (Schwartz); Neurology: Child & Adult, PC, Mobile, AL (Williamson); University of Florida and VA RR&D Brain Rehabilitation Center, Gainesville, FL (Rothi, Heilman); and University of Florida, Gainesville, FL (Heilman). Accepted April 3, 2000. Supported by the National Institutes of Health, Department of Veterans Affairs Medical Research Service and Rehabilitation Research and Development Service (grant no. NS 25675). Presented in part at the annual meeting of the International Neuropsychological Society, Orlando, FL, February 1997. No commercial party having a direct financial interest in the results of the research supporting this article has or will confer a benefit upon the authors or upon any organization with which the authors are associated. Reprint requests to Anastasia M. Raymer, PhD, Dept of Early Childhood/Speech Pathology/Special Education, Child Study Center, 45th St & Hampton Blvd, Old Dominion University, Norfolk, VA 23529-0136, e-mail: sraymer@odu.edu. 0003-9993/01/8201-5998$35.00/0 doi:10.1053/apmr.2001.18056 139 IN RECENT YEARS, researchers have explored the effectiveness of pharmacologic treatments in the restitution of language in some patients with aphasia. 1 For example, the nonfluency or reduced verbal generativity often seen in patients with left frontal lesions and either Broca s or transcortical motor aphasia may relate in part to dynamic intentional impairments caused by reduced dopaminergic input to the frontal lobe. 2 Therefore, the use of dopaminergic agents may induce improved fluency in some patients with nonfluent aphasia, particularly transcortical motor aphasia. 3-6 However, studies using the dopaminergic agent bromocriptine have yielded mixed results. Although some studies have reported dramatic improvements associated with the administration of bromocriptine, 3 others have reported limited or no verbal improvement. 7-11 The positive effects of bromocriptine have included improved naming accuracy and latency, 4,5 fewer pauses and hesitations in connected speech, 2,4 and increased initiation of verbalization. 6 However, other studies have found no reduction in hesitancy or naming improvement. 8 In general, studies have shown little improvement in standardized aphasia tests and other language measures such as length of utterances and grammatical complexity. 5,8-10 One reason for the divergent findings may relate in part to the neurologic basis for nonfluent aphasia. Whereas poorer effects may be anticipated in patients for whom critical language regions are destroyed (eg, Broca s aphasia), bromocriptine would be expected to be more effective in patients whose nonfluent aphasia may relate partly to a disconnection of frontal regions from subcortical dopaminergic input (ie, transcortical motor aphasia). 2 Occasionally, patients who are right handed may present with aphasia after a lesion of the nondominant right hemisphere (crossed aphasia). Researchers estimate that the incidence of crossed aphasia is 1% to 2% of all right-handed patients with aphasia. 12 These crossed aphasias may parallel the patterns of aphasia seen in left hemisphere lesioned patients because nonfluency may be associated with frontal lesions. 12,13 In an extensive review of patients with crossed aphasia, Coppens and Hungerford 14 noted that more patients have been described with nonfluent crossed aphasia than with fluent forms of aphasia. Although some recovery is possible in patients with crossed aphasia, 14 little is known about its treatment. With regard to pharmacologic treatment in nonfluent crossed aphasia, a dopamine agonist not only may help activate the dysfunctional right frontal lobe, but also may activate dormant left hemisphere language areas that typically mediate language production. We propose that pharmacologic treatment of the nonfluency in crossed aphasia by using bromocriptine may be beneficial, especially because these patients have an intact left hemisphere that typically mediates processes related to language fluency. We examined this possibility in a patient with crossed nonfluent aphasia.

