Issues Related to the Development of Drugs for the Treatment of Patients with Vascular Dementia (VaD)

1
Issues Related to the Development of Drugs for
the Treatment of Patients with Vascular
Dementia (VaD)

• Andrew Satlin, M.D.
• Director, Clinical Research
• Novartis Pharmaceuticals Corporation
• March 14, 2001

2
Is VaD Clearly Definable Clinically?

• Maximize reliability and validity of diagnosis
  using
• NINDS-AIREN criteria requiring focal signs on
  examination neuroimaging evidence and causal
  relationship

3
Is VaD Clearly Definable Clinically?

• Maximize reliability and validity of diagnosis
  using
• NINDS-AIREN criteria requiring focal signs on
  examination neuroimaging evidence and causal
  relationship
• several studies (e.g., Chui et al. Arch Neurol
  200057191-196) suggest that these criteria are
  conservative for diagnosis (i.e. less sensitive
  but likely to yield more homogeneous population
  than other criteria)

4
Is VaD Clearly Definable Clinically?

• Maximize reliability and validity of diagnosis
  using
• NINDS-AIREN criteria requiring focal signs on
  examination neuroimaging evidence and causal
  relationship
• several studies (e.g., Chui et al. Arch Neurol
  200057191-196) suggest that these criteria are
  conservative for diagnosis (i.e. less sensitive
  but likely to yield more homogeneous population
  than other criteria)
• misclassify only 9 of AD and 29 of mixed as VaD
  (Gold et al. Neurology 199749690-694)

5
Is VaD Clearly Definable Clinically?

• Maximize reliability and validity of diagnosis
  using
• NINDS-AIREN criteria requiring focal signs on
  examination neuroimaging evidence and causal
  relationship
• several studies (e.g., Chui et al. Arch Neurol
  200057191-196) suggest that these criteria are
  conservative for diagnosis (i.e. less sensitive
  but likely to yield more homogeneous population
  than other criteria)
• misclassify only 9 of AD and 29 of mixed as VaD
  (Gold et al. Neurology 199749690-694)
• have at least moderate inter-rater reliability
  (kappa 0.42 - 0.72 Chui Lopez et al.
  Neurology 1994441240-1245)

6
Is VaD Clearly Definable Clinically?

• Maximize reliability and validity of diagnosis
  using
• MRI scanning for all screened patients
• eliminating requirement for temporal relationship
  in cases of pure subcortical VaD by MRI criteria
• inter-rater training in applying NINDS-AIREN
  criteria

7
Is VaD distinguishable from AD?

• Patients who meet clinical criteria for VaD may
  also have AD pathology but
• specificity of NINDS-AIREN criteria are high
  (91) for exclusion of AD (Gold et al., Neurology
  199749690-694)

8
Is VaD distinguishable from AD?

• Patients who meet clinical criteria for VaD may
  also have AD pathology but
• specificity of NINDS-AIREN criteria are high
  (91) for exclusion of AD (Gold et al., Neurology
  199749690-694)
• other evidence suggests that dementia in patients
  with VaD (even due to subcortical ischemic
  vascular disease) does not simply indicate the
  presence of AD (Fein et al., Neurology
  2000551626-1635)

9
Is VaD distinguishable from AD?

• Patients who meet clinical criteria for VaD may
  also have AD pathology but
• specificity of NINDS-AIREN criteria are high
  (91) for exclusion of AD (Gold et al., Neurology
  199749690-694)
• other evidence suggests that dementia in patients
  with VaD (even due to subcortical ischemic
  vascular disease) does not simply indicate the
  presence of AD (Fein et al., Neurology
  2000551626-1635)
• requirement for statistical and clinically
  relevant effect in VaD treatment study will
  preclude possibility that effect is entirely due
  to treatment of AD

10
Outcome Measures in the Design of VaD Clinical
Drug Trials

• Two primary outcome measures, as in AD trials
• VaDAS - cognitive
• includes ADAS-Cog, plus items targeting deficits
  found more commonly in VaD (attention/concentratio
  n executive function verbal fluency working
  memory psychomotor speed)
• additional items not validated in VaD, but each
  valid in assessing AD patients
• recommended by expert committee (Desmond, Ferris,
  Mohs see Ferris, Alz Dis Assoc Disord
  199913S140-2)
• ADCS-CGIC - global rating for clinical relevance

11
Other Features in the Design of VaD Clinical Drug
Trials

• Trials need to be comparable in length to AD
  trials
• Disease is characterized by a slow step-wise
  decline
• Adequate duration needed to ensure decline in
  placebo group and to reduce variability
• Longer duration also provides more safety data in
  a population that may be more medically ill
• Monitor for changes in vascular risk factors,
  e.g., hypertension smoking hyperlipidemia
  diabetes

12
Conclusions

• A properly designed clinical trial in VaD should
  select as homogenous a population as possible to
  ensure that the overall effect is driven by the
  population of interest

13
Conclusions

• A properly designed clinical trial in VaD should
  select as homogenous a population as possible to
  ensure that the overall effect is driven by the
  population of interest
• Study must use outcome measures with demonstrated
  validity and reliability

14
Conclusions

• A properly designed clinical trial in VaD should
  select as homogenous a population as possible to
  ensure that the overall effect is driven by the
  population of interest
• Study must use outcome measures with demonstrated
  validity and reliability
• Study must be of adequate duration e.g.,
  comparable to AD studies