Bench to Bedside: Clinical Trials

IV. Clinical Trials

A. Prevention and Treatment Trials

Hormones and CHD: Discrepancies Between Laboratory Studies, Observational Studies, and Clinical Trials—Dr. Jacques Rossouw

Dr. Rossouw's presentation outlined the following:

• Why are there differences between observational data and clinical trials for cardiac events?
  • Healthier
  • During HT
    • Compliance bias
    • Those who take it are healthier
  • Early coronary heart disease (CHD) events are missed in cohort studies
• Primary prevention
  • WHI
  • Women's International Study of Long Duration Oestrogen after Menopause (WISDOM)
  • RUTH-raloxifene (Raloxifene [RLX] for the heart)
• Secondary prevention
  • All negative for E, E+P
• Why are there differences?
  • Nurses Health Study (e.g., were younger and were on cyclic MPA)
• Why is there a difference between basic science and clinical studies?
  • Hormones may retard atherosclerosis if there is little or no atherosclerosis when women start therapy
  • Not helpful with raised lesions
  • Harmful with complicated lesions
• Conclusion: short-term relief of menopausal symptoms will come at a small CAD or PE price
• Transdermal E2, unlike oral, does not increase triglycerides, etc.
• Future testing is likely to use:
  • Transdermal E2
  • Low dose

Design and Analysis of the Women's Health Initiative Memory Study (WHIMS)—Dr. Steve Rapp

Dr. Rapp's presentation highlighted the following:

1. Design and case ascertainment in WHIMS
2. Results of WHIMS for dementia, mild cognitive impairment, and global cognitive functioning
3. What's next for WHIMS: The WHIMS-Magnetic Resonance Imaging (MRI) study, Supplemental Case Ascertainment Protocol, Extended Followup, and mild cognitive impairment (MCI) analysis
4. Critical analysis: MRI and supplemental case study

Specific points included:

• Worse result for dementia and MCI on HT
• Modified minimental status exam (MMSE) also worsened over time

Questions posed were:

• Why is there an increase in dementia?
• Is the risk sustained over time?
• What is the effect of withdrawal?

Implications of WHIMS—Dr. Sally Shumaker

Dr. Shumaker's presentation outlined the following:

1. What do we know?
2. How did the design of WHIMS influence what was found?
3. What questions remain answered?
4. What questions will be answered by studying other cohorts, and what questions will never be answered?
5. What needs to be done?

Specific points included:

• Further analyses
  • Refining MCI category to amnestic
  • Does HT affect amnestic MCI?
• Supplemental case ascertainment protocol for followup on approximately 900 women who are deceased or not using the Dementia Questionnaire
• Would standardize the instrument and allow phone collection of data
• MRI cross-sectional study on 1,450 women formerly in WHIMS to look for infarcts on HT
  • Secondary:
    • Effect of different treatments on events
    • Duration of treatment effect, etc.
• MRI scans to be conducted from September 2004 to December 2005
• WHIMS extension study (NHLBI funded)
• Determine effect of cessation of HT on cognition and incidence of dementia and MCI
• Cognition in the study of tamoxifen [TAM] and raloxifen (Co-STAR)
  • Funded by NIA
  • Still enrolling
  • Uses same NP measures
• Testing the effects of HT on younger women using the simplified ascertainment
• Weakness is the absence of a baseline measure
• Biases: selection
• Discussion:
  • When does one randomize?
    •   Issues of dosage and timing
    •   Optimization of design: genotypes and phenotypes—mood, and menopause stage
    •   How far back does one go to find a cognitively normal woman?
    •   Use of neurologically intact cohort as a baseline
    •   Cannot study the effects of HT and dementia in younger women—they do not have it
    •   How does one know when women cross over into cognitive decline?
    •   HT exacerbated the underlying pre-existing DX
    •   What is the prevalence of subclinical dementia?
    •   What is the effect size expected for accelerating disease?

PREPARE: Preventing Postmenopausal Alzheimer's Disease With Replacement Estrogen—Dr. Mary Sano

Dr. Sano's presentation outlined the following:

1. Justification of the trial and how it differs from WHIMS
2. Study design
3. Recruitment problems
4. Changes in protocol as a result of WHI/WHIMS
5. What answers does this study provide that other studies did not or do not?
6. What is ongoing now?