140 BROMOCRIPTINE IN CROSSED APHASIA, Raymer METHODS Case Report Our patient was a 50-year-old right-handed salesman with 15 years of education who had sustained a right hemisphere infarction that resulted in left hemiparesis, nonfluent aphasia, severe limb apraxia, emotional aprosodia, and neglect of left space. When questioned about manual activities based on versions of handedness questionnaires, the patient reported that he had used his preferred right hand all his life. Both his parents and siblings also were fully right-handed. A magnetic resonance image of his right hemisphere lesion, completed 2 months poststroke, was mapped onto computerized images. 15 The lesion involved a large portion of right dorsolateral frontal cortex (Brodmann s areas 4, 44, 45, 47), extending posteriorly to the postcentral gyrus (Brodmann s area 3, 1, 2), as well as patchy involvement of subcortical white matter. (See Raymer et al 16 for the scans.) Standardized aphasia assessment by using the Western Aphasia Battery 17 (WAB) 1 month after his stroke indicated an aphasia quotient of 79.8 of 100, with reduced performance in fluency and naming subtests. The patient s performance in comprehension and repetition was excellent. He showed moderate anomia, scoring 33 of 60 on the Boston Naming Test 18, BNT which is significantly poorer than a score of 45 of 60 that is expected for persons of his age and education level. 19 On the Controlled Oral Word Association Test 20 (COWAT) (letters F, A, S), 20 he produced 7 words in 3 minutes, which is far below expected levels (approximately 12 words/min). His spontaneous speech was nonfluent because of significantly reduced verbal initiation and elaboration leading to reduced amount of verbal output. However, he produced no phonologic errors or agrammatism such as is observed in some forms of left frontal aphasia. Overall, his aphasia pattern was consistent with transcortical motor aphasia. On the Florida Apraxia Screening Test Revised, 21 in which subjects produce 30 gestures to verbal command, the patient correctly produced only 9 of 30 gestures with his nonparetic right arm and hand, indicating a significant crossed apraxia compared with normal levels of performance (cutoff score 15 of 30). 21 (See Raymer 16 for an extensive description of our patient s crossed apraxia). In addition, consistent with a right frontal lesion, he showed left spatial neglect and emotional aprosodia in spontaneous speech. The patient gave his consent to participate in all experimental procedures. Experimental Procedures Because of the patient s severe nonfluent aphasia, at 2 months poststroke we initiated a trial of bromocriptine to evaluate its effects on his reduced verbal fluency. Our study incorporated a within-subject ABAB withdrawal design in which we alternated phases A (no drug) and B (bromocriptine), withdrawing the drug for an extended time between treatments. We monitored his performance on 3 experimental tasks for which we predicted that bromocriptine might improve his impaired performance, and a control task for which we anticipated relatively little improvement. Whereas earlier studies of the use of bromocriptine treatment of aphasia often used standardized aphasia assessment tools, the time series design that we incorporated in this experiment dictated that we use repeated measures that were replicable in the context of multiple 60-minute probe sessions. The 3 tasks were administered in sessions 1 to 5 times per week over the phases of the experiment. Performance in standardized measures was evaluated before and after treatment. Two examiners administered daily probe measures. Gesture to command. Because we were measuring improvement in the acute phase of the patient s language recovery, we monitored his performance in a control praxis task that we anticipated was less likely to improve in conjunction with the pharmacologic treatment. In this gesture to verbal command task, he produced 20 transitive gestures (tool-object pantomimes, eg, use a hammer to pound a nail into the wall), as instructed by the examiner. All gesture responses were videotaped for later scoring. Three trained examiners viewed the videotaped gestures and scored accuracy by using a coding system sensitive to gesture content and temporospatial dimensions of gestures. 21 Correct responses were those gestures identified as correct in both content and temporospatial accuracy by at least 2 of the 3 raters. In a 30-item version of this task, neurologically normal individuals accurately show both content and temporospatial dimensions of movement with greater than 50% accuracy. Procedural discourse. To elicit a consistent corpus of conversational speech, we asked the patient to describe the procedures involved in 6 different daily living activities (eg, making lunch, changing a flat tire), speaking for 1.5 to 2 minutes per item. In later phases of the experiment, we added 3 new topics for discourse for comparison with the original 6 topics that had been repeatedly rehearsed. All discourse was audiotape recorded. One examiner completed all transcription of discourse over all phases of the experiment. From the corpus of connected speech, we calculated words spoken per minute, a measure that considered the large number of pauses he made in connected speech. We also calculated measures of lexical retrieval (proportions of nouns, verbs, adjectives/adverbs, function words) to evaluate whether