This article was downloaded by: [Northwestern University] On: 14 February 2015, At: 09:33 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK
Neurocase: The Neural Basis of Cognition Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/nncs20
Language disturbances associated to insular and entorhinal damage: study of a patient affected by herpetic encephalitis a
b
a
Elena Baratelli , Marcella Laiacona & Erminio Capitani a
Health Sciences Department, Neurology Unit, Milan University, S.Paolo Hospital, Milan, Italy b
S.Maugeri Foundation, IRCCS, Aphasia Unity, Neurology, Veruno Scientific Institute, Veruno (NO), Italy Published online: 04 Mar 2014.
Click for updates To cite this article: Elena Baratelli, Marcella Laiacona & Erminio Capitani (2014): Language disturbances associated to insular and entorhinal damage: study of a patient affected by herpetic encephalitis, Neurocase: The Neural Basis of Cognition, DOI: 10.1080/13554794.2014.892623 To link to this article: http://dx.doi.org/10.1080/13554794.2014.892623
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Neurocase, 2014 http://dx.doi.org/10.1080/13554794.2014.892623
Language disturbances associated to insular and entorhinal damage: study of a patient affected by herpetic encephalitis Elena Baratellia, Marcella Laiaconab and Erminio Capitania* a
Health Sciences Department, Neurology Unit, Milan University, S.Paolo Hospital, Milan, Italy; bS.Maugeri Foundation, IRCCS, Aphasia Unity, Neurology, Veruno Scientific Institute, Veruno (NO), Italy
Downloaded by [Northwestern University] at 09:33 14 February 2015
(Received 11 July 2013; accepted 5 February 2014) The herpes simplex encephalitis (HSE) patient reported in this study presented a left hemisphere lesion limited to the left insula and to the left anterior parahippocampal region. The patient was followed longitudinally, focusing on the aphasia type, the language recovery, and the integrity of semantic representations. The language deficit was of fluent type, without phonological impairment, and showed a good but incomplete recovery after four months. A semantic impairment was possible at the onset, but recovered quickly and did not present a disproportionate impairment of living categories. Keywords: herpetic encephalitis; insula; parahippocampal region; aphasia; semantics
The cognitive deficits associated to herpes simplex encephalitis (HSE) are of special interest for neuropsychologists. HSE may cause a variety of disturbances, for instance, amnesic syndromes, semantic impairment, and visual agnosia, which are less commonly seen in stroke patients. These disturbances may co-occur in the same patient or present as rather selective disorders. For instance, Kapur et al. (1994) in a study focused on amnesia and its structural correlates, described 10 HSE patients and reported naming deficits in only a limited number of cases (4/10). In contrast, Barbarotto, Capitani, and Laiacona (1996) found that 7/8 cases (87.5%) in their HSE series were impaired on picture naming: the discrepancy with Kapur et al. is probably due to the extent of left hemisphere involvement in different samples. Also the findings regarding visual agnosia as an effect of HSE are not homogeneous: a deficit of visual object recognition not dependent on visual loss or language impairment, assessed with visual reality decision on chimeric pictures, has been found spared in some cases and defective in other cases, sometimes only for biological categories (for a comprehensive review of this aspect see Capitani, Laiacona, Mahon, & Caramazza, 2003). The heterogeneous cognitive profile of HSE patients is, at least in part, a function of the variable site of brain damage. To fully explain the heterogeneity of the HSE population, a first point to consider is that encephalitis may affect with varying severity left and/or right hemispheres, and that different areas can be damaged within each hemisphere. The realm of damaged structures, reviewed in detail by Kapur et al. (1994), includes the hippocampus, parahippocampal region, amygdala, *Corresponding author. Email: [email protected] © 2014 Taylor & Francis
temporal poles and lateral temporal gyri, fusiform gyrus, insular cortex, thalamus, medial and basal frontal cortex, occipital cortex, and possibly other areas. A more recent study of the lesion topography of five HSE patients (Gitelman, Ashburner, Friston, Tyler, & Price, 2001) reports that “the areas of most prominent gray matter abnormality were in the temporal lobes, including the anterior aspects of the lateral temporal cortex, the temporal poles, and the medial temporal cortex including the amygdala, hippocampus, entorhinal cortex, and insular cortex.” To uncover the role of each of the above cortical areas, we may also consider that HSE is not the sole cause of neurological damage of these structures, and also different diseases can affect the same areas damaged by HSE. The areas affected by posterior cerebral artery (PCA) infarcts partly overlap with those damaged by HSE, but PCA strokes never impinge upon the temporal poles and quite frequently affect the more posterior temporal cortex and occipital cortex. Hence, a comparative discussion of the effects of HSE and PCA strokes is quite informative in this context (Capitani et al., 2009). Another source of evidence relevant for the study of anatomo-functional correspondences is functional imaging, although Capitani et al. (2009) noted that lesion studies and functional imaging data have sometimes yielded inconsistent findings. Nevertheless, functional imaging can provide a sharp localization of cognitive tasks and should not be overlooked when trying to interpret the patient’s impairment, also because neurological lesions are seldom so selective as to affect just a single functional brain region. In the present study, which is mainly focused on language and semantic memory, we report clinical,
Downloaded by [Northwestern University] at 09:33 14 February 2015
2
E. Baratelli et al.
anatomical, and functional data of a HSE patient, case RMR, followed for four months after the disease onset. The interest of this case derives from the localization of her lesion, which was strictly limited to the left hemisphere and affected the insular cortex and anterior parahippocampal region, substantially sparing the fusiform gyrus, the most anterior temporal pole areas, posterior parahippocampal gyrus, and frontal and occipital structures. This lesion pattern potentially enables us to explore the role played by the entorhinal-perirhinal cortex in the representation of semantic knowledge. In addition, a clearly defined insular lesion, can help in verifying whether insular lesions are sufficient per se for impairing some aspects of language processing that are debated in the current literature. Before describing the patient, it is useful to briefly summarize the literature about the cognitive effects of damage to the insular cortex and parahippocampal region. The neurological and neuropsychological literature does not offer a large record of cases with circumscribed insular damage and sparing of the classical language areas. Lemieux et al. (2012) published a review of 23 insular cases (7 from their Institution and 16 from the literature): 14/23 patients suffered from left insular infarcts, and 10/23 presented language deficits. Although different types of aphasia are reported in the review, nonfluent speech output and phonemic disturbances seem predominant (eight cases); in the two cases with fluent aphasia (cases 1 and 4 of the series, see supplementary material of the paper by Lemieux et al. (2012)) the insular lesion affected the posterior long insular gyri (B1–B2); in another case (case 15), a lesion of the same structures was associated to nonfluent aphasia, but the brain damage also affected the frontal operculum. In general, language deficits were transient but in one case the deficit persisted for several months (Shuren, 1993). In a more general discussion of the effects of insular damage in humans, Ibañez, Gleichgerrcht, and Manes (2010) noted that the anterior insula has been considered by some authors (e.g. Price, 2000) a critical site for articulatory planning and part of the phonological network (e.g. Price, Devlin, Moore, Morton, & Laird, 2005), with specific reference to the left anterior insula. However, Ibañez and colleagues also argued that the insular cortex has wide connections with other language areas and the effects of its lesion might, at least in part, depend on the distant inactivation of functionally connected neural structures. Nonetheless, a lesion of the left insula has been most often associated to nonfluent aphasia with phonemic distortions, and the middle left insula was identified as a crucial structure for apraxia of speech (Dronkers, 1996). In contrast, in a detailed study of a large series of nonlacunar acute stroke patients, Hillis et al. (2004) found no association between the presence of apraxia of speech and lesions of the left insula. Moreover, a neuroimaging study
by Gitelman, Nobre, Sonty, Parrish, and Mesulam (2005) did not find activation of the anterior insula in phonological tasks. Concerning the parahippocampal region, the other structure of interest concerning our patient, this area can be subdivided into three anatomical sectors: the entorhinal, perirhinal, and postrhinal cortex; the latter section, well identifiable in rodents, coincides with the parahippocampal gyrus of the primate and human brain (Furtak, Wei, Agster, & Burwell, 2007). Evidence from animal experiments (see Eichenbaum, 2007, for a review) suggests that neurons in the perirhinal and lateral entorhinal cortex are involved in the representation of individual perceptual stimuli (“what” information) whereas the parahippocampal cortex and medial entorhinal cortex are specialized in processing spatial stimuli (“where” information). On this basis, some authors have suggested that neurons in the left perirhinal cortex play “a crucial role in object identification by integrating information from different sensory systems into more complex polymodal feature conjunctions” (Moss, Rodd, Stamatakis, Bright, & Tyler, 2005; Tyler et al., 2004). The latter authors also claim that objects belonging to the realm of biological categories (e.g. animals, fruit, and vegetables) should “place greater demands on the process of integration of complex conjunction of features and fine-grained discrimination supported by antero-medial temporal cortex.” Moss et al. (2005) found with a neuroimaging study in normal subjects that the left entorhinal cortex was more active when participants were asked to name animals and plant life stimuli than when the same subjects were requested to name tools and vehicles. Their conclusion was that anterior–medial structures of the temporal lobe are also “critical for semantic processing and not just for visual object recognition, as these areas would support the combination of many types of features into meaningful multimodal conceptual representations.” Other authors have investigated the relationship between the parahippocampal region and lexical–semantic abilities. Venneri et al. (2008), examining a group of early Alzheimer patients, observed a significant correlation between lexical competence and the integrity of the perirhinal and parahippocampal cortices. In a recent study, Kivisaari, Tyler, Monsch, and Taylor (2012) analyzed patients with mild cognitive impairment and Alzheimer’s disease and claimed that “the human medial perirhinal cortex is necessary for the disambiguation of perceptually and semantically confusable objects,” with an explicit reference to the living things category; their study presents an updated review of the role of these areas based on different experimental approaches. The above theoretical and experimental account implies that damage to the left parahippocampal region, that includes the lateral entorhinal and the perirhinal cortex, should be sufficient to cause a semantic deficit more severe for biological categories. In a stronger
Downloaded by [Northwestern University] at 09:33 14 February 2015
Neurocase sense, one might even claim that lesions of this area are necessary for there to be a semantic deficit of living category and that semantic knowledge about living entities is specifically and exclusively represented in the parahippocampal region (for a discussion on this point, see also Capitani and Laiacona (2011) and Gainotti (2010)). The HSE patient reported in the present study, case RMR, offers an opportunity to improve our understanding of the functions of the insular cortex and anterior parahippocampal region. Her lesion was substantially limited to these areas, and this is interesting because the role of the same structures has often been investigated in patients affected by larger lesions that also included insula or parahippocampus in the context of a wider brain damage. For instance, in the review by Lemieux et al. (2012) a condition for inclusion was that the lesion of areas adjacent to the insular cortex had to account for less than 50% of the total infarct volume. In turn, Moss et al. (2005) included semantic dementia and HSE cases whose lesions were centered on the left temporal pole, but the affected temporal areas, especially for HSE patients, were far more extended. A caveat is that in our case parahippocampal and insular lesions were associated; therefore, it might be problematic to univocally associate a specific cognitive deficit with a given structure. However, if the perirhinal cortex were really necessary for the semantic processing of living categories, its inactivation would invariably produce a deficit of this function, irrespective of the associated insular lesion. In the light of these considerations, the main questions for our study of RMR are: (1) what type of language disorders are present? (2) was there a semantic category dissociation? and (3) to what extent did the patient recover during the clinical follow-up?
