Isolated amygdala enlargement in temporal lobe epilepsy: A systematic review

Isolated amygdala enlargement in temporal lobe epilepsy: A systematic review

Epilepsy & Behavior 60 (2016) 33–41 Contents lists available at ScienceDirect Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh R...

826KB Sizes 7 Downloads 31 Views

Recommend Documents

No documents
Epilepsy & Behavior 60 (2016) 33–41

Contents lists available at ScienceDirect

Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh

Review

Isolated amygdala enlargement in temporal lobe epilepsy: A systematic review S.M. Jessica Beh ⁎, Mark J. Cook, Wendyl J. D'Souza a b

The Department of Medicine, St. Vincent's Hospital Melbourne, St Vincent's PO Box 2900, Fitzroy, VIC 3065, Australia The University of Melbourne, Parkville, VIC 3010, Australia

a r t i c l e

i n f o

Article history: Received 19 December 2015 Revised 5 February 2016 Accepted 4 April 2016 Available online xxxx Keywords: Amygdala enlargement Temporal lobe epilepsy Seizures Amygdala volume Amygdala volumetry

a b s t r a c t Objective: The objective of this study was to compare the seizure characteristics and treatment outcomes in patient groups with temporal lobe epilepsy (TLE) identified with isolated amygdala enlargement (AE) on magnetic resonance imaging studies. Methods: PubMed, Embase, and the Cochrane Library were searched for relevant studies using the keywords ‘amygdala enlargement’, ‘epilepsy’, and ‘seizures’ in April 2015. Human studies, written in English, that investigated cohorts of patients with TLE and AE were included. Results: Of 204 abstracts initially identified using the search strategy, 14 studies met the inclusion criteria (11 epilepsy studies and 3 psychiatry studies). Ultimately, 8 full studies on AE and TLE involving 107 unique patients were analyzed. Gender distribution consisted of 50 males and 57 females. Right amygdala enlargement was seen in 39 patients, left enlargement in 58 patients, and bilateral enlargement in 7 patients. Surgical resection was performed in 28 patients, with the most common finding being dysplasia/hamartoma or focal cortical dysplasia. Most studies involved small samples of less than 12 patients. There was a wide discrepancy in the methods used to measure amygdala volume, in both patients and controls, hindering comparisons. Most TLE with AE studies observed a later age of seizure onset (mean: 32.2 years) compared with studies involving TLE with HS (mean of mid- to late childhood). A higher frequency of complex partial seizures compared with that of convulsive seizures is seen in patients with AE (67–100% vs. 26–47%), and they have an excellent response to antiepileptic drugs (81.8%–100% of seizure-free patients). All studies that included controls also found a significant difference in frequency of seizure types between their cases and controls. Conclusions: Reliable assessment of amygdala volume remains a critical issue hindering better understanding of the clinical management and research of this focal epilepsy syndrome. Within these limitations, the literature suggests characteristics of an older age of epilepsy onset, a greater tendency to nonconvulsive seizures, and a good response to antiepileptic drugs in this interesting group of epilepsies. Crown Copyright © 2016 Published by Elsevier Inc. All rights reserved.

1. Introduction The most common lesions identified in focal epilepsy are hippocampal sclerosis (40%), long-term epilepsy-associated tumors (27%), and malformations of cortical development (13%) [1]. Historically, the

Abbreviations: TLE, temporal lobe epilepsy; AE, amygdala enlargement; HS, hippocampal sclerosis; MRI, magnetic resonance imaging; FDG-PET, [18F]-fluorodeoxyglucose positron emission tomography; AMV, amygdala volume; VBM, voxel-based morphometry; AEDs, antiepileptic drugs; EEG, electroencephalography; SPECT, interictal single photon emission computed tomography; MET-PET, [11C]-methionine positron emission tomography; FCD, focal cortical dysplasia; LGI1, leucine-rich, glioma-inactivated 1 (antibody); CPS, complex partial seizures; CS, convulsive seizures; IEDs, interictal epileptiform discharges. ⁎ Corresponding author at: The Department of Medicine, St. Vincent's Hospital Melbourne, St Vincent's PO Box 2900, Fitzroy, VIC 3065, Australia. Tel.: + 61 452 622 865 (mobile); fax: +61 92883065. E-mail addresses: [email protected] (S.M.J. Beh), [email protected] (M.J. Cook), [email protected] (W.J. D'Souza).

http://dx.doi.org/10.1016/j.yebeh.2016.04.015 1525-5050/Crown Copyright © 2016 Published by Elsevier Inc. All rights reserved.

most common form of focal epilepsy, mesial temporal lobe epilepsy (TLE) due to hippocampal sclerosis (HS), has received the widest surgical attention. Hippocampal sclerosis can be reliably detected on routine magnetic resonance imaging (MRI) studies and is well-known for its epileptogenicity [2]. Mesial TLE due to HS is often medically resistant, and surgical resection can achieve up to 70% long-term seizure freedom in patients [3]. With improved neuroimaging and correlative presurgical techniques, clinicians are becoming increasingly aware of other potentially epileptogenic temporal lobe lesions that may also have good surgical outcome [4,5]. While seizures originating from these other causes might have been classified as ‘nonlesional’ or ‘imaging-negative’ epilepsy previously, a greater awareness of other potentially epileptogenic lesions, as well as improved imaging technology, has resulted in the identification of enlarged amygdalae in a proportion of patients (see Fig. 1). Bower et al. retrospectively found that 64% of ‘imaging-negative’ cases had both significant amygdala asymmetry and amygdala enlargement

34

S.M.J. Beh et al. / Epilepsy & Behavior 60 (2016) 33–41

Fig. 1. Magnetic resonance imaging study demonstrating left-sided amygdala enlargement.

