- Email: [email protected]
Volumetry of amygdala and hippocampus and memory performance in Alzheimer's disease
Psychiatry Research: Neuroimaging 146 (2006) 251 – 261 www.elsevier.com/locate/psychresns
Volumetry of amygdala and hippocampus and memory performance in Alzheimer's disease Michael Bassoa , John Yanga , Lauren Warrena , Martha G. MacAvoya , Pradeep Varmac , Richard A. Bronenc , Christopher H. van Dycka,b,⁎ a
Alzheimer's Disease Research Unit, Department of Psychiatry, Yale University School of Medicine, New Haven, CT, 06510, USA b Department of Neurobiology, Yale University School of Medicine, New Haven, CT, USA c Department of Diagnostic Radiology, Yale University School of Medicine, New Haven, CT, USA Received 30 July 2005; received in revised form 3 January 2006; accepted 8 January 2006
Abstract Magnetic resonance imaging (MRI) is showing increased utility in examining medial temporal lobe atrophy and its relationship to memory performance in Alzheimer's disease (AD). We studied 56 AD patients and 42 older healthy subjects with neuropsychological assessment and MRI. Hippocampal and amygdaloid volumes (normalized to intracranial volume) were contrasted between AD patients and healthy controls and correlated with neuropsychological performance. Comparisons between AD patients and healthy controls revealed highly significant differences in the normalized volume of hippocampus and amygdala by analysis of covariance. Group differences tended to be at least as large for amygdaloid as hippocampal volume, including when the subset of AD patients with the mildest symptoms was considered separately. Within the AD group, performance on the Memory–Orientation subscale of the Alzheimer's Disease Assessment Scale-Cognition (ADAS-Cog) was significantly correlated with normalized amygdaloid volume but not with normalized hippocampal volume. Other ADAS-Cog subscales (Language, Praxis) were uncorrelated with either volume. In the healthy control sample, neither hippocampal nor amygdaloid volumes were significant predictors of any neuropsychological measure. While a substantial literature continues to justify the focus on the hippocampus in MRI studies of AD, these results suggest that the amygdala should receive similar attention, including in studies of the prodromal stages of AD. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Alzheimer's disease; Magnetic resonance imaging; Amygdala; Hippocampus; Memory
1. Introduction The neuropathological hallmarks of Alzheimer's disease (AD) include neurofibrillary tangles, β-amyloid ⁎ Corresponding author. Alzheimer's Disease Research Unit, Department of Psychiatry, Yale University School of Medicine, One Church Street, Suite 600, New Haven, CT 06510, USA. Tel.: +1 203 764 8100; fax: +1 203 764 8111. E-mail address: [email protected] (C.H. van Dyck).
plaques, and neuronal and synaptic loss. Although these findings occur in a range of brain structures, the site of earliest and most severe pathology in AD is the medial temporal lobe (Hyman et al., 1990; Braak and Braak, 1991). These pathological features have been found to correlate with the severity of dementia (Wilcock and Esiri, 1982; Terry et al., 1991), and their early occurrence in the medial temporal lobe may account for the memory impairment that is typically the earliest and most prominent symptom in AD (Hyman et al., 1990).
0925-4927/$ - see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.pscychresns.2006.01.007
252
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
Although a link between the medial temporal lobes and memory is well established (Squire et al., 2004), the precise role of specific medial temporal lobe structures in memory functions continues to be elucidated. The hippocampus is believed to play a central role in declarative memory (Squire et al., 2004), and this is borne out by most human (Zola-Morgan et al., 1986; Press et al., 1989) and animal (Alvarez et al., 1995; Beason-Held et al., 1999; Zola et al., 2000) lesion studies. The role of the amygdala in emotional memory is well established (LeDoux, 1993), but that in declarative memory is less clear. Amygdala lesions produce declarative memory impairment in humans (Tranel and Hyman, 1990; Markowitsh et al., 1994) and nonhuman primates (Murray and Gaffan, 1994), although one primate study suggests that lesions of the amygdala that spare adjacent cortical regions do not impair memory (Zola-Morgan et al., 1989a). In recent years, magnetic resonance imaging (MRI) has been used to examine medial temporal lobe atrophy and its relationship to memory performance in AD. These studies have confirmed volume differences between AD patients and healthy controls for the hippocampus (Jack et al., 1992; Lehéricy et al., 1994; Jack et al., 1997; Mori et al., 1997; Krasuski et al., 1998; Laakso et al., 2000; Mizuno et al., 2000; Pennanen et al., 2004) and/or the amygdala (Cuénod et al., 1993; Lehéricy et al., 1994; Laakso et al., 1995a; Maunoury et al., 1996; Mori et al., 1997; Krasuski et al., 1998; Mauri et al., 1998; Mizuno et al., 2000). Studies of mild cognitive impairment (considered prodromal for AD in most cases) have generally neglected the amygdala and focused on hippocampal volume (Jack et al., 1999; Grundman et al., 2003; Pennanen et al., 2004), despite the fact that those studies that have examined both hippocampal and amygdaloid volumes in the mildest stages of AD suggest that amygdaloid deficits are at least as robust (Lehéricy et al., 1994; Krasuski et al., 1998; Mizuno et al., 2000). Some MRI studies have also shown a correlation between hippocampal (Deweer et al., 1995; Laakso et al., 1995b; Fama et al., 1997; Laakso et al., 2000; Petersen et al., 2000; Kramer et al., 2004) or amygdaloid (Mori et al., 1997; Mizuno et al., 2000) volumes and memory performance in AD. In the present study, we compared normalized hippocampal and amygdaloid volumes by MRI between older healthy subjects and patients with AD, including a subsample of patients with the mildest symptoms. We examined the relationship between these volumes and an index of memory performance in the AD patients, hypothesizing that memory performance would be
correlated with both hippocampal and amygdaloid volumes. We examined the specificity of these correlations by contrasting them with non-memory cognitive performance and also by examining similar correlates in the older control sample. 2. Methods The study sample comprised 56 patients with probable AD who received MRI scanning in research protocols in the Yale Alzheimer's Disease Research Unit (ADRU). Patients had been referred to the ADRU from a variety of sources or were self-referred. Five of these patients have subsequently died and had autopsy confirming definite AD (Mirra et al., 1991). Forty-two healthy elderly control subjects (most spouses of participating AD patients) were recruited for MRI scanning. The demographics and clinical characteristics of patients and controls are displayed in Table 1. Two AD patients were African American; all other patients and controls were of European ancestry. All patients and controls underwent a comprehensive evaluation by a research physician and ancillary staff, including cognitive assessment, medical history, physical and neurological examinations, serum chemistries, thyroid function studies, complete blood count, B12, folate, VDRL, urinalysis, electrocardiogram, and clinical reading of brain MRI. AD patients were required to meet standard diagnostic criteria for probable AD (McKhann et al., 1984), and controls were required to be in good health for their age. Subjects were excluded for any neurological or medical disorder (other than AD Table 1 Subject characteristics Variable
Demographics Age Gender Handedness Education (years) Neuropsychological MMSE ADAS-Cog Disease characteristics Onset age Duration (years) Family history
Healthy (N = 42)
Alzheimer's (N = 56)
Mean ± S.D.
Mean ± S.D.
73.2 ± 6.7 22M, 20F 42R 14.3 ± 3.4
71.2 ± 8.6 28M, 28F 54R, 2L 14.1 ± 3.3
29.0 ± 1.0 5.1 ± 1.8
*18.3 ± 4.3 *27.3 ± 10.5 66.6 ± 8.6 4.6 ± 1.5 25+, 31−
MMSE = Mini-Mental State Examination; ADAS-Cog = Alzheimer's Disease Assessment Scale — Cognitive Subscale. Family history is positive if primary degenerative dementia is present in at least one first degree relative. *Differs from control value, P b 10–7, t-test.
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
in the patient group) that could produce cognitive deterioration or for significant psychiatric illness, alcohol, or substance abuse. Healthy subjects were also required to have no significant evidence of cognitive impairment, as indicated by a Mini-Mental State Examination (MMSE) (Folstein et al., 1975) score ≥ 27 and a normal brain MRI scan. Neuropsychological testing administered for subject characterization (see Table 1) included the MMSE (Folstein et al., 1975) and the cognitive subscale of the Alzheimer's Disease Assessment Scale (ADAS-Cog; range = 0–70, lower score signifies less impairment) (Rosen et al., 1984). The ADAS-Cog was divided into three subscales: (1) Memory–Orientation (the sum of the ratings for “orientation,” “word recall,” “word recognition,” and “recall of test instructions”; range of scores = 0–35), (2) Language (the sum of the ratings for “difficulty making self understood,” “comprehension of spoken language,” “word-finding difficulty,” “commands,” and “naming”; range = 0–25), and (3) Praxis (the sum of the ratings for “constructional praxis” and “ideational praxis”; range = 0–10). These subscales are similar to those previously used (Lawlor et al., 1994) but combined Memory and Orientation as declarative memory functions hypothesized to be related to hippocampal and amygdaloid volumes in AD. Family history of AD was assessed using the Alzheimer Dementia Risk Questionnaire (Breitner and Folstein, 1984) and the Dementia Questionnaire (Silverman et al., 1986), and was considered to be positive if at least one first degree relative met criteria for primary degenerative dementia. No cases suggestive of autosomal dominant inheritance were identified. All subjects (or their responsible next of kin) provided written informed consent, and were studied under protocols approved by the Yale Human Investigation Committee and conducted in accordance with the Declaration of Helsinki. 2.1. MRI methods Subjects were imaged with a 1.5-Tesla General Electric Signa scanner (General Electric, Milwaukee, WI) according to a standardized protocol. The protocol began with a T1-weighted sagittal pulse sequence (5mm contiguous slices, TR = 600 ms, TE = 11 ms, matrix = 256 × 192, number of signal averages = 1, field of view = 24 cm) that was used to measure intracranial volume (ICV) and to localize subsequent acquisitions. The pulse sequence used to measure volumes of the hippocampus and the amygdala was a T1-weighted, three-dimensional volume spoiled gradient echo axial
253
sequence through the temporal lobes (sixty 1.5-mm contiguous slices, TR = 24 ms, TE = 5 ms, matrix = 256 × 192, number of signal averages = 2, field of view = 30 cm, and a 45° flip angle). Images were transferred to a Sun SPARC 2 workstation (Sun Microsystems, Santa Clara, CA) where volumetric measurements were performed using ANALYZE software (Robb et al., 1989) by the same rater (MB) who was blinded to all subject information but was not blinded to hemisphere (left or right). The axial pulse sequence data were rotated and re-sliced into 1.2-mm isotropic voxel coronal sections orthogonal to the long axis of the hippocampus (matrix = 256 × 256). Any head rotation with respect to the orthogonal coronal plane was also corrected during this step. In the resulting orthogonal coronal slices, the borders of the hippocampus and the amygdala were then traced sequentially from posterior to anterior using a mouse track ball. The ICV was delineated along the inner border of the skull from the sagittal pulse sequence. The ICV was defined as all tissue interior to the dura mater, including the pituitary gland and its surrounding cerebrospinal fluid. The foramen magnum defined the inferior border of the brainstem. All bony structures outside of the brain were disarticulated. The region-of-interest module of ANALYZE calculated volumes for each slice, multiplying the cross-sectional area by the slice thickness. Final hippocampal and amygdaloid volumes and ICVs were calculated by summing the volumes for each structure across all slices. Individual hippocampal and amygdaloid volumes were normalized for intersubject variation in head size by dividing by the ICV of each subject. 2.2. Anatomic guidelines Neuroanatomic guidelines for hippocampus and amygdala (see Fig. 1) were adapted from those of Watson et al. (1992) and the atlases of Duvernoy (1991, 1998) in consultation with a neuroradiologist expert in medial temporal lobe anatomy (RB). The hippocampal volume included the CA1 to CA4 fields of the hippocampus proper, the dentate gyrus, the subiculum, the alveus, and the fimbria. To improve reliability, the posterior margin of the hippocampus was arbitrarily defined by the first oblique coronal section that contained the thalamus. At this level the crura of the fornices are usually observable in profile and serve as an additional landmark. The anterior margin of the hippocampus is determined by the alveus that covers the ventricular surface of the hippocampal head at the uncal recess within the temporal horn. When this alveus
254
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
Fig. 1. Anatomic margins of hippocampus and amygdala on MRI. Selected images from a 69-year-old female control subject. Panel (A) Oblique coronal section through the hippocampus (designated by H), which was defined to include the CA1 to CA4 fields of the hippocampus proper, the dentate gyrus, the subiculum, the alveus, and the fimbria. Panel (B) Oblique coronal section through the amygdala (designated by A), which was defined to include the amygdala proper, the semilunar gyrus and a small portion of the ambient gyrus.
