1. Mary Ganguli’s Slides March 13 th Meeting

2. Mild Cognitive Impairment A View from the Trenches

3. FDA’s 5 Questions A.Clinical definition of MCI. B.Validity of MCI Diagnostic Criteria. C.Distinction from AD and other dementias. D.Appropriate outcome measures for MCI clinical trials. E.Special design features for MCI clinical trials.

4. Additional questions What is MCI? How is it similar or different to related concepts such as AAMI, CIND? How is it related to Clinical Dementia Rating 0.5 Memory Only? How is it similar and different to normal aging? Is MCI a single condition? Is MCI a homogenous condition?

5. What is the conceptual issue in question? Clinicians are familiar with patients who do not seem either quite normal or quite demented. Typically we make – or reserve - a judgment as to whether we think they have an incipient dementing disorder. Would we all make the same judgment, given the same patient?

6. Clinical Characterization of MCI Petersen Criteria: 1. Memory Complaint 2. Normal Activities of Daily Living (ADLs) 3. Normal General Cognitive Function 4. Abnormal Memory for Age 5. Not Demented –(Petersen et al 1999)

7. A. Clinical Definition of MCI Can MCI be clearly defined in a clinical setting? Translation: Can the Petersen criteria be clearly applied to patients in the clinic?

8. 1. Memory Complaint Must the patient complain spontaneously or can the physician elicit this complaint by questioning? What if the patient denies memory problems? What if there is no reliable informant? Can patients with anosognosia have MCI?

9. 2. Normal ADLs Assuming this includes normal Instrumental ADLs (IADLs): Is “normality” determined by self-report? What if there is no reliable informant? Does it depend on the patient’s actual IADLs? (e.g., impairment in cooking and keeping house may occur earlier than impairment in drinking beer and watching TV.)

10. 3. Normal general cognitive function Does this mean “normal” scores on brief mental status screening (which is the most that will happen in the average office practice)? How will “normal” be defined? Does it depend on age, sex, education, etc.? Does it depend on whether or not the same patient previously had a higher score?

11. 4. Abnormal memory for age. Must the same standard memory test (e.g. Wechsler Memory Scale) be administered to all patients before the diagnosis of MCI is made? Must appropriate age-norms be available on the given test? (e.g., age-norms by race, sex, and language.) Can MCI be diagnosed without neuro- psychological testing?

12. 5. Not Demented How is dementia being defined? Does this mean “not meeting DSM criteria for dementia,” i.e., not demonstrating decline in two or more cognitive domains, including memory, sufficient to interfere with social and occupational functioning?” Does a Clinical Dementia Rating (CDR) =0.5 mean “not demented?”

13. B. Validity of Clinical Criteria B. Are there valid clinical criteria for the diagnosis of MCI? Translation: Do the criteria in fact measure what they purport to measure?

14. Aspects of Validity Face validity- “makes sense, appears valid.” Content validity – covers the appropriate content.  Criterion-related validity: ? Concurrent validity – associated with an external, independent, gold standard criterion. ? Predictive validity – predicts certain outcomes.

15. Face and Content Validity Seems internally consistent, BUT Is it too exclusive? Is MCI always amnestic? Can another cognitive domain be impaired in isolation?

16. Concurrent Validity Compared to controls, MCI subjects had greater memory loss but were otherwise similar. Compared to mild AD patients, MCI had similar memory loss but were less impaired in other cognitive domains. –(Petersen et al 1999)

17. Circularity vs Concurrent Validity If MCI is defined on the basis of abnormal memory but otherwise normal cognitive and ADL function,  MCI subjects - by definition - will be worse than controls only in memory.  MCI subjects – by definition – will be better than mild AD patients only in domains other than memory.

18. Predictive Validity MCI subjects declined at a rate intermediate between those of controls and mild AD patients. “Conversion” rate to dementia was 12% per year. (Petersen et al 1999)

19. “Conversion” Does “conversion” represent a change in diagnosis or primarily a change in severity of the same condition? Who are the subjects who do not convert? Is it only a matter of time before they all convert? Do some of them convert to conditions other than AD?

20. Potential Interpretations Is MCI a separate entity? Is MCI an intermediate stage on a continuum between normal aging and AD? Is MCI always incipient AD? Is MCI sometimes incipient AD and sometimes something else?

21. C. Differential Diagnosis of MCI Can MCI be distinguished from AD and other causes of dementia? MCI subjects had comparable memory impairment but less impairment in other domains than mild AD patients. MCI subjects suffered less decline over time than mild AD patients. Other Dementias – no data (?)

22. D. Appropriate Outcome Measures for MCI Drug Trials Depends on MCI definition: Raw & change scores (stability, improvement, rate of decline) per unit time on: Memory scores; General mental status scores; Other cognitive domains: attention, orientation, working memory, etc. ADL/IADL scores; “conversion.”

23. Standard Features for MCI Trials Double-blind parallel placebo-controlled; Minimal exclusion criteria; Adequate power to detect small effects (sample size); Adequate length of followup; Intent-to-treat analysis.

24. Additional Special Features for MCI Trials? Normal age-sex-education-matched controls not on drug (normal aging comparison group)? Source of MCI subjects is VERY important: real world “effectiveness” trial may be needed. Randomized start or randomized withdrawal design?

25. Know the Enemy First develop consensus on definition and scope of MCI; Then examine the distribution and outcome of MCI in clinical practice and in the community at large; there may be many high-functioning persons who will, if asked, report isolated memory loss. Then design the intervention trials.