Case description Clinical history Data included in this manuscript were obtained in compliance with regulations of the Milan University and San Paolo Hospital, Milan. RMR, a 63-year-old right-handed woman with 13 years of education, was admitted on November 29, 2012 to the medical ward of the San Paolo Hospital, Milan, on account of the presence of fever, headache, and language difficulties that had lasted for two days. Her previous medical and neurological history were unremarkable. The general medical examination did not reveal pathological data. On the neurological examination, she presented normal alertness and was well oriented in space and time, but was fatuous and slightly impulsive. Her spontaneous language was fluent, without grammatical or prosodic alterations, but her language disorder hampered a communicative exchange showing several anomias and occasional phonemic paraphasias. The same language alterations were also evident
3
when she was asked to name common objects. Verbal comprehension could not be assessed with formal testing; however, she understood without any problem the verbal orders given during the neurological examination. Cranial nerves were unimpaired. There was neither limb weakness nor sensory or coordination disturbances. Tendon reflexes were present, symmetrical and had normal strength. Response to plantar stimulation was in flexion. Standing and gait were normal. The blood tests showed slight neutrophil leukocytosis (WBC = 12,000, neutrophils 82.3%) and elevated C-Reactive Protein (15.2 mg/dl); chest X-rays were normal. The brain CT-scan showed a small hyperdense lesion in the left anterior–temporal region surrounded by edema. The background EEG activity with the patient awake consisted of low-voltage 8–9 Hz alpha-waves on posterior regions, with theta waves on the left side and well-organized activity on the right side. On the left frontal and frontal–temporal areas, there were pseudo-periodic slow abnormalities with theta waves, and interpolated slow spike discharges. Photic stimulation did not produce any abnormalities. Cerebrospinal fluid (CSF) was clear, with 500 WBC/mm3, prevailingly lympho-monocytes. Total protein content was slightly increased (59 mg/dl) and CSF sugar was normal. Polymerase chain reaction was positive for HSV-1 (553,123 cp/ml). Treatment was immediately started with e.v. acyclovir, 750 mg tid, and with levetiracetam, 500 mg bid. During the following week, fever disappeared and blood analysis reverted to normality. The patient presented productive, fluent aphasia: the onset of spontaneous speech was normal, and verbal output was neither agrammatic nor affected by prosodic impairment. There were no longer any phonological errors, and RMR was able to repeat nouns without problems. However, verbal communication was impoverished also because there were several anomias and a tendency to use stereotyped, context-dependent utterances. For instance, when requested to explain why she had been admitted to the Hospital, RMR replied: I did not feel well at home, but nothing really severe … but my … my relatives brought me here and the … they gave me the … – pointing to the intravenous treatment – but now I’m feeling well – repeatedly touching her body and moving her upper limbs. To overcome difficulties on noun retrieval, RMR used gestures as an aid, often successfully. Verbal comprehension, tested informally by asking her to execute simple orders, was quite good, although on one single occasion, on visual presentation, she seemed not to recognize a pen or its function. On the whole, RMR was not very concerned about her condition; she was aware of her deficits, but the emotional involvement was flattened. She never presented behavioral alterations, but complained of anosmia and of lack of appetite. RMR’s clinical status steadily improved, and she was dismissed from the hospital on December 20, 2012.
4
E. Baratelli et al.
Downloaded by [Northwestern University] at 09:33 14 February 2015
MRI imaging On November 30, 2012, RMR underwent a morphological magnetic resonance imaging (MRI) scan. The involved structures will be described with reference to the coronal sections of the atlas by Mai, Paxinos, and Voss (2008); these sections conform to the Talairach coordinates; however, in the atlas the sign of the y-coordinate is reversed (negative for the anterior sections, and positive for the posterior sections). In our description, we will use the Talairach notation, that is, with anterior positive values. On the right hemisphere, there were no lesions at all. On the left, the lesion affected the temporal, hippocampal and parahippocampal regions, and the insular cortex. The MRI study on the coronal plane showed that the temporal pole was spared from its anterior limit backwards to about y = 0. At this level, an area of abnormal signal appeared, that extended backwards until about y = −30 to −35. Posterior to that mark, all brain structures were normal. The anterior part of the lesion (from y = 0 backwards) interested the entorhinal and perirhinal cortices, amygdala nuclei, and piriform and insular cortices; at this site, the signal alteration clearly affected both the medial and lateral perirhinal cortex (for an anatomical reference, see Figure 1c, and the corresponding Figure 1 of the paper by Kivisaari et al., 2012). On more posterior coronal sections (at about y = −15), the entorhinal lesion was still evident and also the hippocampal structures were affected. At this level, where the perirhinal cortex merges into the parahippocampal gyrus, the perirhinal lesion extended backwards to involve the parahippocampal
1a-Coronal T2-weigthed MRI scan at the Talairach y-coordinate +10/+15.The image shows that the temporal poles were not involved.
1e-Axial FLAIR MRI scan at the upper pontine level.The image shows involvement left entorhinal shows impaired ofinsular cortex onand the left. perirhinal cortex and amygdalar nuclei
Figure 1.
1b-Coronal T2-weigthed MRI scan at the Talairach y-coordinate 0 /-5. The image shows involvement of left insula (particularly of the ventral part), entorhinal and perirhinal cortex, amygdalar nuclei and piriform cortex
1f-Axial FLAIR MRI scan at the lower mesencephalon level. The image shows involvement of left entorhinal and perirhinal cortex and amygdalar nuclei
MRI images representing RMR’s brain lesion.
gyrus. The signal alteration affecting the parahippocampal region was no longer detectable posterior to y = −35. The insular cortex was clearly affected: on the coronal sections its damage was more severe in its inferior (caudal) half; on the axial sections, the signal alteration involved the whole insula. The axial MRI images confirmed that the left temporal pole was spared, and that the lesion selectively interested the anterior left parahippocampal structures and the whole antero-posterior extent of the left insula.
Neuropsychological assessment First examination (one week from onset) On December 5, 2012, RMR underwent (one week from onset) a picture-naming battery based on 80 stimuli from the Snoodgrass and Vanderwart (1980) set. This battery includes 10 stimuli each from 8 categories: 3 biological categories (animals, fruit, vegetables), 3 artifact categories (tools, pieces of furniture, vehicles), musical instruments and body parts. For each stimulus a number of control indices are available, among which lexical frequency, age of acquisition of the corresponding word, stimulus familiarity, and visual complexity. During the naming session, the patient was never given feedback on the correctness of her responses. A thorough description of this test and of the statistical methodology used in its analysis can be found in Capitani et al. (2009). RMR was cooperative, but performed the task very slowly and with considerable fatigue; therefore, besides
1c-Coronal T2-weigthed MRI scan at the Talairach y-coordinate-10 /-15. The image shows involvement of left insula (particularly of the ventral part), hippocampus, entorhinal cortex and parahippocampus gyrus.
1g-Axial FLAIR MRI scan at the upper mesencephalon level. The image shows involvement of left entorhinal and perirhinal cortex and amygdalar nuclei
1d-Coronal T2-weigthed MRI scan at the Talairach y-coordinate-35/-40. At this level any structural alterations are no longer evident.
1h-Axial FLAIR MRI scan at the diencephalic level. The image shows involvement of left insular cortex
Downloaded by [Northwestern University] at 09:33 14 February 2015
Neurocase the picture-naming task, she underwent only a brief, informal evaluation of cognitive status. On the whole, she did not present problems with visual perception, including the recognition of known people, nor a deficit of gestural abilities. Memory status could not be reliably investigated due to fatigue and attention impairment. On the formal naming evaluation, she was correct on 24/80 stimuli (30%) (see Table 1). The mean of a group of 60 age- and education-matched controls (Laiacona, Barbarotto, Trivelli, & Capitani, 1993) was 75.40/80 (sd = 4.32), that is, 94.2%. Artifacts and biological stimuli yielded the same number of correct responses (9/30), musical instruments were never given the correct name, and 6/10 body parts stimuli were correct. The statistical analysis of the factors that influenced the naming performance was carried out by means of a logistic regression analysis (McCullagh & Nelder, 1983). Because the main interest of this analysis was to assess whether RMR presented a semantic category effect, as customary in our practice we considered only the 60 stimuli pertaining to the living and nonliving categories, that is, animals, fruit, and vegetables for biological categories, and tools, vehicles and pieces of furniture for artifacts (for a discussion of this point see Barbarotto, Capitani, & Laiacona, 2001). The binary response variable was the naming success observed for each stimulus (coded 1 if the response was correct, and 0 if the response was wrong). The model variables were: (1) the category of the stimulus (living versus nonliving), (2) lexical frequency of each stimulus word in the Italian language, after logarithmic transformation, according to Bortolini, Tagliavini, and Zampolli (1972); (3) age of acquisition of the stimulus word (Barbarotto, Laiacona, & Capitani, 2005), which was directly estimated from the performance of 202 Italian children; (4) familiarity with the stimulus: this index, that yields separate ratings for males and females, was designed to capture how frequently we think or speak of the stimulus, how often we see it represented in the media, and how frequently we are confronted with real exemplars (Mean Familiarity Index, see Albanese, Capitani, Barbarotto, & Laiacona, 2000); (5) prototypicality of the
5
stimulus according to the study of Battig and Montague (1969); and (6) the visual complexity of the image (Snoodgrass & Vanderwart, 1980). We first analyzed whether each model variable, considered alone, did in fact significantly predict the naming success with each stimulus; thereafter we included all the significant predictors in the same model, in order to evaluate their contribution (after removing the overlap with the effect of the other terms of the model). The analysis of each effect yields a chi-square distributed point. The computations were carried out by means of the GLIM Programme (Aitkin, Anderson, Francis, & Hinde, 1989). The statistical analysis did not show a discrepancy between biological categories and artifacts (chi-square = 0.000, df = 1, ns). In the simultaneous logistic regression model, significant predictors of the correct responses were stimulus familiarity (chi-square = 7.755, df = 1, p = .005) and age of acquisition of the target name (chi-square = 4.885, df = 1, p = .027). Visual complexity, lexical frequency, and prototypicality were not significant. The errors of RMR were classified as follows (see Table 2): phonological errors (words and non-words), semantic paraphasias, circumlocutions, non-responses, and other errors (including visual errors).
Second examination (four weeks post onset) Four weeks from the encephalitis onset, RMR was examined as an outpatient. On picture naming, which was assessed using the same battery used in the first examination, correct responses were 43/80 (53.3%): biological stimuli were 13/30 correct (43.3%), and artifacts 19/30 correct (63.3%). One musical instrument and all the 10 body parts were correctly named. The statistical analysis was carried out following the same method described above. In the simultaneous logistic regression model, the only significant predictor of correct responses was the lexical frequency of the target name (chi-square = 5.081, df = 1, p = .024). The raw contrast between biological categories and artifacts yielded a chi-square of 2.430 (df = 1, p = .120, ns); after
Table 1. Correct scores on lexical–semantic tasks carried out from one week to four months post onset. The scores marked with an asterisk should be considered pathological. The scores and the pathology thresholds take into account the influence of the demographic variables. Pathology threshold Picture naming Overall (n = 80) Biological stimuli (n = 30) Artifacts (n = 30) Naming on verbal definition Naming of famous faces Verbal comprehension (word-picture matching) Verbal semantic questionnaire Note: n.e. = Not examined.
63/80 21/30 25/30 33.3/38 53/78 71/80 336/360
(78.8%) (70%) (83.3%) (87.5%) (67.9%) (89%) (93%)
After one week
After four weeks
After four months
24/80 (30%)* 9/30 (30%)* 9/30 (30%)* n.e. n.e. n.e. n.e.
43/80 (53.8%)* 13/30 (43.3%)* 19/30 (63.3%)* n.e. n.e. 74/80 (92.5%) n.e.
