Abstract

Patients affected by brain tumours may show behavioural and emotional regulation deficits, sometimes showing flattened affect and sometimes experiencing a true ‘change’ in personality. However, little evidence is available to the surgeon as to what changes are likely to occur with damage at specific sites, as previous studies have either relied on single cases or provided only limited anatomical specificity, mostly reporting associations rather than dissociations of symptoms. We investigated these aspects in patients undergoing surgery for the removal of cerebral tumours. We argued that many of the problems described can be ascribed to the onset of difficulties in one or more of the different levels of the process of mentalizing (i.e. abstracting and reflecting upon) emotion and intentions, which impacts on everyday behaviour. These were investigated in terms of (i) emotion recognition; (ii) Theory of Mind; (iii) alexithymia; and (iv) self-maturity (personality disorder). We hypothesized that temporo/limbic areas would be critical for processing emotion and intentions at a more perceptual level, while frontal lobe structures would be more critical when higher levels of mentalization/abstraction are required. We administered four different tasks, Task 1: emotion recognition of Ekman faces; Task 2: the Eyes Test (Theory of Mind); Task 3: Toronto Alexithymia Scale; and Task 4: Temperament and Character Inventory (a personality inventory), both immediately before and few days after the operation for the removal of brain tumours in a series of 71 patients (age range: 18–75 years; 33 female) with lesions located in the left or right frontal, temporal and parietal lobes. Lobe-based and voxel-based analysis confirmed that tasks requiring interpretation of emotions and intentions at more basic (less mentalized) levels (Tasks 1 and 2) were more affected by temporo/insular lesions, with emotion recognition (Task 1) being maximally impaired by anterior temporal and amygdala lesions and Task 2 (found to be a ‘basic’ Theory of Mind task involving only limited mentalization) being mostly impaired by posterior temporoparietal lesions. Tasks relying on higher-level mentalization (Tasks 3 and 4) were maximally affected by prefrontal lesions, with the alexithymia scale (Task 3) being mostly associated with anterior/medial lesions and the self-maturity measure (Task 4) with lateral prefrontal ones.

Introduction

Brain tumours can occur in many parts of the brain and it is not unusual that patients affected by brain tumours show an alteration of their emotional behaviour, a blunting or flattening of emotionality or sometimes can even experience apparent changes in personality. From time to time, different regions, ranging from insula to frontal or temporal lobes have been differently associated with one or more of these changes in social cognition (see Stuss et al., 1992; Lane et al., 1997; Gallagher and Frith, 2003; Lindquist et al., 2012 for extensive reviews). However, little evidence is available to the surgeon as to what changes are likely to occur with damage at specific sites.

Modifications in social cognition have tended to be associated with different types of impairment in the emotional or relational domain, ranging from simple emotion recognition to Theory of Mind processes, to emotional flattening, or a combination of two or more of these aspects. In this paper we address whether the underlying brain sites of such processes do in fact differ. We investigated the presence of deficits in emotion recognition, in Theory of Mind, the presence of alexithymia and the risk of personality disorder in a large sample of patients affected by brain tumours to assess whether dissociations at the behavioural and also the anatomical level exist in the processes necessary to implement these functions. We also wanted to see whether these different social cognition skills differ in the levels of abstraction/mentalization they imply.

Emotion recognition

A deficit in recognizing facial emotions might lead to the misinterpretation of the emotional state of others during social interactions, leading to inappropriate reactions and maladaptive behaviour. Emotion recognition difficulties have been associated in a number of patient group studies, with damage to temporal (Schmolck and Squire, 2001; Rosen et al., 2004; Rankin et al., 2006; Shaw et al., 2007) and frontal regions (Hornak et al., 2003; Rosen et al., 2004; Heberlein et al., 2008). Moreover, different specific regions have, at times, been associated with specific emotion processing deficits, for example, amygdala with fear (Adolphs et al., 1994), and insula (Calder et al., 2000; see Craig, 2009 for review of patients data) with disgust (see Fusar-Poli et al., 2009 for a recent meta-analysis of functional MRI data on the topic).

However, a debate exists between ‘natural kind models’ of emotions (Ekman, 1999), according to which, different emotions are biologically based, inherited ‘basic’ states each implemented by a distinct brain region, and ‘psychological constructionist’ models according to which, brain regions traditionally associated with the recognition of emotions do not process emotions per se, but are involved in more general processes. For example brain areas associated with the processing of fear would be activated for all stimuli that signal ‘uncertainty’ and novelty, and are thus motivationally salient in general (Lindquist et al., 2012). Emotion recognition would be achieved by developing a mental representation (core affect) of the ‘motivational stimulus’, a process implying conceptualization skills and relying more on limbic/paralimbic structures (Kober et al., 2008). This process would, however, not require the subject to abstract, reflect and make inferences upon the perceived stimulus. In these respects, for an emotion to be recognized, the degree of ‘mentalizing’ required would be relatively limited.

Theory of Mind

Another key domain of processing required for efficient social interactions is that of Theory of Mind, the ability to conceive of mental states, such as thoughts, beliefs and intentions in oneself and others (Premack and Woodruff, 1978; Fonagy et al., 2003), allowing us to make sense of a person’s behaviour and behave consequently. Failure in Theory of Mind reasoning is held to be the core impairment in autism and is typically investigated by means of tasks (usually stories, but also cartoons or pictures) that vary in the degree to which they depend upon higher order mentalization mechanisms, i.e. those involving reflection, abstraction and symbolic representation of external stimuli (Gallagher and Frith, 2003). Some tasks imply more basic and some more complex interpretations of others’ intentions and belief (e.g. first and second order false beliefs: Frith and Frith, 1999; Perner and Lang, 1999).

Even with some degree of variability, neuroimaging studies have achieved a great deal of consensus in indicating the medial prefrontal cortex (PFC), anterior cingulate cortex and the posterior temporo-parietal areas (superior temporal sulcus and TPJ) as the brain structures most commonly involved in Theory of Mind reasoning (Carrington and Bailey, 2009). The evidence is also supported by the—somewhat more recent—literature on acquired brain lesions, suggesting that lesions to the ventromedial PFC in particular would be critically related to lower Theory of Mind scores (Stuss and Anderson, 2004; Martin-Rodriguez and Leon-Carrion, 2010; Samson and Michel, 2013).

What is still debated but with very little evidence from neurological populations (Samson et al., 2004), however, is the precise role and relationship between frontal (medial PFC and anterior cingulate cortex) and temporal-parietal regions in Theory of Mind and whether the latter is a necessary part of the Theory of Mind circuit (Saxe and Kanwisher, 2003).

Alexithymia

Alexithymia (Sifneos, 1973), is defined as a deficit in the symbolic mental representation of emotions, in verbal behaviour, fantasy or dreams (Lane et al., 1997). Alexithymic individuals are unable to describe their emotions and to identify the causes of their emotional reactions, they show lack of imagination and, behaviourally, they can manifest bland or flattened affect. They show a difficulty in self-regulation and self-care and tend to be impulsive and feel discomfort in social interactions (and consequently avoid them). By definition therefore, it is a deficit in ‘mentalizing’ emotions.

Although frontal lesions have been often associated with blunted emotional responses (Stuss et al., 1992; Larsen et al., 2003), few formal assessments of the presence of alexithymia have been produced in neurological patient populations, and these mainly rely on single case reports (Becerra et al., 2002; Schafer et al., 2007; Fricchione and Howanitz, 2010), whereas groups studies either reported a generally higher incidence of alexithymia in traumatic brain injuries or, at most, found a prevalence of alexithymia in right hemisphere lesions (Spalletta et al., 2001). In neuroimaging investigations, however, alexithymia has been consistently associated with reduced activity or decreased volume, particularly in anterior cingulate cortex in alexithymic healthy subjects (Larsen et al., 2003; Gundel et al., 2004; Ihme et al., 2013).

Personality disorder

Stuss et al. (1992) suggested that a frequent major change after frontal lobe pathology is a true personality change. Indeed, frontal patients often show decreased initiative and emotional control, reduced reliability and foresight, childish behaviour and disinhibition (Milner and Petrides, 1984; Barrash et al., 2000; Stuss et al., 2003), behavioural modifications that may appear, in everyday life, like a real change in the patient’s ‘maturity of the self’. An interesting relation between personality disorder and development of the ‘self’ concept has been proposed in the so-called ‘biosocial’ model of personality developed by Cloninger et al. (1994). According to Cloninger et al. (1994), personality is characterized by two components: temperament and character. Although ‘temperament’ reflects our way of automatically reacting to environmental stimulation (see Supplementary material for a more detailed description of the model), ‘character’ refers to a more ‘conceptualized’ (and therefore ‘mentalized’) knowledge and evaluation of the self (the conceptual idea one has of oneself) and is responsible for efficient behavioural self-regulation. In Cloninger’s terms, to obtain efficient self-regulation and be protected against personality disorder, an appropriate level of maturity has to be reached particularly in terms of self directedness which measures the maturity of the self in terms of ‘self-efficacy’ and ‘self-esteem’ and cooperativeness defining the maturity of the individual as part of a society (compassion, empathy, and tolerance). Individuals with low ‘self maturity’ (both self-directedness and cooperativeness scores) are held to be at high risk of personality disorders by showing immature behaviour, being unable to delay gratification or pursue long-term objectives, being selfish and impulsive and having an external locus of control. Although a link has been suggested between prefrontal dysfunctions and self-directedness scores in some neuropsychological studies involving patients with Parkinson’s disease (McNamara et al., 2007), the Temperament and Character Inventory (TCI) has to date not been much used in patients with more focal brain lesions.

