1109

Journal of Alzheimer’s Disease 41 (2014) 1109–1116 DOI 10.3233/JAD-131916 IOS Press

Initial Memory Deficit Profiles in Patients with a Cerebrospinal Fluid Alzheimer’s Disease Signature Jing Xiea,∗ , Audrey Gabelleb , Aline Doreya , Antoine Garnier-Crussarda , Armand Perret-Liaudetc,f , Floriane Delphin-Combea , Anthony Bathsavanisa , Virginie Dauphinota , Sylvain Lehmannb , Bernadette Mercierd , Virginie Desestretd,e,f , Isabelle Roullet-Solignacd , Alain Vighettod,e,f and Pierre Krolak-Salmona,e,f a Research and Resources Memory Center of Lyon, Hˆ opital des Charpennes, Hospices Civils de Lyon, Lyon, France b Research and Resources Memory Center of Montpellier, University Hospital of Montpellier, and INSERM U1040,

Institute of Biotherapies, France Department, Hospices Civils de Lyon, Lyon, France d Research and Resources Memory Center of Lyon, Hˆ opital Pierre Wertheimer, Hospices Civils de Lyon, Lyon, France e University Lyon I, Lyon, France f INSERM U1028, CNRS UMR5292, Lyon Neuroscience Research Center, Lyon, France c Neurobiology

Accepted 10 March 2014

Abstract. Background: Alzheimer’s disease (AD) clinical onset is usually characterized by a memory complaint and a progressive memory deficit. The proportion of typical medial-temporal amnesia revealing AD remains unknown. Objective: The present study explores the episodic memory impairment profiles by the Free and Cued Selective Recall Reminding Test (FCSRT) in patients with initial memory complaint and a cerebrospinal fluid (CSF) biomarker signature of AD. Methods: Seventy-three patients referred for memory complaint to the Centers for Memory, Resource and Research of Lyon and Montpellier (France) were included consecutively. All patients underwent an extensive neuropsychological examination and had a Mini-Mental State Examination (MMSE) score ≥20 and a positive CSF AD signature. The patients were classified as having mild dementia or prodromal AD. Verbal episodic memory was assessed using the French version of the FCSRT exploring encoding, storage/consolidation, and cued delayed retrieval phases of memorization. Three different memory profiles were identified according to the results of FCSRT. Results: The median age was 72 year-old [interquartiles: 65–76]. The median MMSE score was 23 [interquartiles: 21–25]. 88% of the patients (n = 64) presented with a medial temporal amnesia profile. The dysexecutive amnesia and normal verbal episodic memory profiles represented respectively 5% (n = 4) and 7% (n = 5). There were no significant differences in term of age, gender, and MMSE score between the three profile groups. Conclusion: In a population initially presenting with memory complaints and depicting a CSF AD signature, a high proportion of medial temporal amnesia is disclosed as expected, but also a proportion of dysexecutive amnesia and normal FCSRT. Keywords: Alzheimer’s disease, cerebrospinal fluid, memory

INTRODUCTION ∗ Correspondence

to: Dr. Jing Xie, MD, PhD, Hˆopital des Charpennes, 27, Rue G P´eri, 69100 Villeurbanne, France. Tel.: +33 4 72 43 20 50; Fax: +33 4 72 43 20 54; E-mail: [email protected].

Alzheimer’s disease (AD) is a neurodegenerative disease usually heralded, at early clinical stages, by a progressive episodic memory deficit [1, 2]. Medial

ISSN 1387-2877/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved

1110

J. Xie et al. / What Memory Impairment Initiates AD?

temporal amnesia, characterized by unavailable encoded information due to a deficit in memory consolidation processes, may be the most common type of episodic amnesia at early stages of AD [3–5]. Frontal dysexecutive amnesia leading to defect of the encoding of new information and of successful retrieval may also lead to episodic memory deficits [5–7]. Encoding may also be impaired in the context of attention disorders. However, recent lesion and neuroimaging studies indicate an interaction and overlap in function between the frontal lobe and medial temporal/diencephalic structures engaged in memory formation [5]. At the mild clinical stages of AD, tangles spread from the entorhinal cortex in the parahippocampal region to the hippocampus proper, associated with episodic memory deficits of the “medial temporal profile” [8]. However, the heterogeneity of AD initial neuropsychological profiles suggests the existence of subgroups including patients with progressive focal cortical atrophy involving frontal, temporal, parietal, and occipital lobes, as well as co-existing lesions like Lewy body and vascular pathologies [9–12]. Medial temporal profile patients show faulty storage and consolidation, whereas frontal lobe profile patients show faulty retrieval [5]. Thus, frontal lobe dysfunction may contribute to memory complaint and deficits inaugurating the clinical phase of AD [9, 13]. Epidemiological studies even show that some executive dysfunction may be disclosed years before the clinical diagnosis of AD [14]. Some qualitative diagnosis approaches are now used through neuropsychological testing describing encoding, consolidation, and retrieval phases of memorization and thus, disclosing medial temporal type amnesia [3]. In this study called Pre-Al, the authors suggested that the Free and Cued Selective Recall Reminding Test (FCSRT) would be a good neuropsychological test for positive diagnosis of prodromal AD with a sensitivity and a specificity of nearly 80% and 90%, respectively [3]. Even if the medial temporal amnesia profile is particularly associated with AD, little is known about the proportion of AD patients who start the disease by a frontal dysexecutive amnesia disorder. The recent development in the field of diagnosis biomarkers research has allowed a dramatic improvement of early diagnosis of AD in vivo [15]. A decreased CSF amyloid-␤ (A␤)1-42 combined with an increased CSF total tau (t-tau) and phosphorylated tau (p-tau) protein levels have been established as a “CSF Alzheimer’s disease signature”. Therefore, our aim was to explore qualitatively the encoding, storage/consolidation, and retrieval phases