certain classes of words were affected preferentially by the bromocriptine. Letter fluency. By using a procedure similar to the CO- WAT, 20 the patient was given 1 minute to name words beginning with each of 6 letters (B, P, M, D, T, W), selected from letter categories identified as being from easy to moderate in difficulty. In later phases of the experiment, he provided responses to 3 new letters (L, R, H) for comparison with the other 6 rehearsed letters. The examiner transcribed online the words the patient produced. We measured the mean number of words he generated per letter category. Emotional prosody. The patient was audiotape recorded as he read 12 sentences aloud by using 1 of 4 target emotional intonations indicated by the examiner (happy, sad, angry, neutral). Two raters who were unaware of the prosody targets listened to the recording and judged the emotional prosody the subject had attempted for each sentence. Correct responses were considered to be those sentences for which both raters agreed that the prosody matched the target emotion. We calculated the percentage of correct prosodic production. Treatment Procedures After a baseline period during which we obtained 5 measurements per task (phase A1), the treatment phase was instituted as the patient was escalated to a maximum dose of 20mg of bromocriptine per day (drug phase B1). Phase B1 was maintained for 3 weeks before the drug was gradually withdrawn. After 2 months of withdrawal (phase A2), the patient was again escalated to the maximum dose of bromocriptine for 6 weeks (drug phase B2). The drug was again withdrawn (phase A3) and behaviors were measured for 2 final months. Results of all tasks were tabulated for daily sessions to assess treatment effects across experimental phases. We report results for daily performance in experimental tasks, as well as average performance in each task within each phase of the experiment. Although it is often preferable to compare the performance of

BROMOCRIPTINE IN CROSSED APHASIA, Raymer 141 Table 1: Mean Performance Within Each Phase for 4 Experimental Tasks Task A1 B1-Drug A2 B2-Drug A3 Gesture to command (% correct) 3.0% 4.4% 5.0% 8.3% 5.0% Emotional prosody (% correct) 30% 24% 28% 38% 50% Discourse (mean words/min) Set 1 (6 topics) 11.22 words 18.60 words* 23.74 words* 29.49 words* 29.95 words Set 2 (3 topics) 29.29 words 33.99 words Discourse (proportions of words) Nouns.18.22.24.25.27 Verbs.19.15.13.14.13 Adjectives/adverbs.06.10.10.14.09 Functors.57.54.53.52.52 Thurstone (mean words/letter) Set 1 (6 letters) 2.38 words 5.98 words* 8.63 words 9.85 words 13.97 words Set 2 (3 letters) 7.46 words 12.99 words * Denotes significant change compared with previous phase (C statistic). Abbreviations: A, no drug baseline/withdrawal phases; B, drug phases. an experimental patient with that of control patients on the experimental procedures, we do not report data for control patients who also completed repeated measures with our experimental tasks. We constructed our experiment as a withinsubject design across behaviors, in which we anticipated changes would occur for only selected behaviors, while maintaining control by showing that alternative behaviors do not change in relation to the treatment. For the statistical analyses, we used the C statistic, 22 which is sensitive to changes in slope of performance over time while accounting for variability of performance over phases of a time series experiment. RESULTS Table 1 shows the patient s mean performance within each experimental phase for each task and figures 1 through 4 depict daily results of task performance. Gesture to command. Figure 1 shows the results of the gesture to command control task. In baseline measures (A1), the patient attained a maximum of 5% accuracy, which is far below the normal level of performance. 21 His performance remained at floor levels (0% 15% correct) through all subsequent phases of the experiment. His gestures remained largely unintelligible or affected by multiple aberrations in the temporospatial configuration of the intended gestures. Apparently, neither bromocriptine nor spontaneous recovery had an ameliorating effect on his severe limb apraxia. Emotional prosody. Figure 1 also shows the results of the emotional prosody task in which the subject performed at levels no different from chance (25%) in the first baseline phase (A1). Although a small improvement was noted in the final withdrawal phase (A3), this change was not significant compared with baseline levels (C.490, z 1.59, NS). 22 Thus, we did not see a change in his emotional prosody as a result of the drug treatment. Letter fluency. Figure 2 displays our subject s results for the letter fluency tasks. In the baseline phase, his performance was far below normal levels and was stable over 5 baseline observations with no change in slope (letter fluency: C.0039, z.0110, NS). Marked improvements in word fluency were noted. In drug treatment phase B1, he showed significant fluency improvement in letter category generation (C.895, z 3.496, p.01). In subsequent withdrawal and drug phases, though there was an upward trend in letter fluency performance, these changes were not significant (phase A2: C.073; z.257; NS; phase B2: C.209, z.893, NS; phase A3: C.35, z 1.13, NS). Because of concern that the Fig 1. Percentage correct for gesture production and emotional prosody tasks across no drug (A) and drug (B) phases of the experiment.