Aim of the study

From a purely clinical point of view the aim of this study was to provide more specific and consistent information about the origin of the behavioural, emotional and personality changes that may appear following brain tumours and surgery in different brain areas. Many of the studies on neurological populations discussed above relied either on single cases or small groups of patients or, when larger groups were assessed, only one or two of these aspects have been simultaneously investigated (Henry et al., 2006; Shaw et al., 2007; Broicher et al., 2012). Many of these studies moreover reported associations between symptoms within the same patient group.

By contrast, an overall investigation of all of these aspects together in a large sample of brain damaged patients has not been carried out to our knowledge. Yet, neurological patient studies that investigate a single function are liable to a variety of selection artefacts. It has been argued, however, that through examining a number of different functions in the same cognitive domain, one can have much greater confidence if different localizations are found for different functions (Shallice and Cooper, 2011). We further aimed to test whether dissociations could be found within the same patient population, between different lesion locations and different patterns of social cognition difficulties. Finally, from a cognitive point of view, we wanted to assess whether it is possible to separate processes requiring interpretation of emotions and intentions as perceived (or identified) from external stimuli and therefore at a less abstract/mentalized level (and more linked to temporo-limbic areas) from processes requiring inferences and abstraction of more stable traits (more linked to prefrontal cortex functions) (Van Overwalle, 2009).

Materials and methods

Participants

A consecutive series of 71 patients operated for the removal of brain tumours participated in the study. Inclusion criteria were the presence of brain lesions situated in the left or right frontal, parietal or temporal lobes in patients ranging in age from 18 to 75 years. Patients with probable gliomatosis cerebri as well as recurring or multiple distinct lesions were excluded. All but six of the patients were right-handed. The study was approved by the ethical committee of A.O.U. S. Maria della Misericordia of Udine and all patients gave their written consent before participating in the study.

Histology of the lesions included high (n = 29) and low (n = 21) grade gliomas and meningiomas (n = 16); metastases (n = 5) were also present. Patients (40 right and 31 left hemisphere) were subdivided into three different groups according to the location of lesion centre-of-mass: frontal (n = 31 patients), parietal (n = 20) and temporal (n = 20). Age and education did not differ between the three groups (see Supplementary material for details on anatomical and demographical analysis). Patients were tested in two separate sessions: a few days before (usually the day before) the operation and a few days (usually within a week) after the surgery. On both occasions patients were administered a comprehensive neuropsychological assessment covering the main cognitive domains: language, attention, executive function, memory and perception (patients scores are reported in Supplementary Table 1), together with the experimental tasks.

Experimental investigation

The experimental part of the study consisted in four tasks needing different degrees of mentalization and ranging from simple recognition of facial emotion (word-to-picture matching of stimuli from the Ekman faces corpus) (Ekman and Rosenberg, 1997), to attribution of mental states through the interpretation of eye gaze (Theory of Mind, ‘The Eyes Test’: Baron-Cohen et al., 2001), to evaluation/interpretation of one’s own emotional states (Toronto Alexythimia Scale, TAS-20: Bagby et al., 1994) and up to self-reported personality description (TCI questionnaire: character scales; Cloninger et al., 1994). Detailed descriptions of each task and raw scores for each patient are reported in the Supplementary material.

For all tasks, except Task 1, results were entered into a repeated measures ANOVA design, with ‘Surgery’ (pre- versus post-surgery performance) as a within-subject variable and ‘Hemisphere’ (left versus right) and ‘Location’ (frontal versus temporal versus parietal) as between subjects variables. For Task 1, ‘Emotion Type’ was also added as a further within-subject variable. Tukey test post hoc analysis was then performed to investigate the sources of possible significant main effects and interaction. As surgery effects were never significant for any of the tasks, apart from a general, specific decrement in performance for all patients in Task 1, surgery effects results are given in the Supplementary material.

Task 1: Emotion recognition (Ekman faces)

This was a word-to-picture matching task with a display of six faces of the same person expressing the six basic emotions: happiness, sadness, anger, surprise, fear and disgust. There were 36 stimuli (six emotions for each of the six selected models). Subjects were asked to point to the face depicting the facial emotion requested and were provided free time for the response. Cut-off score was set at two standard deviations (SD) below the mean control subjects’ score (28.99/36).

Task 2: Theory of Mind: the Eyes Test

The revised version of the ‘Reading the Mind in the Eyes’ Test (Baron-Cohen et al., 2001) was administered. The task consisted of a picture-to-word matching task in which participants had to identify which of four words (describing complex mental states and intentions) best matched the intention expressed by a photograph of the eye region of a person.

Task 3: Toronto Alexithymia Scale

The Italian version (Bressi et al., 1996) of the TAS-20 (Bagby et al., 1994) was administered. Participants are asked to rate 20 items on a 5-point Likert, capturing possible difficulties in identifying or describing feelings or in having internally oriented thinking. Both before and after surgery, patients were asked to evaluate how they generally feel, not referring to a particular time frame. Universally accepted cut-off scores for alexithymia identify three performance profiles: scores <51 indicate the absence of alexithymia; scores between 52 and 60 indicate possible alexithymia; scores >61 indicate severe alexithymia.

Task 4: Temperament and Character Inventory

The Italian translation (Battaglia and Bajo, 2000) of the TCI (Cloninger et al., 1994) was administered. Both before and after surgery, patients were asked to evaluate themselves in general, not referring to a particular time frame. Within the character scales, a combined measure of both self-directedness and cooperativness was computed and entered in the statistical analysis in order to obtain a direct measure of ‘self-maturity’ and therefore of ‘personality disorder risk’. There was a higher number of drop-out cases for this task. These occurred for different reasons, mainly linked to the onset of language, attention or executive function problems limiting the ability of the patients to carry out a long and demanding questionnaire properly.

Voxel-based lesion–symptom mapping analysis

The voxel-based lesion–symptom mapping (VLSM) technique (Bates et al., 2003; Rorden et al., 2007) was adopted, in addition to standard lobe analyses, to identify brain areas critically involved in accounting for any behavioural result. In this procedure, patients are classified in two groups according to whether or not the lesion affects a specific voxel. Then, the behavioural performance (e.g. number of errors) of patients in a particular task is compared across groups, for instance by using a t-test for each voxel and with appropriate procedures to take into account that multiple comparisons are being made. With this method it is possible to highlight those voxels that are associated with significantly lower scores in a particular task. VLSM has the advantage of avoiding any a priori lesion location grouping. For all patients, preoperative T1 and (when available) T2-weighted or FLAIR scans were collected to determine tumour location. Only preoperative scans, used for neuronavigation by the neurosurgeon as best indicators of macroscopic tumour extent, were considered in the reconstruction procedure because, after surgery, lesion locus is usually at least partially replaced by healthy neighbouring tissue, possibly creating confusion in the reconstruction of the real lesion boundaries. A complete resection has been obtained in meningiomas and metastases, which are well circumscribed lesions. In high grade gliomas, all the compact, visible part of the tumour has been removed allowing a macroscopically complete resection. For low grade gliomas, the extent of resection, evaluated in terms of percentage of the total volume, using the standard current procedure, was equal to 89.4% (Ius et al., 2012). The 3D regions of interest reconstruction of lesions were drawn for each patient from MRI slices on the horizontal plane, and underwent spatial normalization using SPM8 software. Lesion volume for each region of interest as well as distribution of aggressive and slowly growing types of lesion were then estimated and did not differ between patient groups (Supplementary material).

A separate VLSM analysis was performed for each task of the study. As there were no interactions with pre/post-surgery (Supplementary material), it was considered more reliable to compute VLSM analyses using the average pre-post operation scores for each patient in each task as a behavioural measure. Moreover, as no hemispheric lateralization effects were found in any task (see ‘Results’ section), suggesting an equal involvement of both hemispheres in performance of the tasks, lesion lateralization was ignored and all regions of interest were superimposed onto the same hemisphere (right hemisphere for illustrative purposes). Voxel-by-voxel statistical analyses were performed by means of NPM software (www.MRIcro.com), using t-tests. To minimize possible outlier effects, the analyses were conducted only on voxels damaged in at least three patients (86 262 voxels out of 255 170) with the statistical threshold set at P < 0.05 across all tasks (False Discovery Rate correction applied).

