1. Mild Cognitive Impairment as a Target for Drug Development Steven H. Ferris, Ph.D. Silberstein Aging and Dementia Research Center New York University School of Medicine

2. Aging, AAMI (ARCD), MCI, and AD Cognitive Performance Frequency Young Adapted from Ferris and Kluger. Aging, Neuropsychology and Cognition, 1996. Elderly 1 SD A B ADMCIAAMI

3. Longitudinal Course of MCI AAMI / ARCD MCI AD Age Cognitive Decline

4. Syndrome of Mild Cognitive Impairment (MCI) Mild cognitive decline that is worse than typical for age but less severe than in dementia (Flicker, et al, 1991) –Mild Impairment involves memory and generally other cognitive domains that are more impaired in dementia –Common activities of daily living (ADL) are intact, but there may be subtle impairment in very complex ADL Often a very early stage of dementia (most eventually progress to dementia, 10 - 15% per year, 80% over 10 years) –When selected using “AD” inclusion/exclusion criteria, cases generally have prodromal AD (80% have hippocampal atrophy, 75% have AD neuropathology at autopsy)

5. Decline To Dementia Among Nondemented Elderly (N=213) 1 1 Adapted from Kluger, Ferris, Golomb et al, J. of Geriatric Psychiatry and Neurology, 1999 ***

6. Heterogeneous Syndrome vs. Specific Disease Normal MCI Dementia Aging Normal MCI of Alzheimer’s Aging AD Type Disease Syndrome Disease

7. Normal Cognition Prodromal Dementia Dementia Brain Aging Mild Cognitive Impairment Stable Or Reversible Impairment Other Dementias Alzheimer’s Disease Vascular Dementia Reversible Mixed MCI is Prodromal Dementia

8. Clinical Criteria for MCI of AD Type Mild cognitive decline reported by subject or informant Globally, GDS = 3 or CDR = 0.5 Memory impairment confirmed objectively Cognitive and ADL impairment is insufficient for diagnosis of dementia Inclusion and exclusion criteria for AD, except for severity of cognitive and ADL impairment

9. Normal Group MCI Group Proportion Remaining Nondemented Years Since Baseline Evaluation Conversion to AD in MCI

10. Neuropsychologic Prediction of Decline to Dementia* Study (Nondemented sample) N (% decline) SpecificitySensitivity Negative Predictive Value Positive Predictive Value Masur et al, 1994 31794.0%50.0%88.1%68.1% (Community-based) (20.2%) Tierney et al, 1996 123 † 94.0%76.0% —— (Memory-impaired) (23.6%) Devanand et al, 1997 7576.9%81.0%83.3%73.9% (Memory clinic-based) (41.3%) Dal Forno et al, 1995 196 ——91.0%62.0% (Community-based) (12.2%) Kluger, et al, 1999 21391.4%78.4%88.8%82.9% (Research clinic-based) (34.7%) 179 † 96.8%89.3%95.2%92.6% (31.3%) *N = >75. † Decline to AD.

11. Logistic Regression Analyses: Predicting Decline to AD (N = 179) 1 Order of Entry into Logistic Regression † Specificity (%) Sensitivity (%) 1. Education*** 89.4 41.1 74.3 76.9 63.9 2. GDS grouping (1, 2, or 3)*** 89.4 73.2 84.4 88.0 75.9 3. Psychometric tests (set of four): 96.8 89.3 94.4 95.2 92.6 (Paragraph Delayed Recall**, Paired Associate Initial Recall*; Digit Symbol*; and Digit Span Forward*) *p<0.05; **p<0.001; ***p<0.0001. † Forcing in age, sex, and follow-up interval first in all three steps, next sequentially forcing in education and then GDS, and finally allowing for the possible stepwise entry of the psychometric variables. Overall Accuracy Predictive Value (%) Negative Positive 1 Kluger, Ferris, Golomb et al, J. of Geriatric Psychiatry and Neurology, 1999

12. Predicting Decline from MCI to AD with Delayed Paragraph Recall (N=71) 100 80 60 40 20 0 >12>11>10>9>8>7>6>5>4>3>2>1>0 Specificity Sensitivity Overall accuracy Cut Scores Accuracy of prediction %

13. Regression of Delayed Recall on Hippocampal Size 1 Residualized Hippocampal size (rHF) Residualized Delayed Recall (rDR) r = 0.49 1 Adapted from Golomb et al. Learning and Memory, 1994.

14. MCI Trials As a Bridge to Prevention Primary prevention trials require very large samples and long treatment durations due to low annual rate of conversion to AD Study MCI rather than normal elderly, since conversion rates are high (study duration 2-3 years, N < 1000) Disease progression study rather than primary prevention (most already have prodromal AD) Outcome: Time to clinical diagnosis of AD or rate of cognitive decline –May be confounded by “symptomatic” effects –May be corroborated by effect on MRI atrophy

15. Memory Screening in MCI Trials Objective confirmation of mild memory impairment Increase proportion of cases with prodromal AD Enrich study population for conversion to AD

16. Increasing MCI to AD Conversion Rate with Memory Impairment Criteria* 0 10 20 30 40 50 60 70 80 90 100 0.00.51.01.52.0 2.5 3.0 3.5 Memory above cutoff Memory below cutoff Follow-up Time (Years) Proportion Free of Dementia (AD) *Grundman-ADCS pooled data: Logical Memory II cutoff scores

17. MCI Screening / Enrichment Tests Rey Auditory Verbal Learning Test WMS Logical Memory II NYU Paragraph Recall Test Buschke Cued Recall Selective Reminding Test

18. Outcome Measures for MCI Trials Conversion to AD (survival design) Traditional AD outcome domains –Cognitive function – Global status or change –Functioning (ADL) – Behavior –Quality of Life – Pharmacoeconomics MRI atrophy –whole brain or hippocampal volume –currently best biomarker for supporting disease progression claim

19. Special Requirements for MCI Outcome Measures Cognitive battery must be sensitive to very mild impairments (ADAS-cog is not optimal) Global and ADL instruments must be sensitive to subtle impairments in complex activities Depression is the most relevant behavioral domain Suitable instruments have been developed and are being used in current MCI trials

20. Conclusions MCI is a heterogeneous syndrome Homogeneous groups representing prodromal AD or other subtypes can be identified MCI trials can examine disease progression and provide a bridge to prevention Suitable outcome measures for MCI trials are available Labeling for specific prodromal dementias (e.g., MCI of the AD type) is appropriate