57/80 19/30 21/30 31.5/38 44.7/78 77/80 359/360
(71.3%)* (63.3%)* (70%)* (82.9%)* (57.3%)* (96.3%) (99.7%)
Downloaded by [Northwestern University] at 09:33 14 February 2015
6
E. Baratelli et al.
statistical adjustment for the concomitant effect of lexical frequency, the category effect shrank further to a chi-square of 1.454 (df = 1, ns), which fell far short of significance. The availability of a complete array of psycholinguistic variables for the stimuli of our battery prompted us to use the same material for all the evaluations (see below), and this permitted a direct quantification of the recovery of the naming performance. This advantage is definitely greater than the potential drawbacks deriving from response learning or remembering, considering that no feedback was provided to the patient and that the examination was repeated after several weeks. RMR did not present anarthria and apraxia of speech or other phenomena typical of Broca’s or global aphasia. Phonological errors or neologisms were never observed in spontaneous speech. She had long naming latencies, and her errors were almost entirely of a lexical and/or semantic type (see Table 2). Semantic errors mainly consisted in the name of the superordinate category (e.g. for caterpillar and for ostrich: “an animal”; for onion: “a vegetable thing”); in other cases, RMR gave information about the use (e.g. for piano “it is used for playing”). Sometimes she provided correct, although possibly partial information about the stimulus (for camel: “these are animals living in Africa”; for saw: “it serves for cutting”). With other stimuli, she produced the correct pantomime related to the object’s use (this was observed for harp and guitar). At this point the patient was given also a name-comprehension task, based on the same 80 pictures used for the naming task. This test consisted in pointing to picture on verbal command, with a multiple-choice response display that included four foil pictures belonging to the same semantic category of the stimulus. On the whole battery, RMR was correct on 74/80 stimuli (92.5%), a score above the pathology threshold (71/80, i.e., 88.7%), although rather borderline. Artifacts were 27/30 correct (90.0%) and biological stimuli were 28/30 correct (93.3%), without category effects (chi-square = 0.219, df = 1, ns). Third examination (four months post onset) RMR was examined as an outpatient. On the naming battery, assessed with the same battery used in the first Table 2.
and second examinations, correct responses were 57/80 (71.3%), which was still pathological (overall pathology threshold is 63/80, i.e., 78.8%). Although use of the z-score transformation is questionable since the distribution of control scores was skewed, the z-score corresponding to RMR’s performance would have been – 4.26. Biological stimuli were 19/30 correct (63.3%) with a pathology threshold of 20/30, and artifacts were 21/30 correct (70.0%) with a pathology threshold of 24/30. Seven musical instruments and all 10 body parts gave rise to correct responses. The statistical analysis was carried out following the same method described above. Within the simultaneous logistic regression model, the only significant predictor of correct responses was the age of acquisition of the target name (chi-square = 9.699, df = 1, p = .002). The contrast between biological categories and artifacts was not significant (chi-square = 0.300, df = 1, ns). The improvement from the second to the third naming performance was mainly due to a reduction of semantic errors. On a test of naming after verbal definition (Novelli et al., 1986) RMR obtained 31.5/38, that is, a pathological performance (see Table 1). Our patient was also pathological on a test of famous faces naming, where she scored 44.7 with a pathology threshold of 53.00 (Bizzozero, Lucchelli, Saetti, & Spinnler, 2007): the deficit was due to poor name retrieval, and RMR correctly communicated the associative information necessary for an unequivocal identification of the stimulus photograph. On the name-comprehension task, the performance was almost flawless. RMR was correct on 77/80 stimuli (96.3%), cf. a pathology threshold of 71/80 (89%). Artifacts were 28/30 correct (93.3%) and biological stimuli were 100% correct. On this occasion, RMR was given a totally verbal semantic questionnaire and a picture reality decision test. The questionnaire was based on the same 80 Snodgrass and Vanderwart stimuli described above (Laiacona et al., 1993; Laiacona, Capitani, & Barbarotto, 1997): the examiner provided the name of the item and asked six questions that investigated superordinate information, subordinate associative information, and subordinate perceptual information, with two questions for each subtype. Overall there were 360 questions,
Qualitative differentiation of naming responses over the different sessions. Interval from onset
Correct responses Phonological errors
Words Non-words (phonemic paraphasias)
Semantic paraphasias Circumlocutions conveying appropriate semantic content Unrelated circumlocutions No responses Other errors (including visual errors)
One week
Four weeks
Four months
24 0 0 16 13 4 14 9
43 0 1 21 6 1 6 2
57 0 0 12 2 2 6 1
(30%) (20%) (16.2%) (5%) (17.5%) (11.3%)
(53.8%) (1.3%) (26.3%) (7.5%) (1.25%) (7.5%) (2.5%)
(71.3%) (15%) (2.5%) (2.5%) (7.5%) (1.3%)
Neurocase Table 3. Additional neuropsychological tests carried out four months post onset. The scores marked with an asterisk should be considered pathological. The scores and the pathology thresholds take into account the influence of the demographic variables.
Downloaded by [Northwestern University] at 09:33 14 February 2015
Test Picture reality decision Digit span Verbal learning (paired words) Corsi block tapping span Rey’s figure delayed recall Rey’s figure copy Stroop Color-Word interference (effect on errors) Stroop Color-Word interference (effect on time) Digit cancellation task
RMR’s score
Pathology threshold
137.2/160 4.75 7.5/22.5 4.25 14.75/36 31.75/36 1.5
133.7/160 3.5 6/22.5 3.50 9.46 28.87 4.24
27
36.92
31.1/50
23.9
and RMR was 359/360 correct (99.7%), with a pathology threshold of 331/360 (91.9%). At the same time, she was only 71.3% correct on picture naming. The picture reality decision test (Barbarotto, Laiacona, Macchi, & Capitani, 2002) was based on 80 pictures of real objects of the 8 categories investigated earlier, and 80 foils where 2 parts of different pictures were assembled (“chimeric pictures”): in this task she was 85.7% correct, that is, above the pathology threshold of 83.6%. On this occasion, the patient was given also supplementary tests to check other cognitive domains (see Table 3). She was normal on all tests, which included a memory battery (digit span, verbal learning of paired words, Corsi block tapping span, and Rey’s figure delayed recall), the Rey’s figure copy, and two tests of attentionexecutive abilities (Digit Cancellation and Stroop ColorWord interference).
Discussion Given the clear-cut and uncommon lesion site of RMR, the interest of the present report is twofold. The first question concerns the type of language disorder, and the second the status of the semantic representations, with particular reference to a possibly disproportionate impairment of biological categories. The language disturbance of RMR was of a fluent type, although, especially in the very early stages of encephalitis, long naming latencies caused a considerable slowing of name retrieval. Moreover, phonological processing was spared, and aphasia was characterized by a substantial lexical and/or semantic impairment. In our case, the damage to the insula involved both the anterior and posterior parts. Therefore, the observation of a fluent aphasia does not mean that the anterior insular cortex is necessarily spared.
7
As mentioned in the introduction, many cases of insular damage reported in the literature presented nonfluent aphasia, and this association was perhaps stronger for the lesions of the left anterior insula. The case reported here is at odds with this general rule, and is more consistent with the critical claims of Hillis et al. (2004). The recent literature, however, suggests a more detailed fractionation of the insular cortex as results from a finer grained anatomical study of this area. RMR presented a lesion of the whole insula on the antero-posterior axis, but the ventral sectors were perhaps more extensively damaged than the dorsal sectors. Nieuwenhuis (2012) has recently remarked that the functions of the insula might be analyzed not only on the basis of the so-called “anterior-posterior concept” (anterior agranular cortex versus posterior granular area) but also on the basis of a “concentric structure”, that is, a partition consisting of a ventrally situated agranular zone, separated from a dorsocaudal granular zone by an intermediate “dysgranular” zone. The latter account may provide the basis for a differentiation between the effects of ventral and dorsal lesions, and in our case the ventral insula was more severely damaged than the dorsal insula. However, no independent data suggest that dorso-caudal lesions are associated to nonfluent aphasia, and the relative sparing of the dorsal insula does not appear likely to explain the fluent aphasia of RMR. A second point to consider in the interpretation of RMR’s aphasia is that, in principle, the language symptoms presented by RMR could be attributed to the combination of the injured areas which, besides the insula, included the left parahippocampal region. In any case, as articulatory planning was substantially spared in our patient, we can safely conclude that the areas of the insular cortex damaged in RMR were not necessary for an effective articulatory planning. The naming impairment of RMR did not derive from a visual–perceptual deficit, because the patient was impaired also on naming-to-definition, and some nonlinguistic tests based on visual material were normal at a stage where picture naming was still pathological. The relevant question is: can we differentiate between the lexical and the semantic components of the naming impairment? The main mode for discriminating between a lexical and a semantic source of the naming deficit is by the performance on other tasks. Our data show that after four weeks verbal comprehension was not pathological. At the last assessment (four months post the encephalitis onset) verbal comprehension and a verbal semantic questionnaire were normal and almost flawless, but naming was still impaired. It is very unlikely that RMR had only a semantic impairment from the onset which, for the most part recovered but still remained mildly present after four months, and that her naming errors were only due to a subtle semantic impairment: although the questionnaire might be less sensitive than picture naming, at the last
Downloaded by [Northwestern University] at 09:33 14 February 2015
8
E. Baratelli et al.
examination naming was only 71.3% correct (with a pathology threshold of 78.8% or less), while the questionnaire was almost flawless (359/360 correct, i.e., 99.7%) – and controls are not at ceiling in this test (the pathology threshold is 92%). These data would suggest that, at least at one month from onset and following, the language impairment of RMR was mainly of a lexical nature. To be sure, however, it is possible that a semantic impairment was present in the early stages of her disease and that this was the first component that recovered in the recovery process. A qualitative analysis of the responses at the first naming sessions would be, at least partially, in line with this possibility, because with a couple of stimuli the errors of RMR were clearly cases of semantic confusion: presented with the picture of a rooster, she responded “…these are animals that produce milk,” and with a picture of a train she said “these can be seen on the motorway, they have glasses…” The received neuropsychological theory warns against considering semantic error as a warranty that the origin of these (semantic) errors is a dysfunction of the semantic system (Caramazza & Hillis, 1990). Once again, the only way to safely map error types on processing levels is by simultaneous analysis of multiple tasks. This was not possible with our patient at the time of the first examination due to the general impairment of her clinical condition. Therefore, the possibility cannot be discarded; however, in this event we would need to assume that recovery of semantics was relatively quick, while a lexical impairment was more persistent. Another interesting point concerns language recovery. Many cases in the literature of patients affected by insular stroke show a quick recovery, and a plausible hypothesis is that the early language impairment is mainly due to diaschisis, that is, the inactivation of distant classic language areas related to a damaged insula. The deficit presented by our patient seems not easily explained by mere diaschisis, as naming was still impaired after four months; as such, her deficit was more likely to directly depend on the actual neural damage that involved, besides the insula, also the anterior parahippocampal region. The second question concerns the effects of a parahippocampal lesion, in particular of the left entorhinal and perirhinal cortices. Some authors have posited that these areas should be critical for semantic representations, particularly for biological entities (Moss et al., 2005; Tyler et al., 2004). We cannot precisely quantify the level of semantic impairment during the first month. After this term, however, the good verbal comprehension and, later, a complete sparing of the verbal semantic probes are at odds with the presence of a substantial and lasting semantic impairment, while, at the same time, the output lexicon was still defective. On this basis, our data are not fully in line with the notion that the left parahippocampal region plays a crucial role in the storing of semantic
representations. Since the lesion of our patient was strictly unilateral, our data are mute about the possibility that a bilateral anterior parahippocampal lesion is necessary for determining a semantic deficit. However, it should be borne in mind that the functional MRI (fMRI) studies have suggested a semantic competence of the left perirhinal areas (see, e.g., Moss et al., 2005), and that the literature has reported a disproportionate impairment for biological categories in many patients with a unilateral lesion of the left temporal lobe (see Capitani et al. (2003) and Capitani et al. (2009)). A last relevant point concerns semantic category dissociations. RMR never showed a significant naming discrepancy between biological stimuli and artifact stimuli. In the second examination, the difference might have suggested a lower performance with biological categories (animals, fruit, and vegetables were 13/30 correct, 43.3%) than with artifact categories (tools, vehicles, and furniture were 19/30 correct, 63.3%); however, at statistical analysis this discrepancy did not reach significance. On the first and third assessments, the category effect was negligible. The trend toward a greater impairment of biological categories appeared at a stage of intermediate severity: this is not in line with some theoretical accounts that predict the most clear category effects at one of two extremes of clinical severity (Devlin, Gonnerman, Andersen, & Seidenberg, 1998; Moss, Tyler, DurrantPeatfield, & Bunn, 1998; for a general discussion see Capitani, Laiacona, & Caramazza, 2003). Before accepting the absence of a category dissociation, we should ascertain that this negative outcome did not depend on an insufficient sensitivity of our battery or of our statistical approach. Some evidence comes from a number of published cases, examined with the same battery, that presented a significant category dissociation. The data of these patients were analyzed with the same methodology, and their severity level was comparable with the overall naming severity of RMR at the different stages of her disease: for the naming test at the first examination, see case FI (Barbarotto et al., 1996); for the naming test at the second examination, see case AMA (Capitani et al., 2009); for the third naming examination, see case LF (Barbarotto et al., 1996). Similar evidence can be cited for word– picture matching and for the verbal questionnaire: for verbal comprehension at the second examination, see case AMA, also cited above, and for the verbal questionnaire, see case GIZ (Capitani et al., 2009). This excludes the possibility that the absence of a category dissociation simply derived from a lack of sensitivity. Summing up, the intended contribution of this report was to provide fresh, empirical evidence useful for confirming/disconfirming some accounts of two quite circumscribed facets of language and semantics processes. This case shows that damage to the insular cortex can be associated to fluent aphasia without phonological
Downloaded by [Northwestern University] at 09:33 14 February 2015
Neurocase impairment. While Dronkers (1996) assumed that the insular cortex was a crucial structure for the coordination of speech articulation, Hillis et al. (2004) challenged her position, and our empirical evidence is in line with the latter claim. This conclusion is not undermined by the fact that RMR also presented an entorhinal/perirhinal lesion. The presence of semantic impairment was possible in the early stages of the clinical course, but recovered quickly … More specifically, however, we were interested to check a very specific question: does the left entorhinal cortex play a special role in the semantic representation of living categories? In this study semantic category effects were never significant, and the trend toward a dissociation was less evident than that observed on the same battery in other patients of the literature affected by HSE or by PCA infarcts. This conclusion, also, does not appear to be undermined by a concomitant insular lesion in the case presented here.