Results

Task 1: Emotion Recognition

Behavioural results

Group analysis revealed a significant ‘Location’ effect [F(2,57) = 7.812; P = 0.001] with temporal patients (who obtained average scores below the control’s cut-off) showing greater difficulties in identifying facial emotions with respect to parietal (P = 0.001) patients (who obtained average scores in line with controls) (Fig.1). Indeed 16/20 (80%) of the temporal patients performed worse than controls either before or after surgery (or on both occasions, Table 1). Frontal patients, in turn, were somewhat ‘in between’ the two extremes; indeed they did not significantly differ compared to either temporal (P = 0.143) or parietal (P = 0.076) patients (even if a trend was present). No effects of ‘Hemisphere’ [F(1,57) = 0.914; P = 0.343)] or ‘Location × Hemisphere’ [F(2,57) = 0.687; P = 0.507] interaction were found, suggesting that the effect was present in a similar fashion in left and right hemisphere patients.

Figure 1

Performance of patients in the emotion recognition task. Vertical bars indicate standard error. Temporal patients were significantly impaired with respect to both frontal and parietal patients. VLSM analysis showed that worse emotion recognition was associated with lesions in anterior temporal (including amygdala) and posterior insular lesions. As lateralization effects were absent, all lesions were superimposed on the same hemisphere (right) for illustrative purposes.

Figure 1

Performance of patients in the emotion recognition task. Vertical bars indicate standard error. Temporal patients were significantly impaired with respect to both frontal and parietal patients. VLSM analysis showed that worse emotion recognition was associated with lesions in anterior temporal (including amygdala) and posterior insular lesions. As lateralization effects were absent, all lesions were superimposed on the same hemisphere (right) for illustrative purposes.

Table 1

Number of patients (%) obtaining a score of clinical relevance either before or after surgery (or both) in each of the four experimental tasks.

 Lesion location
 
Frontal Temporal Parietal 
  •     Task 1

  • Emotion recognition: Ekman Facesa

 
14/31 16/20 4/20 
(45.16%) (80.00%) (20.00%) 
  •     Task 2

  • Theory of Mind: The Eyes Testb

 
10/28 10/20 5/19 
(35.71%) (50.00%) (26.32%) 
  •     Task 3

  • Alexithymia: TAS-20c

 
23/31 2/17 5/19 
(75.19%) (11.75%) (26.32%) 
  •     Task 4

  • Self Maturity: TCI-SD+COd

 
12/27 2/13 3/16 
(44.44%) (15.38%) (18.75%) 
 Lesion location
 
Frontal Temporal Parietal 
  •     Task 1

  • Emotion recognition: Ekman Facesa

 
14/31 16/20 4/20 
(45.16%) (80.00%) (20.00%) 
  •     Task 2

  • Theory of Mind: The Eyes Testb

 
10/28 10/20 5/19 
(35.71%) (50.00%) (26.32%) 
  •     Task 3

  • Alexithymia: TAS-20c

 
23/31 2/17 5/19 
(75.19%) (11.75%) (26.32%) 
  •     Task 4

  • Self Maturity: TCI-SD+COd

 
12/27 2/13 3/16 
(44.44%) (15.38%) (18.75%) 

The higher percentage of clinically relevant scores (bold text) was found for temporal patients in Tasks 1 and 2, and for frontal patients in Tasks 3 and 4.

a >2 SD below mean control sample score.

b >1 SD below the mean population.

c >51 (cut-off for possible alexithymia).

d <58 (sum of self-directedness + cooperativeness: cut-off for suspect personality disorder).

Critically, however, there was a general main effect of ‘Emotion type’ [F(5,285) = 28.395; P < 0.001] and more critically, an interaction between ‘Emotion type’ and ‘Location’ [F(10,285) = 2.334; P = 0.011]. Post hoc analysis showed that, when comparing performance across patient groups for each emotion (Fig. 2), temporal lobe patients had significantly more difficulty in identifying fear than parietal patients (P < 0.001), and no other significant difference was found for any other comparison (all P > 0.074). A control correlation between task results and data from the Benton Face Recognition Test (Supplementary material) excluded the possibility that the results could be attributed to prosopagnosia.

Figure 2

Among the different types of emotion, temporal patients were particularly impaired in recognizing ‘fear’ with respect to parietal (unimpaired) patients. Vertical bars indicate standard error.

Figure 2

Among the different types of emotion, temporal patients were particularly impaired in recognizing ‘fear’ with respect to parietal (unimpaired) patients. Vertical bars indicate standard error.

Apart from this, and less surprisingly, within-group comparisons, showed that parietal patients were equally good at identifying all emotions (apart from an advantage of happiness over anger: P = 0.006, and surprise: P-values = 0.014), whereas for frontal patients happiness was again easier to identify than all the other emotions (all P-values < 0.002) and no other difference was found between emotions. Finally, for temporal patients, happiness was, once more, easier to identify than all other emotions (all P-values < 0.001), except for disgust (P = 0.200) and disgust was better recognized than both fear (P = 0.004) and anger (P = 0.009).

Voxel-based lesion–symptom mapping analysis

Table 2 shows the areas significantly associated with more severe facial emotion recognition difficulties. Lower scores were associated with lesions to the anterior portion of the temporal lobes (superior, middle and inferior temporal gyri as well as temporal pole and hippocampus), the insula and the amygdala (96% of its total volume fell within the significant low emotion recognition region).

Table 2

Areas significantly associated with lower scores in the four experimental tasks (Facial Emotion Recognition, the Eyes Test, TAS 20 and TCI self-maturity)

Region AAL label n of voxels % of total Max Z-score Max Z: MNI coordinates
 
x y z 
Emoton recognition 
Anterior temporal lobe Middle temporal gyrus 17669 18.13 4.76 46 −12 −11 
Superior temporal gyrus 13380 13.73 5.16 44 −12 −10 
Inferior temporal gyrus 7926 8.13 4.60 45 −16 −20 
Hippocampus 6552 6.72 5.09 36 −14 −12 
Superior temporal pole 6432 6.60 4.34 37 −23 
Parahippocampal gyrus 6083 6.24 4.21 20 −16 −16 
Middle temporal pole 4361 4.47 4.24 47 −17 
Fusiform gyrus 4319 4.43 4.52 41 −14 −20 
Insula Insula 2959 3.04 5.08 41 −13 −7 
Amygdala (96% amygdala volume) 1894 1.94 4.17 29 −5 −15 
Basal ganglia Putamen 1596 1.64 4.76 32 −15 −4 
Subcortical Subcortical 21749 22.31 5.27 40 −14 −8 
Other (<1%)  2560 2.93 − − − − 
Theory of mind 
Posterior temporal + TPJ Superior temporal gyrus 16796 29.43 4.79 53 −38 20 
Middle temporal gyrus 6241 10.93 3.80 61 −38 
Supramarginal gyrus 3036 5.32 4.52 64 −46 25 
Heschl gyrus 1746 3.06 4.16 44 −14 12 
Hippocampus 418 1.35 3.59 16 −26 −7 
Insula Insula 7189 12.60 4.21 37 −16 13 
Lateral prefrontal cortex Rolandic operculum 4847 8.49 4.16 45 −14 12 
Frontal inferior pars opercularis 938 1.64 4.04 46 20 
Anterior temporal lobe Superior temporal pole 3759 6.59 3.82 56 −8 
Middle temporal pole 847 1.48 3.64 51 −15 
Basal ganglia Putamen 1274 2.23 3.74 28 −12 13 
Thalamus Thalamus 770 1.35 3.62 13 −24 −1 
Subcortical Subcortical 7700 13.49 4.02 48 18 −3 
Other (<1%)  1931 3.38 − − − − 
Alexithymia 
Anterior lateral prefrontal cortex Middle frontal gyrus 25017 14.18 4.90 29 32 33 
Superior frontal gyrus 20656 11.71 4.71 24 47 15 
Inferior frontal pars triangularis 10190 5.78 4.15 29 16 28 
Inferior frontal pars opercularis 3047 1.73 4.10 29 14 32 
Cortical midline Structures Medial superior front gyrus 11436 6.48 4.31 16 48 
Anterior cingulate cortex 9483 5.38 4.84 12 32 17 
Medial suppl. motor area 8480 4.81 4.09 10 49 
Inferior orbitofrontal gyrus 7867 4.46 3.48 32 36 −11 
Middle orbitofrontal gyrus 6565 3.72 3.55 29 49 −4 
Middle cingulate cortex 5726 3.25 4.61 13 32 31 
Medial orbitofrontal cortex 5365 3.04 4.12 13 42 −3 
Superior orbitofrontal gyrus 4783 2.71 4.41 27 49 −3 
Basal ganglia Caudate nucleus 5264 2.98 4.02 15 26 
Subcortical Subcortical 45530 25.82 4.77 25 44 16 
Other (<1%)  6955 3.94 − − − − 
Self–maturity 
Lateral prefrontal cortex Inferior frontal pars triangularis 1372 14.27 4.04 33 16 29 
Middle frontal gyrus 1236 12.86 3.82 33 19 33 
Inferior frontal pars opercularis 1067 11.10 4.09 36 29 
Precentral gyrus 697 7.25 4.21 41 33 
Insula Insula 264 2.75 3.61 33 10 16 
Basal ganglia Putamen 180 1.87 3.56 28 10 13 
Subcortical Subcortical 4757 49.48 4.60 32 −4 40 
Other (<1%)  41 0.42 − − − − 
Region AAL label n of voxels % of total Max Z-score Max Z: MNI coordinates
 
