|
QUEST -
Clinical Trial 
AUTHOR
Craig Chepke
Chairman’s Rounds, April 30, 2007
Citation:
Buchanan, Robert W.,
et al. 2005. “A Summary of the FDA-NIMH-MATRICS Workshop on Clinical
Trial Design for Neurocognitive Drugs for Schizophrenia.”
Schizophrenia Bulletin vol. 31 no. 1 pp. 5–19, 2005.
Clinical Question /
Background Info:
Cognitive
impairments are a core feature of schizophrenia and a major determinant of
poor functional outcome. No current pharmacological treatments expressly
target these impairments. The lack of consensus on clinical trial design to
establish therapeutic efficacy has been a formidable barrier to drug
development, so a steering committee comprised of NIMH, FDA, and MATRICS
scientists selected experts in the areas of cognitive impairments in
schizophrenia, neurocognition, neuropharmacology, clinical trial
methodology, and biostatistics to propose guidelines to facilitate future
research.
Inclusion Criteria
•Question 1—
Diagnosis:
Meta-analyses suggest that patients with schizophrenia are characterized by
a distinctive pattern of cognitive impairment. Schizophrenia and
schizoaffective disorder share a similar pattern of cognitive impairments,
but the FDA requires greater diagnostic specificity, so initial studies
should include schizophrenia only. A further distinctive feature of
cognitive impairments in
schizophrenia is that
several core cognitive impairments in schizophrenia are relatively stable
across fluctuations in clinical symptoms.
•Question 2— Clinical
State and Symptom Severity:
a) have been clinically stable and in the non-acute phase for a specified
period of time (e.g., 8–12 wks). b) No major medication changes for a
specified period of time (e.g. 6-8 wks). c-e) no more than a “moderate”
severity ratings on each of: hallucinations/delusions, positive FTD, and
negative symptoms. f) a minimal level of EPS and depressive symptoms.
Rationale: maximize isolation of the drug effect on cognition from other
concurrent changes in clinical status that also may affect cognitive
performance.
•Question 3—
Antipsychotic Medications:
In stage I, avoid pharmacodynamic or pharmacokinetic interactions between
the adjunctive/co-treatment agent and the antipsychotic. In stage II,
evaluate the impact of potential pharmacodynamic or pharmacokinetic
interactions with a larger, stratified sample, with few if any restrictions
on allowed antipsychotics.
•Question 4—
Polypharmacy:
The suggested guideline is to exclude subjects taking more than one
antipsychotic
•Question 5— Concomitant
Medications:
Similar to
question #3, In stage I, avoid pharmacokinetic and pharmacodynamic
interactions between the adjunctive/co-treatment agent and any concomitant
medications (e.g., SSRIs). In stage II, examine potential pharmacokinetic
and pharmacodynamic interactions on an agent-specific basis
•Question 6— Maximum
Level of Impairment:
Exclude
patients from a trial only if their cognitive impairment severity
compromises the validity of the cognitive outcome measures. There is some
evidence that those with the least impairment benefit most, as well as other
evidence that those with the most severe cognitive deficits may benefit
most.
•Question 7— Minimum
Level of Impairment:
Exclude subjects from a trial if their level of cognitive functioning is so
high that they perform at or near ceiling and therefore cannot demonstrate
improvement. With a properly constructed test battery, this will be very
rare.
•Question 8— Screening
Assessments:
If a
screening assessment must be used, then use an assessment that is different
from the measure used to assess cognitive outcome during the trial. The
screening instrument should address potential practice effects, novelty
effects, and the natural tendency of deviant performances to regress toward
the mean upon further testing.
Outcome Measures
*Primary Outcome—
The MATRICS cognitive battery will be used to
assess the primary outcome measure: change in cognitive performance. The
battery will assess the following seven domains: attention/vigilance,
reasoning and problem solving, speed of processing, social cognition, verbal
learning and memory, visual learning and memory, and working memory.
•Question 9—
Co-primary Outcome Measures:
The FDA
requires concurrent change on a co-primary measure of functional outcome for
approval of a neurocognitive drug for schizophrenia, such as measures that
reflect clinically meaningful improvement. The arguments against a
co-primary measure of functional outcome, including community (e.g., work
and social) outcome are that the reliability and/or validity of self-report
measures of functional status are not well established, and mediating
variables (e.g., coping ability, skill acquisition, social cognition) may
obscure the translation of cognitive-enhancing effects into changes in
functional status.
•Question 10—
Co-Primary Measure Characteristics:
a) good face validity for patient improvement; b) expected to change in
close temporal proximity to changes on cognitive performance measures; c)
not be heavily dependent on range of rehabilitation opportunities and level
of social support; and d) practical for the experimenter and tolerable for
the subject.
•Question 11—
Validity of Proxy Measures:
a) good test-retest reliability; b) demonstrated associations with cognitive
performance measures; and c) demonstrated associations with community
functional status.