Specific points included:

• PREPARE
  • Similar to WHIMS, but its intent is to prevent Alzheimer's disease and memory decline in women 65 years or older with a positive family history of Alzheimer's disease
• Current status
  • All are off medications
  • Statistical advisors and Data and Safety Monitoring Board (DSMB) recommended following them blinded for the full 5 years
• Possible results
  • Describe incident MCI
  • Inform about stopping
• Data
  • N=466
  • MMSE 28.8
  • Twenty-nine percent apolipoprotein E (Apo E) 4+
  • About one-third with hysterectomy
  • Age of onset of dementia in family members may predict onset of dementia

Results of the ADCS Estrogen Treatment Trial for AD—Dr. Ruth Mulnard

Dr. Mulnard's presentation highlighted the following:

1. Estrogen therapy was not effective in delaying the progression of Alzheimer's disease in hysterectomized women.
2. Estrogen therapy did not show stabilization, memory improvement, or any other specific cognitive domain.
3. An "unexplained" improvement in MMSE occurred in treated subjects at the 2-month time point, which was not sustained at the 12-month conclusion of the study.
4. Four cases of deep vein thrombosis (DVT) occurred—representing significant risk for this subject pool—in the consideration of long-term estrogen administration.

Specific points included:

• More decline on Global Clinical Dementia Rating (CDR) at either dose
• Slight separation on MMSE at 2 months for 0.625 mg dose
• No dose effect
• Four DVT on E
• Four baseline uteri missed

Alzheimer's Disease: Efficacy of Transdermal 17 Beta-Estradiol—Dr. Sanjay Asthana

Dr. Asthana's presentation included the following:

1. Discussion of the biological advantages of using transdermal estradiol versus oral conjugated equine estrogen.
2. Demonstration that cognition-enhancing efficacy of estradiol is domain specific and proposition of sensitive neuropsychological tests for future studies.
3. Discussion of the potential impact of various progesterone preparations and cyclic versus continuous administration of HT on cognitive function of women with Alzheimer's disease.

Specific points included:

• Patch uses 17ß estradiol (natural form of estrogen)
• Oral estradiol increases SHBG
• Ratio of estrone/estradiol goes from 1:1 to 5:1 with oral estrogens
• Estrone has much less affinity for the estrogen receptor
• In the postmenopausal state, gonadotrophins are elevated CEE
• CEE contains
  • Forty-five percent estrone
  • Little 17ß estradiol
• Of healthy aging studies, 86 percent of those on estradiol reported positive results in cognitive studies, 50 percent were positive on CEE
• Similar for Alzheimer's disease studies
• Transdermal estradiol results in higher levels of plasma estradiol
• Oral estrogen increases prothrombotic activity
• Advantages of 17ß estradiol
  • More potent
  • Lower incidence of DVT
• Better compliance
• Low dose study done (N=12)
  • Better on Stroop, better on Buschke
• High dose study:
  • Better on attention, Buschke, visual memory

The Effect of SERMs on Cognitive Function and Dementia—Dr. Kris Yaffe

Dr. Yaffe's presentation included the following:

1. Review of the basic mechanisms of Selective Estrogen Receptor Modulators (SERMs)
   • SERMS as estrogens and antiestrogens
2. Presentation of data on the effect of SERMs on cognitive function in older women
   • Prevention of MCI
3. Presentation of data on the effect of SERMs on MCI and dementia in older women
4. Ongoing studies with SERMS

Specific points included:

• SERMS and cognition
• Raloxifene in Multiple Outcomes of Raloxifen Evaluation (MORE) trial
  • No overall effect
  • For women older than 70, better on TMT and word list recall
  • For cognitive decline to 10th percentile, slight reduction in risk on TMTB and word list
• Raloxifene dementia study in MORE
  • 5,386 women
  • 52 dementia
  • Approximately 150 CIND
• On the 120 mg Raloxifene, but not the 60 mg:
  • 33 percent decline in CIND
  • 48 percent decline in Alzheimer's disease (NS)
  • 37 percent Dementia + CIND
• Lasofoxitne trial: negative so far
• Large RCT of arzoxifine
• Co-STAR
  • Tamoxifen (TAM) and Raloxifene (RLX) for dementia

Ethical Framework for the Design and Conduct of Clinical Trials—Dr. Frank Miller

Dr. Miller's presentation outlined the following:

A. An overview of an ethical framework for the design and conduct of clinical trials: seven basic ethical requirements:

1. Scientific/clinical value of the research
2. Scientific validity of research design
3. Fair selection of research subjects
4. Acceptable risk-benefit ratio
5. Independent ethics review
6. Informed consent
7. Protecting rights and welfare of subjects in the course of research

B. Discussion of ethically significant differences between clinical trials and medical care.

C. Discussion of issues relating to enrolling research subjects who may have diminished capacity to give informed consent.

