<< Back to Agenda
SlideMoving From Observational Studies to Clinical Trials: Why Do We
Sometimes Get It Wrong?
María Elena Martínez, M.P.H., Ph.D.
Associate
Professor of Epidemiology & Nutrition
Co-Director, Cancer Prevention
and Control
Arizona Cancer Center
University of Arizona, Tucson, AZ
National Institutes of Health
Bethesda, MD
January 11-12, 2005
Slide 2:Clinical Trial vs. Cohort Study (Chart)
Slide 3: Carotene and Lung Cancer
Slide 4: Carotene and Lung Cancer: What Did We Know Prior to RCTs?
- carotene is found in high concentrations in carrots and dark green
leafy vegetables.
- carotene is postulated to have anticancer effects-among the most
efficient quenchers of singlet oxygen molecules.
- Observational epidemiologic studies: Individuals who consume high
amounts of foods rich in carotene or whose blood levels are high experience
approx. 20-40% lower risk of developing cancer, including lung cancer.
- Because confounding could be likely in observational studies, RCTs
are needed to determine whether carotene prevents cancer.
- Dosage for optimal benefit was speculative.
Slide 5: Carotene and Lung Cancer: What Did We Know Prior to RCTs?
(Chart)
Slide 6: -Carotene: An Unusual Type of Lipid Antioxidant
At higher oxygen levels, ? carotene loses its antioxidant activity and
shows an autocatalytic, pro-oxidant effect, particularly at relatively high
concentrations. Similar oxygen-pressure-dependent behavior may be shown by
other compounds containing many conjugated double bonds.
.
Slide 7: Characteristics of RCTs of Carotene (TABLE)
Slide 8: Secondary Analysis of carotene Trials of Lung Cancer
- CARET
- RR = 0.80 for former smokers
- RR = 1.42 for current smokers
- ??ATBC
- RR = 0.97 for 5-19 cigarettes/day
- RR = 1.25 for 20-29 cigarettes/day
- RR = 1.28 for more >29 cigarettes/day
- ??PHS
- RR = 0.78 for nonsmokers
- RR = 1.00 for former smokers
- RR = 0.90 for current smokers
¢
Slide 9: Carotene and Lung Cancer: What Do We Know Now?
- No overall benefit of carotene for smokers.
- Wide array of molecular responses identified that may account for
adverse events in smokers.
- carotene is not a very potent antioxidant and has various other
actions.
- No anti-cancer effects of carotene for well-nourished populations.
- Plausible that any chemopreventive action of carotene supplementation
is confined to poorly nourished individuals (Linxian).
- Individuals who eat high quantities of fruits and vegetables have
lower risk of death from various chronic diseases.
Slide 10: "Nutritional epidemiology is qualitatively incapable of
identifying a dietary compound(s) that will be efficacious"
F. Meyskens
Slide 11: "As if we were searching for a new therapeutic compound, we
have expected high doses of a single nutrient to reproduce the beneficial
effects of the complex nutrient mixtures found in whole foods. Perhaps this
basic assumption is wrong."
T. Byers, CA Cancer J Clin, 1999
Slide 12: "The clinical trials ought to be seen as representing a
triumph of the scientific process rather than a failure of therapy. We now know
that carotene supplements are not an effective means of lowering the risk of
cancer
No one should discount the importance of epidemiologic studies of
diet and chronic diseases. Persons who eat a relatively large quantity of
vegetables, fruits, and grains have a profoundly lower risk of death,
particularly from cardiovascular disease and cancer."
ER Greenberg, MB Sporn, NEJM 1996
Slide 13:Studies of Adenoma Recurrence
Slide 14; RCTs of Adenoma Recurrence (TABLE)
Slide 15: Strengths and Limitations of Adenoma Recurrence Studies: Is
it the Right Model?