(AE), while Coan et al. demonstrated significant AE in 12% of previously diagnosed patients with ‘nonlesional’ TLE [4,6]. Because nonlesional TLE may be associated with an almost three times lower probability of seizure freedom after surgical resection compared with lesional TLE, this reclassification of newly identified AE lesions in TLE has important implications for patient management [7]. Using parcellation based on cytoarchitecture, Murphy et al. showed in 1987 that the amygdaloid complex is asymmetric in humans, while in 1992, Watson et al. attempted to establish standardized anatomical guidelines for outlining the hippocampus and amygdala on high-resolution MRI scans [8,9]. Although the technology has been available to measure amygdala volumes specifically since the early 1990s, studies linking amygdala pathology and size to seizure foci until recently have been lacking [8,9]. There are a number of clinical, electroencephalographic, radiological, and histopathological observations that have led to an increased focus on the role of the amygdala independent from HS in epilepsy in recent years. Firstly, better outcomes have been observed when the amygdala was included in surgical resection borders [10]. Secondly, studies utilizing MRI T2 relaxometry have shown that a portion of patients with ‘imaging-negative’ TLE have increased T2 signal in the amygdala ipsilateral to the focus detected by electroencephalography (EEG) [11]. Thirdly,

[18F]-fluorodeoxyglucose positron emission tomography (FDG-PET) findings of glucose hypometabolism have been isolated to the amygdala and excluding the hippocampus in some patients [12]. Finally, histopathologic abnormalities have been found in amygdala specimens resected from patients with medically resistant TLE [13]. This review pools studies on TLE and AE to provide better understanding of this imperfectly described focal epilepsy syndrome. Specific focus will be on the seizure characteristics, diagnostic investigations, treatment outcomes, and histopathological results in these cohorts of patients. We aimed to identify knowledge gaps in the literature that can guide future clinical management and research. 2. Methods Our search strategy comprised a comprehensive literature search of PubMed, Embase, and the Cochrane Library, conducted using the terms ‘amygdala enlargement’, ‘epilepsy’, and ‘seizures’, either alone or in varying combinations in April 2015. Papers were excluded for the following reasons: animal studies, irrelevant topics, non-English papers, papers not involving a population with epilepsy, case reports, and those described with coexisting brain pathology. The references of included papers were also reviewed to identify any relevant papers that may

S.M.J. Beh et al. / Epilepsy & Behavior 60 (2016) 33–41

have been missed in the initial search (see Fig. 2). The strengthening the reporting of observational studies' (STROBE) guidelines [14] were subsequently applied to ensure the quality of individual studies included in this review. 3. Results Our search strategy in April 2015 ultimately yielded 14 papers from seven countries published between 1999 and September 2014. Although three abstracts presented only at scientific meetings were unavailable for full review, for reference, they have been included in the tables of the Results section [15–17]. Three papers focusing only on psychiatric domains of TLE and AE cannot be directly compared with the other eight papers focusing on epilepsy and were not formally reviewed but were included in our discussion on symptom characterization [18–20]. Although, in total eight papers that focus on seizure characteristics and treatment outcomes in TLE and AE were found, seven out of nine patients in the Takaya et al. [12] study had originally been a part of the earlier Mitsueda-Ono et al. [21] study. These seven patients have

PUBMED 199 & epilepsy 27 & seizures 17

35

therefore been carefully excluded from statistical analyses in this systematic review, with no double counting occurring. Full characteristics of the eight papers, including Takaya et al. [12], have been, however, included in Tables 1 and 2 for the sake of completeness. In total, these eight papers therefore consist of 107 unique patients with AE, 28 of whom underwent surgical resection of their enlarged amygdalae with available histopathology. The largest patient cohort consists of 33 patients, from Beijing, China [22]. 3.1. Summary of study designs Table 1 summarizes the designs of all eleven studies including the three only in abstract form. Differing types of cohorts were studied across the eight full papers — two were primarily surgical, two medical, and four a mixture of both types of patients. 3.1.1 Ascertainment of cases Studies vary in methods of ascertainment of enlarged amygdalae on imaging modalities. Some studies had protocols for measuring of amygdala volumes (AMVs) in mm3 (n = 6), while others relied on visual

COCHRANE A 3 & epilepsy 0 & seizures 0

EMBASE 31 AE 16 & seizures 20

Records identified through database searching: n = 312 Repeated: 108

Title and abstract screened: n = 204 127 Excluded: 56 Animal, 53 Irrelevant, 18 Non-English

Full paper reviewed: n = 77 63 Excluded: 61 not epilepsy, 1 case report, 1 co-existing brain pathology

Included Studies: n = 14 (Abstracts available only 3) (Psychiatric focus 3) Fig. 2. Summary of literature search strategy.

36

S.M.J. Beh et al. / Epilepsy & Behavior 60 (2016) 33–41

Table 1 Summary of study designs of patients with epilepsy and amygdala enlargement. Study

Bower et al. (2003) [4] Mitsueda-Ono et al. (2010) [21] Kim et al. (2011) [23]

N

Controls

Design

EEG

Neuroimaging

7

154

Retrospective

Y

1.5T MRI

11

46

Retrospective

Y

1.5T/3T MRI, FDG-PET, interictal SPECT

12

0

Retrospective

Y

Takaya et al. (2014) [12]

9

45

Retrospective

Coan et al. (2013) [6] Kimura et al. (2013) [24]

8

82

23

Minami et al. (2014) [25] Lv et al. (2014) [22] Sone et al. (2014) [15]d Taniguchi et al. (2013) [16]d Keezer et al. (2012) [17]d

Confirmation of AE

Exclusion criteria

Treatment AEDs

Surgery

Manual tracing by an unblinded single rater Visual inspection (2 neurologists), manual tracing by a single rater

Imaging: dual pathology, increased signal Imaging: dual pathology, brain tumor highly ‘suspected’a

6

1

9

2

1.5/3T MRI, FDG-PET, ictal SPECT

Visual inspection

0

12

Y

MRI-VBM, FDG-PET

VBM standard routines, visual inspection

8

1

Retrospective

Y

0

Retrospective

Y

VBM standard routines, FreeSurfer, T2 relaxometry FreeSurfer automated software

8

20

MRI-VBM, T2 relaxometry MRI-VBM, FDG-PET, interictal SPECT

21

2

11

26

Retrospective

Y

Visual inspection (2 radiologists)