x y z 
Emoton recognition 
Anterior temporal lobe Middle temporal gyrus 17669 18.13 4.76 46 −12 −11 
Superior temporal gyrus 13380 13.73 5.16 44 −12 −10 
Inferior temporal gyrus 7926 8.13 4.60 45 −16 −20 
Hippocampus 6552 6.72 5.09 36 −14 −12 
Superior temporal pole 6432 6.60 4.34 37 −23 
Parahippocampal gyrus 6083 6.24 4.21 20 −16 −16 
Middle temporal pole 4361 4.47 4.24 47 −17 
Fusiform gyrus 4319 4.43 4.52 41 −14 −20 
Insula Insula 2959 3.04 5.08 41 −13 −7 
Amygdala (96% amygdala volume) 1894 1.94 4.17 29 −5 −15 
Basal ganglia Putamen 1596 1.64 4.76 32 −15 −4 
Subcortical Subcortical 21749 22.31 5.27 40 −14 −8 
Other (<1%)  2560 2.93 − − − − 
Theory of mind 
Posterior temporal + TPJ Superior temporal gyrus 16796 29.43 4.79 53 −38 20 
Middle temporal gyrus 6241 10.93 3.80 61 −38 
Supramarginal gyrus 3036 5.32 4.52 64 −46 25 
Heschl gyrus 1746 3.06 4.16 44 −14 12 
Hippocampus 418 1.35 3.59 16 −26 −7 
Insula Insula 7189 12.60 4.21 37 −16 13 
Lateral prefrontal cortex Rolandic operculum 4847 8.49 4.16 45 −14 12 
Frontal inferior pars opercularis 938 1.64 4.04 46 20 
Anterior temporal lobe Superior temporal pole 3759 6.59 3.82 56 −8 
Middle temporal pole 847 1.48 3.64 51 −15 
Basal ganglia Putamen 1274 2.23 3.74 28 −12 13 
Thalamus Thalamus 770 1.35 3.62 13 −24 −1 
Subcortical Subcortical 7700 13.49 4.02 48 18 −3 
Other (<1%)  1931 3.38 − − − − 
Alexithymia 
Anterior lateral prefrontal cortex Middle frontal gyrus 25017 14.18 4.90 29 32 33 
Superior frontal gyrus 20656 11.71 4.71 24 47 15 
Inferior frontal pars triangularis 10190 5.78 4.15 29 16 28 
Inferior frontal pars opercularis 3047 1.73 4.10 29 14 32 
Cortical midline Structures Medial superior front gyrus 11436 6.48 4.31 16 48 
Anterior cingulate cortex 9483 5.38 4.84 12 32 17 
Medial suppl. motor area 8480 4.81 4.09 10 49 
Inferior orbitofrontal gyrus 7867 4.46 3.48 32 36 −11 
Middle orbitofrontal gyrus 6565 3.72 3.55 29 49 −4 
Middle cingulate cortex 5726 3.25 4.61 13 32 31 
Medial orbitofrontal cortex 5365 3.04 4.12 13 42 −3 
Superior orbitofrontal gyrus 4783 2.71 4.41 27 49 −3 
Basal ganglia Caudate nucleus 5264 2.98 4.02 15 26 
Subcortical Subcortical 45530 25.82 4.77 25 44 16 
Other (<1%)  6955 3.94 − − − − 
Self–maturity 
Lateral prefrontal cortex Inferior frontal pars triangularis 1372 14.27 4.04 33 16 29 
Middle frontal gyrus 1236 12.86 3.82 33 19 33 
Inferior frontal pars opercularis 1067 11.10 4.09 36 29 
Precentral gyrus 697 7.25 4.21 41 33 
Insula Insula 264 2.75 3.61 33 10 16 
Basal ganglia Putamen 180 1.87 3.56 28 10 13 
Subcortical Subcortical 4757 49.48 4.60 32 −4 40 
Other (<1%)  41 0.42 − − − − 

AAL = Automated Anatomical Labeling; TPJ = temporoparietal junction.

Discussion

From a behavioural point of view, temporal lobe patients had greater facial emotion recognition difficulties and were particularly impaired in identifying fear. Frontal patients were, in turn, not significantly different from parietal patients. VLSM analysis confirmed the behavioural results, indicating that the regions mostly associated with more severe deficits were the anterior portion of the temporal lobes and both the insula and the almost entire amygdala.

Taken together, behavioural and anatomical results seem not to support a purely ‘localizationist’ account of emotion processing in the brain (see ‘Introduction’ section), with distinct anatomical regions devoted to the processing of distinct, biologically-based and inherited emotions (Ekman, 1999). Indeed our results indicate that damage to posterior insula was associated with the most severe emotion recognition deficits, but the worst recognized emotion was fear (consistent with amygdala involvement) and not disgust, as would be predicted by the localizationist ‘insula-disgust’ association (Calder et al., 2000). On the contrary, disgust was the second best preserved emotion, not different from happiness and better recognized than fear.

The insula itself has not, however, always been universally associated with disgust, but it has also been often linked, in a variety of studies, to basic emotion recognition processing in general (Kober et al., 2008). The findings therefore support more a ‘psychological constructionist’ account of emotion processing (Lindquist et al., 2012), according to which, temporo-limbic structures (amygdala and insula), are devoted to the processing of motivationally-salient stimuli (which are then interpreted as ‘emotion’). Fear-inducing stimuli, for example, would only fall into the broader class of ‘uncertain’, novel and unusual, and therefore emotionally-salient, stimuli (Holland and Gallagher, 1999), signalling whether any exteroceptive sensory information is motivationally salient. Further support for the role of anterior temporal regions in identifying discrete emotions as a general function comes from a recent study by Lindquist et al. (2014) in which it is shown that patients with semantic dementia (i.e. those with predominant damage to the anterior temporal lobes) have difficulties in identifying emotions but not in judging the pleasantness of a face, so preserving therefore the ability to perceive affects. Finally, our data do not support a strong role of frontal areas in emotion recognition, as sometimes suggested (Hornak et al., 2003). Indeed, while our frontal group of patients was not completely ‘unimpaired’ in the task, lower scores in frontal patients were found in patients having larger ‘fronto-temporal’ lesions and were rarely associated with purely frontal lesions.

Task 2: Theory of Mind

Behavioural results

At a group level a significant ‘Location’ effect was again found [F(2,53) = 6.724; P = 0.002]. Post hoc analysis revealed that both frontal and temporal groups of patients obtained significantly lower Theory of Mind scores with respect to parietal patients (Fig. 3) in terms of their ability to infer the intentions from the eye gaze (P = 0.015 for frontal and P = 0.001 for temporal patients), with no difference between the frontal and temporal groups (P = 0.786). Indeed, 50% (10/20) of the temporal lobe patients performed worse than 1 SD below the reference population, as did 36% (10/28) of the frontal patients, either before or after surgery or both (Table 1). No effects of ‘Hemisphere’ [F(1,53) = 0.002; P = 0.968] or a ‘Location × Hemisphere’ [F(2,53) = 0.385; P = 0.682] interaction were found, suggesting roughly equal impairments in left and right hemisphere patients.

Figure 3

Performance of patients in the Eyes Test. Vertical bars indicate standard error. Temporal and frontal patients showed lower Theory of Mind scores with respect to parietal patients. VLSM analysis showed that posterior temporal areas and temporo-parietal junction (TPJ), together with the insula were associated with lower scores on the task. As lateralization effects were absent, all lesions were superimposed on the same hemisphere (right) for illustrative purposes.

