Differential contributions of the hippocampus and rhinal cortices to verbal memory in epilepsy
Introduction
Over the past 15 years, studies of patients with temporal lobe epilepsy (TLE) have confirmed the nexus between left temporal lobe pathology and verbal memory impairment [1], [2], [3], [4], [5], [6], [7], [8]. There is also evidence suggesting that the medial and lateral aspects of the temporal lobe are differentially involved in the control of verbal memory [2], [7], [9]. Semantic aspects of verbal memory are thought to be more dependent on lateral temporal neocortex. On the other hand, forms of verbal learning and memory that are not facilitated by previously established semantic relationships are more dependent on medial temporal structures [9], [10]. Acquisition of unrelated word pairs (“hard” or arbitrary paired associates) exemplifies the latter form of verbal learning. Arbitrary relational learning is widely regarded as the fundamental building block of the declarative memory system. The role of specific medial structures in arbitrary forms of verbal memory, however, remains unclear.
The most striking evidence for the intratemporal specialization of verbal memory function comes from patients with left mesial temporal lobe epilepsy (mTLE). Left mTLE is associated with a selective impairment in the acquisition of unrelated word pairs (“hard” paired associates) [1], [6], [8], but leaves the ability to learn semantically related word pairs and other forms of semantically based material intact [1], [3], [6], [8]. The selective impairment in arbitrary paired associate learning seen in these studies has been attributed to involvement of the hippocampus proper.
A recent whole-brain positron emission tomography (PET) study [7], however, questions the strength of the relationship between the left hippocampus and verbal arbitrary relational learning in left mTLE. Weintrob et al. correlated out-of-scanner verbal paired associate learning performance with resting glucose uptake [7]. The learning of semantically related word pairs correlated with uptake in the left antero-inferior temporal neocortex, a region known to be associated with semantic function [11]. Semantically unrelated word pair learning, however, mapped onto the inferomedial region of the left temporal lobe. Specifically, a peak correlation between learning and glucose uptake was seen in the perirhinal cortex, but not in the hippocampus. These data raise the possibility that the rhinal cortices make their own functional contribution to arbitrary relational forms of verbal memory in patients with TLE.
The aim of the present study was to test this hypothesis by correlating different aspects of verbal memory with structural integrity in hippocampal as well as rhinal regions of interest, using quantified T2 signal (T2 relaxation time). Structural damage is associated with a nonspecific increment in T2 signal, visualized as a regional T2 hyperintensity on magnetic resonance imaging (MRI). This change can be quantified using the standard technique of T2 relaxometry, which is sensitive to the range of structural changes typically associated with TLE [12], [13], and reflects a regional increase in the ratio of glial cells to neurons [14]. A relationship between left hipppocampal T2 relaxation times and performance on some measures of verbal memory has previously been demonstrated in patients with TLE [6], [15], [16], [17], [18], but these studies have been limited to a single region of interest in the hippocampus, with no specific separation between arbitrary and semantic aspects of verbal learning.
To maximize the range of T2 signal within the mesial temporal region, we studied TLE patients with a variety of epileptogenic lesions, as well as those who were lesion negative on MRI. Patients with right-sided TLE also contribute to variability in left mesial temporal T2 signal: increases in contralateral hippocampal T2 relaxation time are common in patients with TLE [19], reflecting general contralateral mesial temporal damage [20], and in some cases with right TLE the changes can be associated with verbal memory decrements [18]. Although the mechanism underlying this effect is not yet clear, there is some evidence to suggest that the contralateral changes are related to persistent seizure activity [21], [22].
For purposes of comparison, we included a measure of retroactive interference in word list learning. Retroactive interference effects are enhanced in left mTLE [5], [6], presumably reflecting impaired consolidation of the newly learned material. As consolidation processes have long been associated with the functions of the hippocampus proper [23], [24], we hypothesized that susceptibility to the effects of retroactive interference would be more closely related to hippocampal rather than rhinal T2 relaxation times.
Section snippets
Participants
A consecutive series of 22 patients (9 male, 20 right-handed) with a longstanding history (mean duration 14.5 ± 13.1 years, range 3–49 years) of unilateral TLE (left-sided seizure focus or left TLE = 15, right sided seizure focus or right TLE = 7) were retrospectively studied. All patients had been investigated in the Comprehensive Epilepsy Program at Austin Health between April 2002 and November 2004. Diagnosis of TLE and seizure focus lateralization was based on clinical history and ictal
T2 relaxometry
Measurements of T2 relaxation time could not be obtained in all regions of interest in four of the patients because of motion/flow artifacts, or unclear anatomical differentiation. For these participants, data obtained in the remaining regions of interest were included in the analyses.