142 BROMOCRIPTINE IN CROSSED APHASIA, Raymer Fig 2. Mean number of words produced per letter category across no drug (A) and drug (B) phases of the experiment. nonsignificant trend over time reflected practice effects for the 6 rehearsed letter categories, we instituted 3 new letter categories in phase B2. Although those new categories had not been rehearsed, performance on them was commensurate with the performance for prior rehearsed categories. Discourse. Figure 3 displays the patient s results for words per minute calculated during spontaneous discourse. In the baseline phase, his performance of 8 to 18 words per minute was far below normal speaking rates ( 150 words/min) and was stable over 5 baseline observations with no change in slope (C.13, z.37, NS). In the initial baseline, his reduced fluency was marked by a large number of pauses and delays in responding, resulting in the overall amount of verbal output being dramatically reduced. However, there were marked improvements in discourse fluency as measured by increases in words spoken per minute. In drug treatment phase B1, the subject showed significant improvement in words per minute across sessions (C.72, z 2.769, p.01). When bromocriptine was withdrawn in phase A2, discourse fluency continued to improve (C.633, z 2.14, p.05). At the end of withdrawal phase A2, as the discourse fluency measure stabilized, we again instituted drug treatment phase B2. We again observed increasing levels of discourse fluency (C.476, z 1.98, p.05). Finally, when the drug was withdrawn in phase A3, we observed no further significant increases in discourse fluency (C.246, z.794, NS). To determine whether increases in words per minute observed over the phases of the experiment were related to rehearsal effects, we instituted 3 new discourse sets (7 9) in phase B2. The patient s performance for these new sets was at a level commensurate with the 6 rehearsed topics (fig 3). It appeared that the bromocriptine led to significant improvement in discourse fluency. Finally, figure 4 shows the results of our analysis of the proportions of words from different grammatical classes that our subject produced in discourse over phases of the experiment. Although we documented a significant increase in overall numbers of words that he spoke per minute, the proportions of words for each grammatical category that he produced remained stable over treatment phases. The majority of his words were functors (approximately 55% of all words), in keeping with the proportion of words typically spoken by unimpaired subjects. 23 He also produced a good portion of nouns (approximately 25%) and verbs (approximately 15%) Fig 3. Number of words per minute produced in discourse across no drug (A) and drug (B) phases of the experiment.

BROMOCRIPTINE IN CROSSED APHASIA, Raymer 143 Fig 4. Proportions of words across grammatic categories (nouns, verbs, adjective/adverbs, functors) over no drug (A) and drug (B) phases of the experiment. over all phases of the experiment, resulting in a noun:verb ratio of approximately 1.67. This ratio is somewhat higher than is typically observed in unimpaired subjects (1.18.30), 23 because he used more nouns than verbs. Standardized testing. In follow-up aphasia assessment with the WAB 2 months after bromocriptine was discontinued, the patient had an aphasia quotient of 85.8 of 100, associated with improvements in spontaneous speech and word fluency subtests compared with the pretreatment testing (79.8 of 100). Little change was evident on the BNT, where he scored 35 of 60 correct compared with 33 of 60 at the onset of treatment. Thus, little change was evident on general indexes of language function. DISCUSSION We studied the response to treatment with bromocriptine in a right-handed patient with crossed nonfluent aphasia to evaluate its effects on measures of verbal output. We included a measure of skilled movement, gesture to verbal command, to evaluate whether the effects of bromocriptine were specific to verbal fluency or whether factors extraneous to the drug treatment might have influenced performance in experimental tasks during the patient s acute phase of recovery. Our subject showed no significant improvement on the gesture to command task. He also showed a defect in the expression of emotional prosody. Although we anticipated that the production of emotional prosody might improve with bromocriptine treatment, we found no significant improvement over experimental phases. The lack of significant change in both gesture to command and emotional prosody suggests that extraneous factors such as repeated practice on experimental tasks, or spontaneous recovery from cognitive impairments, cannot account entirely for the positive results in the experimental verbal fluency measures. In contrast to his praxis and prosody, the patient experienced marked improvement for the 2 verbal fluency tasks during the drug phases of the experiment. By the final phases, his letter fluency performance approached levels that are normal for his age group. In discourse, he doubled the number of words he produced per minute of connected speech from baseline (max 18 words/min) to final stages of the experiment (max 39 words/min), but his overall performance remained much lower than normal ( 150 words/min) and was commensurate with levels of other patients with nonfluent aphasia. 