was not evident, we used the criteria of Watson et al. (1992) to differentiate the hippocampal head from the amygdala. Thus these measurements included essentially the entire hippocampus from the tail through the head. Neuroanatomic boundaries of the amygdala were based closely on those of Watson et al. (1992) and were delineated to include the amygdala proper (the deep/ basolateral and superficial/corticomedial nuclei, and other nuclei), the semilunar gyrus (corresponding to the cortical amygdaloid nucleus) and a small portion of the ambient gyrus (corresponding to the olfactory portion of the entorhinal cortex). Gray–white matter junctions identified the posterior, superior, and lateral borders of the amygdala. Either the uncal recess of the temporal horn or the alveus covering the hippocampal head defined the inferior border. The medial aspect of the amygdala is covered by a portion of entorhinal cortex, which forms the surface of the ambient gyrus in this region. The entorhinal cortex inferior to the tentorial indentation was excluded from the amygdaloid volume. If this indentation was poorly visualized in the anterior amygdaloid region, then it was approximated by a line drawn directly from the inferior and medial border of the amygdala to the supracellar cistern. This method inevitably included a small portion of the superior entorhinal cortex in the amygdaloid measurement
(Watson et al., 1992). The anterior boundary of the amygdala is poorly visualized on MRI and was operationally defined as the anterior most section before the closure of the lateral sulcus to form the endorhinal sulcus. This procedure potentially excludes part of the anterior amygdala but is thought to yield greater reliability (Watson et al., 1992). 2.3. Inter-rater reliability To evaluate inter-rater reliability, a second investigator (JY) also analyzed a subset of the AD patient scans (n = 50 for amygdala, n = 27 for hippocampus, n = 31 for ICV). This analysis yielded intraclass correlation coefficients (ρ) (Kirk, 1982) of 0.85 for left amygdala, 0.80 for right amygdala, 0.90 for left hippocampus, 0.84 for right hippocampus, and 0.95 for ICV, demonstrating satisfactory reliability. No healthy control scans were included in the inter-rater analysis. 2.4. Statistical analysis Demographic and neuropsychological variables were compared between AD and healthy control groups by chi-square or Student's t test. The effect of diagnostic group and brain structure (hippocampus, amygdala) on
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
normalized volume was determined by two-way analysis of covariance (ANCOVA) with repeated measures on brain structure, and age and education as covariates. To determine whether significant volume deficits were present in the mildest AD patients, the quartile of patients with the highest MMSE scores (range 23–27, n = 13) were compared with the control sample by the same ANCOVA. Volume deficits in the AD patient group were expressed as both mean percentages and mean z-scores. Since some previous researchers have observed significant asymmetries between left and right hemispheres for hippocampal and amygdaloid volumes, we also compared left and right volumes by paired t tests for each diagnostic group. We further examined whether asymmetry differed between AD patients and controls, by comparing an asymmetry index (100 × [right − left] / [right + left] / 2) between groups, using Student's t test. Within the AD group, the relationship between neuropsychological measures (dependent variables) and normalized hippocampal and amygdaloid volumes (combining left and right hemispheres) was examined by stepwise multiple linear regression (with P b 0.10 required for retention in the model), considering the contribution of age, sex, and education. We hypothesized that the Memory–Orientation subscale of the ADASCog (but not the other ADAS-Cog subscales: Language, Praxis) would be associated with smaller normalized hippocampal and amygdaloid volumes in the AD patients (and thus we applied a Bonferroni correction for two planned comparisons α = 0.05 / 2 = 0.025). However, preliminary analysis revealed that the contributions of age, education, and gender were not retained in the models for the hypothesis-driven subscores, thus reducing to simple bivariate correlations (Pearson's r) between neuropsychological measures and normalized hippocampal and amygdaloid volumes. A similar investigation of the relationship between neuropsychological measures and normalized hippocampal and amygdaloid volumes was conducted for the healthy control group. All statistical analyses used the SPSS (SPSS Inc., Chicago, IL) software package and twotailed tests of significance.
255
significantly worse on the MMSE and the ADAS-Cog (P b 10− 7, Student's t test). 3.1. Effect of diagnostic group on normalized hippocampal and amygdaloid volumes Absolute volumes (mean ± S.D.) of hippocampus and amygdala were as follows: healthy controls — left hippocampus 2.158 ± 0.359 ml, right hippocampus 2.247 ± 0.377 ml, left amygdala 2.032 ± 0.435 ml, right amygdala 2.160 ± 0.497 ml; AD patients — left hippocampus 1.737 ± 0.486 ml, right hippocampus 1.813 ± 0.478 ml, left amygdala 1.530 ± 0.410 ml, right amygdala 1.580 ± 0.475 ml. ANCOVA revealed significant main effects of diagnosis (F = 46.69; df = 1,94; P b 0.001) and brain structure (F = 14.84; df = 1,96; P b 0.001). Planned contrasts demonstrated highly significant group differences for hippocampus (combining left and right hemispheres, dividing by the ICVof each subject, P b 0.001; see Fig. 2) and amygdala (P b 0.001; see Fig. 3). The diagnosis by structure interaction term (F = 1.09; df = 1,96; P = 0.30) indicated that there was no significant difference in amygdaloid vs. hippocampal volume reductions in the AD patients compared with the controls. Mean differences between healthy controls and AD patients were 17% for hippocampus (z = − 0.99) and 23% for amygdala (z = − 1.28). The effect of age (t = − 2.70, P = 0.008) but not education (t = − 1.75, P = 0.084) was significant in the ANCOVA model. Normalized volumes of hippocampus and amygdala were significantly correlated with each
3. Results The characteristics of AD patients and healthy controls are shown in Table 1. The two groups did not differ in age (t = 1.24, df = 96, P = 0.22), gender distribution (χ 2 = 0.0009, df = 1, P = 0.98), handedness (χ 2 = 0.27, df = 1, P = 0.61), or years of education (t = 0.23, df = 96, P = 0.82). The AD patients scored
Fig. 2. Scatter plot of normalized hippocampal volume (combined left and right hippocampal volume divided by intracranial volume) in healthy control subjects (n = 42) and patients with Alzheimer's disease (AD; n = 56). Each dash denotes an individual patient; solid circles denote treatment group means (± S.D.). The AD patients had significantly lower normalized hippocampal volume than the healthy subjects, controlling for age and education (P b 0.001, ANCOVA).
256
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
and healthy subjects (t = 1.18, df = 96, P = 0.24, Student's t test). 3.3. Effect of hippocampal and amygdaloid volumes on cognitive measures in AD
Fig. 3. Scatter plot of normalized amygdaloid volume (combined left and right amygdaloid volume divided by intracranial volume) in healthy control subjects (n = 42) and patients with Alzheimer's disease (AD; n = 56). Each dash denotes an individual patient; solid circles denote treatment group means (± S.D.). The AD patients had significantly lower normalized amygdaloid volume than the healthy subjects, controlling for age and education (P b 0.001, ANCOVA).
other in the AD patient group (r = 0.48, n = 56, P b 0.001), but not in the control group (r = 0.21, n = 42, P = 0.18). When only the quartile of AD patients with the highest MMSE scores (range 23–27, n = 13) was compared with the control sample, ANCOVA again revealed significant main effects of diagnosis (F = 7.27; df = 1,51; P = 0.009) and brain structure (F = 4.17; df = 1,53; P b 0.046). Planned contrasts showed significant group differences for amygdala (P b 0.016) but not for hippocampus (P b 0.095). However, the diagnosis by structure interaction term (F = 0.14; df = 1,53; P = 0.71) indicated that the difference in amygdaloid vs. hippocampal volume reductions was not significant. Mean differences between healthy controls and these very mild AD patients were 10% for hippocampus (z = − 0.57) and 13% for amygdala (z = − 0.75).
Within the AD group, the relationship between normalized hippocampal and amygdaloid volumes and neuropsychological measures was examined by bivariate correlational analysis (Pearson's r). The results of that analysis are displayed in Table 2. In the two hypothesisdriven comparisons, normalized hippocampal volume was found to be uncorrelated with the Memory– Orientation subscale of the ADAS-Cog (r = − 0.22, n = 56, P = 0.097), whereas normalized amygdaloid volume was significantly correlated with Memory– Orientation (r = − 0.40, P = 0.002; see Fig. 4). Hippocampal and amygdaloid volumes were uncorrelated with other ADAS-Cog subscales (Language, Praxis). Post hoc analyses were undertaken to understand the unexpected absence of a correlation between normalized hippocampal volume and the Memory–Orientation subscale, and to determine if this was influenced by (1) severity of dementia or (2) specific components of the Memory–Orientation subscale. The AD patient sample was dichotomized by MMSE score with the hypothesis that a correlation would emerge in the more impaired patients. However, the correlation was absent both for patients with MMSE scores ≥ 19 (r = − 0.27, n = 25, P = 0.19) and for patients with MMSE scores b 19 (r = 0.11, n = 31, P = 0.056). Memory–Orientation subtest components (“orientation,” “word recall,” “word recognition,” and “recall of test instructions”) were considered separately, with the hypothesis that “word recognition,” in particular, would invoke a hippocampal substrate when the overall score did not. However, the
3.2. Comparison of left and right hippocampal and amygdaloid volume Right hippocampal volume was significantly larger than left hippocampal volume for both the healthy control sample (t = 2.75, df = 41, P = 0.009, paired t test) and the AD sample (t = 2.48, df = 55, P = 0.016, paired t test). The hippocampal asymmetry index did not differ between AD and healthy subjects (t = 0.37, df = 96, P = 0.72, Student's t test). Right amygdaloid volume was larger than left amygdaloid volume for the healthy control sample (t = 2.89, df = 41, P = 0.006, paired t test) but not for the AD sample (t = 1.30, df = 55, P = 0.20, paired t test). However, the amygdaloid asymmetry index did not differ between AD
Table 2 Correlations between normalized hippocampal and amygdaloid volumes and neuropsychological measures in patients with AD Neuropsychological tests
Hippocampus
Amygdala
Pearson's r P value Pearson's r P value
ADAS-Cog Memory–Orientation − 0.22 Language − 0.08 Praxis − 0.21 Total 0.21 MMSE 0.19
0.097 0.559 0.113 0.120 0.150
− 0.40 − 0.20 − 0.10 − 0.34 0.34
0.002 0.132 0.446 0.011 0.011
ADAS-Cog, Alzheimer's Disease Assessment Scale — Cognitive Subscale (lower score better); MMSE, Mini-Mental State Examination.
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
257
loid normalized volumes entered into the model as significant predictor variables. 4. Discussion
Fig. 4. Memory–Orientation subscores (range = 0–35; lower score signifies less impairment) from the Alzheimer's Disease Assessment Scale (ADAS-Cog) versus normalized amygdaloid volume (combined left and right amygdaloid volume divided by intracranial volume) in Alzheimer's disease patients (n = 56). Amygdaloid volume was significantly correlated with Memory–Orientation (r = − 0.40, P = 0.002) but was uncorrelated with other ADAS-Cog subscores (Language, Praxis).