Discussion—Clinical Trials

• Consider genotyping steroid receptor alleles; no data available on ER polymorphisms versus E levels; receptor polymorphisms that increase CRP, cytokines. Repositories of DNA for analysis.
• NHLBI invites sharing of resources from funded studies: PIs should look for hypotheses arising from WHI; $35 million allocated to initiative.
• Consider the whole woman: bone, breast, uterus, adipose, CNS.
• Consider combination of SERMS and E; TAM increase risk of stroke.
• Consider dose and duration of treatment.
• Consider neuroimaging component.
• Were exercise and weight controlled for in clinical trials? There was no apparent interaction with BMI.
• The WHI study groups need long-term followup for durability of effect.
• Conversion of E1 to E2; liver metabolism to cortisol.
• Hypertension: no interactions with treatment.
• Subgroups of large trials can be used to generate hypotheses.
• Measure E levels; determine therapeutic doses and critical windows.
• Proprietary information leads to clinical trials, but academic scientists do not have access to these data.
• CEE: variability between lots can be a great as threefold.
• Study aromatase inhibitors and cognition in the monkey.
• Can transgenic animal models of Alzheimer's disease be used to advantage to study the effects of E on neuropathology?
• Peripheral A beta and translational research: does not work well in humans.
• For multisite MRI: the technology is not yet widely available, and it is difficult to compare across sites.

Questions Arising From Presentations and Problems To Be Addressed

1. Are plasma concentrations of steroids equivalent to concentrations found in brain tissue?
2. What are the opportunities for the administration of estrogen?
3. Oval versus parenteral administration have not been evaluated for efficacy of cognitive function.
4. Effects of cognition by various types of progesterone are not well studied.
5. The effects of varying doses of estrogen have been studied in rodents only, not in nonhuman primates.
6. More study is needed on the effects of steroids on attention as a measure of cognitive function.
7. Who would potentially stand to benefit (in 30 years) from the early administration of hormone therapy?
8. What genetic factors may affect the way women respond to HT?
9. How can additional risk factors be analyzed with animal models to improve the ability to test factors that may impinge on cognitive function (e.g., cardiovascular, diabetes, ischemia)?
10. Should stress measures be included in future analyses (e.g., cortisol)?
11. Are there early protective effects, and are they enduring?
12. There are no data on the effects of estrogen on the coagulation cascade.
13. The WEST, HERS, and WHI had poor vascular outcomes; vascular markers need to be examined.
14. Targets for therapy must include "critical windows."
15. The question of oral contraceptive use is a "black box:" supraphysiologic levels of E in first pills are commercially available.
16. How can the findings of the Cache County study be reconciled with the Petitti Women's Memory Study? Could both sets of data be right? Different populations, different exposures?

Broad Discussion of the Meeting as a Whole

Data and Studies Needed

• Mechanistic studies for why E+P causes harm.
• Not enough is known to conduct a clinical trial at present.
• Consider lifestyles in data acquisition (e.g., obesity).
• Do estrogens work only in healthy neurons?
• After identifying biomarkers in vitro and in vivo in animal models, can clinicians use surrogate and biomarkers in clinical trials?
• Better delivery systems for E2.
• Are studies needed that use either cycling, infrequent, and/or intermittent estrogen therapy?
• Is the difference attributed to the aging brain or menopause? It is necessary to make young animals menopausal and study the effects of hormones?
• In additional models and humans, it is necessary to confirm whether E in young women and young rodents is helpful, but E in old women and old rats is harmful.

Designs

• None prepared as of yet.

Effects on the Research Question

• Cognitive aging
• MCI
• Dementia

Populations To Study

• Perimenopausal women
  • Parameters:
    •     MRI
    •     Pittsburgh compound-B (PIB) scan in a subset
    •     High-risk women such as Apo E e4, other

Possible Treatments    

• Transdermal E + P
• Low dose E + P
• Testosterone + E
• E alone
• Blood brain barrier (BBB) crossing SERMS
• BBB noncrossing SERMS + E

Recommendation for Formulation of Work Groups

Work Group topic suggestions include molecules, subjects, and endpoints.