23 The improvements in overall fluency were not specific to a grammatic category such as verbs, which have important neural representational networks in the left frontal regions. 24 The proportions of words the patient produced in each category were stable over all phases of the experiment. Therefore, the fluency improvements equally affected all grammatic categories. To attribute our patient s verbal fluency improvements to bromocriptine, it would be helpful had we been able to show stabilization or decline in performance when treatment was discontinued. Letter fluency performance did not significantly increase in the no treatment phases, but discourse verbal fluency improved significantly during the first withdrawal phase. This unexplained improvement may have resulted from several sources. One possibility is that improvements were the effect of repeated practice for the tasks/stimuli over time. To evaluate this possibility, we added new items in the discourse and letter fluency tasks during phases B2 and A3. Fluency performance for the new items was at levels commensurate with the rehearsed items, arguing against a simple practice effect leading to improved verbal fluency for the first categories rehearsed. There appeared to be a more generalized improvement that we also observed in the newly introduced categories and topics. Alternatively, the improvement in verbal fluency across experimental phases might have been caused by spontaneous recovery independent of drug treatment. However, our patient had a stable extended baseline performance before treatment. Moreover, he did not show significant improvement on the gesture to command and emotional prosody tasks. If spontaneous recovery was the basis for improvements in verbal fluency measures, it is likely that we would have also seen improvement in gesture and prosody. These observations lead us to believe that the improvements noted in verbal fluency are not entirely related to spontaneous recovery. Therefore, the changes we observed in verbal fluency appear related partly to the effect of treatment with bromocriptine. However, it is somewhat bothersome that improvements were evident in the withdrawal phase (A2) after bromocriptine was discontinued. Again, this observation is less likely to be an effect of spontaneous recovery, because the patient had no change in 2 other measures. Therefore, it may be that the improvements noted in the drug withdrawal phase seem related in part to bromocriptine. This effect may have occurred because bromocriptine had longer lasting effects than we antici-

144 BROMOCRIPTINE IN CROSSED APHASIA, Raymer pated and bromocriptine may have influenced his performance in withdrawal phases. Alternatively, we consider the possibility that bromocriptine may have altered brain plasticity in uninjured regions. Perhaps the bromocriptine incited alternative neural networks to mediate verbal fluency functions that were impaired by the brain injury. 25,26 Even after the medication was withdrawn, the remaining intact brain regions may have become an integral part of the neural network that mediates language functions that were originally processed by the damaged right hemisphere. Absent of substantive findings, such as changes in functional neuroimaging, remains only speculation. Previous articles have noted mixed results when using bromocriptine to treat impaired verbal fluency in patients with nonfluent aphasi. Whereas results for our patient suggest that bromocriptine had a role in improving his verbal fluency, there are several possible reasons why no treatment may be more beneficial in individuals in whom those areas critical for mediating a specific behavior have not been destroyed. In our patient, his crossed nonfluent aphasia resulted from a right hemisphere lesion. Because mesocortical projection pathways to the right frontal lobe were disrupted by the cortical lesion, a possibility is that his improved verbal fluency resulted in part because of the activity of intact left hemisphere frontal regions that typically mediate language fluency in right-handed individuals. This interpretation is partially supported by the fact that our patient showed little improvement in the expression of emotional prosody, a skill that typically depends on the integrity of the damaged right hemisphere. 27 An additional reason for the mixed results across studies lies in the types of measures researchers have used to assess verbal fluency. For example, some researchers who reported little effect of bromocriptine on verbal fluency focused on the length of utterances produced by patients. 