0

11

33

35

Retrospective

Y

MRI, interictal SPECT, FDG-PET, MET-PET 3.0T MRI, FGD-PET

History of head trauma, imaging: dual pathology, brain tumor ‘suspected’b History of brain infection, head trauma, psychiatric illness; signs of focal neurology; EEG: extratemporal discharges; imaging: amygdala tumor ‘suspected’a, dual pathology Imaging: dual pathology, abnormal MRI on visual inspection History of: brain infections, head trauma, cerebral infarction/haemorrhage/surgery; imaging: dual pathology, tumor ‘suspected’b, vascular abnormalities Imaging: tumor ‘suspected’c

0

NS

NS

Y

MRI, MET-PET

Imaging: dual pathology, tumor ‘suspected’a NS

33

23

Visual inspection (2 epileptologists) manual tracing by a single rater NS

23

0

9

NS

NS

N

NS

NS

9

0

25

NS

NS

Y

MRI, MET-PET, FDG-PET, SPECT MRI-based 3D surface-based shape modeling

Manual tracing

NS

NS

NS

AE = amygdala enlargement; AEDs = antiepileptic drugs; EEG = electroencephalography; MRI = magnetic resonance imaging; VBM = voxel-based morphometry; FDG-PET = [18F]fluorodeoxyglucose positron emission tomography; SPECT = interictal single photon emission computed tomography; MET-PET = [11C]-methionine positron emission tomography; NS = not specified; Y = yes; N = no. a Where enlarged amygdala apparently compressed the adjacent tissue or showed clear intensity changes in T2-weighted or fluid-attenuated inversion recovery MRI. b Calcification, cystic changes, or contrast enhancement seen on MRI. c MRI-enhanced lesions, cysts, calcifications, or well-circumscribed intensity lesions. d Abstract available only.

inspection by radiologists or neurologists to determine whether AE was present (n = 2). In total, 4 main methodologies were utilized in the eight papers included in this review: 1) visual assessment by experienced radiologists or neurologists (usually by 2 individuals, information unavailable with regard to blindedness); 2) manual tracing of the amygdaloid complex using an atlas-based approach and subsequent counting of voxels in an outlined area of interest by an unblinded single rater [9]; 3) voxel-based morphometry (VBM) standard routines resulting in automatic segmentation of tissue types [26]; and 4) automated volume measurement using FreeSurfer software [27]. 3.1.2 Inclusion of controls The eight cohorts are heterogeneous in terms of the types and numbers of control groups included in research design. One study included three separate categories of controls, i.e., non-TLE, TLE + HS, and healthy controls [21]; two studies included two categories of controls, i.e., TLE + HS and healthy volunteers [4,12]; some studies included only one category, i.e., either TLE + HS or healthy volunteers [6,22,24,25]; and one study did not include any control patients [23]. 3.1.3 Exclusion criteria Most of the eight papers applied similar criteria for the exclusion of patients to their groups with TLE and AE, in particular, the exclusion of patients with highly ‘suspicious’ tumoral lesions on imaging, either with calcification, cystic changes, or well-circumscribed increased intensity lesions, or patients with likely (on imaging) or confirmed (pathologically) HS. Investigations used in the identification of isolated

AE were also similar across the papers, with most utilizing 1.5/3T MRI, interictal EEG, FDG-PET, and either ictal or interictal single photon emission computed tomography (SPECT). 3.2. Summary of results across studies (see Table 2) 3.2.1 Amygdala assessment Bower et al. [4] and Kim et al. [23] reported uniformly enlarged amygdalae in their patients, while Takaya et al. [12] reported focal enlargement only in the dorsomedial portion of the amygdalae. Other studies did not provide information on uniform as opposed to localized enlargement. 3.2.2 Localization information Ipsilateral interictal epileptiform discharges (IEDs) to the enlarged amygdalae were found in most patients (Table 2) — with ipsilateral temporal lobe localization in 80% of patients, ipsilateral anterior temporal lobe localization in 12% of patients, and ipsilateral hemisphere localization found in 3% of patients. All of the 6 studies that included FDG-PET in their study designs found 91–100% hypometabolism in the ipsilateral mesial temporal lobe including the enlarged amygdalae. Minami et al. [25] reported more exquisite localization information. They, however, found abnormal sharp waves seemingly originating from the hippocampus rather than from the adjacent enlarged amygdala in the intraoperative electrical recording of all their 11 patients.

Table 2 Summary of semiology, risk factors, management outcomes, and histopathological findings in epilepsy associated with amygdala enlargement. N

Mean Mean AMV Side of AE IPS EEG IPS increased IPS hypometabolism Mean age of CPS (%) CS (%) FS (%) Presence Histopathology (n) in AE (mm3) AMV–non-AE (%) IEDs signal (%) on FDG-PET (%) seizure onset of aura (%) R L B controls (mm3) (years)

Engel class postsurgery (%)/seizure frequency post-AED treatment

7 3194 11 1147

2146 1700

4 5

3 5

0 100 1 91

14 NA

NA 91

29 (14–62) 39.8 (9–72)

71 73

29 27

14 18

NA 27

Glioneuronal hamartoma (1) Gliosis (1) FCD (1)

Kim et al. (2011) [23]

12 NA

NA

4

8

0 100

NA

100

27.4 (10–53)

67

33

8

17

Takaya et al. (2014) [12] Coan et al. (2013) [6] Kimura et al. (2013) [24]

9 NS

NS

3

6

0 100

NA

100

45.4 (19–61)

67

33

11

NA

FCD (8) Ganglioma (2) Oligodendroglioma (1) Astrocytoma (1) FCD (1)

8 NS 23 2244.4

NS 1726.0

NS = 5 3 57 11 11 1 95

0 NA

NA 100

17.9 (8–47) 29.8 (3–75)

NS 74

NS 26

25 NA

NA NA

Minami et al. (2014) [25]

11 NA

NA

NS

NA

91

26.3 (15–56)

73

27

9

18

Lv et al. (2014) [22]

33 1649.2

1250.5

15 16 2 100

NA

100

42.0 (20–75)

