Transcript Developing translational biomarkers for psychotic disorders

Validity with Pragmatism
Specific Issues in experimental design and implementation
II-A
– The “schizophrenias” (Bleuler) are
neurobiological disorders diagnosed solely on
the basis of behavioral changes
– Schizophrenia is characterized by a cluster of
cognitive, affective, and motor phenotypes
• What are the neurocognitive/neurobehavioral
processes underlying these phenotypes?
Translational Cognitive
Neuroscience in Psychiatry
Cognitive Mechanism
Maladaptive Behavior
Neurobehavioral assays!
Neural Circuits
Molecular
Mechanisms
Courtesy of Peter Balsam
– Multiple pathogenic pathways, variably
combined, converge onto common
neurobehavioral processes to produce
these phenotypes
• How do we identify these pathogenic
mechanisms and determine their effect on
cognition?
Translational Neuroscience
(How model potential pathogenic mechanisms)
Behavioral Abnormality
Risk factors
Disease Model
Molecular
Targets
Circuits
Molecular
Mechanisms
Risk Genes
Adapted from Peter Balsam
At a pragmatic level, how do we go
about this?
What issues need to be addressed as we
‘translate’ a construct across species or
assay this construct in a disease model?
Task #1: Operationalize
(and use tasks that are true to our operational definition)
Defining and measuring the construct in your species
– How is the construct operationally defined for your
species? How does that differ from how it is defined
in human studies?
– What properties must a task possess in order to
isolate, manipulate and measure the construct of
interest?
– What behavioral output is used to measure the
operation of the construct?
• Change in Motor Output or Neuronal Output Y as a function of
Task Manipulations X, Z
Parametrics and Reliability
• Can we parametrically vary task demands?
(or is the nature of the construct such that
behavior can report an all-or-none effect)
• Can the construct be manipulated/assessed
in the context of steady-state performance
on aspects of the task not related to our
construct of interest?
• Do we have enough trials to reliably assess
the effect of drugs on performance?
Outcome Measures
• What is the temporal and spatial “structure” of the
behavioral expression system?
• How reliable/stable/spontaneous is the expression of the
behavior (competing behaviors)?
• How sensitive is the behavior to changes in the construct?
• What other cognitive, affective or motor processes impact
the behavior? Are we affecting those processes with our
manipulations?
•What are the neural substrates of the behavioral response ?
Sensitivity to disease mechanisms
and common drug treatments
• How is the neurobehavioral construct
affected by
• hypothesized pathogenic mechanisms of
the disease?
• Drugs commonly prescribed to the
patients?
Assessed by measuring the construct in
disease and pharmacological models
Sensitivity to disease mechanisms
and common drug treatments (con’t)
• How does the hypothesized disease mechanism
affect the behavioral expression system?
• How do candidate drugs and - commonly-prescribed
drugs – impact the behavioral expression system?
• How is the behavior reinforced, with positive or
negative reinforcement? Is the construct differentially
sensitive to positive versus negative reinforcement?
Does the disease mechanism differentially impact
appetitive versus avoidant behaviors?
Task #2: Find Homologous Neural Substrates
Neural Circuit Homology – conservation across species
of the neural circuits and neurobiological processes that
underlie the cognitive construct
• Which neural circuits are hypothesized to mediate this construct
in humans?
• Do these circuits have homologues in the experimental animal
species ?
Homologous Neural Circuits (con’t)
Which neural substrates are we studying?
Hypothesis 1: Something’s missing – how do we identify it?
Neural circuit recruitment versus necessity for normal function
Which circuits are recruited by task demands that manipulate
the construct?
Which of these are necessary for the operation of the
construct?
Hypothesis 2: Something’s in the way - Active mechanisms of
disruption – neural circuit intrusion
Convergent validity in determining neural circuit homology
Complementary methods of determining homology of neural
circuits
• neurodevelopment
• anatomical hodology (connectivity)
Efficiency and Cost-effectiveness
Can I train sufficient number of animals and
assess the effects of candidate drugs using
a task with construct validity before I die or
at least before I go broke?
• Training time
• Throughput – how many animals can I
test at once?
• Recycling – can I reuse a trained animal?
Neurobehavioral
Construct Validity
Etiologic (a.k.a.
content) Validity
Predictive Validity
Reliable signal of efficacy in patients
The criteria must be met with high reliability and reasonable throughput
QUESTIONS FOR ALL PARTICIPANTS
1. As an animal or human neurocognition researcher who wants to help develop
tasks to measure constructs affected in schizophrenia…
• What additional information do you need from clinical and neurocognitive
studies of schizophrenia patients?
•How would you suggest they change their task designs to get information that
you need to understand what is going on in the patients?
2. As an animal ‘disease modeler’, what information do you need from animal
neurocognition researchers? From studies of schizophrenia patients (incl. risk
factor, neurocognitive, and drug treatment studies)?
3. As a clinical researcher in schizophrenia, what do you need from animal
modelers? What do need less of? More of? (understanding of relationships
between dependent variables and neural and cognitive operations)
4. Who else should be at this meeting – please give us names/locations and we’ll
try hard to make sure they are at the next meeting.
Operationalization :
Task Structure
Accuracy
What properties must a task possess in
order to isolate, manipulate and
measure the construct of interest?
5
4
3
2
1
0
Easy
Con
Hard
Sick
Treatment Targets
DRUGS
Protein function
Pathophysiology
Cognitive processes
Neural Circuit Function
Maladaptive behavior
Experience
Environment
concurrent with
illness
Cognitive and Psychosocial
Therapy
Summary and [of] Challenges
Challenges for Pro-cognitive
Treatments for Schizophrenia
• Understanding of the neuroscience behind cognitive
changes in schizophrenia is partial
• There is no unitary hypothesis for the cause(s) of cognitive deficits
• The diagnostic syndrome may reflect many different etiologies
• No consensus on the underlying neurobiology
• Cognition is not a unitary concept
• 5 – 12 cognitive domains are affected, each with different substrates
• Is it realistic to seek treatments that will improve cognition globally?
• What would be the most relevant domains that need to improve?
• No reliable and valid biomarkers for cognitive dysfunction
have been validated as yet
• No validated drug targets exist for improving cognition that
can be used as positive controls, although many suspected
targets exist
Adapted from Thomas Steckler
How do we do a better job of identifying targets for new drugs, and combining
drug and cognitive/behavior therapies to achieve a sustained improvement in
the function of a neural circuit?
A better treatment strategy
• CNTRICS I
– Identify cognitive processes that are disrupted in patients diagnosed with
schizophrenia
– Use data from research in patients to propose candidate neural substrates
for these processes
• CNTRICS II – ANIMAL MODELS
– Identify homologous neurocognitive processes in animal models
– Use animal models to
•
•
•
Increase understanding of neural mechanisms underlying the construct
Characterize candidate pathophysiologies that can disrupt the construct (prioritizing on the
basis of plausibility in schizophrenia)
Predict the effects of treatment on those neural mechanisms
The Cascade of Pathogenesis
protein
synthesis
and
regulation
susceptibility
factors
neurophysiology
Neurodevelopment
Embryonic
Postnatal
Late Juvenile
environment
Experience
Maladaptive behavior
Periadolescent
Morbidity in adulthood
Development