Epileptic Disorders


The role of the temporal pole in auditory processing Volume 4, supplement 1, Supplement 1, September 2002


See all figures

Working in animals, Masterton [2] conclusively demonstrated that the temporal pole and the amygdala are involved in auditory processing, and it is also known that the geniculo-amygdaloid pathway plays an important functional role in hearing [1-3]. Beginning in 1961, Kimura [4] showed that anterior temporal lobectomy in humans resulted in partial or total abolition of dichotic listening test scores for sounds or messages delivered to the ear contralateral to the operated temporal lobe [4-6]. In 1972, Berlin [7] detected subjective auditory deficits following anterior temporal lobectomy [7-9] but only the investigation of Jacobson [10] included objective electrophysiological data. They studied 12 patients with drug-refractory epilepsy who had been treated by anterior temporal lobectomy: middle-latency but no late auditory evoked potentials were obtained. The Na component was delayed and its amplitude increased which led them to suggest impaired functioning of a cortico-subcortical inhibitory pathway. These authors only acquired middle-latency auditory potentials on three surface channels and did not perform any spatiotemporal mapping [10].

We measured early, middle-latency and late auditory evoked potentials induced by auditory stimulation in 14 patients prior to and following anterior temporal lobectomy carried out for treatment of drug-refractory epilepsy. The aim was to obtain objective criteria to assess auditory processing in these patients.

Materials and methods


The study was carried out on 14 patients (9 men) with a mean age of 36.6 ± 8.2 years (range: 26-52) and included 11 right-handed subjects and 3 left-handed ones. All underwent anterior temporal lobectomy (7 on the right and 7 on the left) for drug-refractory nonlesional mesial temporal lobe epilepsy. All had normal hearing as evaluated by tuning fork tests and none reported any auditory deficit or history of ENT problems. All were examined by MRI prior to and following the operation.

Dichotic listening tests

A dichotic listening test was carried out 3 months after surgery on 9 of the patients (7 left lobectomies and 2 right). Scores were compared with those obtained from a group of 21 control subjects (17 right-handed and 4 left-handed). The test included a verbal part: 30 pairs of words based on different phonemes; 10 pairs of words based on comparable phonemes but semantically quite distinct; and ten pairs of items of 4 phonemes. The other part of the test consisted of 15 pairs of familiar sounds.

Recording of auditory evoked potentials

Subjects were asked to lie down in a quiet room. Signals were recorded using a headset with 19 channels with the electrodes positioned according to the 10-20 international system. Recordings were made on a calibrated Biologic system using a method described elsewhere [11-13].

Electrode impedance below 5 kohm

Middle-latency AEP: reference = the lobe of the stimulated ear; amplification = 50 000; passing band = 3-1 500 Hz; stimulation by clicks down to 90 dBHL and a repeat frequency of 9.2/s; analysis time = 64 ms; mean of 2 000 sample points per series.

Late AEP: reference = the nose; amplification = 50 000; passing band = 1-100 Hz; stimulation by 50 ms toneburst; ascent and descent time = 3 ms; analysis time = 512 ms; mean of 500 sample points per series.

For MLAEPs and LAEPs, two or three mean series were acquired to check the reproducibility of the data. The latency of the Na and Pa components of the MLAEPs, and of the N100 and P200 of the LAEPs was measured on the basis of the stimulation and the amplitudes normalized to the pre-stimulation phase baseline signal.

Statistical methods

The Mann-Whitney test was used to compare the two groups (9 surgical subjects versus 21 controls) after single-factor analysis of variance (controls/patients) to identify trends. The significance threshold was set at P = 0.05.


Dichotic listening tests

The only significant results emerging from the analysis of variance were in those patients who had undergone left lobectomy among whom postoperative performance levels were lower for both 2-phoneme words (P = 0.0203) and 4-phoneme sentences (P = 0.0203); with respect to familiar sounds, epilepsy was associated with a close to significant effect (P = 0.0622). According to the Mann-Whitney test comparing patient and control scores for each different listening test, the patients performed less well than controls with respect to familiar sounds delivered to the right ear (P = 0.0367).

Evoked potentials

In order to make it possible to analyze the data on the basis of ipsilateral and contralateral hemispheres, and stimulation of the ears ipsilateral and contralateral to the lesion, we only included in the patient group those who had undergone lobectomy on the left side (with the evoked potential data from patients who had undergone right lobectomy reorganized in consequence). This was to enhance the robustness of the statistical tests.

Middle-latency auditory evoked potentials

Both Na and Pa components were significantly delayed and the amplitude of the Pa component was lower in patients than in controls; this was true both prior to and following operation (Figure 1). No significant differences were recorded in patients prior to and after operation.

Late auditory evoked potentials

N100 and P200 components were delayed and the amplitudes of both were reduced in patients compared with controls, and this was true both prior to and following operation (Figure 2). In the patients, the amplitude of the N100 component ipsilateral to the lobectomy was increased following operation compared to the preoperative measurement.


Dichotic listening tests were performed, and middle-latency and late auditory evoked potentials recorded in patients who had undergone anterior temporal lobectomy for drug-refractory temporal lobe epilepsy; in all cases the operation had included the amygdala and the hippocampus but excluded the posterior part of T1 and Heschl's gyrus. Statistically significant changes were observed in the latency and amplitude of the Pa component (peaking at around 30 ms) and the later N100 and P200 components: latencies were delayed and amplitudes reduced. These changes relative to control subjects were observed in recordings made both before and after operation. No significant differences were observed in the patients themselves before and after operation apart from a postoperative increase in the N100 component ipsilateral to the lobectomy. This abnormality should be interpreted with caution because it may simply be due to a problem with conduction secondary to the craniotomy. On the other hand, the preexisting abnormalities cannot be attributed to the surgery and they were in fact unexpected. It would seem that they must be linked in some way to the temporal lobe epilepsy. Preoperative MRI showed that all the patients had hippocampal atrophy and abnormal amygdala and hippocampal signals (abnormalities confirmed by pathological analysis of the resected tissue). Such histological abnormalities are well characterized in the literature which contains numerous reports of neuronal loss and gliosis involving the amygdala, the uncus, the hippocampus and the inferomedial temporal lobe [14].

The origins of the Pa, N100 and P200 components are known to be principally Heschl's gyrus and the planum temporale. It might therefore be proposed that the delay in the Pa, N100 and P200 components observed in these patients derives from the amygdala, the hippocampus and possibly the temporal pole which would support the idea that these structures are involved in auditory activation-inhibition in humans, as they are in animals. These structures probably have a major impact on activity in the auditory cortex.

Therefore, using objective criteria, we have identified a preoperative deficit in auditory processing function in patients suffering from mesial temporal epilepsy. In this light, it is worth considering whether the neuropsychological problems and cognitive deficits which so often affect these patients might not be related ­ at least in part ­ to difficulties with the processing of auditory information.