The Journal of Nutrition, Health & Aging© Volume 12, Number 8, 2008
THE JOURNAL OF NUTRITION, HEALTH & AGING©
1 st Conference
Clinical Trials on Alzheimer’s Disease September 17-18-19, 2008 School of Medecine Montpellier, France
ORAL COMMUNICATIONS 01 INTEREST OF CSF ABETA/TAU INDEX IN ALZHEIMER’S POSITIVE DIAGNOSIS. A. GABELLE 1,2,5 , S. ROCHE 1,3,5 , C. GÉNY 2 , F. PORTET 2,4 , J. TOUCHON2,4, S. LEHMANN1,3 (1. CNRS, Institut de Génétique Humaine UPR 1142, Montpellier, France; 2. Service de Neurologie, CHU Gui de Chauliac, Montpellier, France; 3. Laboratoire de Biochimie, CHU Saint Eloi, Montpellier, France; 4. Faculté de Médecine, Montpellier, France; 5. The first two authors contributed equally to this work) Background : Alzheimer’s disease (AD) is a devastating neurodegenerative affection that is approaching epidemic proportions in the industrialized world due to aging of the populations. Recently, new revisited AD diagnosis criteria point out the major interest of CSF biomarkers. The dosage of tau, its phosphorylated form p-tau181 and amyloïd Aβ42 peptide in CSF have come to the fore. Based on Aβ42 and tau values obtained using Innogenetics ELSIA kits, a index called IATI could be calculated and seemed very interesting to discriminate AD from other dementias. Aim : To investigate the diagnostic value of CSF IATI in AD positive diagnosis. Method : CSF tau, p-tau181 and Aβ42 biomarkers were analysed in a consecutive cohort of 167 patients with neurological disease. Eighteen AD and 46 non Alzheimer’s dementias were identified in this population. The AD diagnosis was based on NINCDS/ADRDA criteria. Lombar punctures were performed after informed consent was obtained from the patient or the legal representative. The IATI was defined as Aβ42(240+1.18 x tau). The sensitivity, sensibility and ROC analysis were computed. Results : The IATI is considered pathologic if inferior to 1. The IATI sensitivity was above 90% in our series while its specificity was lower, around 60%. On the other hand, p-tau had a better specificity (>90%) but a lower sensitivity (close to 80%). Discussion : Depending on the clinical situation, the utilization of IATI, p-tau or both appears to be very interesting. The IATI 15 at baseline visit. Random and fixed mixed coefficient longitudinal regression analysis was conducted to determine whether 13 SNPs within GOLM1 were associated with change in mean MMSE or DRS-2 scores over time in the study, adjusted for significant covariates (age, AD duration, education and sex for the MMSE analysis; age, AD duration, and education for the DRS-2 analysis). Time in the study and intercept were modeled as random effects. Significance of the genotype by time interaction indicated whether the SNP influenced the trajectory of test scores over time. We also evaluated whether the number of APOE 4 alleles affected the rate of cognitive decline as measured by these scales. Results: SNPs within GOLM1 were not associated with the rate of decline of MMSE over time. However, five SNPs within a region of high linkage disequilibrium in GOLM1 significantly influenced the trajectory of decline by DRS-2, as judged using the co-dominant genetic model: rs7862366, rs10868366, rs6559900, rs6559901, and rs6559902 (likelihood ratio test p-values for genotype by time interaction terms < .05). The DRS-2 score in individuals carrying the major allele homozygote (the risk genotype in the AD GWA) tended to decline slower at first and then at an increasing rate. The pooled heterozygotes and minor-allele homozygotes tended to start at a similar baseline value of the DRS-2, then decline at a steadier pace. Rate of progression as indicated by MMSE or DRS-2 was not influenced by APOE genotype. Discussion: Genetic polymorphisms such as these identified in GOLM1 may explain a proportion of the variability in clinical progression of AD. Evaluating candidate SNPs from genomewide association studies for additional phenotypic effects in AD cases (in this case, rate of progression) may therefore be a valuable approach to prioritizing SNPs for
542
target development and pharmacogenetics studies. Acknowledgments: Study funded by GlaxoSmithKline. P28 A TWO-YEAR RANDOMIZED TRIAL OF THE IMPACT OF A SPECIFIC CARE PLAN IN 1121 AD PATIENTS (PLASA STUDY): PRELIMINARY RESULTS. F. NOURHASHÉMI 1,2, S. GILLETTE-GUYONNET 1,2, S. ANDRIEU 2,3, Y. ROLLAND 1,2, P.J. OUSSET 1,2, B. VELLAS 1,2 and the PLASA Group (1. CHU Toulouse, F-31059 Toulouse, France.