This study compared normalized hippocampal and amygdaloid volumes between patients with AD and older healthy subjects and examined neuropsychological correlates with MRI measures in the AD patients. Highly significant group differences were observed in normalized hippocampal and amygdaloid volumes. Group differences tended to be at least as large for amygdaloid as hippocampal volume, including when the subset of AD patients with the mildest symptoms (MMSE ≥ 23) was considered separately. Within the AD group, the Memory–Orientation subscale of the ADAS-Cog (but not other ADAS-Cog subscales) was significantly correlated with amygdaloid volume, but it only showed a nonsignificant trend-level correlation with hippocampal volume. 4.1. Previous studies of patient–control differences
“word recognition” item was also uncorrelated with hippocampal volume (r = − 0.16, n = 56, P = 0.24). Normalized amygdaloid volume was also correlated with total ADAS-Cog score (r = −0.34, P = 0.011) and MMSE score (r = 0.34, P = 0.011); whereas normalized hippocampal volume was uncorrelated with either measure. Although, as previously mentioned, the model of ADAS-Cog subscales was unimproved by the inclusion of age, sex, or education in a multiple linear regression analysis, the models of total ADASCog and MMSE scores were improved slightly by the addition of education in multiple linear regression models. However, the overall results did not differ with the inclusion of education (data not shown). When left and right hippocampal and amygdaloid volumes were considered separately in an exploratory analysis, the results did not differ (i.e., all significant findings for amygdala were true for both left and right amygdala, and no significant correlations emerged for left or right hippocampus). In the healthy control sample, the relationship between neuropsychological measures (dependent variables) and normalized hippocampal and amygdaloid volumes was also examined by stepwise multiple linear regression, considering the contribution of age, sex, and education. For the Memory–Orientation subscale of the ADAS-Cog (and all other neuropsychological measures: Language, Praxis subscales, total ADAS-Cog score, and MMSE), neither hippocampal nor amygda-
Numerous previous studies have established volume differences between AD patients and healthy controls for hippocampus (Jack et al., 1992; Lehéricy et al., 1994; Laakso et al., 1995b; Jack et al., 1997; Mori et al., 1997; Krasuski et al., 1998; Laakso et al., 2000; Mizuno et al., 2000; Pennanen et al., 2004) and/or amygdala (Cuénod et al., 1993; Lehéricy et al., 1994; Laakso et al., 1995a; Maunoury et al., 1996; Mori et al., 1997; Krasuski et al., 1998; Mauri et al., 1998; Mizuno et al., 2000). In agreement with the present results, studies that have measured hippocampal and amygdaloid volumes in the same samples have consistently found deficits in amygdaloid volume that are at least as large as those for hippocampus (Lehéricy et al., 1994; Mori et al., 1997; Krasuski et al., 1998; Mizuno et al., 2000), including in the mild (Lehéricy et al., 1994; Krasuski et al., 1998) or very mild (Mizuno et al., 2000) stages of AD. These results suggest that MRI studies of prodromal AD (mild cognitive impairment) should give greater attention to the amygdala. Such studies have focused almost exclusively on hippocampal volume as an earlier biomarker (Jack et al., 1999; Grundman et al., 2003; Pennanen et al., 2004). We found significantly larger right than left hippocampal volume in the healthy control group as did some (Jack et al., 1989; Krasuski et al., 1998; Geroldi et al., 2000; Pruessner et al., 2000; Jack et al., 2003; Sullivan et al., 2005) but not all (Raz et al., 2004) previous investigators. In contrast to one previous study (Geroldi
258
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
et al., 2000), we also observed the same asymmetry in the AD sample. We found significantly larger right than left amygdaloid volume in the healthy control group as did some (Watson et al., 1992; Filipek et al., 1994; Krasuski et al., 1998) but not all (Pruessner et al., 2000; Whitwell et al., 2005) previous investigators. While this “normal asymmetry” was not present in the AD group, the extent of amygdaloid asymmetry did not differ significantly between AD and healthy subjects. Other investigators have reported a reversed left N right asymmetry in both controls (Laakso et al., 1995b; Convit et al., 1999) and AD patients (Laakso et al., 1995b). 4.2. Previous studies of neuropsychological correlates We observed a significant correlation between an index of recent memory performance (Memory–Orientation subscale of the ADAS-Cog) and amygdaloid but not hippocampal volume. Our findings are therefore similar to those of Mori et al. (1997) who also observed memory performance (Wechsler Memory Scale-Revised, WMS-R subtests) to be correlated with amygdaloid but not hippocampal volumes in 46 mild to moderate AD patients. However, unlike Mori et al. (1997) who observed memory performance to be correlated with right amygdaloid volume, we found memory performance to be equally correlated with left and right amygdaloid volume (data not shown). Our results are also somewhat similar to those of Mizuno et al. (2000), who observed that WMS-R measures were correlated with amygdaloid but not hippocampal volume in a pooled sample of healthy control subjects and patients with very mild AD. However, these authors did not present separate analyses of patients and controls. Heun et al. (1997) also observed that memory dysfunction was highly correlated with atrophy of the combined amygdala–hippocampal complex in AD patients but did not segregate amygdaloid and hippocampal volumes. However, our results are at odds with Deweer et al. (1995) and Mauri et al. (1998) who observed no correlations between normalized volume of amygdala and several measures of memory performance, although Deweer et al. (1995) employed 5-mm slice thickness, which is probably inadequate for volumetry. Although we found only a nonsignificant trend for a correlation between Memory–Orientation and hippocampal volume, other investigators have demonstrated significant correlations between hippocampal volume and memory performance in AD patients (Deweer et al., 1995; Laakso et al., 1995b; Fama et al., 1997; Laakso et
al., 2000; Petersen et al., 2000; Kramer et al., 2004) and even in patients with mild cognitive impairment (Grundman et al., 2003). The absence of any significant correlation between Memory–Orientation and MRI measures in the control group is consistent with previous studies reporting no significant relationship between hippocampal volume (Petersen et al., 2000) and cognitive measures in healthy older subjects. 4.3. Relationship to human and animal lesion studies Although the link between the medial temporal lobes and memory is well established, human and animal lesion studies have provided an increasingly complex understanding of the tissue damage that is necessary and sufficient to produce impaired memory performance (Squire et al., 2004). Lesions of the amygdala result in memory deficits in humans (Tranel and Hyman, 1990; Markowitsh et al., 1994) and nonhuman primates (Murray and Gaffan, 1994). However, studies in nonhuman primates suggest that lesions of the amygdala that spare adjacent cortical regions do not impair memory (Zola-Morgan et al., 1989a). Damage restricted to the hippocampus has been reported to result in memory deficits in humans (Zola-Morgan et al., 1986; Press et al., 1989) and nonhuman primates (Alvarez et al., 1995; Beason-Held et al., 1999; Zola et al., 2000); however, one study observed no significant memory impairment in monkeys with restricted lesions of amygdala and hippocampus (Murray and Mishkin, 1998). Damage to adjacent (perirhinal and parahippocampal (Zola-Morgan et al., 1989b; Suzuki et al., 1993) or to entorhinal and perirhinal (Meunier et al., 1993; Eacott et al., 1994)) cortices that spare the amygdala and hippocampal formation consistently produces severe memory impairment. Thus a significant correlation between normalized amygdaloid volume and memory performance in AD patients might not necessarily be expected from lesion studies. However, because the amygdaloid volume we measured includes some adjacent cortical regions – the semilunar gyrus (corresponds to the cortical amygdaloid nucleus) and a small portion of the ambient gyrus (which corresponds to the olfactory portion of the entorhinal cortex) – its correlation with memory performance may be more robust than predicted from animal lesions localized to the amygdala proper. It is possible that a software package that allowed simultaneous threedimensional visualization (Pruessner et al., 2000) would have permitted a more precise delineation of the amygdala proper. Furthermore, in AD damage to amygdala is likely correlated with damage to
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
hippocampus (e.g., normalized volumes of hippocampus and amygdala were significantly correlated in the present study: r = 0.48, n = 56, P b 0.001) as well as with other medial temporal lobe structures that were not measured in this study (e.g., entorhinal, perirhinal, and parahippocampal cortex). Therefore, greater atrophy of amygdala may be a surrogate for atrophy of other structures necessary for memory performance. The fact that normalized hippocampal volume was not significantly correlated with memory performance or global measures is somewhat contrary to human (Zola-Morgan et al., 1986; Press et al., 1989) and most (Alvarez et al., 1995; Beason-Held et al., 1999; Zola et al., 2000) animal studies. However, this finding is also at variance with many other human MRI studies of AD (Deweer et al., 1995; Laakso et al., 1995b; Fama et al., 1997; Laakso et al., 2000; Petersen et al., 2000; Kramer et al., 2004) and mild cognitive impairment (Grundman et al., 2003), and it may simply reflect a Type II error. Finally, the memory scale used in the present study, the Memory–Orientation Subscore of the ADASCog, is not a test of selective component processes of memory, e.g., storage versus retrieval or verbal versus visual memory. Therefore, its ability to relate specific memory processes to brain structures is limited. In summary, our results confirm previous reports of highly significant differences in normalized hippocampal and amygdaloid volumes between AD patients and healthy controls and suggest that amygdaloid volume deficits are at least as large as hippocampal deficits in the very mild stage of disease. Within the AD group, amygdaloid but not hippocampal volume was significantly correlated with the Memory–Orientation subscale of the ADAS-Cog. While a substantial literature continues to justify the focus on the hippocampus in MRI studies of AD, these results suggest that the amygdala should receive similar attention, including in studies of the prodromal stages of AD. Acknowledgments The authors thank Hedy Sarofin and Terri Hickey for assistance with MRI scanning, Larry Staib, Ph.D., for technical assistance, and Jung Kim, M.D., for histopathological diagnosis in a subset of the AD cases. References Alvarez, P., Zola-Morgan, S., Squire, L.R., 1995. Damage limited to the hippocampal region produces long-lasting memory impairment in monkeys. Journal of Neuroscience 15, 3796–3807.
259
Beason-Held, L., Rosene, D.L., R.J., K., Moss, M.B., 1999. Hippocampal formation lesions produce memory impairment in the rhesus monkey. Hippocampus 9, 562–574. Braak, H., Braak, E., 1991. Neuropathological stageing of Alzheimerrelated changes. Acta Neuropathologica 82, 239–259. Breitner, J.C.S., Folstein, M.F., 1984. Familial Alzheimer's disease: a prevalent disorder with specific clinical features. Psychological Medicine 14, 63–80. Convit, A., McHugh, P., Wolf, O.T., de Leon, M.J., Bobinski, M., De Santi, S., Roche, A., Tsui, W., 1999. MRI volume of the amygdala: a reliable method allowing separation from the hippocampal formation. Psychiatry Research: Neuroimaging 90, 113–123. Cuénod, C.A., Denys, A., Michot, J.L., Jehenson, P., Forette, F., Kaplan, D., Syrota, A., Boller, F., 1993. Amygdala atrophy in Alzheimer's disease. An in vivo magnetic resonance imaging study. Archives of Neurology 50, 941–945. Deweer, B., Lehericy, S., Pillon, B., Baulac, M., Chiras, J., Marsault, C., Agid, Y., Dubois, B., 1995. Memory disorders in probable Alzheimer's disease: the role of hippocampal atrophy as shown with MRI. Journal of Neurology, Neurosurgery and Psychiatry 58, 590–597. Duvernoy, H.M., 1991. The Human Brain: Surface, Three-dimensional Sectional Anatomy with MRI, and Blood Supply, 2nd edition. Springer-Verlag, Wien. Duvernoy, H.M., 1998. The Human Hippocampus: Functional Anatomy, Vascularization and Series Sections with MRI, 2nd edition. Springer-Verlag, Berlin. Eacott, M.J., Gaffan, D., Murray, E.A., 1994. Preserved recognition memory for small sets, and impaired stimulus identification for large sets, following rhinal cortex ablations in monkeys. European Journal of Neuroscience 6, 1466–1478. Fama, R., Sullivan, E.V., Shear, P.K., Marsh, L., Yesavage, J.A., Tinklenberg, J.R., Lim, K.O., Pfefferbaum, A., 1997. Selective cortical and hippocampal volume correlates of Mattis Dementia Rating Scale in Alzheimer disease. Archives of Neurology 54, 719–728. Filipek, P.A., Richelme, C., Kennedy, D.N., Caviness, V.S., 1994. The young adult human brain: an MRI-based morphometric analysis. Cerebral Cortex 4, 344–360. Folstein, M.F., Fostein, S.E., McHugh, P.R., 1975. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research 12, 189–198. Geroldi, C., Laakso, M.P., DeCarli, C., Beltramello, A., Bianchetti, A., Soininen, H., Trabucchi, M., Frisoni, G.B., 2000. Apolipoprotein E genotype and hippocampal asymmetry in Alzheimer's disease: a volumetric MRI study. Journal of Neurology, Neurosurgery and Psychiatry 68, 93–96. Grundman, M., Jack, C.R., Petersen, R.C., Kim, H.T., Taylor, C., Datvian, M., Weiner, M.F., DeCarli, C., DeKosky, S.T., van Dyck, C.H., Darvesh, S., Yaffe, K., Kaye, J., Ferris, S.H., Thomas, R.G., Thal, L.J., 2003. Hippocampal volume is associated with memory but not nonmemory cognitive performance in patients with mild cognitive impairment. Journal of Molecular Neuroscience 20, 241–248. Heun, R., Mazanek, M., Atzor, K.R., Tintera, J., Gawehn, J., Burkart, M., Gansicke, M., Falkai, P., Stoeter, P., 1997. Amygdala– hippocampal atrophy and memory performance in dementia of Alzheimer type. Dementia and Geriatric Cognitive Disorders 8, 329–336. Hyman, B.T., Van Hoesen, G.W., Damasio, A.R., 1990. Memoryrelated neural systems in Alzheimer's disease: an anatomic study. Neurology 40, 1721–1730.