Acknowledgments We are grateful to the colleagues of the Infectious Disease Unit of Milan University, San Paolo Hospital, for having referred RMR to us for neurological and neuropsychological examination. Rosemary Allpress revised the English language. No conflicts of interest are declared.
References Aitkin, M., Anderson, D., Francis, B., & Hinde, J. (1989). Statistical modelling in GLIM. Oxford: Oxford Science. Albanese, E., Capitani, E., Barbarotto, R., & Laiacona, M. (2000). Semantic category dissociations, familiarity and gender. Cortex, 36, 733–746. doi:10.1016/S0010-9452(08) 70549-5 Barbarotto, R., Capitani, E., & Laiacona, M. (1996). Naming deficit in herpes simplex encephalitis. Acta Neurologica Scandinavica, 93, 272–280. doi:10.1111/j.1600-0404.1996. tb00520.x Barbarotto, R., Capitani, E., & Laiacona, M. (2001). Living musical instruments and inanimate body parts? Neuropsychologia, 39, 406–414. doi:10.1016/S0028-3932(00)00128-7 Barbarotto, R., Laiacona, M., & Capitani, E. (2005). Objective versus estimated age of word acquisition: A study of 202 Italian children. Behavior Research Methods, 37, 644–650. doi:10.3758/BF03192735 Barbarotto, R., Laiacona, M., Macchi, V., & Capitani, E. (2002). Picture reality decision, semantic categories and gender. A new set of pictures with norms and an experimental study. Neuropsychologia, 40, 1637–1653. doi:10.1016/S0028-3932 (02)00029-5 Battig, W. F., & Montague, W. E. (1969). Category norms of verbal items in 56 categories: A replication and extension of the Connecticut category norms. Journal of Experimental Psychology Monographs, 80(part 2). doi:10.1037/h0027577 Bizzozero, I., Lucchelli, F., Saetti, M. C., & Spinnler, H. (2007). “Whose face is this?”: Italian norms of naming celebrities. Neurological Sciences, 28, 315–322. doi:10.1007/s10072007-0845-6
9
Bortolini, V., Tagliavini, C., & Zampolli, A. (1972). Lessico di frequenza della lingua italiana contemporanea. Milano: Garzanti. Capitani, E., & Laiacona, M. (2011). Facts and hypotheses relevant for contrasting animal and plant life semantics. A comment on Gainotti (2010). Cortex, 47, 259–264. doi:10.1016/j. cortex.2010.07.012 Capitani, E., Laiacona, M., Mahon, B., & Caramazza, A. (2003). What are the facts of semantic category specific deficits? A critical review of the clinical evidence. Cognitive Neuropsychology, 20, 213–261. doi:10.1080/ 02643290244000266 Capitani, E., Laiacona, M., Pagani, R., Capasso, R., Zampetti, P., & Miceli, G. (2009). Posterior cerebral artery infarcts and semantic category dissociations: A study of 28 patients. Brain, 132, 965–981. doi:10.1093/brain/awp013 Caramazza, A., & Hillis, A. E. (1990). Where do semantic errors come from? Cortex, 26, 95–122. doi:10.1016/S0010-9452 (13)80077-9 Devlin, J. T., Gonnerman, L. M., Andersen, E. S., & Seidenberg, M. S. (1998). Category-specific semantic deficits in focal and widespread brain damage: A computational account. Journal of Cognitive Neuroscience, 10, 77–94. doi:10.1162/ 089892998563798 Dronkers, N. (1996). A new brain region for co-ordinating speech articulation. Nature, 384, 159–161. doi:10.1038/ 384159°0 Eichenbaum, H. (2007). Comparative cognition, hippocampal function, and recollection. Comparative Cognition and Behavior Reviews, 2, 47–66. Furtak, S. C., Wei, S.-M., Agster, K. L., & Burwell, R. D. (2007). Functional neuroanatomy of the parahippocampal region in the rat: The perirhinal and postrhinal cortices. Hippocampus, 17, 709–722. doi:10.1002/hipo.20314 Gainotti, G. (2010). The influence of anatomical locus of lesion and of gender-related familiarity factors in category-specific semantic disorders for animals, fruits and vegetables: A review of single-case studies. Cortex, 46, 1072–1087. doi:10.1016/j.cortex.2010.04.002 Gitelman, D. R., Ashburner, J., Friston, K. J., Tyler, L. K., & Price, C. J. (2001). Voxel-based morphometry of Herpes Simplex encephalitis. Neuroimage, 13, 623–631. doi:10.1006/nimg.2000.0734 Gitelman, D. R., Nobre, A. C., Sonty, S., Parrish, T. B., & Mesulam, M. M. (2005). Language network specializations: An analysis with parallel task designs and functional magnetic resonance imaging. Neuroimage, 26, 975–985. doi:10.1016/j.neuroimage.2005.03.014 Hillis, A. E., Work, M., Barker, P. B., Jacobs, M. A., Breese, E. L., & Maurer, K. (2004). Re-examining the brain regions crucial for orchestrating speech articulation. Brain, 127, 1479–1487. doi:10.1093/brain/awh172 Ibañez, A., Gleichgerrcht, E., & Manes, F. (2010). Clinical effects of insular damage in humans. Brain Structure and Function, 214, 397–410. doi:10.1007/s00429-010-0256-y Kapur, N., Barker, S., Burrows, E. H., Ellison, D., Brice, J., Illis, L. S., … Loates, M. (1994). Herpes simplex encephalitis: Long term magnetic resonance imaging and neuropsychological profile. Journal of Neurology, Neurosurgery and Psychiatry, 57, 1334–1342. doi:10.1136/jnnp.57.11.1334 Kivisaari, S. L., Tyler, L. K., Monsch, A. U., & Taylor, K. I. (2012). Medial perirhinal cortex disambiguates confusable objects. Brain, 135, 3757–3769. doi:10.1093/brain/aws277 Laiacona, M., Barbarotto, R., Trivelli, C., & Capitani, E. (1993). Dissociazioni semantiche categoriali: Descrizione di una
Downloaded by [Northwestern University] at 09:33 14 February 2015
10
E. Baratelli et al.
batteria standardizzata e dati normativi. Archivio di Psicologia, Neurologia e Psichiatria, 54, 209–248. Laiacona, M., Capitani, E., & Barbarotto, R. (1997). Semantic category dissociations: A longitudinal study of two cases. Cortex, 33, 441–461. doi:10.1016/S0010-9452(08)70229-6 Lemieux, F., Lanthier, S., Chevrier, M.-C., Gioia, L., Rouleau, I., Cereda, C., & Nguyen, D. K. (2012). Insular ischemic stroke: Clinical presentation and outcome. Cerebrovascular Diseases Extra, 2, 80–87. doi:10.1159/000343177 Mai, J. K., Paxinos, G., & Voss, T. (2008). Atlas of the human brain (3rd ed.). San Diego, CA: Academic Press. McCullagh, P., & Nelder, J. A. (1983). Generalized linear models. London: Chapman and Hall. Moss, H. E., Rodd, J. M., Stamatakis, E. A., Bright, P., & Tyler, L. K. (2005). Anteromedial temporal cortex supports finegrained differentiation among objects. Cerebral Cortex, 15, 616–627. doi:10.1093/cercor/bhh163 Moss, H. E., Tyler, L. K., Durrant-Peatfield, M., & Bunn, E. M. (1998). “Two eyes of a see-through”: Impaired and intact semantic knowledge in a case of selective damage for living things. Neurocase, 4, 291–310. Nieuwenhuis, R. (2012). The insular cortex: A review. Progress in Brain Research, 195, 123–163. Novelli, G., Papagno, C., Capitani, E., Laiacona, M., Vallar, G., & Cappa, S.F. (1986). Tre test di produzione e ricerca lessicale [Three tests of lexical retrieval].
Archivio di Psicologia, Neurologia e Psichiatria, 47, 477–506. Price, C. J. (2000). The anatomy of language: Contributions from functional neuroimaging. Journal of Anatomy, 197, 335–359. doi:10.1046/j.1469-7580.2000.19730335.x Price, C. J., Devlin, J. T., Moore, C. J., Morton, C., & Laird, A. R. (2005). Meta-analyses of object naming: Effect of baseline. Human Brain Mapping, 25, 70–82. doi:10.1002/ hbm.20132 Shuren, J. (1993). Insula and aphasia. Journal of Neurology, 240, 216–218. doi:10.1007/BF00818707 Snoodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174–215. Tyler, L. K., Stamatakis, E. A., Bright, P., Acres, K., Abdallah, S., Rodd, J. M., & Moss, H. E. (2004). Processing objects at different levels of specificity. Journal of Cognitive Neuroscience, 16, 351–362. doi:10.1162/ 089892904322926692 Venneri, A., McGeown, W. J., Hietanen, H. M., Guerrini, C., Ellis, A. W., & Shanks, M. (2008). The anatomical bases of semantic retrieval deficits in early Alzheimer’s disease. Neuropsychologia, 46, 497–510. doi:10.1016/j. neuropsychologia.2007.08.026
Neurocase: The Neural Basis of Cognition Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/nncs20
Language disturbances associated to insular and entorhinal damage: study of a patient affected by herpetic encephalitis a
b
a
Elena Baratelli , Marcella Laiacona & Erminio Capitani a
Health Sciences Department, Neurology Unit, Milan University, S.Paolo Hospital, Milan, Italy b
S.Maugeri Foundation, IRCCS, Aphasia Unity, Neurology, Veruno Scientific Institute, Veruno (NO), Italy Published online: 04 Mar 2014.
