MRI-derived entorhinal and hippocampal atrophy in incipient and very mild Alzheimer���s disease B.C. Dickersona,f, I. Goncharovaa,g, M.P. Sullivana,h, C. Forchettia,b,i, R.S. Wilsona,b,c,d, D.A. Bennetta,b,d, L.A. Beckettd,e, L. deToledo-Morrella,b,c,* aDepartment of Neurological Sciences bRush Alzheimer���s Disease Center cDepartment of Psychology dRush Institute for Healthy Aging eDepartment of Medicine, Rush University, Chicago, IL 60612, USA fNow at Departments of Neurology, Massachusetts General and Brigham and Women���s Hospitals, Harvard Medical School, Boston, Massachusetts, USA gNow at Wadsworth Center, New York State Department of Health, Albany, New York, USA hNow at the Biomedical Department, University of Alaska, Anchorage, Alaska, USA iNow at the Chicago Center for Clinical Research, Chicago, Illinois, USA Received 15 August 2000 received in revised form 20 June 2001 accepted 20 June 2001 Abstract With high resolution, quantitative magnetic resonance imaging (MRI) techniques, it is now possible to examine alterations in brain anatomy in vivo and to identify regions affected in the earliest stages of Alzheimer���s disease (AD). In this study, we compared MRI-derived entorhinal and hippocampal volume in healthy elderly controls, patients who presented at the clinic with cognitive complaints, but did not meet criteria for dementia (non-demented), and patients with very mild AD. The two patient groups differed significantly from controls in entorhinal volume, but not from each other in contrast, they differed from each other, as well as from controls, in hippocampal volume, with the mild AD cases showing the greatest atrophy. Follow-up clinical evaluations available on 23/28 non-demented patients indicated that 12/23 had converted to AD within 12���77 months from the baseline MRI examination. Converters could be best differentiated from non-converters on the basis of entorhinal, but not hippocampal volume. These data suggest that although both the EC and hippocampal formation degenerate before the onset of overt dementia, EC volume is a better predictor of conversion. �� 2001 Elsevier Science Inc. All rights reserved. Keywords: Mesial temporal lobe Parahippocampal gyrus Aging Mild cognitive impairment Memory 1. Introduction Quantitative, high resolution magnetic resonance imag- ing (MRI) techniques provide a unique tool for examining alterations in brain anatomy in vivo during healthy aging and in age-related degenerative diseases. Such techniques are especially useful in identifying the anatomical origins of Alzheimer���s disease (AD). The entorhinal cortex (EC) and the hippocampal forma- tion (HF) are part of the mesial temporal lobe memory system [42,50] the EC connects the neocortex with the HF via the perforant path, thereby providing the latter with multimodal sensory information. These brain regions have received special attention in investigations on the patho- physiology of AD, since memory dysfunction is one of its earliest hallmarks. Post mortem pathological studies have implicated the EC and the transentorhinal region as early sites of involvement in AD and in individuals with mild cognitive impairment [4 ��� 6,18,21,35,46]. However, the EC has received less at- tention in in vivo investigations, partly due to the fact that MRI-derived quantitative protocols for this structure were only recently developed [2,22]. This research was supported by grants P01 AG09466 and P30 AG10161 from the National Institute on Aging, National Institutes of Health. * Corresponding author. Tel.: 1-312-942-5399 fax: 1-312-942- 2238. E-mail address: ldetoled@rush.edu (L. deToledo-Morrell). www.elsevier.com/locate/neuaging Neurobiology of Aging 22 (2001) 747���754 0197-4580/01/$ ��� see front matter �� 2001 Elsevier Science Inc. All rights reserved. PII: S0197-4580(01)00271-8

Atrophy of the HF, a region important for the acquisition of certain types of new information (i.e., declarative knowl- edge), has been well documented in AD using quantitative volumetric MRI analyses [7,12���14,23,24,29 ���32,38,40,48, 49]. More recently, atrophy of the EC was demonstrated in AD with MRI-based quantitation of its volume [27,49]. Yet, while in vivo anatomical investigations using structural MRI are rapidly proliferating, relatively few volumetric studies have focused strictly on patients in the earliest stages of AD or on those with mild cognitive impairment (see, for exam- ple, 31,49). Two studies that measured the entire parahippocampal gyrus failed to detect significant atrophy in this structure in patients with a clinical diagnosis of very mild AD [14,24]. In these mild cases, however, there was significant hip- pocampal atrophy. These results were initially somewhat surprising, given the early pathological involvement of the entorhinal and transentorhinal