260
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
Jack, C.R., Twomey, C.K., Zinsmeister, A.R., Sharbrough, F.W., Petersen, R.C., Cascino, G.D., 1989. Anterior temporal lobes and hippocampal formations: normative volumetric measurements from MR images in young adults. Radiology 172, 549–554. Jack, C.R., Petersen, R.C., O'Brien, P.C., Tangalos, E.G., 1992. MRbased hippocampal volumetry in the diagnosis of Alzheimer's disease. Neurology 42, 183–188. Jack, C.R., Petersen, R.C., Xu, Y.C., Waring, S.C., O'Brien, P.C., Tangalos, E.G., Smith, G.E., Ivnik, R.J., Kokmen, E., 1997. Medial temporal atrophy on MRI in normal aging and very mild Alzheimer's disease. Neurology 49, 786–794. Jack, C.R., Petersen, R.C., Xu, Y.C., O'Brien, P.C., Smith, G.E., Ivnik, R.J., Boeve, B.F., Waring, S.C., Tangalos, E.G., Kokmen, E., 1999. Prediction of AD with MRI-based hippocampal volume in mild cognitive impairment. Neurology 52, 1397–1403. Jack, C.R.J., Slomkowski, M., Gracon, S., Hoover, T.M., Felmlee, J.P., Stewart, K., Xu, Y., Shiung, M., O'Brien, P.C., Cha, R., Knopman, D., Petersen, R.C., 2003. MRI as a biomarker of disease progression in a therapeutic trial of milameline for AD. Neurology 60, 253–260. Kirk, R., 1982. Experimental Design: Procedures for the Behavioral Sciences. Brooks/Cole Publishing, Co., Pacific Grove, CA. Kramer, J.H., Schuff, N., Reed, B.R., Mungas, D., Du, A.T., Rosen, H.J., Jagust, W.J., Miller, B.L., Weiner, M.W., Chui, H.C., 2004. Hippocampal volume and retention in Alzheimer's disease. Journal of the International Neuropsychological Society 10, 639–643. Krasuski, J.S., Alexander, G.E., Horwitz, B., Daly, E.M., Murphy, D.G.M., Rapoport, S.I., Schapiro, M.B., 1998. Volumes of medial temporal lobe structures in patients with Alzheimer's disease and mild cognitive impairment (and in healthy control subjects). Biological Psychiatry 43, 60–68. Laakso, M.P., Partanen, K., Lehtovirta, M., Hallikainen, M., Hänninen, T., Vainio, P., Riekkinen, P., Soininen, H., 1995a. MRI of amygdala fails to diagnose early Alzheimer's disease. Neuroreport 6, 2414–2418. Laakso, M.P., Soininen, H., Partanen, K., Helkala, E.L., Hartikainen, P., Vainio, P., Hallikainen, M., Hanninen, T., Riekkinen Sr., P.J., 1995b. Volumes of hippocampus, amygdala and frontal lobes in the MRI-based diagnosis of early Alzheimer's disease: correlation with memory functions. Journal of Neural Transmission: Parkinson's Disease Dementia Section 9, 73–86. Laakso, M.P., Hallikainen, M., Hänninen, T., Partanen, K., Soininen, H., 2000. Diagnosis of Alzheimer's disease: MRI of the hippocampus vs delayed recall. Neuropsychologia 38, 579–584. Lawlor, B.A., Ryan, T.M., Schmeidler, J., Mohs, R.C., Davis, K.L., 1994. Clinical symptoms associated with age at onset in Alzheimer's disease. American Journal of Psychiatry 151, 1646–1649. LeDoux, J.E., 1993. Emotional memory systems in the brain. Behavioral Brain Research 58, 69–79. Lehéricy, S., Baulac, M., Chiras, J., Piérot, L., Martin, N., Pillon, B., Deweer, B., Dubois, B., Marsault, C., 1994. Amygdalohippocampal MR volume measurements in the early stages of Alzheimer disease. American Journal of Neuroradiology 15, 927–937. Markowitsh, H.J., Calabrese, P., Würker, M., Durwen, H.F., Kessler, J., Babinsky, R., Brechtelsbauer, D., Heuser, L., Gehlen, W., 1994. The amygdala's contribution to memory—a study on two patients with Urbach–Wiethe disease. Neuroreport 5, 1349–1352. Maunoury, C., Michot, J.L., Caillet, H., Parlato, V., Leroy-Willig, A., 1996. Specificity of temporal amygdala atrophy in Alzheimer's disease: quantitative assessment with magnetic resonance imaging. Dementia 7, 10–14.
Mauri, M., Sibilla, L., Bono, G., Carlesimo, G.A., Sinforiani, E., Martelli, A., 1998. The role of morpho-volumetric and memory correlations in the diagnosis of early Alzheimer dementia. Journal of Neurology 245, 525–530. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., Stadlan, E.M., 1984. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA work group under the auspices of Department of Health and Human Services Task Force on Alzheimer's disease. Neurology 34, 939–944. Meunier, M., Bachevalier, J., Mishkin, M., Murray, E.A., 1993. Effects on visual recognition of combined and separate ablations of the entorhinal and perirhinal cortex in rhesus monkeys. Journal of Neuroscience 13, 5418–5432. Mirra, S.S., Heyman, A., McKeel, D., Sumi, S.M., Crain, B.J., Brownlee, L.M., Vogel, F.S., Hughes, J.P., van Belle, G., Berg, L., 1991. The Consortium to Establish a Registry for Alzheimer's Disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer's disease. Neurology 41, 479–486. Mizuno, K., Wakai, M., Takeda, A., Sobue, G., 2000. Medial temporal atrophy and memory impairment in early stage of Alzheimer's disease: an MRI volumetric and memory assessment study. Journal of Neurological Science 173, 18–24. Mori, E., Yoneda, Y., Yamashita, H., Hirono, N., Ikeda, M., Yamadori, A., 1997. Medial temporal structures relate to memory impairment in Alzheimer's disease: an MRI volumetric study. Journal of Neurology, Neurosurgery and Psychiatry 63, 214–221. Murray, E.A., Gaffan, D., 1994. Removal of the amygdala plus adjacent cortex disrupts the retention of both intramodal and crossmodal associative memories in monkeys. Behavioral Neuroscience 108, 494–500. Murray, E.A., Mishkin, M., 1998. Object recognition and location memory in monkeys with excitotoxic lesions of the amygdala and hippocampus. Journal of Neuroscience 18, 6568–6582. Pennanen, C., Kivipelto, M., Tuomainen, S., Hartikainen, P., Hänninen, T., Laakso, M.P., Hallikainen, M., Vanhanen, M., Nissinen, A., Eeva-Liisa, H., Vainio, P., Vanninen, R., Partanen, K., Soininen, H., 2004. Hippocampus and entorhinal cortex in mild cognitive impairment and early AD. Neurobiology of Aging 25, 303–310. Petersen, R.C., Jack Jr., C.R., Xu, Y.C., Waring, S.C., O'Brien, P.C., Smith, G.E., Ivnik, R.J., Tangalos, E.G., Boeve, B.F., Kokmen, E., 2000. Memory and MRI-based hippocampal volumes in aging and AD. Neurology 54, 581–587. Press, G.A., Amaral, D.G., Squire, L.R., 1989. Hippocampal abnormalities in amnestic patients revealed by high-resolution magnetic resonance imaging. Nature 341, 54–57. Pruessner, J.C., Li, L.M., Serles, W., Pruessner, M., Collins, D.L., Kabani, N., Lupien, S., Evans, A.C., 2000. Volumetry of hippocampus and amygdala with high-resolution MRI and threedimensional analysis software: minimizing the discrepancies between laboratories. Cerebral Cortex 10, 433–442. Raz, N., Gunning-Dixon, F., Head, D., Rodrigue, K.M., Williamson, A., Acker, J.D., 2004. Aging, sexual dimorphism, and hemispheric asymmetry of the cerebral cortex: replicability of regional differences in volume. Neurobiology of Aging 25, 377–396. Robb, R., Hanson, D., Karwoski, R., Larson, A., Workman, E., Stacy, M., 1989. ANALYZE: a comprehensive, operator-interactive software package for multidimensional image display and analysis. Computerized Medical Imaging and Graphics 13, 433–454. Rosen, W.G., Mohs, R.C., Davis, K.L., 1984. A new rating scale for Alzheimer's disease. American Journal of Psychiatry 141, 1356–1364.
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261 Silverman, J.M., Breitner, J.C.S., Mohs, R.C., Davis, K.L., 1986. Reliability of the family history method in genetic studies of Alzheimer's disease and related dementias. American Journal of Psychiatry 143, 1279–1282. Squire, L.R., Stark, C.E., Clark, R.E., 2004. The medial temporal lobe. Annual Review of Neuroscience 27, 279–306. Sullivan, E.V., Marsh, L., Pfefferbaum, A., 2005. Preservation of hippocampal volume throughout adulthood in healthy men and women. Neurobiology of Aging 26, 1093–1098. Suzuki, W.A., Zola-Morgan, S., Squire, L.R., Amaral, D.G., 1993. Lesions of the perirhinal and parahippocampal cortices in the monkey produce long-lasting memory impairment in the visual and tactual modalities. Journal of Neuroscience 13, 2430–2451. Terry, R.D., Masliah, E., Salmon, D.P., Butters, N., DeTeresa, R., Hill, R., Hansen, L.A., Katzman, R., 1991. Physical basis of cognitive alterations in Alzheimer's disease: synapse loss is the major correlate of cognitive impairment. Annals of Neurology 30, 572–580. Tranel, D., Hyman, B.T., 1990. Neuropsychological correlates of bilateral amygdala damage. Archives of Neurology 47, 349–355. Watson, C., Andermann, F., Gloor, P., Jones-Gotman, M., Peters, T., Evans, A., Oliviera, A., Melanson, D., Leroux, G., 1992. Anatomic basis of amygdaloid and hippocampal volume measurement by magnetic resonance imaging. Neurology 42, 1743–1750. Whitwell, J.L., Sampson, E.L., Watt, H.C., Harvey, R.J., Rossor, M.N., Fox, N.C., 2005. A volumetric magnetic resonance imaging study
261
of the amygdala in frontotemporal lobar degeneration and Alzheimer's disease. Dementia and Geriatric Cognitive Disorders 20, 238–244. Wilcock, G.K., Esiri, M.M., 1982. Plaques, tangles and dementia: a quantitative study. Journal of Neurological Science 56, 343–356. Zola, S.M., Squire, L.R., Teng, E., Stefanacci, L., Buffalo, E.A., Clark, R.E., 2000. Impaired recognition memory in monkeys after damage limited to the hippocampal region. Journal of Neuroscience 20, 451–463. Zola-Morgan, S., Squire, L.R., Amaral, D.G., 1986. Human amnesia and the medial temporal region: enduring memory impairment following a bilateral lesion limited to field CA1 of the hippocampus. Journal of Neuroscience 6, 2950–2967. Zola-Morgan, S., Squire, L.R., Amaral, D.G., 1989a. Lesions of the amygdala that spare adjacent cortical regions do not impair memory or exacerbate the impairment following lesions of the hippocampal formation. Journal of Neuroscience 9, 1922–1936. Zola-Morgan, S., Squire, L.R., Amaral, D.G., Suzuki, W.A., 1989b. Lesions of perirhinal and parahippocampal cortex that spare the amygdala and hippocampal formation produce severe memory impairment. Journal of Neuroscience 9, 4355–4370.