Hippocampal relaxation times were bilaterally elevated relative to normative values in the left and right TLE groups: left HC—F(2,97) = 44.72, P < 0.001; left TLE > normal: t(91) = 8.95, P < 0.001; right TLE > normal: t(84) =
Discussion
In this study, an index of structural integrity, derived from three mesial temporal regions of interest, was correlated with measures of verbal learning to investigate the relative contributions of the hippocampus and rhinal cortices to arbitrary relational learning in patients with TLE. Our data provide converging support for the hypothesis that the perirhinal cortex is a key neural substrate for arbitrary relational learning. Specifically, we found a substantial relationship between the
Acknowledgments
The authors gratefully acknowledge the support of the National Health and Medical Research Council, Neurosciences Victoria, and the Brain Imaging Research Foundation, Australia.
References (41)
- et al.
The interrelationship between language function and verbal learning/memory performance in patients with complex partial seizures
Cortex
(1988) - et al.
T2 relaxometry can lateralize mesial temporal lobe epilepsy in patients with normal MRI
NeuroImage
(2000) - et al.
Learning and retention of words and designs following excision from medial or lateral temporal-lobe structures
Neuropsychologia
(1997) - et al.
Differential involvement of left temporolateral and temporomesial structures in verbal declarative learning and memory: evidence from temporal lobe epilepsy
Brain Cogn
(1997) - et al.
Perirhinal cortex ablation impairs configural learning and paired-associate learning equally
Neuropsychologia
(1998) - et al.
Hippocampal neuron loss and memory scores before and after temporal lobe surgery for epilepsy
Arch Neurol
(1993) - et al.
Cognitive consequences of two-thirds anterior temporal lobectomy on verbal memory in 144 patients: a three-month follow-up study
Epilepsia
(1996) - et al.
Verbal memory impairment resulting from hippocampal neuron loss among epileptic patients with structural lesions
Neurology
(1995) - et al.
Wada memory testing and hippocampal volume measurements in the evaluation for temporal lobectomy
Neurology
(1993) - et al.
Components of verbal learning and hippocampal damage assessed by T2 relaxometry
J Int Neuropsychol Soc
(2000)
Verbal memory in left temporal lobe epilepsy: evidence for task-related localization
Ann Neurol
Lateralization of verbal memory and unilateral hippocampal sclerosis: evidence of task-specific effects
J Clin Exp Neuropsychol
Medial and lateral contributions to verbal memory: evidence from temporal lobe epilepsy
Imaging and neuropsychology
Differing patterns of temporal atrophy in Alzheimer’s disease and semantic dementia
Neurology
Hippocampal sclerosis without detectable hippocampal atrophy
Neurology
Quantitative neuropathology and quantitative magnetic resonance imaging of the hippocampus in temporal lobe epilepsy
Ann Neurol
The relationship between quantitative T2 relaxometry and memory in nonlesional temporal lobe epilepsy
Epilepsia
The relationship between quantitative MRI and neuropsychological functioning in temporal lobe epilepsy
Epilepsia
MRI-based hippocampal volumetry and T2 relaxometry: correlation to verbal memory performance in newly diagnosed epilepsy patients with left-sided temporal lobe focus
Neurology
Cited by (26)
Rhinal and hippocampal contributions to spontaneous inter-item binding and verbal memory recall: Evidence from temporal lobe epilepsy
2020, CortexCitation Excerpt :In mTLE, structural alterations can also be found in the rhinal cortex in excess of hippocampal pathology (Bernasconi et al., 2003; Mueller et al., 2006; Yilmazer-Hanke et al., 2000). In line with its proposed role in memory formation, some studies have suggested an association between the structural integrity of the rhinal cortex with measures of verbal memory performance, namely prose recall and learning of paired associates in patients with mTLE (Bonilha et al., 2007; Lillywhite et al., 2007). Likewise, the extent of surgically removed rhinal cortex volumes predicted the outcome in a free recall measure of a word list learning paradigm in patients with mTLE having undergone temporal lobe resections (Liu et al., 2017).
List-learning and verbal memory profiles in childhood epilepsy syndromes
2016, Epilepsy and BehaviorDistinct functional connectivity of the hippocampus during semantic and phonemic fluency
2015, NeuropsychologiaCitation Excerpt :These models (Friederici, 2002, 2011; Friederici and Alter, 2004) relate a particular function to a particular brain region within the language system, leaving open the option that the same brain region serves another function in a domain other than language. It has already been established that in the memory domain, the hippocampus’ domain-specific function involves, for example, also the rhinal cortices (Lillywhite et al., 2007). In the present study we show that in the language domain, the hippocampus interacts with a set of regions underpinning semantic fluency.
The relationship between hippocampal volumes and nonverbal memory in patients with medial temporal lobe epilepsy
2014, Epilepsy ResearchCitation Excerpt :Memory can be divided into verbal memory and nonverbal memory. The left perirhinal region is known to be involved in the uptake of verbal arbitrary linkages that underlie new learning (Lillywhite et al., 2007). Another study proved that verbal memory impairments in patients with left-side MTLE were associated with atrophy of the hippocampus, EC, and perirhinal cortex (Bonelli et al., 2009).