of memorization by the FCSRT in patients with initial memory complaint, in which a CSF biomarker signature is in favor, in vivo, of AD pathology. In this population with a CSF AD signature, this study provides the proportion of patients with a typical medial temporal amnesia profile and those with a frontal dysexecutive profile. MATERIALS AND METHODS Study design and subjects Between January 1, 2004 and December 31, 2011, community patients consulting for an isolated progressive memory complaint were selected consecutively at the Center for Memory, Resources and Research (CMRR) of Lyon (Center 1) and Montpellier (Center 2) (France). The two CMRRs performed the same neuropsychological assessments widely used in France. All included patients had an extensive neuropsychological examination, presented a Mini-Mental State Examination (MMSE) score ≥20, and a positive CSF Alzheimer’s disease signature. The patients were classified as having mild dementia or prodromal AD. We excluded patients with insufficient French language abilities, abnormal neurological examination except cognition, disabling somatic disease, severe depression or any uncontrolled psychiatric axis I disorder, patients with antipsychotic drugs, patients with antidepressant or benzodiazepine drug unstabilized for at least 3 months, severe sub-cortical microvascular disease (Fazekas score of 3 for deep white matter hyperintensities [16]), expansive or any focal lesion other than atrophy on brain MRI. All patients gave their written informed consent to participate in this research study. A local ethic committee approved the use of human subjects for this study. Neuropsychological assessments All included patients underwent a neuropsychological evaluation by specially trained neuropsychologists within the week before the lumbar puncture. The verbal episodic memory was assessed using the French version of the FCSRT in all patients [17]. The FCSRT was administered according to the Grober and Buschke paradigm [17]. Sixteen words, each belonging to a different semantic category, are presented on successive cards. After an immediate recall, 3 successive trials combining a free recall of the 16 items (each score under 16) is followed by 3 cued recalls for the items not spontaneously recalled. After 20 minutes, a delayed

J. Xie et al. / What Memory Impairment Initiates AD?

1111

free and cued recall is performed (score under 16). During this time period, the Trail Marking Test – parts A and B (TMT-A, TMT-B) [18], the Wechsler Adult intelligence Scale (WAIS III) Digit Span [19] backward Test, and the Benton Visual Retention Test (BVRT) [20] were performed. The total recall score is calculated from the three free and cued recalls (score under 16 × 3 = 48). The cutoff scores for diagnosis of medial temporal type amnesia according to the Pre-Al study are 17/48 for free recall and 40/48 for total recall. Executive function and attention were assessed using TMTA and TMT-B. We considered a flexibility impairment in patients who failure these tests. The WAIS III Digit Span (forward and backward) test was used for working memory in all patients. The visual memory was variously assessed using the DMS 48 [21], the BVRT, and the immediate free recall and delayed recall of the Signoret’s complex geometric figure [22]. The language was assessed using denomination tests: the DO 80, Bachy and the Lexis, according to the patients. Three memory profiles were distinguished: the medial temporal profile with faulty storage and consolidation, the dysexecutive profile with faulty retrieval, and the normal profile without any faulty storage, consolidation, and retrieval [5], as depicted by the FCSRT neuropsychological test by using a cut off scores proposed by Pre-Al study [3] criteria:

supplier’s procedures (Innogenetics Ghent Belgium). CSF biomarker analyses were performed blindly to the clinical diagnosis in the Neurobiology Department of the University Hospital of Lyon or in Biochemistry Department of the University Hospital of Montpellier which are participating both in The Alzheimer’s Association Quality Control program. All biomarker levels were measured in duplicate. Since the intra-assay coefficient of variation was less than 10% and the concentrations obtained were in the range of assay, samples were not retested. We considered positive cutoff values for positive AD diagnosis as follows: t-tau >350 ng/L, p-taul81 >60 ng/L, and A␤1-42