cortices in the disease pro- cess. In the two studies cited above, parahippocampal gyrus volume included both white and gray matter and, in the case of the report from our laboratory [14], measurements con- tinued beyond the anatomical boundaries of the entorhinal and perirhinal cortices. As a result, any changes in the EC itself may have been masked. It was, therefore, important to use newly developed protocols for the quantitation of the EC in order to re-examine its involvement in patients with very mild or incipient AD. In addition to patients with very mild AD, individuals at high risk for AD are now being studied in order to identify anatomical changes that precede a clinical diagnosis and to develop sensitive in vivo markers that might be predictive of conversion to AD. One such group consists of those patients who have cognitive complaints, cognitive impairment, or both, but who do not meet diagnostic criteria for dementia after a thorough evaluation. Such groups are of special interest, since they provide information on the transitional state between normal aging and AD, although definitions and designations of this subgroup have varied [8,15,25,36, 37]. A number of investigations have suggested that when followed longitudinally, these individuals are at increased risk for developing AD [1,3,9,15,31,36,43]. Furthermore, there is increased risk for incident dementia among elderly people with subjective (or informant corroborated) memory complaints, even if their baseline cognitive assessment is normal [17,26,39,44,45]. Of the few MRI studies that have examined the volumes of mesial temporal lobe structures in elderly patients with mild cognitive impairment or age-associated memory im- pairment, some have found hippocampal atrophy [7,10,25, 49], while others have not [32,41]. Two recent investiga- tions that assessed MRI-based entorhinal volume in such patients found significant EC atrophy compared to controls [31,49]. The present research was undertaken to compare the extent of MRI-derived hippocampal and entorhinal atrophy in very mild and incipient AD in an attempt to determine the earliest sites of pathological involvement in the disease process and to develop in vivo anatomical markers. 2. Materials and methods 2.1. Subjects Data reported here were obtained from the following three groups of participants: 1) 34 healthy elderly normal control subjects (NC), 2) 28 patients who were evaluated for cognitive complaints, but who did not meet clinical criteria for dementia they are referred to here as non-demented patients or ND, and 3) 16 patients with very mild probable AD. The major difference between the ND patients and the elderly controls is that the patients were recruited from a clinic where they were being evaluated for possible demen- tia, whereas the elderly controls were recruited from the general population and had to meet the criteria described below. 2.2. Clinical work-up All evaluations were carried out at the Rush Alzheimer���s Disease Center (RADC, Chicago, IL) as previously de- scribed [14,48]. Briefly, the evaluation incorporated the Consortium to Establish a Registry for Alzheimer���s Disease (CERAD, 34) procedures and included a medical history, neurological examination, neuropsychological testing, in- formant interview and blood tests. The clinical diagnosis of probable AD followed NINCDS/ADRDA guidelines [33] it required a history of cognitive decline and neuropsychological test evidence of impairment in at least two cognitive domains, one of which had to be memory. In the present study, we only included patients with a diagnosis of probable AD whose Mini Men- tal State Examination (MMSE, 17) score was 26 (i.e., those with very mild AD). Patients in the non-demented group presented at the clinic with cognitive complaints, received the same standard evaluation as the AD patients, but did not meet clinical criteria for dementia. Exclusion criteria for both AD and ND patients were evidence of other neurologic, psychiatric or systemic conditions that could cause cognitive impairment (e.g., stroke, alcoholism, major depression). Neuropsycho- logical test results indicated that the ND group included 13 members with isolated memory impairment, two members with impairment of attention, one with language deficits and 12 individuals without demonstrable cognitive deficits on clinical testing. It should be noted that no individual patient in this group had a deficit in more than one domain. Because elderly individuals with complaints of memory are at high risk of developing AD even if they do not show significant deficits on formal testing, we did not exclude those ND patients without demonstrable cognitive deficits for the pur- pose of this study. 748 B.C. Dickerson et al. / Neurobiology of Aging 22 (2001) 747���754