Mary Ganguli

1
Mary Gangulis Slides

• March 13th Meeting

2
Mild Cognitive Impairment

• A View from the Trenches

3
FDAs 5 Questions

• Clinical definition of MCI.
• Validity of MCI Diagnostic Criteria.
• Distinction from AD and other dementias.
• Appropriate outcome measures for MCI clinical
  trials.
• 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 patients 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.