100

42

0

18

NA NA

13 9 NS

NS NS

NS NS

46.9 44% N 60

NS NS

NS NS

9 NS

NS NS

12.5% seizure-free Engel's I (100) 100% seizure-free or dramatic improvement Closely aggregated hypertrophic Engel's I (91) neurons with vacuolization of Engel's II (9) background matrix (10) Aggregated oligodendroglia-atypia (1) NA 66% seizure-free 33% continuing CPS NA 57% seizure-free NA NS

NS

NS

NS

NS

NS

NS

NS

NS

NS

NS

Sone et al. (2014) [15]a 23 NS 9 NS Taniguchi et al. (2013) [16]a Keezer et al. 25 NS (2012) [17]a

NS

1 100 89 NS

NA Engel's I (50) Engel's II (50) 89% seizure-free or dramatic improvement Engel's I (92) Engel's II (8)

NA

NA Cortical dysplasia not isolated to amygdala (2)

S.M.J. Beh et al. / Epilepsy & Behavior 60 (2016) 33–41

Bower et al. (2003) [4] Mitsueda-Ono et al. (2010) [21]

NS

AE = amygdala enlargement; AMV = amygdala volume; EEG = electroencephalography; FDG-PET = [18F]-fluorodeoxyglucose positron emission tomography; CPS = complex partial seizures; CS = convulsive seizures; FS = febrile seizures; IEDs = interictal epileptiform discharges; FCD = focal cortical dysplasia; IPS = ipsilateral; R = right; L = left; B = bilateral; NS = not specified; NA = not assessed. Engel classification —postoperative outcomes for epilepsy surgery: class I — free of disabling seizures; class II — rare disabling seizures; class III — worthwhile improvement; class IV — no worthwhile improvement [28]. a Abstract available only.

37

38

S.M.J. Beh et al. / Epilepsy & Behavior 60 (2016) 33–41

3.2.3 Histopathology Of the 28 resected amygdala specimens, the histopathological results were the following: 12 dysplasia/hamartoma, 11 focal cortical dysplasia (FCD), 1 nonspecific gliosis, 2 ganglioma, 1 oligodendroglioma, and 1 astrocytoma. The most frequent histology result was that of dysplasia, confined entirely to the amygdala (82%). Only 14% revealed tumor as the cause for the AE. 3.2.4 Clinical features Patients with AE demonstrated onset of seizures with a mean of 32.2 years (range: 17.9–45.4). Gender distribution consisted of 50 males and 57 females. The right amygdala was enlarged in 39 patients and the left amygdala in 54 patients, with bilateral enlargement seen in seven patients. Patients had a higher frequency of complex partial seizures (67–100%) compared with that of convulsive seizures (26–47%). Most studies found a significant difference in frequency of seizure types between their cases and controls, irrespective of the types of controls they had recruited. Auras were reported to be present in 17–27% of patients. One patient described facial pallor [25], three described a ‘nervous fluttering in their heart’ [22], and two described gastric aura [23]. Specific auras were otherwise either not reported in studies or did not involve the abovementioned features. Patients had an excellent response to antiepileptic drugs (AEDs) (81.8–100%) with only 26% proceeding to surgical resection. Kim et al. [23] was the only study that compared the clinical features between tumorous and nontumorous causes of AE. They found that the clinical features and MRI findings were mostly indistinguishable when comparing the patients with brain tumors with those with FCD, with the exception of a tendency for a younger onset of seizures in the group with tumor [23]. 3.2.5 Outcomes The six papers that included outcomes are heterogeneous. All include seizure frequency, with some classifying patients into Engel classes [21,23–25]. In addition, two studies interestingly also included parameters on interval change of AMVs for ≥6-month duration and depression screens [21,22]. A total of 25 out of 38 patients (66%) displayed a decrease in AMVs with interval MRI scans. Lv et al. [22] involved the most comprehensive assessments of treatment outcomes among all the studies, including memory testing, anxiety/depression scales, and paraneoplastic/autoimmune antibody screening in their patient follow-up. They found depression or anxiety in 90% and memory decline in 30%, and 3% were positive for the leucine-rich, glioma-inactivated 1 (LGI1) antibody [22]. 3.2.6 Interval changes in amygdala volumes While only assessed in two studies [21,22], interval changes in AMVs were observed in two-thirds of included patients. Using the same method of ascertaining AMVs between intervals, Mitsueda-Ono et al. [21] found an improvement in AMVs in three out of five of their patients, with no change in the other two, while Lv et al. [22] found significant improvements in AMVs in 22 out of 33 of their patients. There was no association with seizure frequency and decreases in AMVs in the former study [21], but the latter found an association with response to AEDs and AMVs [22]. The remaining 11 patients who did not show a significant decline in CPS frequency over 6–12 months in Lv et al.'s study were the same 11 who did not show an improvement in their AMVs on repeat imaging scans [22].