Click for updates To cite this article: Elena Baratelli, Marcella Laiacona & Erminio Capitani (2014): Language disturbances associated to insular and entorhinal damage: study of a patient affected by herpetic encephalitis, Neurocase: The Neural Basis of Cognition, DOI: 10.1080/13554794.2014.892623 To link to this article: http://dx.doi.org/10.1080/13554794.2014.892623
PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions
Neurocase, 2014 http://dx.doi.org/10.1080/13554794.2014.892623
Language disturbances associated to insular and entorhinal damage: study of a patient affected by herpetic encephalitis Elena Baratellia, Marcella Laiaconab and Erminio Capitania* a
Health Sciences Department, Neurology Unit, Milan University, S.Paolo Hospital, Milan, Italy; bS.Maugeri Foundation, IRCCS, Aphasia Unity, Neurology, Veruno Scientific Institute, Veruno (NO), Italy
Downloaded by [Northwestern University] at 09:33 14 February 2015
(Received 11 July 2013; accepted 5 February 2014) The herpes simplex encephalitis (HSE) patient reported in this study presented a left hemisphere lesion limited to the left insula and to the left anterior parahippocampal region. The patient was followed longitudinally, focusing on the aphasia type, the language recovery, and the integrity of semantic representations. The language deficit was of fluent type, without phonological impairment, and showed a good but incomplete recovery after four months. A semantic impairment was possible at the onset, but recovered quickly and did not present a disproportionate impairment of living categories. Keywords: herpetic encephalitis; insula; parahippocampal region; aphasia; semantics
The cognitive deficits associated to herpes simplex encephalitis (HSE) are of special interest for neuropsychologists. HSE may cause a variety of disturbances, for instance, amnesic syndromes, semantic impairment, and visual agnosia, which are less commonly seen in stroke patients. These disturbances may co-occur in the same patient or present as rather selective disorders. For instance, Kapur et al. (1994) in a study focused on amnesia and its structural correlates, described 10 HSE patients and reported naming deficits in only a limited number of cases (4/10). In contrast, Barbarotto, Capitani, and Laiacona (1996) found that 7/8 cases (87.5%) in their HSE series were impaired on picture naming: the discrepancy with Kapur et al. is probably due to the extent of left hemisphere involvement in different samples. Also the findings regarding visual agnosia as an effect of HSE are not homogeneous: a deficit of visual object recognition not dependent on visual loss or language impairment, assessed with visual reality decision on chimeric pictures, has been found spared in some cases and defective in other cases, sometimes only for biological categories (for a comprehensive review of this aspect see Capitani, Laiacona, Mahon, & Caramazza, 2003). The heterogeneous cognitive profile of HSE patients is, at least in part, a function of the variable site of brain damage. To fully explain the heterogeneity of the HSE population, a first point to consider is that encephalitis may affect with varying severity left and/or right hemispheres, and that different areas can be damaged within each hemisphere. The realm of damaged structures, reviewed in detail by Kapur et al. (1994), includes the hippocampus, parahippocampal region, amygdala, *Corresponding author. Email: [email protected] © 2014 Taylor & Francis
temporal poles and lateral temporal gyri, fusiform gyrus, insular cortex, thalamus, medial and basal frontal cortex, occipital cortex, and possibly other areas. A more recent study of the lesion topography of five HSE patients (Gitelman, Ashburner, Friston, Tyler, & Price, 2001) reports that “the areas of most prominent gray matter abnormality were in the temporal lobes, including the anterior aspects of the lateral temporal cortex, the temporal poles, and the medial temporal cortex including the amygdala, hippocampus, entorhinal cortex, and insular cortex.” To uncover the role of each of the above cortical areas, we may also consider that HSE is not the sole cause of neurological damage of these structures, and also different diseases can affect the same areas damaged by HSE. The areas affected by posterior cerebral artery (PCA) infarcts partly overlap with those damaged by HSE, but PCA strokes never impinge upon the temporal poles and quite frequently affect the more posterior temporal cortex and occipital cortex. Hence, a comparative discussion of the effects of HSE and PCA strokes is quite informative in this context (Capitani et al., 2009). Another source of evidence relevant for the study of anatomo-functional correspondences is functional imaging, although Capitani et al. (2009) noted that lesion studies and functional imaging data have sometimes yielded inconsistent findings. Nevertheless, functional imaging can provide a sharp localization of cognitive tasks and should not be overlooked when trying to interpret the patient’s impairment, also because neurological lesions are seldom so selective as to affect just a single functional brain region. In the present study, which is mainly focused on language and semantic memory, we report clinical,
Downloaded by [Northwestern University] at 09:33 14 February 2015
2
E. Baratelli et al.
anatomical, and functional data of a HSE patient, case RMR, followed for four months after the disease onset. The interest of this case derives from the localization of her lesion, which was strictly limited to the left hemisphere and affected the insular cortex and anterior parahippocampal region, substantially sparing the fusiform gyrus, the most anterior temporal pole areas, posterior parahippocampal gyrus, and frontal and occipital structures. This lesion pattern potentially enables us to explore the role played by the entorhinal-perirhinal cortex in the representation of semantic knowledge. In addition, a clearly defined insular lesion, can help in verifying whether insular lesions are sufficient per se for impairing some aspects of language processing that are debated in the current literature. Before describing the patient, it is useful to briefly summarize the literature about the cognitive effects of damage to the insular cortex and parahippocampal region. The neurological and neuropsychological literature does not offer a large record of cases with circumscribed insular damage and sparing of the classical language areas. Lemieux et al. (2012) published a review of 23 insular cases (7 from their Institution and 16 from the literature): 14/23 patients suffered from left insular infarcts, and 10/23 presented language deficits. Although different types of aphasia are reported in the review, nonfluent speech output and phonemic disturbances seem predominant (eight cases); in the two cases with fluent aphasia (cases 1 and 4 of the series, see supplementary material of the paper by Lemieux et al. (2012)) the insular lesion affected the posterior long insular gyri (B1–B2); in another case (case 15), a lesion of the same structures was associated to nonfluent aphasia, but the brain damage also affected the frontal operculum. In general, language deficits were transient but in one case the deficit persisted for several months (Shuren, 1993). In a more general discussion of the effects of insular damage in humans, Ibañez, Gleichgerrcht, and Manes (2010) noted that the anterior insula has been considered by some authors (e.g. Price, 2000) a critical site for articulatory planning and part of the phonological network (e.g. Price, Devlin, Moore, Morton, & Laird, 2005), with specific reference to the left anterior insula. However, Ibañez and colleagues also argued that the insular cortex has wide connections with other language areas and the effects of its lesion might, at least in part, depend on the distant inactivation of functionally connected neural structures. Nonetheless, a lesion of the left insula has been most often associated to nonfluent aphasia with phonemic distortions, and the middle left insula was identified as a crucial structure for apraxia of speech (Dronkers, 1996). In contrast, in a detailed study of a large series of nonlacunar acute stroke patients, Hillis et al. (2004) found no association between the presence of apraxia of speech and lesions of the left insula. Moreover, a neuroimaging study
by Gitelman, Nobre, Sonty, Parrish, and Mesulam (2005) did not find activation of the anterior insula in phonological tasks. Concerning the parahippocampal region, the other structure of interest concerning our patient, this area can be subdivided into three anatomical sectors: the entorhinal, perirhinal, and postrhinal cortex; the latter section, well identifiable in rodents, coincides with the parahippocampal gyrus of the primate and human brain (Furtak, Wei, Agster, & Burwell, 2007). Evidence from animal experiments (see Eichenbaum, 2007, for a review) suggests that neurons in the perirhinal and lateral entorhinal cortex are involved in the representation of individual perceptual stimuli (“what” information) whereas the parahippocampal cortex and medial entorhinal cortex are specialized in processing spatial stimuli (“where” information). On this basis, some authors have suggested that neurons in the left perirhinal cortex play “a crucial role in object identification by integrating information from different sensory systems into more complex polymodal feature conjunctions” (Moss, Rodd, Stamatakis, Bright, & Tyler, 2005; Tyler et al., 2004). The latter authors also claim that objects belonging to the realm of biological categories (e.g. animals, fruit, and vegetables) should “place greater demands on the process of integration of complex conjunction of features and fine-grained discrimination supported by antero-medial temporal cortex.” Moss et al. (2005) found with a neuroimaging study in normal subjects that the left entorhinal cortex was more active when participants were asked to name animals and plant life stimuli than when the same subjects were requested to name tools and vehicles. Their conclusion was that anterior–medial structures of the temporal lobe are also “critical for semantic processing and not just for visual object recognition, as these areas would support the combination of many types of features into meaningful multimodal conceptual representations.” Other authors have investigated the relationship between the parahippocampal region and lexical–semantic abilities. Venneri et al. (2008), examining a group of early Alzheimer patients, observed a significant correlation between lexical competence and the integrity of the perirhinal and parahippocampal cortices. In a recent study, Kivisaari, Tyler, Monsch, and Taylor (2012) analyzed patients with mild cognitive impairment and Alzheimer’s disease and claimed that “the human medial perirhinal cortex is necessary for the disambiguation of perceptually and semantically confusable objects,” with an explicit reference to the living things category; their study presents an updated review of the role of these areas based on different experimental approaches. The above theoretical and experimental account implies that damage to the left parahippocampal region, that includes the lateral entorhinal and the perirhinal cortex, should be sufficient to cause a semantic deficit more severe for biological categories. In a stronger
Downloaded by [Northwestern University] at 09:33 14 February 2015
Neurocase sense, one might even claim that lesions of this area are necessary for there to be a semantic deficit of living category and that semantic knowledge about living entities is specifically and exclusively represented in the parahippocampal region (for a discussion on this point, see also Capitani and Laiacona (2011) and Gainotti (2010)). The HSE patient reported in the present study, case RMR, offers an opportunity to improve our understanding of the functions of the insular cortex and anterior parahippocampal region. Her lesion was substantially limited to these areas, and this is interesting because the role of the same structures has often been investigated in patients affected by larger lesions that also included insula or parahippocampus in the context of a wider brain damage. For instance, in the review by Lemieux et al. (2012) a condition for inclusion was that the lesion of areas adjacent to the insular cortex had to account for less than 50% of the total infarct volume. In turn, Moss et al. (2005) included semantic dementia and HSE cases whose lesions were centered on the left temporal pole, but the affected temporal areas, especially for HSE patients, were far more extended. A caveat is that in our case parahippocampal and insular lesions were associated; therefore, it might be problematic to univocally associate a specific cognitive deficit with a given structure. However, if the perirhinal cortex were really necessary for the semantic processing of living categories, its inactivation would invariably produce a deficit of this function, irrespective of the associated insular lesion. In the light of these considerations, the main questions for our study of RMR are: (1) what type of language disorders are present? (2) was there a semantic category dissociation? and (3) to what extent did the patient recover during the clinical follow-up?