Figure 3

Performance of patients in the Eyes Test. Vertical bars indicate standard error. Temporal and frontal patients showed lower Theory of Mind scores with respect to parietal patients. VLSM analysis showed that posterior temporal areas and temporo-parietal junction (TPJ), together with the insula were associated with lower scores on the task. As lateralization effects were absent, all lesions were superimposed on the same hemisphere (right) for illustrative purposes.

Voxel-based lesion–symptom mapping analysis

Areas associated with lower scores in the Eyes Test are shown in Table 2. Lower Theory of Mind scores were associated with lesions involving more superior and posterior areas within the temporal lobes with respect to emotion recognition, in particular the temporo-parietal junction (TPJ: posterior superior temporal gyrus, Heschl and supramarginal gyri) and the insula, accounting for >50% of total significant volume. Parts of the inferolateral prefrontal cortex (Rolandic operculum and pars opercularis) were also involved. However, no involvement of orbitofrontal cortex or medial prefrontal structures (anterior cingulate cortex or medial PFC) was detected in this Theory of Mind task.

Discussion

Behaviourally, patients with both temporal and frontal lobe lesions performed significantly less efficiently in identifying the intentions of others from their eye gaze. This is in accordance with a large part of Theory of Mind literature (Carrington and Bailey, 2009) which points to both frontal and posterior temporal regions as being critical regions for Theory of Mind. However, VLSM analysis mainly indicated temporo-parietal junction, superior temporal gyrus and again the insula, but not the typical locations (anterior cingulate cortex, medial PFC and orbitofrontal cortex), found to be active (or damaged) in the large majority of previous studies on Theory of Mind.

In his meta-analysis of functional MRI studies of Theory of Mind, Van Overwalle (2009) divides social cognition processes (which include Theory of Mind) into two broad categories according to the different levels of mentalization skills required, namely inferences on transitory states (goals and intentions) and inferences on enduring characteristics (personality traits). The former are held to be more perceptual in nature and to be mainly processed in temporo-parietal junction, whereas the latter would require higher levels of abstraction and mentalization and would be more linked to medial PFC. If we consider the task we used, these results are in-line with this position; Baron-Cohen et al. (2001) indeed, describes the Eyes Test as a Theory of Mind task which only involves the first stage of Theory of Mind (attribution of a mental state to somebody), making the Eyes Test a more ‘basic’ Theory of Mind task involving only limited reflection and requiring only a minor role for mentalization (Gallagher and Frith, 2003).

An alternative explanation for the involvement of temporo-parietal junction in a Theory of Mind task like the Eyes Test, is that it requires that intentions be interpreted more perceptually through eye gaze, and eye gaze monitoring has often been associated with Theory of Mind processes (Allison et al., 2000; Hoffman and Haxby, 2000) as gaze does not just indicate where someone is looking, but also signals that something attracted his/her attention. Posterior superior temporal sulcus and temporo-parietal junction have often been associated with eye gaze monitoring in general (Carrington and Bailey, 2009), suggesting that these regions may be part of a ‘post-perceptual’ system for processing the actions of external agents (Calder et al., 2002).

Interestingly, similar results on a small sample of patients, using the Eyes Test, were obtained by Samson et al. (2004). However, as far as we know, the present results constitute the first evidence of the role of superior temporal sulcus and temporo-parietal junction in Theory of Mind in a large sample of brain damaged patients.

Task 3: Alexithymia

Behavioural results

Behaviourally, a significant ‘Location’ effect was again found [F(2,56) = 13.031; P < 0.001] with the frontal group, this time, showing significantly higher alexithymia scores (and being above cut-off for ‘possible alexithymia’) than both the other two groups, the temporal (P < 0.001) and parietal (P = 0.001) patients who, in turn, obtained scores overlapping with the reference population (Fig. 4). Indeed, the large majority of frontal lobe patients (23/31, 74%; Table1) obtained a score of 52 or more, being in the range of possible alexithymia, with 10/31 (32%), scoring >61 and being clearly alexithymic, either before or after the surgery or on both occasions.

Figure 4

Alexithymia (TAS-20) scores for the group of patients. Higher scores indicate higher alexithymia. Vertical bars indicate standard error. Alexithymia scores were significantly higher for the frontal than the other two (unimpaired) groups of patients. VLSM analysis clearly indicated that higher alexithymia scores were found with lesions involving the cingulum and the medial and lateral prefrontal cortex. As lateralization effects were absent, all lesions were superimposed on the same hemisphere (right) for illustrative purposes.

Figure 4

Alexithymia (TAS-20) scores for the group of patients. Higher scores indicate higher alexithymia. Vertical bars indicate standard error. Alexithymia scores were significantly higher for the frontal than the other two (unimpaired) groups of patients. VLSM analysis clearly indicated that higher alexithymia scores were found with lesions involving the cingulum and the medial and lateral prefrontal cortex. As lateralization effects were absent, all lesions were superimposed on the same hemisphere (right) for illustrative purposes.

No effect of ‘Hemisphere’ [F(1,56) = 0.353; P = 0.555] or ‘Hemisphere × Location’ [F(2,56) = 0.411; P = 0.665] interaction was again found suggesting that the left and right hemispheres were equally involved. Finally, a non-significant correlation between task scores and scores from the Frontal Assessment Battery (Supplementary material) showed that the results were not a consequence of a dysexecutive syndrome in frontal patients, which might have limited their ability to provide graded responses.

Voxel-based lesion–symptom mapping analysis

Higher alexithymia scores were associated with completely different areas from those associated with the preceding tasks. More severe difficulties were found predominantly (Table 2) with lesions in the antero-medial and lateral surfaces of the frontal lobes, in particular lateral and medial portions of Brodmann areas 8 and 9 (middle and superior frontal gyri), and several medial prefrontal regions, i.e. anterior and middle cingulate cortices, medial superior frontal gyrus and also portions of the orbitofrontal cortex. This time, no temporal lobes regions were significantly associated with clinically relevant results.

Discussion

Unlike the previous tasks, frontal patients were the only group, in the alexithymia task, showing impaired performance. Indeed there was a high incidence of cases in the clinical range for possible alexithymia in the frontal group, with very few in the temporal or parietal groups. Accordingly, VLSM results associated higher alexithymia scores with lesions involving the antero-lateral and medial surfaces of the frontal lobes, largely involving anterior cingulate cortex and medial PFC and the orbitofrontal cortices.

The present results are important since they constitute the first (to our knowledge) attempt to specify, in a large sample of brain damaged patients, the precise brain regions involved in acquired alexithymia. Our findings nicely overlap with areas identified in previously reviewed functional MRI literature, especially regarding the involvement of anterior cingulate cortex and medial PFC (Larsen et al., 2003; Moriguchi et al., 2006). It seems therefore that these structures play a key role in mentalizing one’s own emotional states. The anterior cingulate cortex is particularly relevant, in that it is commonly associated with the conscious awareness of cognitive stimuli (Frith et al., 1991; Posner and Rothbart, 1998; Jack and Shallice, 2001) and with consciousness in general (Cotterill, 1995). Medial prefrontal regions have also been associated with the processes of directing attention and being aware of one’s own emotional states in a recent functional MRI study (Satpute et al., 2013). Moreover, in a meta-analysis of functional MRI evidence, anterior cingulate cortex activation has been consistently found during the processing and evaluation of one’s own emotional states with respect to both that of others and of the level of ‘emotionality’ of certain situations (Lee and Siegle, 2012). It is moreover worth noting that besides anterior cingulate cortex, medial PFC and orbitofrontal cortices (as well as more antero-lateral regions) were also largely involved, in accordance with anatomical localization of higher level mentalization processes in traditional Theory of Mind tasks (Carrington and Bailey, 2009).

Task 4: Maturity of self

Behavioural results

When considering the measures of self-maturity at a group level, the results did not reveal significant effects of any of the variables considered; however a trend was present toward a significant ‘Location’ effect [F(2,42) = 2.439; P = 0.099] with frontal patients again showing lower scores (but not significantly so) (Fig. 5). A high number of frontal patients (12/27, 44%; Table1) obtained a combined raw score of self-maturity (self-directedness + cooperativeness) below 58 (cut-off score indicated by Cloninger as highly predictive of the presence of personality disorder) either before or after surgery or both.

Figure 5

Self-maturity scores (TCI questionnaire). Vertical bars indicate standard error. Frontal patients showed a trend toward a lower level of self maturity with respect to the other two groups of patients. VLSM analysis indicated that a portion of the lateral prefrontal cortex was associated with lower self-maturity scores. As lateralization effects were absent, all lesions were superimposed on the same hemisphere (right) for illustrative purposes.