Volumetry of amygdala and hippocampus and memory performance in Alzheimer's disease Michael Bassoa , John Yanga , Lauren Warrena , Martha G. MacAvoya , Pradeep Varmac , Richard A. Bronenc , Christopher H. van Dycka,b,⁎ a
Alzheimer's Disease Research Unit, Department of Psychiatry, Yale University School of Medicine, New Haven, CT, 06510, USA b Department of Neurobiology, Yale University School of Medicine, New Haven, CT, USA c Department of Diagnostic Radiology, Yale University School of Medicine, New Haven, CT, USA Received 30 July 2005; received in revised form 3 January 2006; accepted 8 January 2006
Abstract Magnetic resonance imaging (MRI) is showing increased utility in examining medial temporal lobe atrophy and its relationship to memory performance in Alzheimer's disease (AD). We studied 56 AD patients and 42 older healthy subjects with neuropsychological assessment and MRI. Hippocampal and amygdaloid volumes (normalized to intracranial volume) were contrasted between AD patients and healthy controls and correlated with neuropsychological performance. Comparisons between AD patients and healthy controls revealed highly significant differences in the normalized volume of hippocampus and amygdala by analysis of covariance. Group differences tended to be at least as large for amygdaloid as hippocampal volume, including when the subset of AD patients with the mildest symptoms was considered separately. Within the AD group, performance on the Memory–Orientation subscale of the Alzheimer's Disease Assessment Scale-Cognition (ADAS-Cog) was significantly correlated with normalized amygdaloid volume but not with normalized hippocampal volume. Other ADAS-Cog subscales (Language, Praxis) were uncorrelated with either volume. In the healthy control sample, neither hippocampal nor amygdaloid volumes were significant predictors of any neuropsychological measure. While a substantial literature continues to justify the focus on the hippocampus in MRI studies of AD, these results suggest that the amygdala should receive similar attention, including in studies of the prodromal stages of AD. © 2006 Elsevier Ireland Ltd. All rights reserved. Keywords: Alzheimer's disease; Magnetic resonance imaging; Amygdala; Hippocampus; Memory
1. Introduction The neuropathological hallmarks of Alzheimer's disease (AD) include neurofibrillary tangles, β-amyloid ⁎ Corresponding author. Alzheimer's Disease Research Unit, Department of Psychiatry, Yale University School of Medicine, One Church Street, Suite 600, New Haven, CT 06510, USA. Tel.: +1 203 764 8100; fax: +1 203 764 8111. E-mail address: [email protected] (C.H. van Dyck).
plaques, and neuronal and synaptic loss. Although these findings occur in a range of brain structures, the site of earliest and most severe pathology in AD is the medial temporal lobe (Hyman et al., 1990; Braak and Braak, 1991). These pathological features have been found to correlate with the severity of dementia (Wilcock and Esiri, 1982; Terry et al., 1991), and their early occurrence in the medial temporal lobe may account for the memory impairment that is typically the earliest and most prominent symptom in AD (Hyman et al., 1990).
0925-4927/$ - see front matter © 2006 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.pscychresns.2006.01.007
252
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
Although a link between the medial temporal lobes and memory is well established (Squire et al., 2004), the precise role of specific medial temporal lobe structures in memory functions continues to be elucidated. The hippocampus is believed to play a central role in declarative memory (Squire et al., 2004), and this is borne out by most human (Zola-Morgan et al., 1986; Press et al., 1989) and animal (Alvarez et al., 1995; Beason-Held et al., 1999; Zola et al., 2000) lesion studies. The role of the amygdala in emotional memory is well established (LeDoux, 1993), but that in declarative memory is less clear. Amygdala lesions produce declarative memory impairment in humans (Tranel and Hyman, 1990; Markowitsh et al., 1994) and nonhuman primates (Murray and Gaffan, 1994), although one primate study suggests that lesions of the amygdala that spare adjacent cortical regions do not impair memory (Zola-Morgan et al., 1989a). In recent years, magnetic resonance imaging (MRI) has been used to examine medial temporal lobe atrophy and its relationship to memory performance in AD. These studies have confirmed volume differences between AD patients and healthy controls for the hippocampus (Jack et al., 1992; Lehéricy et al., 1994; Jack et al., 1997; Mori et al., 1997; Krasuski et al., 1998; Laakso et al., 2000; Mizuno et al., 2000; Pennanen et al., 2004) and/or the amygdala (Cuénod et al., 1993; Lehéricy et al., 1994; Laakso et al., 1995a; Maunoury et al., 1996; Mori et al., 1997; Krasuski et al., 1998; Mauri et al., 1998; Mizuno et al., 2000). Studies of mild cognitive impairment (considered prodromal for AD in most cases) have generally neglected the amygdala and focused on hippocampal volume (Jack et al., 1999; Grundman et al., 2003; Pennanen et al., 2004), despite the fact that those studies that have examined both hippocampal and amygdaloid volumes in the mildest stages of AD suggest that amygdaloid deficits are at least as robust (Lehéricy et al., 1994; Krasuski et al., 1998; Mizuno et al., 2000). Some MRI studies have also shown a correlation between hippocampal (Deweer et al., 1995; Laakso et al., 1995b; Fama et al., 1997; Laakso et al., 2000; Petersen et al., 2000; Kramer et al., 2004) or amygdaloid (Mori et al., 1997; Mizuno et al., 2000) volumes and memory performance in AD. In the present study, we compared normalized hippocampal and amygdaloid volumes by MRI between older healthy subjects and patients with AD, including a subsample of patients with the mildest symptoms. We examined the relationship between these volumes and an index of memory performance in the AD patients, hypothesizing that memory performance would be
correlated with both hippocampal and amygdaloid volumes. We examined the specificity of these correlations by contrasting them with non-memory cognitive performance and also by examining similar correlates in the older control sample. 2. Methods The study sample comprised 56 patients with probable AD who received MRI scanning in research protocols in the Yale Alzheimer's Disease Research Unit (ADRU). Patients had been referred to the ADRU from a variety of sources or were self-referred. Five of these patients have subsequently died and had autopsy confirming definite AD (Mirra et al., 1991). Forty-two healthy elderly control subjects (most spouses of participating AD patients) were recruited for MRI scanning. The demographics and clinical characteristics of patients and controls are displayed in Table 1. Two AD patients were African American; all other patients and controls were of European ancestry. All patients and controls underwent a comprehensive evaluation by a research physician and ancillary staff, including cognitive assessment, medical history, physical and neurological examinations, serum chemistries, thyroid function studies, complete blood count, B12, folate, VDRL, urinalysis, electrocardiogram, and clinical reading of brain MRI. AD patients were required to meet standard diagnostic criteria for probable AD (McKhann et al., 1984), and controls were required to be in good health for their age. Subjects were excluded for any neurological or medical disorder (other than AD Table 1 Subject characteristics Variable
Demographics Age Gender Handedness Education (years) Neuropsychological MMSE ADAS-Cog Disease characteristics Onset age Duration (years) Family history
Healthy (N = 42)
Alzheimer's (N = 56)
Mean ± S.D.
Mean ± S.D.
73.2 ± 6.7 22M, 20F 42R 14.3 ± 3.4
71.2 ± 8.6 28M, 28F 54R, 2L 14.1 ± 3.3
29.0 ± 1.0 5.1 ± 1.8
*18.3 ± 4.3 *27.3 ± 10.5 66.6 ± 8.6 4.6 ± 1.5 25+, 31−
MMSE = Mini-Mental State Examination; ADAS-Cog = Alzheimer's Disease Assessment Scale — Cognitive Subscale. Family history is positive if primary degenerative dementia is present in at least one first degree relative. *Differs from control value, P b 10–7, t-test.
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
in the patient group) that could produce cognitive deterioration or for significant psychiatric illness, alcohol, or substance abuse. Healthy subjects were also required to have no significant evidence of cognitive impairment, as indicated by a Mini-Mental State Examination (MMSE) (Folstein et al., 1975) score ≥ 27 and a normal brain MRI scan. Neuropsychological testing administered for subject characterization (see Table 1) included the MMSE (Folstein et al., 1975) and the cognitive subscale of the Alzheimer's Disease Assessment Scale (ADAS-Cog; range = 0–70, lower score signifies less impairment) (Rosen et al., 1984). The ADAS-Cog was divided into three subscales: (1) Memory–Orientation (the sum of the ratings for “orientation,” “word recall,” “word recognition,” and “recall of test instructions”; range of scores = 0–35), (2) Language (the sum of the ratings for “difficulty making self understood,” “comprehension of spoken language,” “word-finding difficulty,” “commands,” and “naming”; range = 0–25), and (3) Praxis (the sum of the ratings for “constructional praxis” and “ideational praxis”; range = 0–10). These subscales are similar to those previously used (Lawlor et al., 1994) but combined Memory and Orientation as declarative memory functions hypothesized to be related to hippocampal and amygdaloid volumes in AD. Family history of AD was assessed using the Alzheimer Dementia Risk Questionnaire (Breitner and Folstein, 1984) and the Dementia Questionnaire (Silverman et al., 1986), and was considered to be positive if at least one first degree relative met criteria for primary degenerative dementia. No cases suggestive of autosomal dominant inheritance were identified. All subjects (or their responsible next of kin) provided written informed consent, and were studied under protocols approved by the Yale Human Investigation Committee and conducted in accordance with the Declaration of Helsinki. 2.1. MRI methods Subjects were imaged with a 1.5-Tesla General Electric Signa scanner (General Electric, Milwaukee, WI) according to a standardized protocol. The protocol began with a T1-weighted sagittal pulse sequence (5mm contiguous slices, TR = 600 ms, TE = 11 ms, matrix = 256 × 192, number of signal averages = 1, field of view = 24 cm) that was used to measure intracranial volume (ICV) and to localize subsequent acquisitions. The pulse sequence used to measure volumes of the hippocampus and the amygdala was a T1-weighted, three-dimensional volume spoiled gradient echo axial
253
sequence through the temporal lobes (sixty 1.5-mm contiguous slices, TR = 24 ms, TE = 5 ms, matrix = 256 × 192, number of signal averages = 2, field of view = 30 cm, and a 45° flip angle). Images were transferred to a Sun SPARC 2 workstation (Sun Microsystems, Santa Clara, CA) where volumetric measurements were performed using ANALYZE software (Robb et al., 1989) by the same rater (MB) who was blinded to all subject information but was not blinded to hemisphere (left or right). The axial pulse sequence data were rotated and re-sliced into 1.2-mm isotropic voxel coronal sections orthogonal to the long axis of the hippocampus (matrix = 256 × 256). Any head rotation with respect to the orthogonal coronal plane was also corrected during this step. In the resulting orthogonal coronal slices, the borders of the hippocampus and the amygdala were then traced sequentially from posterior to anterior using a mouse track ball. The ICV was delineated along the inner border of the skull from the sagittal pulse sequence. The ICV was defined as all tissue interior to the dura mater, including the pituitary gland and its surrounding cerebrospinal fluid. The foramen magnum defined the inferior border of the brainstem. All bony structures outside of the brain were disarticulated. The region-of-interest module of ANALYZE calculated volumes for each slice, multiplying the cross-sectional area by the slice thickness. Final hippocampal and amygdaloid volumes and ICVs were calculated by summing the volumes for each structure across all slices. Individual hippocampal and amygdaloid volumes were normalized for intersubject variation in head size by dividing by the ICV of each subject. 2.2. Anatomic guidelines Neuroanatomic guidelines for hippocampus and amygdala (see Fig. 1) were adapted from those of Watson et al. (1992) and the atlases of Duvernoy (1991, 1998) in consultation with a neuroradiologist expert in medial temporal lobe anatomy (RB). The hippocampal volume included the CA1 to CA4 fields of the hippocampus proper, the dentate gyrus, the subiculum, the alveus, and the fimbria. To improve reliability, the posterior margin of the hippocampus was arbitrarily defined by the first oblique coronal section that contained the thalamus. At this level the crura of the fornices are usually observable in profile and serve as an additional landmark. The anterior margin of the hippocampus is determined by the alveus that covers the ventricular surface of the hippocampal head at the uncal recess within the temporal horn. When this alveus
254
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
Fig. 1. Anatomic margins of hippocampus and amygdala on MRI. Selected images from a 69-year-old female control subject. Panel (A) Oblique coronal section through the hippocampus (designated by H), which was defined to include the CA1 to CA4 fields of the hippocampus proper, the dentate gyrus, the subiculum, the alveus, and the fimbria. Panel (B) Oblique coronal section through the amygdala (designated by A), which was defined to include the amygdala proper, the semilunar gyrus and a small portion of the ambient gyrus.