8 In our patient s verbal discourse, he used sentence-length utterances from the start of our studies. What we documented was a reduction in the numbers of pauses and aborted attempts to produce utterances. In addition, he was able to convey a number of additional ideas through larger number of words. Therefore, studies that have analyzed pause time and words per minute have documented the effects of bromocriptine in improving verbal output. In contrast, when we analyzed the linguistic features of our patient s recovery, we found no change in the proportions of words across grammatic categories that he was able to produce, contrary to our prediction that verbs might be affected somewhat differentially by increased dopaminergic input to the frontal lobe. CONCLUSION On the basis of these preliminary findings in 1 patient, we suggest that bromocriptine s effects occur at a more dynamic, intentional stage in language processing that is independent of linguistic formulation and retrieval stages. 28 Acknowledgments: The authors thank our patient for his willingness to participate in this study. References 1. Small SL. Pharmacotherapy of aphasia: a critical review. Stroke 1994;25:1282-9. 2. Sabe L, Leiguarda R, Starkstein SE. An open-label trial of bromocriptine in nonfluent aphasia. Neurology 1992;42:1637-8. 3. Albert ML. Neurobiological aspects of aphasia therapy. Aphasiology 1988;2:215-8. 4. Albert ML, Bachman DL, Morgan A, Helm-Estabrooks N. Pharmacotherapy for aphasia. Neurology 1988;38:877-9. 5. Bachman DL, Morgan A. The role of pharmacotherapy in the treatment of aphasia: preliminary results. Aphasiology 1988;2: 225-8. 6. Stewart JT, Leadon M, Gonzalez Rothi LJ. Treatment of a case of akinetic mutism with bromocriptine. J Clin Neuropsychiatr Clin Neurosci 1990;2:462-3. 7. Gupta SR, Mlcoch AG. Bromocriptine treatment of nonfluent aphasia. Arch Phys Med Rehabil 1992;73:373-6. 8. Gupta SR, Mlcoch AG, Scolaro C, Moritz T. Bromocriptine treatment of nonfluent aphasia. Neurology 1995;45:2170-3. 9. MacLennan DL, Nicholas LE, Morley GK, Brookshire RH. The effects of bromocriptine on speech and language function in a man with transcortical motor aphasia. In: Prescott TH, editor. Clinical aphasiology. vol 20. Austin (TX): Pro-Ed; 1992. p 145-55. 10. Ozeren A, Sarica Y, Mavi H, Demirkiran M. Bromocriptine is ineffective in the treatment of chronic nonfluent aphasia. Acta Neurol Belg 1995;95:235-8. 11. Sabe L, Salvarezza F, Garcia Cuerva A, Leiguarda R, Starkstein S. A randomized, double-blind, placebo-controlled study of bromocriptine in nonfluent aphasia. Neurology 1995;45:2272-4. 12. Alexander MP, Fischette MR, Fischer RS. Crossed aphasias can be mirror image or anomalous. Brain 1989;112:953-73. 13. Basso A, Capitani E, Laiacona M, Zanobio ME. Crossed aphasia: one or more syndromes? Cortex 1985;21:25-45. 14. Coppens P, Hungerford S. Crossed aphasia. In: Coppens P, Lebrun Y, Basso A, editors. Aphasia in atypical populations. Mahwah (NJ): Lawrence Erlbaum; 1998. p 203-60. 15. Damasio H. Human brain anatomy in computerized images. New York: Oxford Univ Pr; 1995. 16. Raymer AM, Merians AS, Adair JC, Schwartz RL, Williamson DJG, Rothi LJG, et al. Crossed apraxia: implications for handedness. Cortex 1999;35:183-99. 17. Kertesz A. The Western Aphasia Battery. Orlando (FL): Grune & Stratton; 1982. 18. Kaplan E, Goodglass H, Weintraub S. The Boston Naming Test. Philadelphia: Lea & Febiger; 1983. 19. Nicholas L, Brookshire R, MacLennan J, Porrazzo S. Revised administration and scoring procedures for the Boston Naming Test and norms for non-brain-damaged adults. Aphasiology 1989;3: 569-80. 20. Borkowski JG, Benton AL, Spreen O. Word fluency and brain damage. Neuropsychologia 1967;5:135-40. 21. Rothi LJG, Raymer AM, Heilman KM. Limb praxis assessment. In: Rothi LJG, Heilman KM, editors. Apraxia: the neuropsychology of action. East Sussex (UK): Psychology Pr; 1997. p 61-73. 22. Tryon WW. A simplified time-series analysis for evaluating treatment interventions. J Appl Behav Anal 1982;15:423-9. 23. Saffran EM, Berndt RS, Schwartz MF. The quantitative analysis of agrammatic production: procedure and data. Brain Lang 1989; 37:440-79. 24. Damasio AR, Tranel D. Nouns and verbs are retrieved with differently distributed neural systems. Proc Natl Acad Sci U S A 1993;90:4957-60. 25. Kolb B. Brain plasticity and behavior. Mahwah (NJ): Lawrence Erlbaum; 1995. 26. Walker-Batson D, Smith P, Curtis S, Unwin H, Greenlee R. Amphetamine paired with physical therapy accelerates motor recovery after stroke. Further evidence. Stroke 1995;26:2254-9. 27. Heilman KM, Bowers D, Valenstein E. Emotional disorders associated with neurological diseases. In: Heilman KM, Valenstein E, editors. Clinical neuropsychology. 3rd ed. New York: Oxford Univ Pr; 1993; 461-97. 28. Gold M, Nadeau SE, Jacobs DH, Adair JC, Rothi LJG, Heilman KM. Adynamic aphasia: a transcortical motor aphasia with defective semantic strategy formation. Brain Lang 1997;57:374-93.