Figure 5

Self-maturity scores (TCI questionnaire). Vertical bars indicate standard error. Frontal patients showed a trend toward a lower level of self maturity with respect to the other two groups of patients. VLSM analysis indicated that a portion of the lateral prefrontal cortex was associated with lower self-maturity scores. As lateralization effects were absent, all lesions were superimposed on the same hemisphere (right) for illustrative purposes.

A similar trend was also present in the interaction ‘Hemisphere × Location’ [F(2,42) = 3.062; P = 0.057]: left and right frontal patients obtained equally low scores, whereas right parietal patients obtained slightly lower scores than left, and the contrary was true for temporal patients. Changes in personality profiles were thus not dramatic or evident at the group level of behavioural analysis.

Voxel-based lesion–symptom mapping analysis

Figure 5 and Table 2 show the areas associated with lower self-maturity scores. Patients obtaining lower self-maturity scores had lesions maximally associated with inferolateral prefrontal cortex, in a similar, partially overlapping region with respect to alexithymic patients, but more postero-lateral. No involvement of the temporal lobe was found.

Discussion

The results obtained from TCI questionnaires revealed that 44% of frontal patients obtained a self-maturity score below the cut-off for the presence of personality disorder. In contrast, only a marginal percentage of parietal (18.75%) and temporal (15.38%) patients did. Although this difference did not reach significance on the group analysis, VLSM clearly pointed to a circumscribed set of ventrolateral prefrontal areas as mostly associated with lower self-maturity scores.

Similar to the alexithymia questionnaire, good performance on the character scales of the TCI questionnaire requires cognitive evaluation and reflection upon one’s own goals, intentions, habits and behaviours, involving high levels of mentalization. However, while both mentalization tasks were linked exclusively to the frontal lobes, the two regions overlapped only marginally, with self-maturity areas being more postero-lateral compared to those associated with alexithymia. The lack of involvement of medial prefrontal structures might appear surprising, since these areas have also been linked on more than one occasion to the processing and storing of personality traits in general (Gillihan and Farah, 2005; Northoff et al., 2006; Ma et al., 2013). However, what we found, in analysing TCI SELF measures, is not that frontal patients become unable to evaluate themselves (or to remember their personality traits), instead they consistently tend to describe themselves in a specific way (as more immature, impulsive, reactive, unable to delay gratification and lacking empathy), which is the profile of a person affected by a personality disorder (Cloninger et al., 1994). And this is also a frequent complaint of the relatives of people who sustained frontal lobe damage about their actual behaviour. If one takes this perspective in interpreting the behavioural results, the finding of the involvement of more lateral regions within the prefrontal cortex becomes much less surprising. Lateral PFC, indeed, is involved in general in executive or supervisory control, providing a fundamental role in strategy selection, inhibitory control and in general in regulating goal-directed behaviour (Ridderinkhof et al., 2004; Shallice and Cooper, 2011; Shenhav et al., 2013).

General discussion

A first aim of this study was to investigate, for clinical reasons, the changes in emotionality and personality that occur in patients affected by brain tumours and undergoing surgery for the removal of the lesion. Following brain tumours, some patients become emotionally labile, others show blunted affect, some are unable to control their emotional reactions, some report deterioration in social relations or the loss of interests and some report a genuine personality change. As these changes can occur with lesions located in very different parts of the brain, our aim was to try to specify why damage to different brain regions can give rise to social and emotional dysfunctions, by investigating the processes that can be damaged in these situations and the brain regions responsible for their implementation.

We administered four different tasks (emotion recognition of Ekman faces, The Eyes Test, the TAS-20 and the TCI) both immediately before and a few days after operation for the removal of brain tumours in a series of 71 patients with frontal, temporal or parietal lobes lesions. For many of the tasks used (apart probably from emotion recognition), findings from a large population of patients with focal brain lesions are largely lacking in the neuropsychological literature. Moreover, only rarely have these different aspects of emotional/behavioural regulation been investigated simultaneously in neurological populations and when this was done (Henry et al., 2006; Shaw et al., 2007; Broicher et al., 2012) only one or two aspects were investigated (particularly emotion recognition and Theory of Mind) and associations of impairments were found.

In contrast we focused our attention on possible cognitive as well as anatomical dissociations. We hypothesized that from both a cognitive and anatomical point of view, temporo/limbic areas would be critical for processing social information (emotion and intentions) at a more perceptual level, while frontal lobe structures would be more heavily engaged when higher levels of mentalization (i.e. ‘reflection’ about one’s own or other’s stable mental states and feelings) are required, as is suggested also by the most recent neuroimaging findings (Fusar-Poli et al., 2009; Van Overwalle, 2009; Lee and Siegle, 2012; Schilbach et al., 2012).

The results found at the behavioural level, which were made more specific by VLSM analysis, confirmed that tasks requiring more limited mentalization were more affected by temporo/insular lesions (anterior temporal and amygdala for emotion recognition and temporo-parietal junction and insula for the Eyes Test). Complementarily, tasks relying more heavily on mentalization processes were maximally affected by prefrontal lesions (antero-medial prefrontal cortex in the case of alexithymia and ventro-lateral prefrontal cortex in the case of low self-maturity). It is important that the different pattern of behaviours and behavioural dysfunctions observed was found to dissociate within the same sample population (see Fig. 6 for single cases), thus reducing the possibility that any one of the results might be related to some excessive task difficulty (and therefore to some general decrease of cognitive functioning in brain tumour patients), or to lesion volume or any of other possible confounding variables (age, education or aggressiveness of lesion) (Supplementary material).

Figure 6

Examples of patients with dissociating performances among tasks. Cut-offs for clinically relevant scores are Task 1: < 29/36; Task 2: z-score < −1; Task 3: > 51/100; Task 4: < 58/86. Bold text indicates scores outside the cutoff.

Figure 6

Examples of patients with dissociating performances among tasks. Cut-offs for clinically relevant scores are Task 1: < 29/36; Task 2: z-score < −1; Task 3: > 51/100; Task 4: < 58/86. Bold text indicates scores outside the cutoff.

Some of the results indirectly corroborate previous findings on both healthy and neurological populations on similar topics, such as those from the emotion recognition domain, which confirm the importance of anterior temporal structures as well as the amygdala and insula in perceiving and deciphering other people’s emotions. However, our findings did not support strict localizationist accounts of emotion perception being more compatible with a more distributed ‘constructionist’ view.

The findings from the Eyes Test, on the other hand, the first on a large sample of patients, provide evidence on the different subcomponents of Theory of Mind; they suggest that certain brain areas which have traditionally been associated with this general type of cognitive process (medial PFC and anterior cingulate cortex) might not be required for more ‘basic’ Theory of Mind tasks, which would rely on brain structures more involved in the ‘on-line’ inference of intention from perceivable external cues.

Anatomical results related to alexithymia have not previously been reported on a large sample of brain damaged patients and allow us to specify more precisely the brain regions involved in the process of mentalizing (reflecting and abstracting) emotional states. Again these show an anatomical correspondence with findings from neuroimaging studies on healthy populations. Critically, they suggest that alexithymia can also be an acquired deficit and is not just a ‘cognitive style’. Furthermore, while changes in personality have been very often observed following frontal lobe damage, we have been able to ‘quantify’ this change to some degree, by showing that lower self-maturity and therefore a higher risk of personality disorder seems linked to lateral prefrontal lesions. The findings also suggest that anterior cingulate cortex is an important structure for mentalizing and reflecting upon emotions but not necessarily when abstracting and evaluating goals and needs.

A potential limitation, intrinsic to the use of self-report measures, relates to the possibility that the personality changes detected might only reflect the use of a self-report scale, and not actual changes in everyday life. Thus, prefrontal patients might just be giving a picture of themselves as ‘non socially-desirable’, while this does not accurately characterize their actual behaviour. The present findings cannot unambiguously tell whether the reported self-descriptions are also reflected in the everyday life behaviour of such patients or just reflect a ‘pure’ deficit in mentalizing and abstracting their behaviour (which could instead be normal in everyday life), and this needs to be clarified in future investigations. However in the light of the frequent complaints of the relatives of people suffering from frontal damage about similar changes in the actual behaviour of the patients, the hypothesis that the self-reports are consistent with actual behaviour seems plausible.

Finally from a clinical methods point of view, these results give further support to the validity of using brain tumour patients as a population for reliably studying and localizing brain functions, since the reported data give largely convergent evidence with respect to those found in other, previously investigated, neurological and healthy populations (Van Overwalle, 2009; Shallice et al., 2010; Karnath and Steinbach, 2011; Shallice and Skrap, 2011).

Funding

F.C. was supported by a Post-Doctoral research fellowship from A.O.U.S. Maria della Misericordia “Progetto a valenza regionale di ricerca clinica e di base per utilizzo Tomografo a Risonanza Magnetica ad alto campo (3Tesla)”. Regional basic and clinincal research project for the use of High Field Magnetic Resonance Tomograph (3 tesla).