by the same group involving patients with TLE showed significantly larger amygdalae bilaterally in females and patients with depression and in patients with dysthymia, respectively. 4. Discussion Methodological issues limit comparisons between studies on isolated amygdala enlargement in temporal lobe epilepsy. Population cohorts are relatively small and ascertained through mixed surgical or medical sources with varying control group comparisons. However, the key methodological issue remains the varying identification techniques for amygdala enlargement from automated software (voxel-based morphometry or FreeSurfer), to manual tracing techniques to visual inspection by blinded or unblinded, single or dual raters. These limitations aside, general patterns emerge of a focal epilepsy syndrome characterized by the following: a later age of seizure onset from late adolescence to middle age, greater frequency of complex partial rather than convulsive seizures but with few auras, an excellent response to antiepileptic drugs compared with other typical lesional focal epilepsies, and consequently, only a few resected amygdalae with reassuringly uncommon malignant causes. Systematic reviews can reveal meaningful comparisons that are not evident in individual studies because of their limited sample sizes; this is especially so when investigating conditions that are relatively uncommon, as is the case with isolated amygdala enlargement in temporal lobe epilepsy. We followed the steps outlined by Khan et al. to produce this systematic review on amygdala enlargement in epilepsy [29]. 4.1. Amygdala concordance With one exception [6], in the studies that included information on AE concordance with other findings, mostly ipsilateral IEDs were observed (Table 2). Interictal epileptiform discharges in TLE have been shown to correlate well with seizure-onset zones [30], supporting the enlarged amygdalae as likely being the seizure focus in the patients included in the studies. More exquisite localization to the amygdala specifically would provide even more robust evidence of the epileptogenicity of this enlarged structure. At present, the possibility of a true seizure focus in an area adjacent or in close proximity to the enlarged amygdalae cannot be excluded. An exception to this trend is Coan et al. [6], who observed ipsilateral IEDs in only half of their patients (Table 2). Even though they did the most comprehensive volumetry assessments, i.e., VBM followed by confirmation using FreeSurfer software, followed by a secondary VBM analysis, there remains suspicion regarding the accurate classification of patients with AE vs. control patients given their 50% discordant IED results. This is possibly due to their differing methodology compared with that of the other studies, including smoothing gray matter images to remove large signal discrepancies while using VBM. Regarding Minami et al.'s [25] intraoperative results, they postulate that this could be explained by the synchronization of electrical discharges from the hippocampus and amygdala. With virtually all of the patients achieving Engel's class I outcomes after surgery, which included resection of hippocampi in all cases, it is possible that the hippocampi were the foci of seizures, rather than the adjacent enlarged amygdala. Glucose hypometabolism using FDG-PET has also been shown to be highly sensitive in detecting epileptic foci, even in patients with ‘imaging-negative’ TLE [5]. This being observed in the ipsilateral mesial temporal lobes including the enlarged amygdale in all of the studies that included FDG-PET analyses increases our confidence of amygdala concordance.

3.3. Psychiatric comorbidities in TLE and AE 4.2. Exclusion criteria across studies On comparing 26 patients with psychosis of epilepsy with 24 patients with TLE without psychopathology and 20 healthy volunteers, Tebartz et al. [18] found that the patients with psychosis had a 16– 18% enlargement of amygdalae bilaterally. Two earlier studies [19,20]

With the exception of varying methodologies used to identify enlarged amygdalae, similar inclusion and exclusion criteria were used across all studies, improving confidence in case homogeneity. However,

S.M.J. Beh et al. / Epilepsy & Behavior 60 (2016) 33–41

it should be noted that some types of highly epileptogenic brain tumors occur frequently in the mesial temporal lobes, including the amygdala, and tend to lack radiological features of tumors that investigators were trying to screen for [31]. We unfortunately cannot completely rule out the possibility of a number of neoplastic causes of AE in the 79 medically managed patients with TLE and AE. 4.3. Techniques for measuring amygdalar volumes Accurate, standardized, and validated assessment of AMV remains the cornerstone of the definition of the AE epilepsy syndrome, and the lack of a standard method of measurement is the single greatest methodological challenge limiting its well-defined descriptive analysis and clinical management. While volumetric studies involving the hippocampus are now well established to be relatively accurate in detecting even subtle degrees of hippocampal atrophy in patients with epilepsy [32,33], the same cannot be said for the amygdaloid complex. Volumetric MRI studies of the hippocampi have also been shown to highly predict the presence of histopathological HS [34], but imaging techniques with regard to the amygdala have yet to reach the same level of sophistication. Bower et al. [4] were the first to attempt to estimate AMVs in ‘imaging-negative’ TLE in 2003, and since then, various groups have adopted different techniques to try and ascertain AMVs in their patient cohorts. This challenge stems from the amygdalae having poorly demarcated anatomical boundaries, with its relatively indistinct borders with adjacent structures such as the hippocampus, putamen, and parahippocampal gyrus [9,35]. With nonstandardized methodology across studies, there remains uncertainty that the patient cohorts are homogenous. The radiological identification techniques used in the studies have poor interrater reliability for manual tracing methods, and error in automated tissue classification using VBM has also been well documented [11,36]. Although some groups have shown that volumetry estimates by manual tracing or automated techniques can both deliver highly consistent results [37,38], we do not see the same trends in this review. Current literature reports normal AMVs to range from 1240–1630 mm3 [39]. The normally reported AMVs for controls in four studies included in this review range from 1250 mm3 to 2146 mm3 [4,21,22,24] (Table 2). Even taking into account normal variation in amygdalar volume and correcting for individual brain size, three quarters of the studies had ‘normal’ AMVs outside the currently accepted range for the amygdala. Unlike the hippocampus where absolute and relative normative values exist [40], there remains no standard definition as to what specific values of AMVs constitute AE. The reported values of AMVs for patients with AE span a wide range, with average measurements ranging from 1700 mm3 to 3194 mm3 (Table 2). Even studies that used the same methods for the measurement of AMV ended up with values that were dissimilar to each other [4,21]. The studies too do not concur on the specific areas of the amygdalae that are enlarged vs. uniform enlargement, again bringing into question the reliability of their measurement methodologies. Without a clear definition of what constitutes an increased total volume and standardized validated methods for estimating amygdalar architecture boundaries, contamination of cases and controls remains an ongoing source of bias. Given these challenges, blinded independent interrater visual inspection by neuroradiologists or epilepsy specialists with appropriate expertise coupled with concordance from electroencephalographic and functional imaging multimodalities is likely a more valid approach for determining the presence of AE.

