Introduction
The use of awake surgery for removal of brain tumours within the eloquent area can improve the quality of tumour resection and lead to longer survival (Reference Duffau, Lopes and Arthuis1–Reference Shinoura, Yoshida and Yamada3). However, patients can experience significant stress during awake surgery, although the elevation in blood pressure during awake surgery might be related to withdrawal of anaesthetic as well as intra-operative stress (Reference Conte, Guzzetti, Porta, Tobaldini, Baratta and Stocchetti4).
The amygdala plays a critical role in coping with stress (Reference Rauch, Whalen and Shin5,Reference Herpertz, Dietrich and Wenning6). For example, the amygdala is activated in patients with post-traumatic stress disorder (PTSD) and borderline personality disorder (BPD). The left and right amygdala have different roles in emotion processing (Reference Baas, Aleman and Kahn7), and although the amygdala has been implicated in stress-induced aggression or escape behaviour, it is not clear which side of the amygdala induces aggression or escape (Reference Machado, Kazama and Bachevalier8,Reference New, Hazlett and Newmark9). Thus, the goal of the present study was to evaluate whether the left or right amygdala was involved in aggressive or escape behaviour in patients undergoing awake surgery for temporal lobe brain tumours.
Patients and methods
Patients
This study included six patients who underwent awake surgery between 2007 and 2009 at our hospital. Brain tumours were primarily located in the left temporal lobe in three patients (cases 1–3) and in the right temporal lobe in the other three patients (cases 4–6). Patient data are summarised in Table 1. The brain tumour invaded into the left amygdala in cases 1 and 2 (arrow in Fig. 1a, b) and was near the left amygdala in case 3 (Fig. 1c). Similarly, the brain tumour invaded into the right amygdala in cases 4 and 5 (arrow in Fig. 2a, b), and was near the right amygdala in case 6 (Fig. 2c). Aphasia was present in cases 1 and 2 but not in cases 4–6, suggesting that the left hemisphere was dominant in those patients. In case 3, propofol injection to the right or left carotid arteries was used to show that the right hemisphere was dominant. Notably, as the degree of aphasia in cases 1 and 2 was mild, patients were still able to comply with intra-operative language testing and other cognitive tests. Informed consent to perform awake surgery was obtained from all patients prior to surgery and surgery was performed in accordance with current ethical standards of our hospital and the guidelines of the Japanese Awake Surgery Association.
Table 1 Patient data
GA, general anaesthesia; GTR, gross total removal; Lt., left; NC, not clear; ND, not determined; PR, partial removal; Rt., right; +, amygdala was definitely invaded by the tumour or disease.
Fig. 1. Axial MRI with contrast enhancement before surgery in cases 1–3. Axial MRI with contrast enhancement [panel (a), case 1; panel (b), case 2; panel (c), case 3] shows a brain lesion of unknown etiology located in the left temporal lobe, including the amygdala (arrow) in case 1; glioblastoma in the left temporal lobe, including the amygdala (arrow) in case 2 and glioblastoma in the left temporal lobe in case 3.
Fig. 2. Axial MRI with contrast enhancement before surgery in cases 4–6. Axial MRI with contrast enhancement [panel (a), case 4; panel (b), case 5; panel (c), case 6] shows glioblastoma located in the right temporal lobe, including the amygdala (arrow) in case 4; glioblastoma in the right temporal lobe, including the amygdala (arrow) in case 5 and glioblastoma in the right temporal lobe in case 6.
Tumour resection
In all cases, awake surgery was performed as described previously (Reference Shinoura, Yamada, Kodama, Suzuki, Takahashi and Yagi10). Briefly, patients were positioned in the supine lateral position with rigid head fixation (Sugita headrest; Mizuho Medical, Tokyo, Japan) after administration of local anaesthetic agents (1% xylocaine and 0.75% anapain) at pin sites and regional field block sites. Under intravenous anaesthesia with propofol and remifentanil, the skin was infiltrated with the same local anaesthetic agent and incised, and neuronavigated craniotomy and incision of the dura were performed. Following cessation of intravenous anaesthesia, the patient woke and a modified Ojemann stimulator was used for cortical mapping (Reference Berger, Kincaid, Ojemann and Lettich11). To avoid inducing intra-operative seizures, a low-stimulus setting (3–5 mA, 60-Hz biphasic square wave pulse of 1 ms/phase for 4 s duration) was used. All patients were continuously observed by a neurosurgeon and a neuropsychologist. Testing included object naming, repetition, auditory comprehension (for patients with left temporal lesions) and memory and face recognition (for patients with right temporal lesions); all results were reported to the surgeon. Corticotomy was performed, avoiding sites that might induce aphasia by cortical mapping in the left temporal lobe. The tumour was removed in the usual fashion. Adequacy of language function (for patients with left temporal lesions) and memory and face recognition (for patients with right temporal lesions) was continuously assessed during tumour removal (Reference Shinoura, Yamada, Kodama, Suzuki, Takahashi and Yagi10). Tumour removal was assisted by neuronavigation. Following completion of tumour resection, intravenous anaesthesia was administered using propofol. After closure of the dura, the bone flap was replaced and the skin was closed in the usual manner.