Case description Clinical history Data included in this manuscript were obtained in compliance with regulations of the Milan University and San Paolo Hospital, Milan. RMR, a 63-year-old right-handed woman with 13 years of education, was admitted on November 29, 2012 to the medical ward of the San Paolo Hospital, Milan, on account of the presence of fever, headache, and language difficulties that had lasted for two days. Her previous medical and neurological history were unremarkable. The general medical examination did not reveal pathological data. On the neurological examination, she presented normal alertness and was well oriented in space and time, but was fatuous and slightly impulsive. Her spontaneous language was fluent, without grammatical or prosodic alterations, but her language disorder hampered a communicative exchange showing several anomias and occasional phonemic paraphasias. The same language alterations were also evident
3
when she was asked to name common objects. Verbal comprehension could not be assessed with formal testing; however, she understood without any problem the verbal orders given during the neurological examination. Cranial nerves were unimpaired. There was neither limb weakness nor sensory or coordination disturbances. Tendon reflexes were present, symmetrical and had normal strength. Response to plantar stimulation was in flexion. Standing and gait were normal. The blood tests showed slight neutrophil leukocytosis (WBC = 12,000, neutrophils 82.3%) and elevated C-Reactive Protein (15.2 mg/dl); chest X-rays were normal. The brain CT-scan showed a small hyperdense lesion in the left anterior–temporal region surrounded by edema. The background EEG activity with the patient awake consisted of low-voltage 8–9 Hz alpha-waves on posterior regions, with theta waves on the left side and well-organized activity on the right side. On the left frontal and frontal–temporal areas, there were pseudo-periodic slow abnormalities with theta waves, and interpolated slow spike discharges. Photic stimulation did not produce any abnormalities. Cerebrospinal fluid (CSF) was clear, with 500 WBC/mm3, prevailingly lympho-monocytes. Total protein content was slightly increased (59 mg/dl) and CSF sugar was normal. Polymerase chain reaction was positive for HSV-1 (553,123 cp/ml). Treatment was immediately started with e.v. acyclovir, 750 mg tid, and with levetiracetam, 500 mg bid. During the following week, fever disappeared and blood analysis reverted to normality. The patient presented productive, fluent aphasia: the onset of spontaneous speech was normal, and verbal output was neither agrammatic nor affected by prosodic impairment. There were no longer any phonological errors, and RMR was able to repeat nouns without problems. However, verbal communication was impoverished also because there were several anomias and a tendency to use stereotyped, context-dependent utterances. For instance, when requested to explain why she had been admitted to the Hospital, RMR replied: I did not feel well at home, but nothing really severe … but my … my relatives brought me here and the … they gave me the … – pointing to the intravenous treatment – but now I’m feeling well – repeatedly touching her body and moving her upper limbs. To overcome difficulties on noun retrieval, RMR used gestures as an aid, often successfully. Verbal comprehension, tested informally by asking her to execute simple orders, was quite good, although on one single occasion, on visual presentation, she seemed not to recognize a pen or its function. On the whole, RMR was not very concerned about her condition; she was aware of her deficits, but the emotional involvement was flattened. She never presented behavioral alterations, but complained of anosmia and of lack of appetite. RMR’s clinical status steadily improved, and she was dismissed from the hospital on December 20, 2012.
4
E. Baratelli et al.
Downloaded by [Northwestern University] at 09:33 14 February 2015
MRI imaging On November 30, 2012, RMR underwent a morphological magnetic resonance imaging (MRI) scan. The involved structures will be described with reference to the coronal sections of the atlas by Mai, Paxinos, and Voss (2008); these sections conform to the Talairach coordinates; however, in the atlas the sign of the y-coordinate is reversed (negative for the anterior sections, and positive for the posterior sections). In our description, we will use the Talairach notation, that is, with anterior positive values. On the right hemisphere, there were no lesions at all. On the left, the lesion affected the temporal, hippocampal and parahippocampal regions, and the insular cortex. The MRI study on the coronal plane showed that the temporal pole was spared from its anterior limit backwards to about y = 0. At this level, an area of abnormal signal appeared, that extended backwards until about y = −30 to −35. Posterior to that mark, all brain structures were normal. The anterior part of the lesion (from y = 0 backwards) interested the entorhinal and perirhinal cortices, amygdala nuclei, and piriform and insular cortices; at this site, the signal alteration clearly affected both the medial and lateral perirhinal cortex (for an anatomical reference, see Figure 1c, and the corresponding Figure 1 of the paper by Kivisaari et al., 2012). On more posterior coronal sections (at about y = −15), the entorhinal lesion was still evident and also the hippocampal structures were affected. At this level, where the perirhinal cortex merges into the parahippocampal gyrus, the perirhinal lesion extended backwards to involve the parahippocampal
1a-Coronal T2-weigthed MRI scan at the Talairach y-coordinate +10/+15.The image shows that the temporal poles were not involved.
1e-Axial FLAIR MRI scan at the upper pontine level.The image shows involvement left entorhinal shows impaired ofinsular cortex onand the left. perirhinal cortex and amygdalar nuclei
Figure 1.
1b-Coronal T2-weigthed MRI scan at the Talairach y-coordinate 0 /-5. The image shows involvement of left insula (particularly of the ventral part), entorhinal and perirhinal cortex, amygdalar nuclei and piriform cortex
1f-Axial FLAIR MRI scan at the lower mesencephalon level. The image shows involvement of left entorhinal and perirhinal cortex and amygdalar nuclei
MRI images representing RMR’s brain lesion.
gyrus. The signal alteration affecting the parahippocampal region was no longer detectable posterior to y = −35. The insular cortex was clearly affected: on the coronal sections its damage was more severe in its inferior (caudal) half; on the axial sections, the signal alteration involved the whole insula. The axial MRI images confirmed that the left temporal pole was spared, and that the lesion selectively interested the anterior left parahippocampal structures and the whole antero-posterior extent of the left insula.
Neuropsychological assessment First examination (one week from onset) On December 5, 2012, RMR underwent (one week from onset) a picture-naming battery based on 80 stimuli from the Snoodgrass and Vanderwart (1980) set. This battery includes 10 stimuli each from 8 categories: 3 biological categories (animals, fruit, vegetables), 3 artifact categories (tools, pieces of furniture, vehicles), musical instruments and body parts. For each stimulus a number of control indices are available, among which lexical frequency, age of acquisition of the corresponding word, stimulus familiarity, and visual complexity. During the naming session, the patient was never given feedback on the correctness of her responses. A thorough description of this test and of the statistical methodology used in its analysis can be found in Capitani et al. (2009). RMR was cooperative, but performed the task very slowly and with considerable fatigue; therefore, besides
1c-Coronal T2-weigthed MRI scan at the Talairach y-coordinate-10 /-15. The image shows involvement of left insula (particularly of the ventral part), hippocampus, entorhinal cortex and parahippocampus gyrus.
1g-Axial FLAIR MRI scan at the upper mesencephalon level. The image shows involvement of left entorhinal and perirhinal cortex and amygdalar nuclei
1d-Coronal T2-weigthed MRI scan at the Talairach y-coordinate-35/-40. At this level any structural alterations are no longer evident.
1h-Axial FLAIR MRI scan at the diencephalic level. The image shows involvement of left insular cortex
Downloaded by [Northwestern University] at 09:33 14 February 2015
Neurocase the picture-naming task, she underwent only a brief, informal evaluation of cognitive status. On the whole, she did not present problems with visual perception, including the recognition of known people, nor a deficit of gestural abilities. Memory status could not be reliably investigated due to fatigue and attention impairment. On the formal naming evaluation, she was correct on 24/80 stimuli (30%) (see Table 1). The mean of a group of 60 age- and education-matched controls (Laiacona, Barbarotto, Trivelli, & Capitani, 1993) was 75.40/80 (sd = 4.32), that is, 94.2%. Artifacts and biological stimuli yielded the same number of correct responses (9/30), musical instruments were never given the correct name, and 6/10 body parts stimuli were correct. The statistical analysis of the factors that influenced the naming performance was carried out by means of a logistic regression analysis (McCullagh & Nelder, 1983). Because the main interest of this analysis was to assess whether RMR presented a semantic category effect, as customary in our practice we considered only the 60 stimuli pertaining to the living and nonliving categories, that is, animals, fruit, and vegetables for biological categories, and tools, vehicles and pieces of furniture for artifacts (for a discussion of this point see Barbarotto, Capitani, & Laiacona, 2001). The binary response variable was the naming success observed for each stimulus (coded 1 if the response was correct, and 0 if the response was wrong). The model variables were: (1) the category of the stimulus (living versus nonliving), (2) lexical frequency of each stimulus word in the Italian language, after logarithmic transformation, according to Bortolini, Tagliavini, and Zampolli (1972); (3) age of acquisition of the stimulus word (Barbarotto, Laiacona, & Capitani, 2005), which was directly estimated from the performance of 202 Italian children; (4) familiarity with the stimulus: this index, that yields separate ratings for males and females, was designed to capture how frequently we think or speak of the stimulus, how often we see it represented in the media, and how frequently we are confronted with real exemplars (Mean Familiarity Index, see Albanese, Capitani, Barbarotto, & Laiacona, 2000); (5) prototypicality of the
5
stimulus according to the study of Battig and Montague (1969); and (6) the visual complexity of the image (Snoodgrass & Vanderwart, 1980). We first analyzed whether each model variable, considered alone, did in fact significantly predict the naming success with each stimulus; thereafter we included all the significant predictors in the same model, in order to evaluate their contribution (after removing the overlap with the effect of the other terms of the model). The analysis of each effect yields a chi-square distributed point. The computations were carried out by means of the GLIM Programme (Aitkin, Anderson, Francis, & Hinde, 1989). The statistical analysis did not show a discrepancy between biological categories and artifacts (chi-square = 0.000, df = 1, ns). In the simultaneous logistic regression model, significant predictors of the correct responses were stimulus familiarity (chi-square = 7.755, df = 1, p = .005) and age of acquisition of the target name (chi-square = 4.885, df = 1, p = .027). Visual complexity, lexical frequency, and prototypicality were not significant. The errors of RMR were classified as follows (see Table 2): phonological errors (words and non-words), semantic paraphasias, circumlocutions, non-responses, and other errors (including visual errors).
Second examination (four weeks post onset) Four weeks from the encephalitis onset, RMR was examined as an outpatient. On picture naming, which was assessed using the same battery used in the first examination, correct responses were 43/80 (53.3%): biological stimuli were 13/30 correct (43.3%), and artifacts 19/30 correct (63.3%). One musical instrument and all the 10 body parts were correctly named. The statistical analysis was carried out following the same method described above. In the simultaneous logistic regression model, the only significant predictor of correct responses was the lexical frequency of the target name (chi-square = 5.081, df = 1, p = .024). The raw contrast between biological categories and artifacts yielded a chi-square of 2.430 (df = 1, p = .120, ns); after
Table 1. Correct scores on lexical–semantic tasks carried out from one week to four months post onset. The scores marked with an asterisk should be considered pathological. The scores and the pathology thresholds take into account the influence of the demographic variables. Pathology threshold Picture naming Overall (n = 80) Biological stimuli (n = 30) Artifacts (n = 30) Naming on verbal definition Naming of famous faces Verbal comprehension (word-picture matching) Verbal semantic questionnaire Note: n.e. = Not examined.
63/80 21/30 25/30 33.3/38 53/78 71/80 336/360
(78.8%) (70%) (83.3%) (87.5%) (67.9%) (89%) (93%)
After one week
After four weeks
After four months
24/80 (30%)* 9/30 (30%)* 9/30 (30%)* n.e. n.e. n.e. n.e.
43/80 (53.8%)* 13/30 (43.3%)* 19/30 (63.3%)* n.e. n.e. 74/80 (92.5%) n.e.
