Longitudinal analysis of structural connectivity in patients with newly diagnosed focal epilepsy of unknown origin

Highlights

• We evaluate alterations in structural connectivity occur over time in focal epilepsy.

• The volumes of the thalamus and brainstem on the second MRI scan have an increased tendency.

• In terms of connectivity, small-worldness index are increased on the second.

Abstract

Objectives

The aim of this longitudinal study was to clarify whether significant alterations in structural connectivity occur over time in patients with newly diagnosed focal epilepsy of unknown origin.

Methods

A total of 40 patients with newly diagnosed focal epilepsy of unknown origin and with normal brain magnetic resonance imaging (MRI) on visual inspection were enrolled. All subjects underwent MRI twice involving three-dimensional volumetric T1-weighted imaging, which were suitable for structural volume analysis. Gray matter volumes were obtained using the FreeSurfer image analysis suite, and structural connectivity analyses were performed using Matlab-based BRain Analysis using graPH theory software.

Results

The median interval between the two MRI scans was 18.5 months in patients with epilepsy. There was a general tendency toward decreased gray matter volumes on the second scan compared with the initial scan. However, the volumes of the right and left thalamus and brainstem on the second MRI scan had an increased tendency compared with those on the initial MRI scan. In measures of connectivity, there were significant differences between the two MRI scans. The mean clustering coefficient, global efficiency, local efficiency, and the small-worldness index were significantly increased, whereas the characteristic path length was decreased on the second MRI scan compared with the initial MRI scan.

Conclusions

The structural connectivity in patients with newly diagnosed focal epilepsy of unknown origin increases over time in the initial stage. These alterations and increases in structural connectivity may be related to underlying epileptogenicity in the initial stages of epilepsy.

Introduction

Epilepsy of unknown cause, which, in the past, has been termed “cryptogenic,” accounts for at least one-third of all epilepsy diagnoses.¹ In approximately 25 % of patients with focal epilepsy, previously referred to as “partial epilepsy,” no specific cause can be identified using neuroimaging studies and, as such, is classified as focal epilepsy of unknown cause [1].

Mounting evidence has suggested that epilepsy is a network disease, with focal epilepsy also exhibiting widespread alterations in the network beyond the epileptogenic zone [2]. This is in agreement with findings that patients with focal epilepsy exhibit gray matter atrophy beyond the putative epileptogenic focus [3]. Quantitative magnetic resonance imaging (MRI) provides biologically valid imaging markers and high test-retest reliability, thus representing an objective framework from which to investigate disease extension, which is more sensitive in detecting gray matter atrophy than visual analysis. In patients with temporal lobe epilepsy involving hippocampal sclerosis, the most common refractory focal epilepsy in adults, quantitative MRI analysis has revealed extra-temporal volume reduction as well as extra-hippocampal atrophy [3,4]. In addition, many studies have reported significant volume reduction, even in patients with focal epilepsy of unknown cause [[5], [6], [7]].

Despite the traditional hypothesis that “seizures beget seizures,” controversy remains as to whether epilepsy is a progressive disease. In the fields of neuroimaging research, gray matter atrophy could reflect the progression of epilepsy, especially targeted structures involved in the epilepsy network. Several studies have investigated the relationship between the duration of epilepsy and changes in gray matter volume in focal epilepsy. Some research has demonstrated a significant correlation between gray matter volume reduction and duration of epilepsy in several regions [3,4,[8], [9], [10]], whereas other studies did not reveal any such association [[11], [12], [13], [14]]. These discrepancies are most likely due to methodological differences, such as the method of analysis in cross-sectional [3,8,9,11] versus longitudinal follow-up studies [4,10,[12], [13], [14]]. Cross-sectional studies cannot determine whether neuronal damage occurs progressively through repeated seizures. One basic assumption of cross-sectional studies is that the brain volumes of all patients are same at the same age, which may be incorrect. Thus, longitudinal studies with serial MRI scans better enable the quantification of morphological changes over time than cross-sectional investigations. Furthermore, most previous studies have shared the problem of enrolling heterogeneous patient groups, thereby mixing individuals with newly diagnosed and chronic epilepsy [3,4,6,[8], [9], [10], [11]13]. There are differences between newly diagnosed and chronic epilepsy in many respects, such as response to antiepileptic drugs, prognosis, and mortality rate [15,16]. To address these issues, we performed a longitudinal quantitative MRI study involving patients with newly diagnosed focal epilepsy, who had no structural lesions on their brain MRI.

There have been few reports describing changes in the brain network over time. One cross-sectional study observed a progressive increase in connection strength in the hippocampus over time in patients with mesial temporal lobe epilepsy [17]. Another study reported that interhemispheric hippocampal connectivity exhibited a significant negative correlation with the duration of epilepsy [18]. However, to date, neither longitudinal follow-up studies nor research investigating focal epilepsy of unknown origin have been conducted.

The aim of this longitudinal study, therefore, was to clarify whether there significant alterations in structural volumes and connectivity occur over time in patients with focal epilepsy of unknown origin. We only included patients with newly diagnosed focal epilepsy of unknown origin who had normal brain MRI results on visual inspection. We hypothesized that there are significant alterations in structural volumes and connectivity observed over time in this patient population compared to healthy normal controls.