39

included, as well as that currently reported in the HS literature, with an onset usually in mid- to late childhood [41,42]. All studies observed a higher frequency of complex partial seizures (CPS) compared with that of convulsive seizures (CS) in patients with AE, consistent with Bruton's series where a similar pattern of seizure frequency was observed when developmental abnormalities were found in resected amygdaloid specimens [43]. Overall in the eight studies, more than two-thirds of patients experienced CPS, while a quarter to almost half experienced one or more CS in their lifetime (Table 2). Wieser reported that amygdalar epilepsy is characterized by ictal fear, painful gastric sensations, and marked autonomic symptoms [44]. However, only 5% of patients in these studies report aura types that hint at these limbic phenomena. Interestingly, Coan et al. have also published a study on AE in patients with TLE and HS [45]. They describe finding AE in 14% of patients with TLE and HS using automated volumetric methods, contralateral to the epileptogenic zone and signs of HS on MRI. The patients with AE and HS had significantly earlier ages of onset of epilepsy (median of 6 years) compared with the patients with HS without enlarged amygdala (median of 11 years) [45]. Coan et al. did not find any other significant differences in clinical features between patients with HS and AE and patients with isolated HS [45]. The patients included in this review, however, had isolated AE ipsilateral to the epileptogenic zone, with a much later mean age of onset of 32.2 years. These 2 groups of patients with AE, one with HS and one without, thus appear to have distinct clinical features. 4.5. Etiology While the etiology of AE remains unknown, hypotheses for the causes underpinning this older onset age include neurodevelopmental abnormalities, e.g., FCD; benign tumors, such as hamartoma or lowgrade glioma; or a focal neuroinflammatory process [12,25]. Although neuroinflammatory processes, in particular, are increasingly reported in adult-onset TLE [46], results in these eight studies have only revealed FCD and dysplasia/hamartoma from resected specimens. With the exception of one study [6] where no patients reported childhood febrile seizures, less than 20% of patients with AE report an antecedent history of childhood febrile seizures compared with up to 50% of patients with HS [47,48]. Although two of the papers included in this review focused solely on surgical patients, the authors acknowledge that surgical candidates in their patient cohort were the minority because of the high response rate to AEDs [23,25]. The remaining six papers describe rates of up to one-third with medically refractory seizures (Table 2), considerably lower than those observed in cohorts with HS epilepsy [49,50]. Twenty-eight patients underwent surgical resection of their amygdalae and sometimes their surrounding temporal lobe structures, and 4 tumors were found. Because seven of the eight studies used MRI features to exclude suspicious tumoral lesions including those involving the amygdala, the causes of AE may include a greater number of tumors than observed in these studies. Although Kim et al. [23] found tumors in a third of their patients with AE, we cannot be certain that this trend applies to cohorts with amygdala epilepsy in general, given their small sample size of 12 patients. Future studies should explore the true incidence of tumors in AE as opposed to benign nontumorous causes. Further corroboration of Kim et al.'s results would be useful in possibly guiding clinicians to identify patients with tumors from those with benign pathology in TLE and AE.

4.4. Clinical features of amygdala enlargement in temporal lobe epilepsy 4.6. Decrease in amygdala volumes after antiepileptic treatment Cohorts with AE remain small, and findings warrant confirmation in larger cohorts. With one exception where no differences were found [6], the mean observed age of seizure onset was 32.2 years across the eight studies. This is significantly older compared with all types of controls

Two studies [21,22] that assessed interval changes in AMVs found significant improvements in two-thirds of patients. Although interesting, further studies with larger sample sizes are needed to confirm

40

S.M.J. Beh et al. / Epilepsy & Behavior 60 (2016) 33–41

that AMVs reduce over time and are associated, causally or otherwise, with AED treatment and improved seizure frequencies.

department has received research funding from GlaxoSmithKline, SciGen, Pfizer, Novartis, and UCB Pharma.

4.7. Psychiatric comorbidities in epilepsy with amygdalar enlargement

References

The observations of Tebartz et al. around TLE and AE and an association with depression and dysthymia raise the possibility that the presence of AE is in some way causal to mood disturbance [19,20]. It remains unknown whether psychiatric comorbidities are specifically increased in patients with epilepsy with enlarged amygdalae. However, depression was only seen when amygdalae were bilaterally enlarged. While up to half of patients with TLE have been observed to have psychiatric comorbidities, especially depression, this number can be as high as 70% in patients with refractory epilepsy [51,52]. Only Lv et al. investigated psychiatric illnesses in TLE and AE and found that 90% of patients reported either depression or anxiety or both [22]. Unlike the estimates of Tebartz et al. [18], to date, the presence of ictal or postictal psychotic symptoms had not yet been reported in any of the 107 patients included across the eight studies.