57/80 19/30 21/30 31.5/38 44.7/78 77/80 359/360
(71.3%)* (63.3%)* (70%)* (82.9%)* (57.3%)* (96.3%) (99.7%)
Downloaded by [Northwestern University] at 09:33 14 February 2015
6
E. Baratelli et al.
statistical adjustment for the concomitant effect of lexical frequency, the category effect shrank further to a chi-square of 1.454 (df = 1, ns), which fell far short of significance. The availability of a complete array of psycholinguistic variables for the stimuli of our battery prompted us to use the same material for all the evaluations (see below), and this permitted a direct quantification of the recovery of the naming performance. This advantage is definitely greater than the potential drawbacks deriving from response learning or remembering, considering that no feedback was provided to the patient and that the examination was repeated after several weeks. RMR did not present anarthria and apraxia of speech or other phenomena typical of Broca’s or global aphasia. Phonological errors or neologisms were never observed in spontaneous speech. She had long naming latencies, and her errors were almost entirely of a lexical and/or semantic type (see Table 2). Semantic errors mainly consisted in the name of the superordinate category (e.g. for caterpillar and for ostrich: “an animal”; for onion: “a vegetable thing”); in other cases, RMR gave information about the use (e.g. for piano “it is used for playing”). Sometimes she provided correct, although possibly partial information about the stimulus (for camel: “these are animals living in Africa”; for saw: “it serves for cutting”). With other stimuli, she produced the correct pantomime related to the object’s use (this was observed for harp and guitar). At this point the patient was given also a name-comprehension task, based on the same 80 pictures used for the naming task. This test consisted in pointing to picture on verbal command, with a multiple-choice response display that included four foil pictures belonging to the same semantic category of the stimulus. On the whole battery, RMR was correct on 74/80 stimuli (92.5%), a score above the pathology threshold (71/80, i.e., 88.7%), although rather borderline. Artifacts were 27/30 correct (90.0%) and biological stimuli were 28/30 correct (93.3%), without category effects (chi-square = 0.219, df = 1, ns). Third examination (four months post onset) RMR was examined as an outpatient. On the naming battery, assessed with the same battery used in the first Table 2.
and second examinations, correct responses were 57/80 (71.3%), which was still pathological (overall pathology threshold is 63/80, i.e., 78.8%). Although use of the z-score transformation is questionable since the distribution of control scores was skewed, the z-score corresponding to RMR’s performance would have been – 4.26. Biological stimuli were 19/30 correct (63.3%) with a pathology threshold of 20/30, and artifacts were 21/30 correct (70.0%) with a pathology threshold of 24/30. Seven musical instruments and all 10 body parts gave rise to correct responses. The statistical analysis was carried out following the same method described above. Within the simultaneous logistic regression model, the only significant predictor of correct responses was the age of acquisition of the target name (chi-square = 9.699, df = 1, p = .002). The contrast between biological categories and artifacts was not significant (chi-square = 0.300, df = 1, ns). The improvement from the second to the third naming performance was mainly due to a reduction of semantic errors. On a test of naming after verbal definition (Novelli et al., 1986) RMR obtained 31.5/38, that is, a pathological performance (see Table 1). Our patient was also pathological on a test of famous faces naming, where she scored 44.7 with a pathology threshold of 53.00 (Bizzozero, Lucchelli, Saetti, & Spinnler, 2007): the deficit was due to poor name retrieval, and RMR correctly communicated the associative information necessary for an unequivocal identification of the stimulus photograph. On the name-comprehension task, the performance was almost flawless. RMR was correct on 77/80 stimuli (96.3%), cf. a pathology threshold of 71/80 (89%). Artifacts were 28/30 correct (93.3%) and biological stimuli were 100% correct. On this occasion, RMR was given a totally verbal semantic questionnaire and a picture reality decision test. The questionnaire was based on the same 80 Snodgrass and Vanderwart stimuli described above (Laiacona et al., 1993; Laiacona, Capitani, & Barbarotto, 1997): the examiner provided the name of the item and asked six questions that investigated superordinate information, subordinate associative information, and subordinate perceptual information, with two questions for each subtype. Overall there were 360 questions,
Qualitative differentiation of naming responses over the different sessions. Interval from onset
Correct responses Phonological errors
Words Non-words (phonemic paraphasias)
Semantic paraphasias Circumlocutions conveying appropriate semantic content Unrelated circumlocutions No responses Other errors (including visual errors)
One week
Four weeks
Four months
24 0 0 16 13 4 14 9
43 0 1 21 6 1 6 2
57 0 0 12 2 2 6 1
(30%) (20%) (16.2%) (5%) (17.5%) (11.3%)
(53.8%) (1.3%) (26.3%) (7.5%) (1.25%) (7.5%) (2.5%)
(71.3%) (15%) (2.5%) (2.5%) (7.5%) (1.3%)
Neurocase Table 3. Additional neuropsychological tests carried out four months post onset. The scores marked with an asterisk should be considered pathological. The scores and the pathology thresholds take into account the influence of the demographic variables.
Downloaded by [Northwestern University] at 09:33 14 February 2015
Test Picture reality decision Digit span Verbal learning (paired words) Corsi block tapping span Rey’s figure delayed recall Rey’s figure copy Stroop Color-Word interference (effect on errors) Stroop Color-Word interference (effect on time) Digit cancellation task
RMR’s score
Pathology threshold
137.2/160 4.75 7.5/22.5 4.25 14.75/36 31.75/36 1.5
133.7/160 3.5 6/22.5 3.50 9.46 28.87 4.24
27
36.92
31.1/50
23.9
and RMR was 359/360 correct (99.7%), with a pathology threshold of 331/360 (91.9%). At the same time, she was only 71.3% correct on picture naming. The picture reality decision test (Barbarotto, Laiacona, Macchi, & Capitani, 2002) was based on 80 pictures of real objects of the 8 categories investigated earlier, and 80 foils where 2 parts of different pictures were assembled (“chimeric pictures”): in this task she was 85.7% correct, that is, above the pathology threshold of 83.6%. On this occasion, the patient was given also supplementary tests to check other cognitive domains (see Table 3). She was normal on all tests, which included a memory battery (digit span, verbal learning of paired words, Corsi block tapping span, and Rey’s figure delayed recall), the Rey’s figure copy, and two tests of attentionexecutive abilities (Digit Cancellation and Stroop ColorWord interference).
Discussion Given the clear-cut and uncommon lesion site of RMR, the interest of the present report is twofold. The first question concerns the type of language disorder, and the second the status of the semantic representations, with particular reference to a possibly disproportionate impairment of biological categories. The language disturbance of RMR was of a fluent type, although, especially in the very early stages of encephalitis, long naming latencies caused a considerable slowing of name retrieval. Moreover, phonological processing was spared, and aphasia was characterized by a substantial lexical and/or semantic impairment. In our case, the damage to the insula involved both the anterior and posterior parts. Therefore, the observation of a fluent aphasia does not mean that the anterior insular cortex is necessarily spared.
7
As mentioned in the introduction, many cases of insular damage reported in the literature presented nonfluent aphasia, and this association was perhaps stronger for the lesions of the left anterior insula. The case reported here is at odds with this general rule, and is more consistent with the critical claims of Hillis et al. (2004). The recent literature, however, suggests a more detailed fractionation of the insular cortex as results from a finer grained anatomical study of this area. RMR presented a lesion of the whole insula on the antero-posterior axis, but the ventral sectors were perhaps more extensively damaged than the dorsal sectors. Nieuwenhuis (2012) has recently remarked that the functions of the insula might be analyzed not only on the basis of the so-called “anterior-posterior concept” (anterior agranular cortex versus posterior granular area) but also on the basis of a “concentric structure”, that is, a partition consisting of a ventrally situated agranular zone, separated from a dorsocaudal granular zone by an intermediate “dysgranular” zone. The latter account may provide the basis for a differentiation between the effects of ventral and dorsal lesions, and in our case the ventral insula was more severely damaged than the dorsal insula. However, no independent data suggest that dorso-caudal lesions are associated to nonfluent aphasia, and the relative sparing of the dorsal insula does not appear likely to explain the fluent aphasia of RMR. A second point to consider in the interpretation of RMR’s aphasia is that, in principle, the language symptoms presented by RMR could be attributed to the combination of the injured areas which, besides the insula, included the left parahippocampal region. In any case, as articulatory planning was substantially spared in our patient, we can safely conclude that the areas of the insular cortex damaged in RMR were not necessary for an effective articulatory planning. The naming impairment of RMR did not derive from a visual–perceptual deficit, because the patient was impaired also on naming-to-definition, and some nonlinguistic tests based on visual material were normal at a stage where picture naming was still pathological. The relevant question is: can we differentiate between the lexical and the semantic components of the naming impairment? The main mode for discriminating between a lexical and a semantic source of the naming deficit is by the performance on other tasks. Our data show that after four weeks verbal comprehension was not pathological. At the last assessment (four months post the encephalitis onset) verbal comprehension and a verbal semantic questionnaire were normal and almost flawless, but naming was still impaired. It is very unlikely that RMR had only a semantic impairment from the onset which, for the most part recovered but still remained mildly present after four months, and that her naming errors were only due to a subtle semantic impairment: although the questionnaire might be less sensitive than picture naming, at the last
Downloaded by [Northwestern University] at 09:33 14 February 2015
8
E. Baratelli et al.
examination naming was only 71.3% correct (with a pathology threshold of 78.8% or less), while the questionnaire was almost flawless (359/360 correct, i.e., 99.7%) – and controls are not at ceiling in this test (the pathology threshold is 92%). These data would suggest that, at least at one month from onset and following, the language impairment of RMR was mainly of a lexical nature. To be sure, however, it is possible that a semantic impairment was present in the early stages of her disease and that this was the first component that recovered in the recovery process. A qualitative analysis of the responses at the first naming sessions would be, at least partially, in line with this possibility, because with a couple of stimuli the errors of RMR were clearly cases of semantic confusion: presented with the picture of a rooster, she responded “…these are animals that produce milk,” and with a picture of a train she said “these can be seen on the motorway, they have glasses…” The received neuropsychological theory warns against considering semantic error as a warranty that the origin of these (semantic) errors is a dysfunction of the semantic system (Caramazza & Hillis, 1990). Once again, the only way to safely map error types on processing levels is by simultaneous analysis of multiple tasks. This was not possible with our patient at the time of the first examination due to the general impairment of her clinical condition. Therefore, the possibility cannot be discarded; however, in this event we would need to assume that recovery of semantics was relatively quick, while a lexical impairment was more persistent. Another interesting point concerns language recovery. Many cases in the literature of patients affected by insular stroke show a quick recovery, and a plausible hypothesis is that the early language impairment is mainly due to diaschisis, that is, the inactivation of distant classic language areas related to a damaged insula. The deficit presented by our patient seems not easily explained by mere diaschisis, as naming was still impaired after four months; as such, her deficit was more likely to directly depend on the actual neural damage that involved, besides the insula, also the anterior parahippocampal region. The second question concerns the effects of a parahippocampal lesion, in particular of the left entorhinal and perirhinal cortices. Some authors have posited that these areas should be critical for semantic representations, particularly for biological entities (Moss et al., 2005; Tyler et al., 2004). We cannot precisely quantify the level of semantic impairment during the first month. After this term, however, the good verbal comprehension and, later, a complete sparing of the verbal semantic probes are at odds with the presence of a substantial and lasting semantic impairment, while, at the same time, the output lexicon was still defective. On this basis, our data are not fully in line with the notion that the left parahippocampal region plays a crucial role in the storing of semantic
representations. Since the lesion of our patient was strictly unilateral, our data are mute about the possibility that a bilateral anterior parahippocampal lesion is necessary for determining a semantic deficit. However, it should be borne in mind that the functional MRI (fMRI) studies have suggested a semantic competence of the left perirhinal areas (see, e.g., Moss et al., 2005), and that the literature has reported a disproportionate impairment for biological categories in many patients with a unilateral lesion of the left temporal lobe (see Capitani et al. (2003) and Capitani et al. (2009)). A last relevant point concerns semantic category dissociations. RMR never showed a significant naming discrepancy between biological stimuli and artifact stimuli. In the second examination, the difference might have suggested a lower performance with biological categories (animals, fruit, and vegetables were 13/30 correct, 43.3%) than with artifact categories (tools, vehicles, and furniture were 19/30 correct, 63.3%); however, at statistical analysis this discrepancy did not reach significance. On the first and third assessments, the category effect was negligible. The trend toward a greater impairment of biological categories appeared at a stage of intermediate severity: this is not in line with some theoretical accounts that predict the most clear category effects at one of two extremes of clinical severity (Devlin, Gonnerman, Andersen, & Seidenberg, 1998; Moss, Tyler, DurrantPeatfield, & Bunn, 1998; for a general discussion see Capitani, Laiacona, & Caramazza, 2003). Before accepting the absence of a category dissociation, we should ascertain that this negative outcome did not depend on an insufficient sensitivity of our battery or of our statistical approach. Some evidence comes from a number of published cases, examined with the same battery, that presented a significant category dissociation. The data of these patients were analyzed with the same methodology, and their severity level was comparable with the overall naming severity of RMR at the different stages of her disease: for the naming test at the first examination, see case FI (Barbarotto et al., 1996); for the naming test at the second examination, see case AMA (Capitani et al., 2009); for the third naming examination, see case LF (Barbarotto et al., 1996). Similar evidence can be cited for word– picture matching and for the verbal questionnaire: for verbal comprehension at the second examination, see case AMA, also cited above, and for the verbal questionnaire, see case GIZ (Capitani et al., 2009). This excludes the possibility that the absence of a category dissociation simply derived from a lack of sensitivity. Summing up, the intended contribution of this report was to provide fresh, empirical evidence useful for confirming/disconfirming some accounts of two quite circumscribed facets of language and semantics processes. This case shows that damage to the insular cortex can be associated to fluent aphasia without phonological
Downloaded by [Northwestern University] at 09:33 14 February 2015
Neurocase impairment. While Dronkers (1996) assumed that the insular cortex was a crucial structure for the coordination of speech articulation, Hillis et al. (2004) challenged her position, and our empirical evidence is in line with the latter claim. This conclusion is not undermined by the fact that RMR also presented an entorhinal/perirhinal lesion. The presence of semantic impairment was possible in the early stages of the clinical course, but recovered quickly … More specifically, however, we were interested to check a very specific question: does the left entorhinal cortex play a special role in the semantic representation of living categories? In this study semantic category effects were never significant, and the trend toward a dissociation was less evident than that observed on the same battery in other patients of the literature affected by HSE or by PCA infarcts. This conclusion, also, does not appear to be undermined by a concomitant insular lesion in the case presented here.