Supplementary material

Supplementary material is available at Brain online.

Abbreviations

    Abbreviations
  • PFC

    prefrontal cortex

  • TAS-20

    Toronto Aleithymia Scale (20 items)

  • TCI

    Temperament and Character Inventory

  • VLSM

    voxel-based lesion–symptom mapping

References

Adolphs
R
Tranel
D
Damasio
H
Damasio
A
Impaired recognition of emotion in facial expressions following bilateral damage to the human amygdala
Nature
 , 
1994
, vol. 
372
 (pg. 
669
-
72
)
Allison
T
Puce
A
McCarthy
G
Social perception from visual cues: role of the STS region
Trends Cogn Sci
 , 
2000
, vol. 
4
 (pg. 
267
-
78
)
Bagby
RM
Parker
JD
Taylor
GJ
The twenty-item Toronto Alexithymia Scale–I. Item selection and cross-validation of the factor structure
J Psychosom Res
 , 
1994
, vol. 
38
 (pg. 
23
-
32
)
Baron-Cohen
S
Wheelwright
S
Hill
J
Raste
Y
Plumb
I
The “Reading the Mind in the Eyes” Test revised version: a study with normal adults, and adults with Asperger syndrome or high-functioning autism
J Child Psychol Psychiatry
 , 
2001
, vol. 
42
 (pg. 
241
-
51
)
Barrash
J
Tranel
D
Anderson
SW
Acquired personality disturbances associated with bilateral damage to the ventromedial prefrontal region
Dev Neuropsychol
 , 
2000
, vol. 
18
 (pg. 
355
-
81
)
Bates
E
Wilson
SM
Saygin
AP
Dick
F
Sereno
MI
Knight
RT
, et al.  . 
Voxel-based lesion-symptom mapping
Nat Neurosci
 , 
2003
, vol. 
6
 (pg. 
448
-
50
)
Battaglia
M
Bajo
S
Temperament and character inventory
Repertorio delle scale di valutazione in psichiatria
 , 
2000
Firenze
SEE
(pg. 
1375
-
88
)
Becerra
R
Amos
A
Jongenelis
S
Organic alexithymia: a study of acquired emotional blindness
Brain Inj
 , 
2002
, vol. 
16
 (pg. 
633
-
45
)
Bressi
C
Taylor
G
Parker
J
Bressi
S
Brambilla
V
Aguglia
E
, et al.  . 
Cross validation of the factor structure of the 20-item Toronto Alexithymia Scale: an Italian multicenter study
J Psychosom Res
 , 
1996
, vol. 
41
 (pg. 
551
-
9
)
Broicher
SD
Kuchukhidze
G
Grunwald
T
Kramer
G
Kurthen
M
Jokeit
H
Tell me how do I feel: emotion recognition and theory of mind in symptomatic mesial temporal lobe epilepsy
Neuropsychologia
 , 
2012
, vol. 
50
 (pg. 
118
-
28
)
Calder
AJ
Keane
J
Manes
F
Antoun
N
Young
AW
Impaired recognition and experience of disgust following brain injury
Nat Neurosci
 , 
2000
, vol. 
3
 (pg. 
1077
-
8
)
Calder
AJ
Lawrence
AD
Keane
J
Scott
SK
Owen
AM
Christoffels
I
, et al.  . 
Reading the mind from eye gaze
Neuropsychologia
 , 
2002
, vol. 
40
 (pg. 
1129
-
38
)
Carrington
SJ
Bailey
AJ
Are there theory of mind regions in the brain? A review of the neuroimaging literature
Hum Brain Mapp
 , 
2009
, vol. 
30
 (pg. 
2313
-
35
)
Cloninger
CR
Przybeck
T
Svrakic
D
The Temperament and Character Inventory (TCI): a guide to its development and use
In: Center for Psychobiology of Personality
 , 
1994
St. Louis, MO
Washington University
Cotterill
RM
On the unity of conscious experience
J Conscious Stud
 , 
1995
, vol. 
2
 (pg. 
290
-
311
)
Craig
AD
How do you feel–now? The anterior insula and human awareness
Nat Rev Neurosci
 , 
2009
, vol. 
10
 (pg. 
59
-
70
)
 
Ekman P, Rosenberg E. What the face reveals: basic and applied studies of spontaneous expression using the Facial Action Coding System (FACS). In: Ekman P, Rosenberg E, editors. New York, NY: Oxford University Press; 1997
Ekman
P
Dalgleish
T
Basic emotions
Handbook of cognition and emotion
 , 
1999
Chichester, UK
Wiley
(pg. 
5
-
60
)
Fonagy
P
Gergely
G
Jurist
EL
Affect regulation, mentalization and the development of the self
London: Karnac Books
 , 
2003
Fricchione
G
Howanitz
E
Aprosodia and alexithymia: a case report
Psychother Psychosom
 , 
2010
, vol. 
43
 (pg. 
156
-
60
)
Frith
CD
Friston
K
Liddle
PF
Frackowiak
RS
Willed action and the prefrontal cortex in man: a study with PET
Proc Biol Sci
 , 
1991
, vol. 
244
 (pg. 
241
-
6
)
Frith
CD
Frith
U
Interacting minds–a biological basis
Science
 , 
1999
, vol. 
286
 (pg. 
1692
-
5
)
Fusar-Poli
P
Placentino
A
Carletti
F
Landi
P
Allen
P
Surguladze
S
, et al.  . 
Functional atlas of emotional faces processing: a voxel-based meta-analysis of 105 functional magnetic resonance imaging studies
J Psychiatry Neurosci
 , 
2009
, vol. 
34
 pg. 
418
 
Gallagher
HL
Frith
CD
Functional imaging of ‘theory of mind’
Trends Cogn Sci
 , 
2003
, vol. 
7
 (pg. 
77
-
83
)
Gillihan
SJ
Farah
MJ
Is self special? A critical review of evidence from experimental psychology and cognitive neuroscience
Psychol Bull
 , 
2005
, vol. 
131
 (pg. 
76
-
97
)
Gundel
H
Lopez-Sala
A
Ceballos-Baumann
AO
Deus
J
Cardoner
N
Marten-Mittag
B
, et al.  . 
Alexithymia correlates with the size of the right anterior cingulate
Psychosom Med
 , 
2004
, vol. 
66
 (pg. 
132
-
40
)
Heberlein
AS
Padon
AA
Gillihan
SJ
Farah
MJ
Fellows
LK
Ventromedial frontal lobe plays a critical role in facial emotion recognition
J Cogn Neurosci
 , 
2008
, vol. 
20
 (pg. 
721
-
33
)
Henry
JD
Phillips
LH
Crawford
JR
Ietswaart
M
Summers
F
Theory of mind following traumatic brain injury: the role of emotion recognition and executive dysfunction
Neuropsychologia
 , 
2006
, vol. 
44
 (pg. 
1623
-
8
)
Hoffman
EA
Haxby
JV
Distinct representations of eye gaze and identity in the distributed human neural system for face perception
Nat Neurosci
 , 
2000
, vol. 
3
 (pg. 
80
-
4
)
Holland
PC
Gallagher
M
Amygdala circuitry in attentional and representational processes
Trends Cogn Sci
 , 
1999
, vol. 
3
 (pg. 
65
-
73
)
Hornak
J
Bramham
J
Rolls
ET
Morris
RG
O'Doherty
J
Bullock
PR
, et al.  . 
Changes in emotion after circumscribed surgical lesions of the orbitofrontal and cingulate cortices
Brain
 , 
2003
, vol. 
126
 (pg. 
1691
-
712
)
Ihme
K
Dannlowski
U
Lichev
V
Stuhrmann
A
Grotegerd
D
Rosenberg
N
, et al.  . 
Alexithymia is related to differences in gray matter volume: a voxel-based morphometry study
Brain Res
 , 
2013
, vol. 
1491
 (pg. 
60
-
7
)
Ius
T
Isola
M
Budai
R
Pauletto
G
Tomasino
B
Fadiga
L
, et al.  . 
Low-grade glioma surgery in eloquent areas: volumetric analysis of extent of resection and its impact on overall survival. A single-institution experience in 190 patients: clinical article
J Neurosurg
 , 
2012
, vol. 
117
 (pg. 
1039
-
52
)
Jack
AI
Shallice
T
Introspective physicalism as an approach to the science of consciousness
Cognition
 , 
2001
, vol. 
79
 (pg. 
161
-
96
)
Karnath
HO
Steinbach
JP
Do brain tumours allow valid conclusions on the localisation of human brain functions?–Objections
Cortex
 , 
2011
, vol. 
47
 (pg. 
1004
-
6
)
Kober
H
Barrett
LF
Joseph
J
Bliss-Moreau
E
Lindquist
K
Wager
TD
Functional grouping and cortical−subcortical interactions in emotion: a meta-analysis of neuroimaging studies
Neuroimage
 , 
2008
, vol. 
42
 pg. 
998
 