[1] Blumcke I. Neuropathology of focal epilepsies: a critical review. Epilepsy Behav 2009;15:34–9. [2] Berkovic SF, Andermann F, Olivier A, Ethier R, Melanson D, Robitaille Y, et al. Hippocampal sclerosis in temporal lobe epilepsy demonstrated by magnetic resonance imaging. Ann Neurol 1991;29:175–82. [3] Mathon B, Bedos Ulvin L, Adam C, Baulac M, Dupont S, Navarro V, et al. Surgical treatment for mesial temporal lobe epilepsy associated with hippocampal sclerosis. Rev Neurol (Paris) 2015;171:315–25. [4] Bower SP, Vogrin SJ, Morris K, Cox I, Murphy M, Kilpatrick CJ, et al. Amygdala volumetry in “imaging-negative” temporal lobe epilepsy. J Neurol Neurosurg Psychiatry 2003;74:1245–9. [5] Carne RP, O'Brien TJ, Kilpatrick CJ, MacGregor LR, Hicks RJ, Murphy MA, et al. MRI-negative PET-positive temporal lobe epilepsy: a distinct surgically remediable syndrome. Brain 2004;127:2276–85. [6] Coan AC, Morita ME, Campos BM, Yasuda CL, Cendes F. Amygdala enlargement in patients with mesial temporal lobe epilepsy without hippocampal sclerosis. Front Neurol 2013 Oct 25;4:166. [7] Tellez-Zenteno JF, Hernandez Ronquillo L, Moien-Afshari F, Wiebe S. Surgical outcomes in lesional and non-lesional epilepsy: a systematic review and meta-analysis. Epilepsy Res 2010;89:310–8. [8] Murphy Jr GM, Inger P, Mark K, Lin J, Morrice W, Gee C, et al. Volumetric asymmetry in the human amygdaloid complex. J Hirnforsch 1987;28:281–9. [9] Watson C, Andermann F, Gloor P, Jones-Gotman M, Peters T, Evans A, et al. Anatomic basis of amygdaloid and hippocampal volume measurement by magnetic resonance imaging. Neurology 1992;42:1743–50. [10] Kullmann DM. What's wrong with the amygdala in temporal lobe epilepsy? Brain 2011;134:2800–1. [11] Van Paesschen W, Connelly A, Johnson CL, Duncan JS. The amygdala and intractable temporal lobe epilepsy: a quantitative magnetic resonance imaging study. Neurology 1996;47:1021–31. [12] Takaya S, Ikeda A, Mitsueda-Ono T, Matsumoto R, Inouchi M, Namiki C, et al. Temporal lobe epilepsy with amygdala enlargement: a morphologic and functional study. J Neuroimaging 2014;24:54–62. [13] Aroniadou-Anderjaska V, Fritsch B, Qashu F, Braga MF. Pathology and pathophysiology of the amygdala in epileptogenesis and epilepsy. Epilepsy Res 2008;78:102–16. [14] von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med 2007; 147:573–7. [15] Sone DTG, Mogi T, Murata Y, Watanabe Y, Okazaki M, et al. Clinical characteristics of mesial temporal lobe epilepsy with amygdala enlargement: association with lateonset temporal lobe epilepsy? Epilepsia 2014;55:127. [16] Taniguchi GIK, Murata Y, Sone D, Watanabe Y, Okazaki M, et al. Evaluation of amygdala enlargement in symptomatic localization-related epilepsy patient using 11Cmethionine-pet. Epilepsy Curr 2013;13:408. [17] Keezer MRKH, Bernhardt BC, Bernasconi A, Bernasconi N. Shape analysis in temporal lobe epilepsy with amygdalar enlargement. Epilepsy Curr 2012;12. [18] Tebartz Van Elst L, Baeumer D, Lemieux L, Woermann FG, Koepp M, Krishnamoorthy S, et al. Amygdala pathology in psychosis of epilepsy: a magnetic resonance imaging study in patients with temporal lobe epilepsy. Brain 2002;125:140–9. [19] Tebartz van Elst L, Woermann F, Lemieux L, Trimble MR. Increased amygdala volumes in female and depressed humans. A quantitative magnetic resonance imaging study. Neurosci Lett 2000;281:103–6. [20] Tebartz van Elst L, Woermann FG, Lemieux L, Trimble MR. Amygdala enlargement in dysthymia—a volumetric study of patients with temporal lobe epilepsy. Biol Psychiatry 1999;46:1614–23. [21] Mitsueda-Ono T, Ikeda A, Inouchi M, Takaya S, Matsumoto R, Hanakawa T, et al. Amygdalar enlargement in patients with temporal lobe epilepsy. J Neurol Neurosurg Psychiatry 2011;82:652–7. [22] Lv RJ, Sun ZR, Cui T, Guan HZ, Ren HT, Shao XQ. Temporal lobe epilepsy with amygdala enlargement: a subtype of temporal lobe epilepsy. BMC Neurol 2014;14:194. [23] Kim DW, Lee SK, Chung CK, Koh YC, Choe G, Lim SD. Clinical features and pathological characteristics of amygdala enlargement in mesial temporal lobe epilepsy. J Clin Neurosci 2012;19:509–12. [24] Kimura Y, Sato N, Saito Y, Ito K, Kamiya K, Nakata Y, et al. Temporal lobe epilepsy with unilateral amygdala enlargement: morphometric MR analysis with clinical and pathological study. J Neuroimaging 2015;25:175–83. [25] Minami N, Morino M, Uda T, Komori T, Nakata Y, Arai N, et al. Surgery for amygdala enlargement with mesial temporal lobe epilepsy: pathological findings and seizure outcome. J Neurol Neurosurg Psychiatry 2015;86:887–94. [26] Keller SS, Roberts N. Voxel-based morphometry of temporal lobe epilepsy: an introduction and review of the literature. Epilepsia 2008;49:741–57. [27] Pardoe HR, Pell GS, Abbott DF, Jackson GD. Hippocampal volume assessment in temporal lobe epilepsy: how good is automated segmentation? Epilepsia 2009;50: 2586–92. [28] Engel J. Surgical treatment of the epilepsies. Lippincott Williams & Wilkins; 1993. [29] Khan KS, Kunz R, Kleijnen J, Antes G. Five steps to conducting a systematic review. J R Soc Med 2003;96:118–21. [30] Blume WT, Borghesi JL, Lemieux JF. Interictal indices of temporal seizure origin. Ann Neurol 1993;34:703–9.

4.8. Limitations in current study designs Of the eight papers, two-thirds reported treatment outcomes in their patients. All these estimated seizure frequency, with some classifying patients into Engel classes. Despite numerous studies assessing quality-of-life outcomes in TLE and HS and some even comparing outcomes between medical or surgical treatments [53–55], no quality-oflife estimates have been reported to date for patients with epilepsy with AE beyond that of seizure freedom. Given the small reported numbers, the tendency for seizures in patients with AE to more likely be controlled on AEDs compared with those with HS and difficulty in ethically justifying, withholding, or delaying surgical intervention in the relatively small subgroup with refractory epilepsy, it is not surprising that surgical versus medical outcome studies in this cohort of patients have also not been reported to date. While seizure characteristics in this group of patients have been analyzed, there is also a lack of information in the current literature with regard to the specific seizure or radiological features that may predict a better response to AEDs or surgery. 5. Conclusions This review on isolated AE in epilepsy included eight key papers and, in total, 107 unique patients. Most papers provide reasonable evidence of the enlarged amygdalae as the epileptic focus and provide consistent syndrome characteristics with a later age of onset, more CPS compared with CS, and a high likelihood of being responsive to AED treatment. Overall, however, sample sizes were small, suggesting that caution should be applied when attempting to draw conclusions with regard to patients with epilepsy with AE. A major limitation of these studies is the lack of homogeneity in ascertainment of AE or in the measured values of AMVs. An operational definition of this epilepsy syndrome of isolated AE in TLE cannot yet be made at this stage given these limitations. Further research is also warranted with respect to associated comorbidity, especially psychiatric, underlying etiology, and the selection and timing of candidates for surgical intervention in this group of patients. Disclosure of conflict of interest Wendyl D'Souza has received travel, investigator-related, speaker honoraria and has served on a scientific advisory board from UCB Pharma; educational grants from Novartis Pharmaceuticals, Pfizer Pharmaceuticals and Sanofi–Synthelabo; educational, travel, and fellowship grants from GSK Neurology Australia; has served on scientific advisory boards and received honoraria from SciGen Pharmaceuticals. His