[email protected]; 2. Inserm U558, F31073 Toulouse, France; 3. CHU Toulouse, Service d’épidémiologie et de santé publique, F-31073 Toulouse, France) Objectives: To evaluate the effect of a multicomponent specific care and assistance plan (PLASA) in AD primarily looking at change in functional capacity. Methods: Twoyear prospective randomized controlled trial comparing PLASA and usual care in 1121 community-dwelling AD included in 49 memory centers in France. Patients in the intervention group were evaluated bi-annually using a standardized comprehensive global assessment. In the case of decline in any one domain (cognitive or non cognitive) a standardized study protocol recommends specific physician directed intervention in addition to information and training for the caregiver. The main objective of the study is to examine whether a well-defined specific care and assistance plan can slow the progression of dependency in AD using ADCS-ADL (Alzheimer Disease Cooperative Study-Activities of Daily Living). The secondary objectives of this project are to study the effect of the PLASA plan on resource use and health care consumption (Resource Utilization in Dementia scale) and overall clinical evaluation of change (Clinical Global Impression of Change). Results: We will present preliminary results concerning effect of PLASA on ADCS-ADL after 2-year of follow-up. At baseline, the two groups were similar regarding patient and caregiver characteristics. The mean patient age was 79.62 + 5.72 years and the mean MMSE 19.73 + 4.00 for the whole cohort. Time since dementia diagnosis was about 1.37 + 1.64 years in the whole cohort. Almost a third of the patients lived alone at baseline. Conclusion: Persons with dementia suffer different losses at different stages of the disease and therefore accurate assessment of abilities and losses is critical to assist the person in planning for their future and for care needs. P29 A WORLD WIDE MULTICENTER COMPARISON OF ASSAYS FOR AD CSF BIOMARKERS. N.A. VERWEY1, K. BLENNOW2, C. CLARK3, G.M. COLE4, P.P. DE DEYN 5, D. GALASKO 6, H. HAMPEL 7, T. HARTMANN 8, E. KAPAKI 9, L. LANNFELT10, P.D. MEHTA11, L. PARNETTI12, A. PETZOLD13, T. PIRTTILA14, L. SALEH 15 , A. SKINNINGSRUD 16 , J.C.V. SWIETEN 17 , M.M. VERBEEK 18 , J. WILTFANG19, S. YOUNKIN20, M.A. BLANKENSTEIN1 (1. Departments Clinical Chemistry and Neurology, VUMC, Amsterdam, The Netherlands; 2. Neurochemistry lab, Möldal hospital, Möldal, Sweden; 3. The Penn Ralston Center, University of Pennsylvania, Philadelphia, USA; 4. UCLA School of Nursing,Los Angeles, USA; 5. Instituur BornBunge, University of Antwerp, Antwerp, Belgium; 6. Department of Biochemical Research, University of San Diego, La Jolla, USA; 7. Psychiatrische ik der LMU, Munchen, Germany; 8. Molecular Biology, University of Heidelberg, Heidelberg, Germany; 9. Department of Neurology, Eginition Hospital, Athene, Greece; 10. Institutionen för folkhälso-och vardnetnkap; Rudbecklab, Uppsala, Sweden; 11. Institute for basic research in development disabilities, New York, USA; 12. Department of Neurology, Ospedale S.M. della Misericordia, Perugia, Italy; 13. Institute of Neurology, Clinical Neurosciences, UCL, London, United Kingdom; 14. Department of Neurology, University of Kuopio, Kuopio, Finland; 15. Department of Clinical Chemistry, University Hospital of Basel, Basel, Switserland; 16. Departmnet of Clinical Chemistry, Akerhus University Hospital, Lorenskog, Norway; 17. Department of Neurology, Erasmus Medical Center, Rotterdam, The Netherlands; 18. Laboratory of Neurology, Radboud University, Nijmegen, The Netherlands; 19. Neurobiology Laboratory, University of Erlangen-Neurenberg, Erlangen, Germany; 20. Mayo Clinic, Jacksonville, USA) Background: CSF Amyloid-beta 1-42 (Aβ1-42), Tau (Tau) and phosporylated Tau (PTau) are being established as biomarkers for Alzheimer’s disease (AD). Different marker levels are reported, but quality control (QC) programs are currently lacking. Methods: Three CSF-pool samples, with different Aβ1-42, Tau and P-Tau levels, were distributed at -80°C in a blinded fashion to 13 laboratories in 2004 and to 18 labs in 2008. Nine labs participated in both time points. Results of the three markers levels for the three samples were reported to the central facility (VUmc) together with information of the assay used and the condition in which the specimens arrived. Results: In 2004 six laboratories measured all 3 markers, 7 measured 1 or 2 markers and all measurements were performed by ELISA. In 2008 12 laboratories measured all 3 markers, 3 measured 1 or 2 markers by ELISA and three laboratories measured all 3 markers by Luminex. Sample N Ab42 pg/ml; mean (SD) Tau pg/ml; mean (SD) P-Tau pg/ml; mean (SD)