was not evident, we used the criteria of Watson et al. (1992) to differentiate the hippocampal head from the amygdala. Thus these measurements included essentially the entire hippocampus from the tail through the head. Neuroanatomic boundaries of the amygdala were based closely on those of Watson et al. (1992) and were delineated to include the amygdala proper (the deep/ basolateral and superficial/corticomedial nuclei, and other nuclei), the semilunar gyrus (corresponding to the cortical amygdaloid nucleus) and a small portion of the ambient gyrus (corresponding to the olfactory portion of the entorhinal cortex). Gray–white matter junctions identified the posterior, superior, and lateral borders of the amygdala. Either the uncal recess of the temporal horn or the alveus covering the hippocampal head defined the inferior border. The medial aspect of the amygdala is covered by a portion of entorhinal cortex, which forms the surface of the ambient gyrus in this region. The entorhinal cortex inferior to the tentorial indentation was excluded from the amygdaloid volume. If this indentation was poorly visualized in the anterior amygdaloid region, then it was approximated by a line drawn directly from the inferior and medial border of the amygdala to the supracellar cistern. This method inevitably included a small portion of the superior entorhinal cortex in the amygdaloid measurement
(Watson et al., 1992). The anterior boundary of the amygdala is poorly visualized on MRI and was operationally defined as the anterior most section before the closure of the lateral sulcus to form the endorhinal sulcus. This procedure potentially excludes part of the anterior amygdala but is thought to yield greater reliability (Watson et al., 1992). 2.3. Inter-rater reliability To evaluate inter-rater reliability, a second investigator (JY) also analyzed a subset of the AD patient scans (n = 50 for amygdala, n = 27 for hippocampus, n = 31 for ICV). This analysis yielded intraclass correlation coefficients (ρ) (Kirk, 1982) of 0.85 for left amygdala, 0.80 for right amygdala, 0.90 for left hippocampus, 0.84 for right hippocampus, and 0.95 for ICV, demonstrating satisfactory reliability. No healthy control scans were included in the inter-rater analysis. 2.4. Statistical analysis Demographic and neuropsychological variables were compared between AD and healthy control groups by chi-square or Student's t test. The effect of diagnostic group and brain structure (hippocampus, amygdala) on
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
normalized volume was determined by two-way analysis of covariance (ANCOVA) with repeated measures on brain structure, and age and education as covariates. To determine whether significant volume deficits were present in the mildest AD patients, the quartile of patients with the highest MMSE scores (range 23–27, n = 13) were compared with the control sample by the same ANCOVA. Volume deficits in the AD patient group were expressed as both mean percentages and mean z-scores. Since some previous researchers have observed significant asymmetries between left and right hemispheres for hippocampal and amygdaloid volumes, we also compared left and right volumes by paired t tests for each diagnostic group. We further examined whether asymmetry differed between AD patients and controls, by comparing an asymmetry index (100 × [right − left] / [right + left] / 2) between groups, using Student's t test. Within the AD group, the relationship between neuropsychological measures (dependent variables) and normalized hippocampal and amygdaloid volumes (combining left and right hemispheres) was examined by stepwise multiple linear regression (with P b 0.10 required for retention in the model), considering the contribution of age, sex, and education. We hypothesized that the Memory–Orientation subscale of the ADASCog (but not the other ADAS-Cog subscales: Language, Praxis) would be associated with smaller normalized hippocampal and amygdaloid volumes in the AD patients (and thus we applied a Bonferroni correction for two planned comparisons α = 0.05 / 2 = 0.025). However, preliminary analysis revealed that the contributions of age, education, and gender were not retained in the models for the hypothesis-driven subscores, thus reducing to simple bivariate correlations (Pearson's r) between neuropsychological measures and normalized hippocampal and amygdaloid volumes. A similar investigation of the relationship between neuropsychological measures and normalized hippocampal and amygdaloid volumes was conducted for the healthy control group. All statistical analyses used the SPSS (SPSS Inc., Chicago, IL) software package and twotailed tests of significance.
255
significantly worse on the MMSE and the ADAS-Cog (P b 10− 7, Student's t test). 3.1. Effect of diagnostic group on normalized hippocampal and amygdaloid volumes Absolute volumes (mean ± S.D.) of hippocampus and amygdala were as follows: healthy controls — left hippocampus 2.158 ± 0.359 ml, right hippocampus 2.247 ± 0.377 ml, left amygdala 2.032 ± 0.435 ml, right amygdala 2.160 ± 0.497 ml; AD patients — left hippocampus 1.737 ± 0.486 ml, right hippocampus 1.813 ± 0.478 ml, left amygdala 1.530 ± 0.410 ml, right amygdala 1.580 ± 0.475 ml. ANCOVA revealed significant main effects of diagnosis (F = 46.69; df = 1,94; P b 0.001) and brain structure (F = 14.84; df = 1,96; P b 0.001). Planned contrasts demonstrated highly significant group differences for hippocampus (combining left and right hemispheres, dividing by the ICVof each subject, P b 0.001; see Fig. 2) and amygdala (P b 0.001; see Fig. 3). The diagnosis by structure interaction term (F = 1.09; df = 1,96; P = 0.30) indicated that there was no significant difference in amygdaloid vs. hippocampal volume reductions in the AD patients compared with the controls. Mean differences between healthy controls and AD patients were 17% for hippocampus (z = − 0.99) and 23% for amygdala (z = − 1.28). The effect of age (t = − 2.70, P = 0.008) but not education (t = − 1.75, P = 0.084) was significant in the ANCOVA model. Normalized volumes of hippocampus and amygdala were significantly correlated with each
3. Results The characteristics of AD patients and healthy controls are shown in Table 1. The two groups did not differ in age (t = 1.24, df = 96, P = 0.22), gender distribution (χ 2 = 0.0009, df = 1, P = 0.98), handedness (χ 2 = 0.27, df = 1, P = 0.61), or years of education (t = 0.23, df = 96, P = 0.82). The AD patients scored
Fig. 2. Scatter plot of normalized hippocampal volume (combined left and right hippocampal volume divided by intracranial volume) in healthy control subjects (n = 42) and patients with Alzheimer's disease (AD; n = 56). Each dash denotes an individual patient; solid circles denote treatment group means (± S.D.). The AD patients had significantly lower normalized hippocampal volume than the healthy subjects, controlling for age and education (P b 0.001, ANCOVA).
256
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
and healthy subjects (t = 1.18, df = 96, P = 0.24, Student's t test). 3.3. Effect of hippocampal and amygdaloid volumes on cognitive measures in AD
Fig. 3. Scatter plot of normalized amygdaloid volume (combined left and right amygdaloid volume divided by intracranial volume) in healthy control subjects (n = 42) and patients with Alzheimer's disease (AD; n = 56). Each dash denotes an individual patient; solid circles denote treatment group means (± S.D.). The AD patients had significantly lower normalized amygdaloid volume than the healthy subjects, controlling for age and education (P b 0.001, ANCOVA).
other in the AD patient group (r = 0.48, n = 56, P b 0.001), but not in the control group (r = 0.21, n = 42, P = 0.18). When only the quartile of AD patients with the highest MMSE scores (range 23–27, n = 13) was compared with the control sample, ANCOVA again revealed significant main effects of diagnosis (F = 7.27; df = 1,51; P = 0.009) and brain structure (F = 4.17; df = 1,53; P b 0.046). Planned contrasts showed significant group differences for amygdala (P b 0.016) but not for hippocampus (P b 0.095). However, the diagnosis by structure interaction term (F = 0.14; df = 1,53; P = 0.71) indicated that the difference in amygdaloid vs. hippocampal volume reductions was not significant. Mean differences between healthy controls and these very mild AD patients were 10% for hippocampus (z = − 0.57) and 13% for amygdala (z = − 0.75).
Within the AD group, the relationship between normalized hippocampal and amygdaloid volumes and neuropsychological measures was examined by bivariate correlational analysis (Pearson's r). The results of that analysis are displayed in Table 2. In the two hypothesisdriven comparisons, normalized hippocampal volume was found to be uncorrelated with the Memory– Orientation subscale of the ADAS-Cog (r = − 0.22, n = 56, P = 0.097), whereas normalized amygdaloid volume was significantly correlated with Memory– Orientation (r = − 0.40, P = 0.002; see Fig. 4). Hippocampal and amygdaloid volumes were uncorrelated with other ADAS-Cog subscales (Language, Praxis). Post hoc analyses were undertaken to understand the unexpected absence of a correlation between normalized hippocampal volume and the Memory–Orientation subscale, and to determine if this was influenced by (1) severity of dementia or (2) specific components of the Memory–Orientation subscale. The AD patient sample was dichotomized by MMSE score with the hypothesis that a correlation would emerge in the more impaired patients. However, the correlation was absent both for patients with MMSE scores ≥ 19 (r = − 0.27, n = 25, P = 0.19) and for patients with MMSE scores b 19 (r = 0.11, n = 31, P = 0.056). Memory–Orientation subtest components (“orientation,” “word recall,” “word recognition,” and “recall of test instructions”) were considered separately, with the hypothesis that “word recognition,” in particular, would invoke a hippocampal substrate when the overall score did not. However, the
3.2. Comparison of left and right hippocampal and amygdaloid volume Right hippocampal volume was significantly larger than left hippocampal volume for both the healthy control sample (t = 2.75, df = 41, P = 0.009, paired t test) and the AD sample (t = 2.48, df = 55, P = 0.016, paired t test). The hippocampal asymmetry index did not differ between AD and healthy subjects (t = 0.37, df = 96, P = 0.72, Student's t test). Right amygdaloid volume was larger than left amygdaloid volume for the healthy control sample (t = 2.89, df = 41, P = 0.006, paired t test) but not for the AD sample (t = 1.30, df = 55, P = 0.20, paired t test). However, the amygdaloid asymmetry index did not differ between AD
Table 2 Correlations between normalized hippocampal and amygdaloid volumes and neuropsychological measures in patients with AD Neuropsychological tests
Hippocampus
Amygdala
Pearson's r P value Pearson's r P value
ADAS-Cog Memory–Orientation − 0.22 Language − 0.08 Praxis − 0.21 Total 0.21 MMSE 0.19
0.097 0.559 0.113 0.120 0.150
− 0.40 − 0.20 − 0.10 − 0.34 0.34
0.002 0.132 0.446 0.011 0.011
ADAS-Cog, Alzheimer's Disease Assessment Scale — Cognitive Subscale (lower score better); MMSE, Mini-Mental State Examination.
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
257
loid normalized volumes entered into the model as significant predictor variables. 4. Discussion
Fig. 4. Memory–Orientation subscores (range = 0–35; lower score signifies less impairment) from the Alzheimer's Disease Assessment Scale (ADAS-Cog) versus normalized amygdaloid volume (combined left and right amygdaloid volume divided by intracranial volume) in Alzheimer's disease patients (n = 56). Amygdaloid volume was significantly correlated with Memory–Orientation (r = − 0.40, P = 0.002) but was uncorrelated with other ADAS-Cog subscores (Language, Praxis).