S.M.J. Beh et al. / Epilepsy & Behavior 60 (2016) 33–41 [31] Phi JH, Kim SK, Cho BK, Lee SY, Park SY, Park SJ, et al. Long-term surgical outcomes of temporal lobe epilepsy associated with low-grade brain tumors. Cancer 2009;115: 5771–9. [32] Cook MJ, Fish DR, Shorvon SD, Straughan K, Stevens JM. Hippocampal volumetric and morphometric studies in frontal and temporal lobe epilepsy. Brain 1992; 115(Pt 4):1001–15. [33] Cascino GD, Jack Jr CR, Parisi JE, Sharbrough FW, Hirschorn KA, Meyer FB, et al. Magnetic resonance imaging-based volume studies in temporal lobe epilepsy: pathological correlations. Ann Neurol 1991;30:31–6. [34] Cascino GD. Clinical correlations with hippocampal atrophy. Magn Reson Imaging 1995;13:1133–6. [35] Cendes F, Andermann F, Gloor P, Evans A, Jones-Gotman M, Watson C, et al. MRI volumetric measurement of amygdala and hippocampus in temporal lobe epilepsy. Neurology 1993;43:719–25. [36] Burton EJ, Karas G, Paling SM, Barber R, Williams ED, Ballard CG, et al. Patterns of cerebral atrophy in dementia with Lewy bodies using voxel-based morphometry. Neuroimage 2002;17:618–30. [37] Bertoni MA, Sclavi NE. Isovoxel-based morphology of hippocampi and amygdalae: a comparison of manual and automated volume measurements. Neuroradiol J 2010;23:711–7. [38] Morey RA, Petty CM, Xu Y, Hayes JP, Wagner 2nd HR, Lewis DV, et al. A comparison of automated segmentation and manual tracing for quantifying hippocampal and amygdala volumes. Neuroimage 2009;45:855–66. [39] Brabec J, Rulseh A, Hoyt B, Vizek M, Horinek D, Hort J, et al. Volumetry of the human amygdala — an anatomical study. Psychiatry Res 2010;182:67–72. [40] Bhatia S, Bookheimer SY, Gaillard WD, Theodore WH. Measurement of whole temporal lobe and hippocampus for MR volumetry: normative data. Neurology 1993;43:2006–10. [41] Menzler K, Thiel P, Hermsen A, Chen X, Benes L, Miller D, et al. The role of underlying structural cause for epilepsy classification: clinical features and prognosis in mesial temporal lobe epilepsy caused by hippocampal sclerosis versus cavernoma. Epilepsia 2011;52:707–11. [42] Baulac M. MTLE with hippocampal sclerosis in adult as a syndrome. Rev Neurol (Paris) 2015;171:259–66.

41

[43] Bruton C. The neuropathology of temporal lobe epilepsy. New York: Oxford University Press; 1988. [44] Wieser HG. Mesial temporal lobe epilepsy versus amygdalar epilepsy: late seizure recurrence after initially successful amygdalotomy and regained seizure control following hippocampectomy. Epileptic Disord 2000;2:141–52. [45] Coan AC, Morita ME, Campos BM, Bergo FP, Kubota BY, Cendes F. Amygdala enlargement occurs in patients with mesial temporal lobe epilepsy and hippocampal sclerosis with early epilepsy onset. Epilepsy Behav 2013;29:390–4. [46] Bien CG, Urbach H, Schramm J, Soeder BM, Becker AJ, Voltz R, et al. Limbic encephalitis as a precipitating event in adult-onset temporal lobe epilepsy. Neurology 2007; 69:1236–44. [47] Harvey AS, Grattan-Smith JD, Desmond PM, Chow CW, Berkovic SF. Febrile seizures and hippocampal sclerosis: frequent and related findings in intractable temporal lobe epilepsy of childhood. Pediatr Neurol 1995;12:201–6. [48] Kuks JB, Cook MJ, Fish DR, Stevens JM, Shorvon SD. Hippocampal sclerosis in epilepsy and childhood febrile seizures. Lancet 1993;342:1391–4. [49] Semah F, Picot MC, Adam C, Broglin D, Arzimanoglou A, Bazin B, et al. Is the underlying cause of epilepsy a major prognostic factor for recurrence? Neurology 1998; 51:1256–62. [50] Stephen LJ, Kwan P, Brodie MJ. Does the cause of localisation-related epilepsy influence the response to antiepileptic drug treatment? Epilepsia 2001;42:357–62. [51] Gaitatzis A, Trimble MR, Sander JW. The psychiatric comorbidity of epilepsy. Acta Neurol Scand 2004;110:207–20. [52] DeToffol BCP, Praline J, Mondon K. The burden of psychiatric comorbidity in epilepsy. Epilepsies 2007;19:231–41. [53] Pauli C, Thais ME, Claudino LS, Bicalho MA, Bastos AC, Guarnieri R, et al. Predictors of quality of life in patients with refractory mesial temporal lobe epilepsy. Epilepsy Behav 2012;25:208–13. [54] Elsharkawy AE, May T, Thorbecke R, Koch-Stoecker S, Villagran A, Urak L, et al. Longterm outcome and determinants of quality of life after temporal lobe epilepsy surgery in adults. Epilepsy Res 2009;86:191–9. [55] Lowe AJ, David E, Kilpatrick CJ, Matkovic Z, Cook MJ, Kaye A, et al. Epilepsy surgery for pathologically proven hippocampal sclerosis provides long-term seizure control and improved quality of life. Epilepsia 2004;45:237–42.