Strengths
- Prospective nature
- Investigation of etiology of adenoma formation
- Detection bias minimized
- Relatively short-follow-up period needed for end point analysis
- Selectivity of population and limited generalizability
- Assessment of premalignant lesion rather than invasive cancer
- Investigating risk factors related to early events in carcinogenesis
pathway
- Short duration in follow-
- up from exposure to end point
Limitations
- Selectivity of population and limited generalizability
- Assessment of premalignant lesion rather than invasive cancer
- Investigating risk factors related to early events in carcinogenesis
pathway
- Short duration in follow-up from exposure to end point
Slide 16: NSAIDs and Colorectal Neoplasia
Slide 17: Cohort Studies of NSAIDs and Colorectal Cancer (Graph)
Slide 18: Colorectal Cancer and Adenomas Have High Incidence of
COX-2
- 85% of cases positive at both the RNA and protein levels
- 40-50% of colorectal adenomas showed significant elevation of
COX-2
Slide 19; Effect of Celecoxib in Familial Adenomatous Polyposis:
Percent Change in Number of Polyps After Six Months of Treatment (bar chart)
Slide 20: Celecoxib/Selenium Trial Schema (Pre-December 20, 2004) (Flow
Chart)
Slide 21: FDA Statement
For Immediate Release
Media Inquiries: 301-827-6242
December 17, 2004
Consumer Inquiries: 888-INFO
FDA Statement on the Halting of a Clinical Trial of the Cox-2 Inhibitor
Celebrex
?
The FDA today released the following statement on the halting of a
clinical trial of the Cox-2 inhibitor Celebrex (celecoxib):
The Food and Drug Administration (FDA) learned last night from the
National Cancer Institute (NCI) and Pfizer, Inc., that NCI has stopped drug
administration in an ongoing clinical trial investigating a new use of Celebrex
(celecoxib) to prevent colon polyps because of an increased risk of
cardiovascular (CV) events in patients taking Celebrex versus those taking a
placebo.
Patients in the clinical trial taking 400 mg. of Celebrex twice daily
had a 3.4 times greater risk of CV events compared to placebo. For patients in
the trial taking 200 mg. of Celebrex twice daily, the risk was 2.5 times
greater. The average duration of treatment in the trial was 33 months.
Slide 22: Celecoxib/Selenium Trial Schema (Post-December 20, 2004)
(FLOW Chart)
Slide 23: A question of Balance: Chemoprevention Trials Require Both a
High Level of Efficacy and Safety (Picture)
Slide 24: Safety of Chemoprevention Trials
- What evidence on toxicity of the agent should preclude the launching
of a RCT?
- Previous trials might not have considered this adequately
- Phase II trial data on toxicity needed
- How much toxicity are we willing to tolerate? None? Minimal?
Slide 25: Themis and Justice (Picture)
Slide 26: Removing the Blindfold
.Moving Forward (Picture)
Slide 27: Moving Forward
.
- Systematic development of chemopreventive agents is a long process,
multi-factorial process
- All available evidence must be considered and any missing pieces must
be disclosed or filled in
- Development of studies that will identify safe and efficacious agents
that can be integrated into routine preventive medical practice
- As a research community, we need guidelines to inform that process in
the most useful way
- Models have been proposed (i.e., Meysken's algorithm). Such models
will need to be validated using existing RCTs
- Next generation of RCTs will need to incorporate lessons learned so
far
Slide 28: Lessons Learned
- Must test effects of nutritional supplements given in broader
combinations, and in more modest doses, thereby simulating the micronutrient
combinations in the matrix of whole foods
- We should critically appraise observational methodology as well as
limitations of RCT design
- Trials should be designed to be long-term, testing nutrients over
many years among people at average risk
- Must acknowledge agent's anti-carcinogenic potential as well as
disruption of normal homeostasis
Slide 29: Future Directions
- Do we continue searching for the "magic bullet"?
- Should we consider a pill as an alternative to lifestyle practices
(i.e., diet, exercise, tobacco)?
- Do we continue to believe that a RCT is necessary before fully
accepting a factor as protective? Some lifestyle factors are not amenable to
double-blind RCTs.
- All agents must be suspected of adverse effects; don't just focus on
their potential benefits.
<< Back to Agenda