In cases 3 and 6, awake surgery was terminated because of the development of aggressive behaviour and tumour haemorrhage, respectively, and both procedures were ultimately completed under general anaesthesia. In case 5, awake surgery was terminated because of the development of escape behaviour and the procedure was ultimately completed under intravenous anaesthesia with propofol and remifentanil.
Neuropsychological evaluation
Aggressive or escape behaviour was assessed by two neurosurgeons and two neuropsychologists during awake surgery. Symptoms before and after surgery were evaluated to assess for the tendency for aggression or escape behaviours. These behaviours were assessed by observation alone, because any change in emotion during awake surgery was difficult to evaluate by any mechanism other than observation. In addition to the observation of emotional state, neuropsychologists continuously evaluated language and cognitive function during awake surgery.
Results
In case 1, the patient showed aggressive behaviour before partial removal of the left amygdala during awake surgery. He spoke with an irritable voice, complained of urinary urgency and attempted to move. Just after partial removal of the left amygdala for biopsy, he showed modest behaviour, apologised for his previous rude behaviour, expressed gratitude for the doctor's effort and spoke with quiet voice. After surgery, he complained of negative mood as compared to his pre-operative mood.
In case 2, the patient showed aggressive behaviour when we removed the tumour near the left amygdala. She suddenly spoke with an irritable voice and enquired whether the surgeons were killing her. During removal of other lesions, including a lesion near the left thalamus, she spoke with quiet voice or showed aphasia.
In case 3, the patient showed aggressive behaviour upon awakening during surgery. His aggressive and violent behaviour precluded continuation of awake surgery and general anaesthesia was instituted to complete the procedure.
In case 4, the patient showed escape behaviour during initiation of tumour resection (located near the right amygdala) and complained of profound sleepiness during tumour removal. This complaint of sudden onset of sleepiness was similar to the symptoms he experienced when performing work activities over a 3-month period before surgery.
In case 5, the patient showed escape behaviour during initiation of tumour resection (located in the temporal lobe). He did not speak, but tried to stand up on the operating table. This behaviour precluded continuation of awake surgery and intravenous anaesthesia was instituted to complete the procedure.
In case 6, the patient showed escape behaviour during initiation of tumour resection (located in the temporal lobe). She quietly stated, ‘I want to go home'. Subsequent bleeding from the tumour necessitated cessation of awake surgery. The procedure was ultimately completed under general anaesthesia.
Discussion
Awake surgery allows direct demonstration of neuroanatomical localisation of function by observing changes in neurological function in response to surgical manipulation of specific brain tissue. The present study described six cases in which the patients showed aggressive or escape behaviour during awake surgery. In case 1, the patient showed aggressive behaviour before partial removal of the left amygdala but not just after partial removal of the left amygdala. In case 2, the patient showed aggressive behaviour during removal of the tumour near the left amygdala. In case 3, the patient showed aggressive behaviour upon awakening during surgery. The tumour had invaded into the left amygdala in cases 1 and 2 and was near the left amygdala in case 3. In case 4, the patient showed escape behaviour, namely profound sleepiness, when we started to remove the tumour near the right amygdala. In cases 5 and 6, the patient showed escape behaviour after awakening during surgery and during removal of the tumour in temporal lobe. The tumour had invaded into the right amygdala in cases 4 and 5 and was near the right amygdala in case 6. These observations suggest that the left and right temporal lesions are involved in aggression or escape, respectively. Moreover, behavioural changes of aggression or escape occurred during removal of the tumour within the amygdala (case 1) or near the amygdala (cases 2, 4, 5 and 6), indicating that the amygdala on the respective side may play a role in these behaviours.
Of note, in all six cases, it appeared as though the amygdala was not completely invaded by the tumour and that normal nerve cells were still present within the amygdala. This is corroborated by the fact that only a small region of the amygdala was enhanced on MRI (Figs 1 and 2). Furthermore, in all cases in which tumour within or near the amygdala was removed, patients displayed aggression or escape behaviour during awake surgery. In contrast, when tumours that were not located within or near amygdala (i.e. those located in the temporal) were removed, patients did not display such behaviour during awake surgery. These observations suggest that aggression or escape behaviour may be mediated by the amygdala. Indeed, we have never encountered these behaviours during awake surgery when removing tumour located in other brain regions (with the exception of 2 of 140 cases in which tumours were removed from the cingulate cortex). Regardless, fMRI paradigms that target the amygdala may be of use in characterising the functionality of the amygdala before surgery and further investigation would be of benefit to evaluate the function of amygdala in cases with brain tumours within or near amygdala.