Other Design and
Statistical Issues
•Question 12—
Choice of Primary Measure:
A complicated issue, but briefly, pre-specify a single reliable and valid
primary cognitive outcome measure, either global or domain-specific, based
on its psychometric properties and results of pilot studies.
•Question 13— Testing
Occasions:
Use more occasions to reduce the impact of attrition & capture change in
symptom severity over time
•Question 14—
Heterogeneity of Severity and Response:
In order to reduce baseline within-group heterogeneity and to increase the
chance of detecting a therapeutic effect, include subjects in the residual
(non-acute) phase of their illness and use one primary efficacy measure.
•Question 15— Concurrent
Change in Symptoms:
Statistical approaches cannot be used to rule out pseudospecificity (i.e.,
an artificially narrow claim of cognitive enhancement that could result from
post-baseline confounding, such as reduction in other aspects of the
illness). Pseudospecificity is best dealt with by restricting symptom
severity prior to randomization.
•Question 16— Comparison
Group: To
study an adjunctive/co-treatment agent, use placebo as the comparator. The
choice of comparator for a broad spectrum agent poses a more substantial
challenge, but should be, at worst, cognitively neutral.
•Question 17— Trial
Duration:
The trial needs to be of sufficient duration to show an enduring effect on
cognition (i.e., at least 6 months). Longer duration studies should use
multiple testing occasions, which require the existence of parallel forms of
the outcome measure.
Return
to top of the page
Chairman’s
Rounds
11/22/04
Andrew Kaufman
Research Question:
How does a clinical investigator of antidepressants design a trial
to increase the chance that the results show efficacy?
Khan A, Kolts R, Thase M, Krishnan K, Brown W: Research
Design Features and Patient Characteristics Associated With the Outcome of
Antidepressant Clinical Trials. Am J Psychiatry 2004; 161:2045-2049.
Background: AD’s show no benefit over placebo in half of recent FDA
trials: Why? It is partly because of increased response to placebo (and
treatment to a lesser extent). Why
a bigger placebo response? Is it types of patients in trials i.e. severity,
gender, age, etc? Is it trial characteristics i.e. number of tx arms, length
of trial, flexible vs. fixed dosing, etc? Other possibilities not examined:
rater bias at baseline, do AD’s really work?
Methods:
- Data
– FDA Summary Basis of Approval trials for 9 AD’s (see Table 1), 4/52
trials excluded because of insufficient data and 1 for different outcome
measure. Only mean HAM-D scores available, no individual scores. Multiple
arms of fixed-dose lumped together (sub-therapeutic doses excluded).
- Groups
of Less Successful and More Successful AD trials defined based on mean
HAM-D difference b/w placebo and tx arms. Mean difference was 3.07 (-2.3
– 9.4) overall. Analysis by
median-split at mean (2 groups of 26 trials) and quartile-split with
analysis of extreme quartiles (2 groups of 13 trials).
- Evaluation
of trial design features and patient characteristics b/w groups. Nine
features found and analyzed: baseline depression severity, trial duration,
flexible versus fixed doses, number of study sites, number of treatment
arms, number of patients in each condition, patient age,
percentage of female patients in the placebo group, and percentage
of female patients in the antidepressant group.
- Statistical
Analysis – t tests for parametric statistics, Mann-Whitney U tests for
non-parametric statistics. Pairwise deletion used for missing data.
Correlational analysis conducted (see Table 4).
Validity:
Is this a meta-analysis/overview or a cohort of AD trials?
- Sensible
question?... Yes for investigators and clinicians (rooting out trial
design bias)
- Relevant
studies included?..... Is the FDA sensible?
- Quality
of the primary studies?… The characteristics of these studies are not
well described
- Are
the assessments reproducible?... Dosing schedule: Neuropsychopharmacology
2003;28:552-7
- Are
the results homogenous?... Cannot completely answer because only means are
reported but unlikely due to some overlap in CI’s and different results
from study to study
Results:
- Validity
of median and quartile splits confirmed by significant different b/w
groups
- Three
features associated with More Successful group: (1) flexible over fixed
dosing, (2) lower percentage of female subjects, (3) higher baseline HAM-D
scores
- Quartile-split
also showed fewer tx arms in Most Successful trial group
- Correlational
analysis confirmed all 4 findings (see Table 4)
Discussion:
- Dosing
schedule – dropout rates and mean dose data not included
- Difficult
to analyze patient characteristic b/c individual patient statistics not
provided.
- Quartile
multiplicity effect increases type 1 error for number of tx arms. Patients
in multiple tx arm trials may have a higher expectation of being in a tx
arm.
- More
patients overall in less successful group. May have been helpful to
calculate weighted means for patient characteristic data.
Ethics:
Is it ethical to design therapy trials with the goal of finding a positive
outcome?
Return
to top of the page
|