This study compared normalized hippocampal and amygdaloid volumes between patients with AD and older healthy subjects and examined neuropsychological correlates with MRI measures in the AD patients. Highly significant group differences were observed in normalized hippocampal and amygdaloid volumes. Group differences tended to be at least as large for amygdaloid as hippocampal volume, including when the subset of AD patients with the mildest symptoms (MMSE ≥ 23) was considered separately. Within the AD group, the Memory–Orientation subscale of the ADAS-Cog (but not other ADAS-Cog subscales) was significantly correlated with amygdaloid volume, but it only showed a nonsignificant trend-level correlation with hippocampal volume. 4.1. Previous studies of patient–control differences
“word recognition” item was also uncorrelated with hippocampal volume (r = − 0.16, n = 56, P = 0.24). Normalized amygdaloid volume was also correlated with total ADAS-Cog score (r = −0.34, P = 0.011) and MMSE score (r = 0.34, P = 0.011); whereas normalized hippocampal volume was uncorrelated with either measure. Although, as previously mentioned, the model of ADAS-Cog subscales was unimproved by the inclusion of age, sex, or education in a multiple linear regression analysis, the models of total ADASCog and MMSE scores were improved slightly by the addition of education in multiple linear regression models. However, the overall results did not differ with the inclusion of education (data not shown). When left and right hippocampal and amygdaloid volumes were considered separately in an exploratory analysis, the results did not differ (i.e., all significant findings for amygdala were true for both left and right amygdala, and no significant correlations emerged for left or right hippocampus). In the healthy control sample, the relationship between neuropsychological measures (dependent variables) and normalized hippocampal and amygdaloid volumes was also examined by stepwise multiple linear regression, considering the contribution of age, sex, and education. For the Memory–Orientation subscale of the ADAS-Cog (and all other neuropsychological measures: Language, Praxis subscales, total ADAS-Cog score, and MMSE), neither hippocampal nor amygda-
Numerous previous studies have established volume differences between AD patients and healthy controls for hippocampus (Jack et al., 1992; Lehéricy et al., 1994; Laakso et al., 1995b; Jack et al., 1997; Mori et al., 1997; Krasuski et al., 1998; Laakso et al., 2000; Mizuno et al., 2000; Pennanen et al., 2004) and/or amygdala (Cuénod et al., 1993; Lehéricy et al., 1994; Laakso et al., 1995a; Maunoury et al., 1996; Mori et al., 1997; Krasuski et al., 1998; Mauri et al., 1998; Mizuno et al., 2000). In agreement with the present results, studies that have measured hippocampal and amygdaloid volumes in the same samples have consistently found deficits in amygdaloid volume that are at least as large as those for hippocampus (Lehéricy et al., 1994; Mori et al., 1997; Krasuski et al., 1998; Mizuno et al., 2000), including in the mild (Lehéricy et al., 1994; Krasuski et al., 1998) or very mild (Mizuno et al., 2000) stages of AD. These results suggest that MRI studies of prodromal AD (mild cognitive impairment) should give greater attention to the amygdala. Such studies have focused almost exclusively on hippocampal volume as an earlier biomarker (Jack et al., 1999; Grundman et al., 2003; Pennanen et al., 2004). We found significantly larger right than left hippocampal volume in the healthy control group as did some (Jack et al., 1989; Krasuski et al., 1998; Geroldi et al., 2000; Pruessner et al., 2000; Jack et al., 2003; Sullivan et al., 2005) but not all (Raz et al., 2004) previous investigators. In contrast to one previous study (Geroldi
258
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
et al., 2000), we also observed the same asymmetry in the AD sample. We found significantly larger right than left amygdaloid volume in the healthy control group as did some (Watson et al., 1992; Filipek et al., 1994; Krasuski et al., 1998) but not all (Pruessner et al., 2000; Whitwell et al., 2005) previous investigators. While this “normal asymmetry” was not present in the AD group, the extent of amygdaloid asymmetry did not differ significantly between AD and healthy subjects. Other investigators have reported a reversed left N right asymmetry in both controls (Laakso et al., 1995b; Convit et al., 1999) and AD patients (Laakso et al., 1995b). 4.2. Previous studies of neuropsychological correlates We observed a significant correlation between an index of recent memory performance (Memory–Orientation subscale of the ADAS-Cog) and amygdaloid but not hippocampal volume. Our findings are therefore similar to those of Mori et al. (1997) who also observed memory performance (Wechsler Memory Scale-Revised, WMS-R subtests) to be correlated with amygdaloid but not hippocampal volumes in 46 mild to moderate AD patients. However, unlike Mori et al. (1997) who observed memory performance to be correlated with right amygdaloid volume, we found memory performance to be equally correlated with left and right amygdaloid volume (data not shown). Our results are also somewhat similar to those of Mizuno et al. (2000), who observed that WMS-R measures were correlated with amygdaloid but not hippocampal volume in a pooled sample of healthy control subjects and patients with very mild AD. However, these authors did not present separate analyses of patients and controls. Heun et al. (1997) also observed that memory dysfunction was highly correlated with atrophy of the combined amygdala–hippocampal complex in AD patients but did not segregate amygdaloid and hippocampal volumes. However, our results are at odds with Deweer et al. (1995) and Mauri et al. (1998) who observed no correlations between normalized volume of amygdala and several measures of memory performance, although Deweer et al. (1995) employed 5-mm slice thickness, which is probably inadequate for volumetry. Although we found only a nonsignificant trend for a correlation between Memory–Orientation and hippocampal volume, other investigators have demonstrated significant correlations between hippocampal volume and memory performance in AD patients (Deweer et al., 1995; Laakso et al., 1995b; Fama et al., 1997; Laakso et
al., 2000; Petersen et al., 2000; Kramer et al., 2004) and even in patients with mild cognitive impairment (Grundman et al., 2003). The absence of any significant correlation between Memory–Orientation and MRI measures in the control group is consistent with previous studies reporting no significant relationship between hippocampal volume (Petersen et al., 2000) and cognitive measures in healthy older subjects. 4.3. Relationship to human and animal lesion studies Although the link between the medial temporal lobes and memory is well established, human and animal lesion studies have provided an increasingly complex understanding of the tissue damage that is necessary and sufficient to produce impaired memory performance (Squire et al., 2004). Lesions of the amygdala result in memory deficits in humans (Tranel and Hyman, 1990; Markowitsh et al., 1994) and nonhuman primates (Murray and Gaffan, 1994). However, studies in nonhuman primates suggest that lesions of the amygdala that spare adjacent cortical regions do not impair memory (Zola-Morgan et al., 1989a). Damage restricted to the hippocampus has been reported to result in memory deficits in humans (Zola-Morgan et al., 1986; Press et al., 1989) and nonhuman primates (Alvarez et al., 1995; Beason-Held et al., 1999; Zola et al., 2000); however, one study observed no significant memory impairment in monkeys with restricted lesions of amygdala and hippocampus (Murray and Mishkin, 1998). Damage to adjacent (perirhinal and parahippocampal (Zola-Morgan et al., 1989b; Suzuki et al., 1993) or to entorhinal and perirhinal (Meunier et al., 1993; Eacott et al., 1994)) cortices that spare the amygdala and hippocampal formation consistently produces severe memory impairment. Thus a significant correlation between normalized amygdaloid volume and memory performance in AD patients might not necessarily be expected from lesion studies. However, because the amygdaloid volume we measured includes some adjacent cortical regions – the semilunar gyrus (corresponds to the cortical amygdaloid nucleus) and a small portion of the ambient gyrus (which corresponds to the olfactory portion of the entorhinal cortex) – its correlation with memory performance may be more robust than predicted from animal lesions localized to the amygdala proper. It is possible that a software package that allowed simultaneous threedimensional visualization (Pruessner et al., 2000) would have permitted a more precise delineation of the amygdala proper. Furthermore, in AD damage to amygdala is likely correlated with damage to
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
hippocampus (e.g., normalized volumes of hippocampus and amygdala were significantly correlated in the present study: r = 0.48, n = 56, P b 0.001) as well as with other medial temporal lobe structures that were not measured in this study (e.g., entorhinal, perirhinal, and parahippocampal cortex). Therefore, greater atrophy of amygdala may be a surrogate for atrophy of other structures necessary for memory performance. The fact that normalized hippocampal volume was not significantly correlated with memory performance or global measures is somewhat contrary to human (Zola-Morgan et al., 1986; Press et al., 1989) and most (Alvarez et al., 1995; Beason-Held et al., 1999; Zola et al., 2000) animal studies. However, this finding is also at variance with many other human MRI studies of AD (Deweer et al., 1995; Laakso et al., 1995b; Fama et al., 1997; Laakso et al., 2000; Petersen et al., 2000; Kramer et al., 2004) and mild cognitive impairment (Grundman et al., 2003), and it may simply reflect a Type II error. Finally, the memory scale used in the present study, the Memory–Orientation Subscore of the ADASCog, is not a test of selective component processes of memory, e.g., storage versus retrieval or verbal versus visual memory. Therefore, its ability to relate specific memory processes to brain structures is limited. In summary, our results confirm previous reports of highly significant differences in normalized hippocampal and amygdaloid volumes between AD patients and healthy controls and suggest that amygdaloid volume deficits are at least as large as hippocampal deficits in the very mild stage of disease. Within the AD group, amygdaloid but not hippocampal volume was significantly correlated with the Memory–Orientation subscale of the ADAS-Cog. While a substantial literature continues to justify the focus on the hippocampus in MRI studies of AD, these results suggest that the amygdala should receive similar attention, including in studies of the prodromal stages of AD. Acknowledgments The authors thank Hedy Sarofin and Terri Hickey for assistance with MRI scanning, Larry Staib, Ph.D., for technical assistance, and Jung Kim, M.D., for histopathological diagnosis in a subset of the AD cases. References Alvarez, P., Zola-Morgan, S., Squire, L.R., 1995. Damage limited to the hippocampal region produces long-lasting memory impairment in monkeys. Journal of Neuroscience 15, 3796–3807.
259
Beason-Held, L., Rosene, D.L., R.J., K., Moss, M.B., 1999. Hippocampal formation lesions produce memory impairment in the rhesus monkey. Hippocampus 9, 562–574. Braak, H., Braak, E., 1991. Neuropathological stageing of Alzheimerrelated changes. Acta Neuropathologica 82, 239–259. Breitner, J.C.S., Folstein, M.F., 1984. Familial Alzheimer's disease: a prevalent disorder with specific clinical features. Psychological Medicine 14, 63–80. Convit, A., McHugh, P., Wolf, O.T., de Leon, M.J., Bobinski, M., De Santi, S., Roche, A., Tsui, W., 1999. MRI volume of the amygdala: a reliable method allowing separation from the hippocampal formation. Psychiatry Research: Neuroimaging 90, 113–123. Cuénod, C.A., Denys, A., Michot, J.L., Jehenson, P., Forette, F., Kaplan, D., Syrota, A., Boller, F., 1993. Amygdala atrophy in Alzheimer's disease. An in vivo magnetic resonance imaging study. Archives of Neurology 50, 941–945. Deweer, B., Lehericy, S., Pillon, B., Baulac, M., Chiras, J., Marsault, C., Agid, Y., Dubois, B., 1995. Memory disorders in probable Alzheimer's disease: the role of hippocampal atrophy as shown with MRI. Journal of Neurology, Neurosurgery and Psychiatry 58, 590–597. Duvernoy, H.M., 1991. The Human Brain: Surface, Three-dimensional Sectional Anatomy with MRI, and Blood Supply, 2nd edition. Springer-Verlag, Wien. Duvernoy, H.M., 1998. The Human Hippocampus: Functional Anatomy, Vascularization and Series Sections with MRI, 2nd edition. Springer-Verlag, Berlin. Eacott, M.J., Gaffan, D., Murray, E.A., 1994. Preserved recognition memory for small sets, and impaired stimulus identification for large sets, following rhinal cortex ablations in monkeys. European Journal of Neuroscience 6, 1466–1478. Fama, R., Sullivan, E.V., Shear, P.K., Marsh, L., Yesavage, J.A., Tinklenberg, J.R., Lim, K.O., Pfefferbaum, A., 1997. Selective cortical and hippocampal volume correlates of Mattis Dementia Rating Scale in Alzheimer disease. Archives of Neurology 54, 719–728. Filipek, P.A., Richelme, C., Kennedy, D.N., Caviness, V.S., 1994. The young adult human brain: an MRI-based morphometric analysis. Cerebral Cortex 4, 344–360. Folstein, M.F., Fostein, S.E., McHugh, P.R., 1975. “Mini-mental state”: a practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research 12, 189–198. Geroldi, C., Laakso, M.P., DeCarli, C., Beltramello, A., Bianchetti, A., Soininen, H., Trabucchi, M., Frisoni, G.B., 2000. Apolipoprotein E genotype and hippocampal asymmetry in Alzheimer's disease: a volumetric MRI study. Journal of Neurology, Neurosurgery and Psychiatry 68, 93–96. Grundman, M., Jack, C.R., Petersen, R.C., Kim, H.T., Taylor, C., Datvian, M., Weiner, M.F., DeCarli, C., DeKosky, S.T., van Dyck, C.H., Darvesh, S., Yaffe, K., Kaye, J., Ferris, S.H., Thomas, R.G., Thal, L.J., 2003. Hippocampal volume is associated with memory but not nonmemory cognitive performance in patients with mild cognitive impairment. Journal of Molecular Neuroscience 20, 241–248. Heun, R., Mazanek, M., Atzor, K.R., Tintera, J., Gawehn, J., Burkart, M., Gansicke, M., Falkai, P., Stoeter, P., 1997. Amygdala– hippocampal atrophy and memory performance in dementia of Alzheimer type. Dementia and Geriatric Cognitive Disorders 8, 329–336. Hyman, B.T., Van Hoesen, G.W., Damasio, A.R., 1990. Memoryrelated neural systems in Alzheimer's disease: an anatomic study. Neurology 40, 1721–1730.