1 2 3 1 2 3 1 2 3
11 11 11 9 9 9 8 8 7
2004 ELISA 498 (100) 387 (185) 743 (195) 851 (109) 367 (106) 206 (41) 107 (12) 50 (5) 36 (6)
2008 N
ELISA
N
AlzBio3
15 15 15 13 13 13 12 12 12
460 (157) 284 (101) 682 (258) 844 (156) 315 (40) 165 (30) 108 (17) 54 (8) 40(6)
3 3 3 3 3 3 3 3 3
160 (21) 144 (11) 251 (36) 161 (15) 79 (3) 44 (5) 74 (15) 27 (2) 23 (2)
The Journal of Nutrition, Health & Aging© Volume 12, Number 8, 2008 In 2004 the inter center coefficients of variance (interCV) were for Aβ1-42 31%, for Tau 21% and for P-Tau 13%. In 2008 the ELISA interCV’s were for Aβ1-42 37%, for Tau 14% and P-Tau 15%. The Luminex interCV’s were for Aβ1-42 20%, for Tau 14%, and PTau 22%. The 9 laboratories that participated in 2004 and in 2008 showed an ELISA intracenter CV (intraCV) of 27%18% and 7%. Conclusions: Comparing ELISA results the interCV of Tau decreased from 2004 to 2008 and the interCV’s for Aβ1-42 and P-Tau remained stable. Due to the large differences in absolute CSF levels between ELISA and Luminex, comparing these two methods in this stage is not useful. The ELISA intraCV’s of Aβ1-42 and Tau are comparable with the ELISA interCV’s of 2008, indicating that there is not only a lot of variation between, but also within the centers. P30 USE OF AN US CLAIMS DATABASE AND AN US ALZHEIMER’S DISEASE REGISTRY TO ASSESS CLINICAL TRIAL ELIGIBILITY CRITERIA. L. ISHIHARA-PAUL 1 , A. VISWANATHAN 2 , J.K. ALLEN 1 , B. T. HYMAN 2 , R. BETENSKY2, J. WEIL1, D. BLACKER2, M.C. IRIZARRY1 (1. WW Epidemiology, GlaxoSmithKline, Research Triangle Park, US and Harlow, UK; 2. Massachusetts Alzheimer’s Disease Research Center, Massachusetts General Hospital, Boston, MA, USA) Background: The eligibility criteria for Alzheimer’s disease (AD) clinical trials influence the characteristics of the study population. Excessively stringent inclusion and exclusion criteria can delay recruitment and reduce generalizability. Lenient criteria may lead to heterogeneity in the clinical trial population that can obscure treatment effects. Aims: To examine the impact of patient inclusion and exclusion criteria on patient recruitment for AD clinical trials using the “projection method”. The method applies the diagnosis, inclusion, and exclusion criteria to a database to obtain the proportion and characteristics of individuals eligible for the study. We compared estimates of patient
recruitment when specified eligibility criteria based on sociodemographic factors, stage of illness, AD-related medications, other medications, and comorbidities were applied to: (1) an US managed-care claims database—Integrated HealthCare Information Services (IHCIS); and (2) an US Alzheimer’s disease patient registry—the National Alzheimer’s Disease Coordinating Center (NACC). Results: The eligibility criteria that excluded the greatest proportion of AD patients in the IHCIS claims database were: use of AD medications (cholinesterase inhibitor, memantine), age restrictions (ages 50-80 years required), and comorbidities. The eligibility criteria that excluded the greatest proportion of AD patients in the NACC database were AD medications, age restrictions, and dementia severity (MMSE 10-26 required). Cholinesterase inhibitor and memantine use were more often reported in AD patients within NACC than those in IHCIS. AD patients in IHCIS showed a greater proportion of diagnosed comorbidities such as diabetes, congestive heart failure, arrhythmia, and stroke, than those in NACC. Discussion: Claims databases capture prescribing patterns and reimbursed comorbidities, but have limitations in the study of AD: the elderly are often underrepresented in managed care programs; mild dementia may be under-diagnosed; the severity of dementia is not captured; and the individuals with AD may not reflect the pool of individuals who participate in clinical trials. The NACC AD disease registry tends to draw from the population from which patients are recruited for clinical trials; it contains AD diagnosed by NINCDS-ADRDA criteria; and captures disease severity. However, the registry may not be representative of the general AD population, since patients are entered from referral centers, and have fewer recorded comorbidities. The projection approach allows quantitation of the potential effects of relaxing eligibility criteria on clinical trial recruitment based on the database used. Acknowledgments: Study funded by GlaxoSmithKline. NACC is funded by NIA U01 AG016976.
543
