ضایعه آمیگدال در افراد صرع لوب تمپورال حالات عاطفی چهره مرتبط با یادگیری را دچار اختلال می کند

Abstract To investigate the role of unilateral amygdala lesions on processing emotions, 22 drug-resistant temporal lobe epilepsy (TLE) subjects (12 with left-sided and ten with right-sided focus) were tested, after anterior temporal lobectomy or selective amygdalo-hippocampectomy on two associative learning tasks containing emotional and neutral facial expressions, respectively. Volumetric lesion analysis was performed on the basis of 3-D MR images. No effects of lesion side were found in TLE subjects. Taken the extent of amygdala damage into account, an interaction effect could be shown between task (learning of neutral facial expressions versus emotional facial expressions) and group (subjects with little versus considerable amygdala damage), indicating worse performance of subjects with considerable amygdala damage in learning emotional facial expressions. Subjects with considerable amygdala damage were also significantly impaired in learning emotional facial expressions when compared with control subjects.

. Introduction Looking for an anatomical correlate of emotion, animal research has pointed to the amygdala as a core structure [16]. The probably best known phenomenon of bilateral amygdala damage in non-human primates is the so-called ‘Klüver-Bucy syndrome’, producing an extraordinary loss of affective and social behaviour [28]. Anatomical studies have shown that the amygdala has numerous cortical and subcortical connections, receiving extensive convergent sensory input, and projecting it back not only upon unimodal, but also to association areas of the cortex, back-projections being most impressive for the visual system. The consequences of its bilateral destruction are discussed both as a disconnection syndrome, and as a disruption in forming stimulus-reward associations [2]. In humans, selective bilateral amygdala damage is extremely rare, and studies mostly rely on single cases. It does not seem to produce the same serious consequences as in non-human primates, but may disturb conditioning of autonomic responses [5], impair memory for emotional material [8], impair auditory recognition of fear and anger [21], and modulate social and emotional behaviour [25]. A number of studies, concentrating on the recognition of emotional facial expressions in humans with bilateral amygdala damage, have also found deficits, especially in recognising fear (e.g. [1] and [9]). Thus, bilateral damage to the amygdala in humans may affect processing of emotional stimuli of different sensory qualities. Unilateral amygdala damage is more common than bilateral amygdala damage in humans, but studies on its consequences regarding processing of emotions are less frequent. Electrophysiological studies on temporal lobe epilepsy subjects with implanted electrodes showed signals from the amygdala, when subjects reported affective experiences evoked by seizure discharge or electrical stimulation [13]. Single case studies with restricted unilateral amygdala damage indicated an impairment in processing emotional material [4] and [29]. While studies based on amygdala-damaged subjects with additional damage to related cortical and subcortical regions found impaired fear conditioning [15], they failed to find an impairment in processing emotional facial expressions [1]. Taken together, there is strong evidence that bilateral amygdala damage in humans has consequences on processing emotional stimuli, while the effect of unilateral amygdala damage on emotion processing is less clear. Taking functional studies (PET, fMRI) with healthy subjects into account, both bilateral [7], and unilateral [18] amygdala activation could be found, while subjects were processing emotional stimuli. The aim of the present study was to examine the influence of unilateral amygdala damage on the processing of emotional facial expressions as opposed to neutral facial expressions. Since the amygdala is closely related to mnemonic functions [2] and [17], and also so far, there are no studies on amygdala damage and memory for emotional facial expressions, two associative learning tasks with emotional and neutral facial expressions were applied. We expected that subjects with major amygdala damage should show deficits in learning associations between emotional facial expressions compared with subjects with minor amygdala damage and control subjects, and compared with the learning of associations of between facial identities.

Results Statistical comparisons showed that groups TLE-A and TLE-a did not differ significantly with respect to any sociodemographic, clinical, or neuropsychological variable. No significant differences were found comparing group performance in recognition of faces (BFRT), or in rating emotional facial expressions by valence and arousal. Both groups scored significantly lower in verbal memory functions and in TMT, Part A compared with control subjects. Performance on visual memory functions of TLE-A was also significantly lower compared with CG (cp. Table 1). The comparison of TLE-a, TLE-A and CG with the Split Plot Test revealed a significant effect for the factor Task (B=19.6, P=0.001), but not for the factor Group (B=3.1, P=0.14). The interaction between Group and Task just failed to reach conventional level of significance (B=4.9, P=0.06). Comparison of TLE-A and CG (U tests) showed no significant difference in the Identity Learning Task (Z=−1.7, P>0.05), but TLE-A scored significantly lower in the Emotion Learning Task (Z=−2.7, P=0.007). Compared with CG, scores of TLE-a did not differ significantly on either task (P>0.10). To investigate patients’ performance closer, another Split Plot Test was done with groups TLE-A and TLE-a, leaving control group data out. Significant effects were obtained for factor Task (B=6.8, P=0.01), and for the interaction of Group and Task (B=5.1, P=0.02), indicating worse performance in TLE-A (cp. Table 2). For intra-group comparison (Wilcoxon's tests), both CG (Z=−4.5, P=0.001) and TLE-A (Z=−2.7, P=0.007) scored significantly better in the Identity Learning Task, than in the Emotion Learning Task. TLE-a performed equally in both tasks (Z=−0.52, P=0.60). Table 2. Performance of all subjects in the experimental tasks Large amygdala lesion Small amygdala lesion Healthy and clinical control group (CG) (TLE-A) (n=16) (TLE-a) (n=6) (n=40) Mean (S.D.) Mean (S.D.) Mean (S.D.) Identity learning task Total score (% correct) 68 (21) 69 (20) 79 (18) Emotion learning task Total score (% correct) 45 (23) and 66 (24) 63 (22) Pairs of emotion task b Pair 1: Anger–happiness 7 (2)c 7 (3) 8 (2)c Pair 2: Sadness–fear 4 (3) 5 (3) 6 (3) Pair 3: Happiness–surprise 5 (3) 8 (2) 6 (3) Pair 4: Disgust–sadness 4 (3)a 8 (3) 7 (2) Pair 5: Surprise–disgust 4 (3)a 7 (3) 6 (3) Pair 6: Fear–anger 5 (3) 7 (2) 6 (3) a Significantly different (U-test; P<0.05) compared with group CG. b Total number of correctly recognised pairs are given. The target facial expression in pair 1–6 of the emotion learning task is underlined. c Significantly different (Wilcoxon's test; P<0.05) compared with all other pairs of the emotion learning task. Table options Like control subjects, TLE-A remembered the happy face better than any other emotional facial expression (Wilcoxon's tests, all P<0.03), and faces expressing disgust or fear were learned worst. Group comparison showed that CG were significantly better than TLE-A recognising the sad (Z=−3.3, P=0.001), and the disgusted (Z=−2.0, P=0.05) face. TLE-a showed a different performance pattern, with a surprised and a sad face being remembered best, and a fearful face being remembered worst. Comparison of CG and TLE-a did not reveal any significant difference recognising a specific emotional facial expression (P>0.10) (cp. Table 2). The comparison of subjects with selective amygdalo-hippocampectomy (sAH; n=6) and subjects with anterior temporal lobectomy (aTL; n=16), in both learning tasks, did not reach significance (Identity Learning Task, Z=−1.1, P=0.25; Emotion Learning Task, Z=0.99, P=0.32). The same was true comparing subjects with left (TLE-L) and right (TLE-R) side lesions (Identity Learning Task, Z=0.07, P=0.95, Emotion Learning Task, Z=0.86, P=0.39).