The amygdala is an integral region that mediates various emotions, including anger and fear, and is controlled by the ventromedial prefrontal cortex (Reference Damasio, Grabowski, Frank, Galaburda and Damasio12–Reference Anderson, Barrash, Bechara and Tranel14). Klüver–Bucy syndrome is a neurobehavioural syndrome associated with bilateral lesions in the anterior temporal horn or amygdala of rhesus monkeys. These lesions result in attenuation of emotion, with dulled facial movements and vocalisations, as well as loss of normal fear responses (Reference Kluver and Bucy15). Other studies have showed that the amygdala is also involved in fear conditioning in rats (Reference Hitchcock and Davis16) and in emotional behaviour in humans (Reference Tranel and Hyman17,Reference Buchanan, Tranel and Adolphs18) and that intracerebral electrical stimulation of the temporal lobe induced fear-related behaviours in most patients with drug-resistant epilepsy (Reference Meletti, Tassi, Mai, Fini, Tassinari and Russo19). Indeed, post-operative mood and anxiety disorders are more common in fear aura patients after temporal lobe epilepsy, suggesting that the amygdala is involved in fear conditioning (Reference Kohler, Carran, Bilker, O'Connor and Sperling20). Elevated activation in the amygdala is observed in BPD patients who show excessive aggression (Reference Herpertz, Dietrich and Wenning6,Reference New, Hazlett and Newmark9,Reference Donegan, Sanislow and Blumberg21) and a significant proportion (28%) of aggressive patients with temporal lobe epilepsy (28%) have temporal lesions affecting either the amygdala or periamygdaloid structures (Reference van Elst, Woermann, Lemieux, Thompson and Trimble22). In another report, patients showed significant improvement in their aggressiveness after amygdalo-hippocampectomy (Reference Sachdev, Smith, Matheson, Last and Blumbergs23). These data are consistent with observations from the present study that surgery for brain lesions within or near the amygdala resulted in disturbances in emotion. Of note, there is no published literature that describes removal of tumours within or near the relatively intact amygdala. This may account for the absence of reports of aggression or escape behaviour during awake surgery, as damage to the intact amygdala likely mediates these behaviours.
The amygdala also plays a critical role in escape behaviours in monkeys exposed to life-threatening stimuli (Reference Machado, Kazama and Bachevalier8). Hyperreactivity in the amygdala is observed in PTSD patients, who show excessive escape behaviours in response to stressors (Reference Rauch, Whalen and Shin5). In animal studies, the amygdala is involved in negative emotion in general and also with aggressive behaviour. Indeed, neurotoxic damage to the amygdala results in reduced fear and aggression, increased submission and more affiliative social interactions in monkeys (Reference Meunier, Bachevalier, Murray, Malkova and Mishkin24,Reference Machado, Emery, Capitanio, Mason, Mendoza and Amaral25). These reports indicate that the amygdala is deeply involved in aggression and escape behaviour.
The functions of the right and left amygdala may differ from one another. Patients with left amygdala damage have impaired memory for emotional stimuli (Reference Adolphs, Tranel and Denburg26), whereas BPD patients have significantly greater left amygdala activation in response to facial expressions of emotions (Reference Donegan, Sanislow and Blumberg21). Furthermore, the right amygdala is preferentially involved in several processes related to the expressions of PTSD symptoms (Reference Smith, Abou-Khalil and Zald27) and may be a part of a neural circuit representing facial expressions of withdrawal (Reference Anderson, Spencer, Fulbright and Phelps28). In one study, suicidal patients had larger right amygdala volumes than non-suicidal female patients with major depressive disorder (Reference Monkul, Hatch and Nicoletti29), while another investigator reported that aversive gustatory stimuli activated the right amygdala (Reference Zald, Lee, Fluegel and Pardo30). In animal studies, rats given lidocaine into the right amygdala have impaired memory for inhibitory avoidance training, indicating that the right amygdala may be important for the expression of memory for aversively motivated training (Reference Coleman-Mesches and McGaugh31). Furthermore, potentiation in the left amygdala afferents and efferents predicts anxiolytic-like effects, while potentiation in the right amygdala afferents predicts anxiogenic-like effects (Reference Adamec, Blundell and Collins32). These data indicate that the left amygdala is related to aggression, whereas the right amygdala is related to escape, which is consistent with observations from the present study.
When performing surgical resection of brain tumour, preservation of neurological functions, including language function in the left temporal lobe and memory function in the right temporal lobe, is critical for optimal patient outcomes. However, aggressive and escape behaviours interfered with our ability to resect brain tumours in the left and right temporal lobes, respectively, during awake surgery. As some of these behaviours occurred upon awakening from anaesthesia during surgery, avoidance of deep anaesthesia is a possible approach for reducing the incidence of these behaviours. Regardless, further investigation into potential approaches to reduce these behaviours during awake surgery would be of benefit.
In conclusion, these results suggest that the left or right temporal lesion might induce aggressive or escape behaviour during awake surgery, respectively, and that the amygdala on the respective side may play a role in these behaviours.
Acknowledgement
This work was supported by the Tokyo Metropolitan Hospital Foundation.