Lane
RD
Ahern
GL
Schwartz
GE
Kaszniak
AW
Is alexithymia the emotional equivalent of blindsight?
Biol Psychiatry
 , 
1997
, vol. 
42
 (pg. 
834
-
44
)
Larsen
JK
Brand
N
Bermond
B
Hijman
R
Cognitive and emotional characteristics of alexithymia: a review of neurobiological studies
J Psychosom Res
 , 
2003
, vol. 
54
 (pg. 
533
-
41
)
Lee
KH
Siegle
GJ
Common and distinct brain networks underlying explicit emotional evaluation: a meta-analytic study
Soc Cogn Affect Neurosci
 , 
2012
, vol. 
7
 (pg. 
521
-
34
)
Lindquist
KA
Gendron
M
Barrett
LF
Dickerson
BC
Emotion perception, but not affect perception, is impaired with semantic memory loss
Emotion
 , 
2014
, vol. 
14
 (pg. 
375
-
87
)
Lindquist
KA
Wager
TD
Kober
H
Bliss-Moreau
E
Barrett
LF
The brain basis of emotion: a meta-analytic review
Behav Brain Sci
 , 
2012
, vol. 
35
 (pg. 
121
-
43
)
Ma
N
Baetens
K
Vandekerckhove
M
Kestemont
J
Fias
W
Van Overwalle
F
Traits are represented in the medial prefrontal cortex: an fMRI adaptation study
Soc Cogn Affect Neurosci
 , 
2013
 
Advance Access published on June 18, 2013, DOI: 10.1093/scan/nst098
Martin-Rodriguez
JF
Leon-Carrion
J
Theory of mind deficits in patients with acquired brain injury: a quantitative review
Neuropsychologia
 , 
2010
, vol. 
48
 (pg. 
1181
-
91
)
McNamara
P
Durso
R
Harris
E
“Machiavellianism” and frontal dysfunction: evidence from Parkinson's disease
Cogn Neuropsychiatry
 , 
2007
, vol. 
12
 (pg. 
285
-
300
)
Milner
B
Petrides
M
Behavioural effects of frontal-lobe lesions in man
Trends Neurosci
 , 
1984
, vol. 
7
 (pg. 
403
-
7
)
Moriguchi
Y
Ohnishi
T
Lane
RD
Maeda
M
Mori
T
Nemoto
K
, et al.  . 
Impaired self-awareness and theory of mind: an fMRI study of mentalizing in alexithymia
Neuroimage
 , 
2006
, vol. 
32
 (pg. 
1472
-
82
)
Northoff
G
Heinzel
A
de
GM
Bermpohl
F
Dobrowolny
H
Panksepp
J
Self-referential processing in our brain–a meta-analysis of imaging studies on the self
Neuroimage
 , 
2006
, vol. 
31
 (pg. 
440
-
57
)
Perner
J
Lang
B
Development of theory of mind and executive control
Trends Cogn Sci
 , 
1999
, vol. 
3
 (pg. 
337
-
44
)
Posner
MI
Rothbart
MK
Attention, selfΓÇôregulation and consciousness
Philos Trans R Soc Lond S B Biol Sci
 , 
1998
, vol. 
353
 (pg. 
1915
-
27
)
Premack
D
Woodruff
G
Chimpanzee problem-solving: a test for comprehension
Science
 , 
1978
, vol. 
202
 (pg. 
532
-
5
)
Rankin
KP
Gorno-Tempini
ML
Allison
SC
Stanley
CM
Glenn
S
Weiner
MW
, et al.  . 
Structural anatomy of empathy in neurodegenerative disease
Brain
 , 
2006
, vol. 
129
 (pg. 
2945
-
56
)
Ridderinkhof
KR
van den Wildenberg
WP
Segalowitz
SJ
Carter
CS
Neurocognitive mechanisms of cognitive control: the role of prefrontal cortex in action selection, response inhibition, performance monitoring, and reward-based learning
Brain Cogn
 , 
2004
, vol. 
56
 (pg. 
129
-
40
)
Rorden
C
Karnath
HO
Bonilha
L
Improving lesion-symptom mapping
J Cogn Neurosci
 , 
2007
, vol. 
19
 (pg. 
1081
-
8
)
Rosen
HJ
Pace-Savitsky
K
Perry
RJ
Kramer
JH
Miller
BL
Levenson
RW
Recognition of emotion in the frontal and temporal variants of frontotemporal dementia
Dement Geriatr Cogn Disord
 , 
2004
, vol. 
17
 (pg. 
277
-
81
)
Samson
D
Apperly
IA
Chiavarino
C
Humphreys
GW
Left temporoparietal junction is necessary for representing someone else's belief
Nat Neurosci
 , 
2004
, vol. 
7
 (pg. 
499
-
500
)
Samson
D
Michel
C
Baron-Cohen
S
Lombardo
M
Tager-Flusberg
H
Cohen
D
Theory of mind: insights from patients with acquired brain damage
Understanding other minds: perspectives from developmental social neuroscience
 , 
2013
Oxford University Press
pg. 
164
 
Satpute
AB
Shu
J
Weber
J
Roy
M
Ochsner
KN
The functional neural architecture of self-reports of affective experience
Biol Psychiatry
 , 
2013
, vol. 
73
 (pg. 
631
-
8
)
Saxe
R
Kanwisher
N
People thinking about thinking people. The role of the temporo-parietal junction in “theory of mind”
Neuroimage
 , 
2003
, vol. 
19
 (pg. 
1835
-
42
)
Schafer
R
Popp
K
Jurgens
S
Lindenberg
R
Franz
M
Seitz
RJ
Alexithymia-like disorder in right anterior cingulate infarction
Neurocase
 , 
2007
, vol. 
13
 (pg. 
201
-
8
)
Schilbach
L
Bzdok
D
Timmermans
B
Fox
PT
Laird
AR
Vogeley
K
, et al.  . 
Introspective minds: using ALE meta-analyses to study commonalities in the neural correlates of emotional processing, social & unconstrained cognition
PLoS One
 , 
2012
, vol. 
7
 pg. 
e30920
 
Schmolck
H
Squire
LR
Impaired perception of facial emotions following bilateral damage to the anterior temporal lobe
Neuropsychology
 , 
2001
, vol. 
15
 (pg. 
30
-
8
)
Shallice
T
Cooper
R
The organisation of mind
2011
Oxford University Press
Shallice
T
Mussoni
A
D'Agostini
S
Skrap
M
Right posterior cortical functions in a tumour patient series
Cortex
 , 
2010
, vol. 
46
 (pg. 
1178
-
88
)
Shallice
T
Skrap
M
Localisation through operation for brain tumour: a reply to Karnath and Steinbach
Cortex
 , 
2011
, vol. 
47
 (pg. 
1007
-
9
)
Shaw
P
Lawrence
E
Bramham
J
Brierley
B
Radbourne
C
David
AS
A prospective study of the effects of anterior temporal lobectomy on emotion recognition and theory of mind
Neuropsychologia
 , 
2007
, vol. 
45
 (pg. 
2783
-
90
)
Shenhav
A
Botvinick
MM
Cohen
JD
The expected value of control: an integrative theory of anterior cingulate cortex function
Neuron
 , 
2013
, vol. 
79
 (pg. 
217
-
40
)
Sifneos
PE
The prevalence of ‘alexithymic’ characteristics in psychosomatic patients
Psychother Psychosom
 , 
1973
, vol. 
22
 (pg. 
255
-
62
)
Spalletta
G
Pasini
A
Costa
A
De Angelis
D
Ramundo
N
Paolucci
S
, et al.  . 
Alexithymic features in stroke: effects of laterality and gender
Psychosom Med
 , 
2001
, vol. 
63
 (pg. 
944
-
50
)
Stuss
DT
Anderson
V
The frontal lobes and theory of mind: developmental concepts from adult focal lesion research
Brain Cogn
 , 
2004
, vol. 
55
 (pg. 
69
-
83
)
Stuss
DT
Gow
CA
Hetherington
CR
“No longer Gage”: frontal lobe dysfunction and emotional changes
J Consult Clin Psychol
 , 
1992
, vol. 
60
 pg. 
349
 
Stuss
DT
Murphy
KJ
Binns
MA
Alexander
MP
Staying on the job: the frontal lobes control individual performance variability
Brain
 , 
2003
, vol. 
126
 (pg. 
2363
-
80
)
Van Overwalle
F
Social cognition and the brain: a meta−analysis
Hum Brain Mapp
 , 
2009
, vol. 
30
 (pg. 
829
-
58
)