Acknowledgments We are grateful to the colleagues of the Infectious Disease Unit of Milan University, San Paolo Hospital, for having referred RMR to us for neurological and neuropsychological examination. Rosemary Allpress revised the English language. No conflicts of interest are declared.
References Aitkin, M., Anderson, D., Francis, B., & Hinde, J. (1989). Statistical modelling in GLIM. Oxford: Oxford Science. Albanese, E., Capitani, E., Barbarotto, R., & Laiacona, M. (2000). Semantic category dissociations, familiarity and gender. Cortex, 36, 733–746. doi:10.1016/S0010-9452(08) 70549-5 Barbarotto, R., Capitani, E., & Laiacona, M. (1996). Naming deficit in herpes simplex encephalitis. Acta Neurologica Scandinavica, 93, 272–280. doi:10.1111/j.1600-0404.1996. tb00520.x Barbarotto, R., Capitani, E., & Laiacona, M. (2001). Living musical instruments and inanimate body parts? Neuropsychologia, 39, 406–414. doi:10.1016/S0028-3932(00)00128-7 Barbarotto, R., Laiacona, M., & Capitani, E. (2005). Objective versus estimated age of word acquisition: A study of 202 Italian children. Behavior Research Methods, 37, 644–650. doi:10.3758/BF03192735 Barbarotto, R., Laiacona, M., Macchi, V., & Capitani, E. (2002). Picture reality decision, semantic categories and gender. A new set of pictures with norms and an experimental study. Neuropsychologia, 40, 1637–1653. doi:10.1016/S0028-3932 (02)00029-5 Battig, W. F., & Montague, W. E. (1969). Category norms of verbal items in 56 categories: A replication and extension of the Connecticut category norms. Journal of Experimental Psychology Monographs, 80(part 2). doi:10.1037/h0027577 Bizzozero, I., Lucchelli, F., Saetti, M. C., & Spinnler, H. (2007). “Whose face is this?”: Italian norms of naming celebrities. Neurological Sciences, 28, 315–322. doi:10.1007/s10072007-0845-6
9
Bortolini, V., Tagliavini, C., & Zampolli, A. (1972). Lessico di frequenza della lingua italiana contemporanea. Milano: Garzanti. Capitani, E., & Laiacona, M. (2011). Facts and hypotheses relevant for contrasting animal and plant life semantics. A comment on Gainotti (2010). Cortex, 47, 259–264. doi:10.1016/j. cortex.2010.07.012 Capitani, E., Laiacona, M., Mahon, B., & Caramazza, A. (2003). What are the facts of semantic category specific deficits? A critical review of the clinical evidence. Cognitive Neuropsychology, 20, 213–261. doi:10.1080/ 02643290244000266 Capitani, E., Laiacona, M., Pagani, R., Capasso, R., Zampetti, P., & Miceli, G. (2009). Posterior cerebral artery infarcts and semantic category dissociations: A study of 28 patients. Brain, 132, 965–981. doi:10.1093/brain/awp013 Caramazza, A., & Hillis, A. E. (1990). Where do semantic errors come from? Cortex, 26, 95–122. doi:10.1016/S0010-9452 (13)80077-9 Devlin, J. T., Gonnerman, L. M., Andersen, E. S., & Seidenberg, M. S. (1998). Category-specific semantic deficits in focal and widespread brain damage: A computational account. Journal of Cognitive Neuroscience, 10, 77–94. doi:10.1162/ 089892998563798 Dronkers, N. (1996). A new brain region for co-ordinating speech articulation. Nature, 384, 159–161. doi:10.1038/ 384159°0 Eichenbaum, H. (2007). Comparative cognition, hippocampal function, and recollection. Comparative Cognition and Behavior Reviews, 2, 47–66. Furtak, S. C., Wei, S.-M., Agster, K. L., & Burwell, R. D. (2007). Functional neuroanatomy of the parahippocampal region in the rat: The perirhinal and postrhinal cortices. Hippocampus, 17, 709–722. doi:10.1002/hipo.20314 Gainotti, G. (2010). The influence of anatomical locus of lesion and of gender-related familiarity factors in category-specific semantic disorders for animals, fruits and vegetables: A review of single-case studies. Cortex, 46, 1072–1087. doi:10.1016/j.cortex.2010.04.002 Gitelman, D. R., Ashburner, J., Friston, K. J., Tyler, L. K., & Price, C. J. (2001). Voxel-based morphometry of Herpes Simplex encephalitis. Neuroimage, 13, 623–631. doi:10.1006/nimg.2000.0734 Gitelman, D. R., Nobre, A. C., Sonty, S., Parrish, T. B., & Mesulam, M. M. (2005). Language network specializations: An analysis with parallel task designs and functional magnetic resonance imaging. Neuroimage, 26, 975–985. doi:10.1016/j.neuroimage.2005.03.014 Hillis, A. E., Work, M., Barker, P. B., Jacobs, M. A., Breese, E. L., & Maurer, K. (2004). Re-examining the brain regions crucial for orchestrating speech articulation. Brain, 127, 1479–1487. doi:10.1093/brain/awh172 Ibañez, A., Gleichgerrcht, E., & Manes, F. (2010). Clinical effects of insular damage in humans. Brain Structure and Function, 214, 397–410. doi:10.1007/s00429-010-0256-y Kapur, N., Barker, S., Burrows, E. H., Ellison, D., Brice, J., Illis, L. S., … Loates, M. (1994). Herpes simplex encephalitis: Long term magnetic resonance imaging and neuropsychological profile. Journal of Neurology, Neurosurgery and Psychiatry, 57, 1334–1342. doi:10.1136/jnnp.57.11.1334 Kivisaari, S. L., Tyler, L. K., Monsch, A. U., & Taylor, K. I. (2012). Medial perirhinal cortex disambiguates confusable objects. Brain, 135, 3757–3769. doi:10.1093/brain/aws277 Laiacona, M., Barbarotto, R., Trivelli, C., & Capitani, E. (1993). Dissociazioni semantiche categoriali: Descrizione di una
Downloaded by [Northwestern University] at 09:33 14 February 2015
10
E. Baratelli et al.
batteria standardizzata e dati normativi. Archivio di Psicologia, Neurologia e Psichiatria, 54, 209–248. Laiacona, M., Capitani, E., & Barbarotto, R. (1997). Semantic category dissociations: A longitudinal study of two cases. Cortex, 33, 441–461. doi:10.1016/S0010-9452(08)70229-6 Lemieux, F., Lanthier, S., Chevrier, M.-C., Gioia, L., Rouleau, I., Cereda, C., & Nguyen, D. K. (2012). Insular ischemic stroke: Clinical presentation and outcome. Cerebrovascular Diseases Extra, 2, 80–87. doi:10.1159/000343177 Mai, J. K., Paxinos, G., & Voss, T. (2008). Atlas of the human brain (3rd ed.). San Diego, CA: Academic Press. McCullagh, P., & Nelder, J. A. (1983). Generalized linear models. London: Chapman and Hall. Moss, H. E., Rodd, J. M., Stamatakis, E. A., Bright, P., & Tyler, L. K. (2005). Anteromedial temporal cortex supports finegrained differentiation among objects. Cerebral Cortex, 15, 616–627. doi:10.1093/cercor/bhh163 Moss, H. E., Tyler, L. K., Durrant-Peatfield, M., & Bunn, E. M. (1998). “Two eyes of a see-through”: Impaired and intact semantic knowledge in a case of selective damage for living things. Neurocase, 4, 291–310. Nieuwenhuis, R. (2012). The insular cortex: A review. Progress in Brain Research, 195, 123–163. Novelli, G., Papagno, C., Capitani, E., Laiacona, M., Vallar, G., & Cappa, S.F. (1986). Tre test di produzione e ricerca lessicale [Three tests of lexical retrieval].
Archivio di Psicologia, Neurologia e Psichiatria, 47, 477–506. Price, C. J. (2000). The anatomy of language: Contributions from functional neuroimaging. Journal of Anatomy, 197, 335–359. doi:10.1046/j.1469-7580.2000.19730335.x Price, C. J., Devlin, J. T., Moore, C. J., Morton, C., & Laird, A. R. (2005). Meta-analyses of object naming: Effect of baseline. Human Brain Mapping, 25, 70–82. doi:10.1002/ hbm.20132 Shuren, J. (1993). Insula and aphasia. Journal of Neurology, 240, 216–218. doi:10.1007/BF00818707 Snoodgrass, J. G., & Vanderwart, M. (1980). A standardized set of 260 pictures: Norms for name agreement, image agreement, familiarity, and visual complexity. Journal of Experimental Psychology: Human Learning and Memory, 6, 174–215. Tyler, L. K., Stamatakis, E. A., Bright, P., Acres, K., Abdallah, S., Rodd, J. M., & Moss, H. E. (2004). Processing objects at different levels of specificity. Journal of Cognitive Neuroscience, 16, 351–362. doi:10.1162/ 089892904322926692 Venneri, A., McGeown, W. J., Hietanen, H. M., Guerrini, C., Ellis, A. W., & Shanks, M. (2008). The anatomical bases of semantic retrieval deficits in early Alzheimer’s disease. Neuropsychologia, 46, 497–510. doi:10.1016/j. neuropsychologia.2007.08.026
![[Postpartum herpetic encephalitis complicated by cerebral hematoma].](/assets/img/hold.png)