260
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261
Jack, C.R., Twomey, C.K., Zinsmeister, A.R., Sharbrough, F.W., Petersen, R.C., Cascino, G.D., 1989. Anterior temporal lobes and hippocampal formations: normative volumetric measurements from MR images in young adults. Radiology 172, 549–554. Jack, C.R., Petersen, R.C., O'Brien, P.C., Tangalos, E.G., 1992. MRbased hippocampal volumetry in the diagnosis of Alzheimer's disease. Neurology 42, 183–188. Jack, C.R., Petersen, R.C., Xu, Y.C., Waring, S.C., O'Brien, P.C., Tangalos, E.G., Smith, G.E., Ivnik, R.J., Kokmen, E., 1997. Medial temporal atrophy on MRI in normal aging and very mild Alzheimer's disease. Neurology 49, 786–794. Jack, C.R., Petersen, R.C., Xu, Y.C., O'Brien, P.C., Smith, G.E., Ivnik, R.J., Boeve, B.F., Waring, S.C., Tangalos, E.G., Kokmen, E., 1999. Prediction of AD with MRI-based hippocampal volume in mild cognitive impairment. Neurology 52, 1397–1403. Jack, C.R.J., Slomkowski, M., Gracon, S., Hoover, T.M., Felmlee, J.P., Stewart, K., Xu, Y., Shiung, M., O'Brien, P.C., Cha, R., Knopman, D., Petersen, R.C., 2003. MRI as a biomarker of disease progression in a therapeutic trial of milameline for AD. Neurology 60, 253–260. Kirk, R., 1982. Experimental Design: Procedures for the Behavioral Sciences. Brooks/Cole Publishing, Co., Pacific Grove, CA. Kramer, J.H., Schuff, N., Reed, B.R., Mungas, D., Du, A.T., Rosen, H.J., Jagust, W.J., Miller, B.L., Weiner, M.W., Chui, H.C., 2004. Hippocampal volume and retention in Alzheimer's disease. Journal of the International Neuropsychological Society 10, 639–643. Krasuski, J.S., Alexander, G.E., Horwitz, B., Daly, E.M., Murphy, D.G.M., Rapoport, S.I., Schapiro, M.B., 1998. Volumes of medial temporal lobe structures in patients with Alzheimer's disease and mild cognitive impairment (and in healthy control subjects). Biological Psychiatry 43, 60–68. Laakso, M.P., Partanen, K., Lehtovirta, M., Hallikainen, M., Hänninen, T., Vainio, P., Riekkinen, P., Soininen, H., 1995a. MRI of amygdala fails to diagnose early Alzheimer's disease. Neuroreport 6, 2414–2418. Laakso, M.P., Soininen, H., Partanen, K., Helkala, E.L., Hartikainen, P., Vainio, P., Hallikainen, M., Hanninen, T., Riekkinen Sr., P.J., 1995b. Volumes of hippocampus, amygdala and frontal lobes in the MRI-based diagnosis of early Alzheimer's disease: correlation with memory functions. Journal of Neural Transmission: Parkinson's Disease Dementia Section 9, 73–86. Laakso, M.P., Hallikainen, M., Hänninen, T., Partanen, K., Soininen, H., 2000. Diagnosis of Alzheimer's disease: MRI of the hippocampus vs delayed recall. Neuropsychologia 38, 579–584. Lawlor, B.A., Ryan, T.M., Schmeidler, J., Mohs, R.C., Davis, K.L., 1994. Clinical symptoms associated with age at onset in Alzheimer's disease. American Journal of Psychiatry 151, 1646–1649. LeDoux, J.E., 1993. Emotional memory systems in the brain. Behavioral Brain Research 58, 69–79. Lehéricy, S., Baulac, M., Chiras, J., Piérot, L., Martin, N., Pillon, B., Deweer, B., Dubois, B., Marsault, C., 1994. Amygdalohippocampal MR volume measurements in the early stages of Alzheimer disease. American Journal of Neuroradiology 15, 927–937. Markowitsh, H.J., Calabrese, P., Würker, M., Durwen, H.F., Kessler, J., Babinsky, R., Brechtelsbauer, D., Heuser, L., Gehlen, W., 1994. The amygdala's contribution to memory—a study on two patients with Urbach–Wiethe disease. Neuroreport 5, 1349–1352. Maunoury, C., Michot, J.L., Caillet, H., Parlato, V., Leroy-Willig, A., 1996. Specificity of temporal amygdala atrophy in Alzheimer's disease: quantitative assessment with magnetic resonance imaging. Dementia 7, 10–14.
Mauri, M., Sibilla, L., Bono, G., Carlesimo, G.A., Sinforiani, E., Martelli, A., 1998. The role of morpho-volumetric and memory correlations in the diagnosis of early Alzheimer dementia. Journal of Neurology 245, 525–530. McKhann, G., Drachman, D., Folstein, M., Katzman, R., Price, D., Stadlan, E.M., 1984. Clinical diagnosis of Alzheimer's disease: report of the NINCDS-ADRDA work group under the auspices of Department of Health and Human Services Task Force on Alzheimer's disease. Neurology 34, 939–944. Meunier, M., Bachevalier, J., Mishkin, M., Murray, E.A., 1993. Effects on visual recognition of combined and separate ablations of the entorhinal and perirhinal cortex in rhesus monkeys. Journal of Neuroscience 13, 5418–5432. Mirra, S.S., Heyman, A., McKeel, D., Sumi, S.M., Crain, B.J., Brownlee, L.M., Vogel, F.S., Hughes, J.P., van Belle, G., Berg, L., 1991. The Consortium to Establish a Registry for Alzheimer's Disease (CERAD). Part II. Standardization of the neuropathologic assessment of Alzheimer's disease. Neurology 41, 479–486. Mizuno, K., Wakai, M., Takeda, A., Sobue, G., 2000. Medial temporal atrophy and memory impairment in early stage of Alzheimer's disease: an MRI volumetric and memory assessment study. Journal of Neurological Science 173, 18–24. Mori, E., Yoneda, Y., Yamashita, H., Hirono, N., Ikeda, M., Yamadori, A., 1997. Medial temporal structures relate to memory impairment in Alzheimer's disease: an MRI volumetric study. Journal of Neurology, Neurosurgery and Psychiatry 63, 214–221. Murray, E.A., Gaffan, D., 1994. Removal of the amygdala plus adjacent cortex disrupts the retention of both intramodal and crossmodal associative memories in monkeys. Behavioral Neuroscience 108, 494–500. Murray, E.A., Mishkin, M., 1998. Object recognition and location memory in monkeys with excitotoxic lesions of the amygdala and hippocampus. Journal of Neuroscience 18, 6568–6582. Pennanen, C., Kivipelto, M., Tuomainen, S., Hartikainen, P., Hänninen, T., Laakso, M.P., Hallikainen, M., Vanhanen, M., Nissinen, A., Eeva-Liisa, H., Vainio, P., Vanninen, R., Partanen, K., Soininen, H., 2004. Hippocampus and entorhinal cortex in mild cognitive impairment and early AD. Neurobiology of Aging 25, 303–310. Petersen, R.C., Jack Jr., C.R., Xu, Y.C., Waring, S.C., O'Brien, P.C., Smith, G.E., Ivnik, R.J., Tangalos, E.G., Boeve, B.F., Kokmen, E., 2000. Memory and MRI-based hippocampal volumes in aging and AD. Neurology 54, 581–587. Press, G.A., Amaral, D.G., Squire, L.R., 1989. Hippocampal abnormalities in amnestic patients revealed by high-resolution magnetic resonance imaging. Nature 341, 54–57. Pruessner, J.C., Li, L.M., Serles, W., Pruessner, M., Collins, D.L., Kabani, N., Lupien, S., Evans, A.C., 2000. Volumetry of hippocampus and amygdala with high-resolution MRI and threedimensional analysis software: minimizing the discrepancies between laboratories. Cerebral Cortex 10, 433–442. Raz, N., Gunning-Dixon, F., Head, D., Rodrigue, K.M., Williamson, A., Acker, J.D., 2004. Aging, sexual dimorphism, and hemispheric asymmetry of the cerebral cortex: replicability of regional differences in volume. Neurobiology of Aging 25, 377–396. Robb, R., Hanson, D., Karwoski, R., Larson, A., Workman, E., Stacy, M., 1989. ANALYZE: a comprehensive, operator-interactive software package for multidimensional image display and analysis. Computerized Medical Imaging and Graphics 13, 433–454. Rosen, W.G., Mohs, R.C., Davis, K.L., 1984. A new rating scale for Alzheimer's disease. American Journal of Psychiatry 141, 1356–1364.
M. Basso et al. / Psychiatry Research: Neuroimaging 146 (2006) 251–261 Silverman, J.M., Breitner, J.C.S., Mohs, R.C., Davis, K.L., 1986. Reliability of the family history method in genetic studies of Alzheimer's disease and related dementias. American Journal of Psychiatry 143, 1279–1282. Squire, L.R., Stark, C.E., Clark, R.E., 2004. The medial temporal lobe. Annual Review of Neuroscience 27, 279–306. Sullivan, E.V., Marsh, L., Pfefferbaum, A., 2005. Preservation of hippocampal volume throughout adulthood in healthy men and women. Neurobiology of Aging 26, 1093–1098. Suzuki, W.A., Zola-Morgan, S., Squire, L.R., Amaral, D.G., 1993. Lesions of the perirhinal and parahippocampal cortices in the monkey produce long-lasting memory impairment in the visual and tactual modalities. Journal of Neuroscience 13, 2430–2451. Terry, R.D., Masliah, E., Salmon, D.P., Butters, N., DeTeresa, R., Hill, R., Hansen, L.A., Katzman, R., 1991. Physical basis of cognitive alterations in Alzheimer's disease: synapse loss is the major correlate of cognitive impairment. Annals of Neurology 30, 572–580. Tranel, D., Hyman, B.T., 1990. Neuropsychological correlates of bilateral amygdala damage. Archives of Neurology 47, 349–355. Watson, C., Andermann, F., Gloor, P., Jones-Gotman, M., Peters, T., Evans, A., Oliviera, A., Melanson, D., Leroux, G., 1992. Anatomic basis of amygdaloid and hippocampal volume measurement by magnetic resonance imaging. Neurology 42, 1743–1750. Whitwell, J.L., Sampson, E.L., Watt, H.C., Harvey, R.J., Rossor, M.N., Fox, N.C., 2005. A volumetric magnetic resonance imaging study
261
of the amygdala in frontotemporal lobar degeneration and Alzheimer's disease. Dementia and Geriatric Cognitive Disorders 20, 238–244. Wilcock, G.K., Esiri, M.M., 1982. Plaques, tangles and dementia: a quantitative study. Journal of Neurological Science 56, 343–356. Zola, S.M., Squire, L.R., Teng, E., Stefanacci, L., Buffalo, E.A., Clark, R.E., 2000. Impaired recognition memory in monkeys after damage limited to the hippocampal region. Journal of Neuroscience 20, 451–463. Zola-Morgan, S., Squire, L.R., Amaral, D.G., 1986. Human amnesia and the medial temporal region: enduring memory impairment following a bilateral lesion limited to field CA1 of the hippocampus. Journal of Neuroscience 6, 2950–2967. Zola-Morgan, S., Squire, L.R., Amaral, D.G., 1989a. Lesions of the amygdala that spare adjacent cortical regions do not impair memory or exacerbate the impairment following lesions of the hippocampal formation. Journal of Neuroscience 9, 1922–1936. Zola-Morgan, S., Squire, L.R., Amaral, D.G., Suzuki, W.A., 1989b. Lesions of perirhinal and parahippocampal cortex that spare the amygdala and hippocampal formation produce severe memory impairment. Journal of Neuroscience 9, 4355–4370.