CLINICAL

TRIALS II
Lars Berglund, M.D., Ph.D.
CRISP (Clinical Research Investigator
Services Program)
U.C. Davis School of Medicine

Clinical Trials
Objectives:
Practical aspects on how to succeed
with a trial
Alternative study designs

Clinical Trials
Recruitment:
A key factor no results without participants.
Balance between being too specific or too
diffuse risk of having too narrow
inclusion/exclusion criteria versus
introduction of confounders
Its not entirely your study best if participants
are genuinely interested in the study

Clinical Trials
Run-in period
Pro:
Provides stabilization and baseline
Tests endurance/reliability of subjects
Con:
Can be perceived as too demanding

Clinical Trials
It all comes down to . . .
Making study participants stakeholders
Protocol as simple to follow as possible
refrain from designing the perfect
study
Developing a good contact between
participants and study personnel
Team spirit

Clinical Trials
Outcome variables
1. Clinical outcomes/endpoints (mortality,
admissions, verified disease)

2. Surrogate outcomes
(risk factors, e.g. LDL chol, CRP, bone
density) leap of faith as changes in
surrogate outcomes assumed to predict
clinical outcomes.
Usually power and cost issues.

Clinical Trials
Endpoints
Balance between being too ambitious or too
conservative a bit of common sense
(The OConnor situation).
Focus on a single or a limited number of
pre-specified endpoints
Just because you can measure 15 variables
doesnt necessarily mean that a
statistically significant finding in any of
these is clinically meaningful
Primary vs secondary endpoints

Clinical Trials
Monitoring
Adverse events
Safety concerns grounds to stop a
study?
For small trials investigators monitor
adverse events, large trials have Data and
Safety Monitoring Boards (DSMBs)
Every GCRC study needs a Data and
Safety Monitoring Plan (DSMP)

Clinical Trials
Examples of early termination of studies:
CARET Study (Beta carotene and alpha tocopherol
study) unanticipated CVD risks
Physicians Health Study benefit from aspirin
CARDS (Collaborative Atorvastatin Diabetes Study)
benefit of atorvastatin for CVD in type II
diabetes

Clinical Trials
Clinical trial phases
Phase I safety test in volunteers
Phase II small randomized, blinded study
to test tolerability and efficacy on
surrogate outcomes
Phase III large, randomized, blinded trial to
test clinical outcome
Phase IV post-marketing study to assess
side effects and/or additional uses

Clinical Trials
Alternative designs
Factorial design two questions asked, examples
Physicians Health Study (aspirin and betacarotene), Heart Protection Study (simvastatin and
vitamins)
Time series study subjects studied twice (no
control group, could be affected by learning trends,
secular trends time of year, regression to the
mean)
Cross-over design wash-out period

Clinical Trials
Examples of studies and practical
problems:
1. Program project dominated by
basic science with one clinical
component
2. Small investigator-initiated study
3. Large, multicenter NIH-sponsored
study

Clinical Trials
1. NIH: Diabetes and accelerated vascular
disease (program project)
Clinical project: Atherogenic TG-rich lipoproteins
in diabetes
Focus: Postprandial response in relation to
coronary artery disease would postprandial
response be associated with presence of
coronary artery disease?

Clinical Trials
Program project dominated by basic science
with one clinical component:
Postprandial study in diabetic patients with
and without coronary artery disease
Design:
Measurement of blood lipids and endothelial
reactivity after a fat-rich high-calorie meal

Clinical Trials
Problems:
Inclusion/exclusion criteria
Safety parameters (risk of too high TG,
hypo/hyperglycemia)
Optimal design relation of fat meal to
endothelial flow measurements

Clinical Trials
Inclusion/exclusion criteria:
How to rule out coronary artery disease in
patients with diabetes?
Initial approach: Negative angiograms within
1 yr of recruitment - Recruitment problems
Modified: Within 2 yrs of recruitment
Modified: Negative thallium exercise test

Clinical Trials
Safety parameters:
High TG:
Patients with DM have commonly hyperTG risk for
very high TG during fat load and potential risk for
pancreatitis need to screen for baseline
parameters

Hyper-hypoglycemia:
Patients fasting following fat load measure fasting
glucose at start of study

Clinical Trials
Optimal design:
Time involved 10 hr follow-up needed
Need for endothelial reactivity measurement
at baseline and at expected TG peak

How to ensure a smooth study for each

patient?

Clinical Trials
2. Small investigator-initiated study:
Influence of estrogen on lipoprotein(a) in
postmenopausal women
Issues:
1. Define length of wash-out phase
2. Practical aspects: Placebo, who will
ensure blinding, randomization in blocks
3. How to deal with subgroups (ethnic,
Lp(a) characteristics)

Clinical Trials
3. Large, multicenter NIH-sponsored
study (4 research centers)
Dietary effects on lipoproteins and
thrombogenic activities (DELTA)
Design: Randomized, double-blind
cross-over study to test
carbohydrates vs monounsaturated
fat as replacement for saturated fat
in insulin-resistant subjects

Clinical Trials
Dietary effects on lipoproteins and
thrombogenic activities (DELTA)
Challenges:

Recruitment retainment (meals provided and

consumed on site)
Meal design same menus for Minnesota and
Louisiana
Weight maintenance, physical exercise
Wash-out period
Between-center variability harmonize teams

Clinical Trials
What resources do you need?
Good teamwork (PI and staff)
Recruitment base
Infrastructure (stats, databases, IRB
protocols, location)

Clinical Trials
Resources at UC Davis:

CRISP

GCRC