Key questions for psychiatric neuroimaging

1
Key questions for psychiatric neuroimaging

• Ed Bullmore

2
Key questions for psychiatry

• Phenomenology, nosology
• what is/are schizophrenia(s) as distinct from
  normality and manic-depressive and other
  psychoses?
• are there disease entities or dimensions of
  psychopathology?
• how does disorder relate to normal variability in
  the population?
• Aetiology, pathogenesis
• what are the causes of psychiatric disorder?
• what is the process by which remote causes, e.g.
  genes, express themselves as disorder?
• Therapeutics
• how do effective drug treatments work?
• how can we identify new treatments both
  clinically and industrially?

3
Key issues for psychiatric fMRI

• what is the functional anatomy of disorders,
  symptoms and subsyndromes?
• can we image functional effects of single genes
  or heritability in general?
• can we image normal neurodevelopmental processes,
  normal adult brain variability, and understand
  pathogenesis in that context?
• can we image functional effects of psychotropic
  drug action? do these predict clinical or
  cognitive benefits in individual patients? can we
  use fMRI to accelerate psychopharmaceutical drug
  development?

4
Case-Control Studies in fMRIConfounded
hypofrontality
Honey et al (2000) 30 cases
(schizophrenia)/ 30 controls Verbal working
memory task (blocked periodic design)
5
Possible Mechanisms for Inconstant Hypofrontality
in Schizophrenia

• Variable extracerebral confounds
• e.g. head movement (Bullmore et al (1999) HBM)
• Variable task type or task difficulty
• easy semantic categorisation does not elicit
  hypofrontality in the same patients scanned in
  same session (Curtis et al (1998) Schizophrenia
  Research Fletcher et al (1999) Archives of
  General Psychiatry)
• Variable symptom state
• Variable antipsychotic drug exposure
• Variable age (between study variation)
• Abnormal frontal connectivity

6
Positive Symptoms Attenuate Activation in Left
Fronto-Temporal Regions
Honey et al (2000) N30 Positive syndrome scores
based on factor analysis of N100
7
Causal Network Modelling by Path Analysis
Fitting a Wernicke-Lichtheim-like diagram to
fMRI data
Bullmore et al (2000) NeuroImage Semantic
categorisation and subvocal rehearsal task
induces correlated activation of 5 main left
brain regions Causal interactions between these
regions are summarised in a path diagram of and
quantified by partial regression coefficients
8
Functional MRI of single gene effects

• There have been very few (Egan et al, PNAS, 2001)
  fMRI studies of single gene effects
• 7 individuals with heterozygous PAX6 mutation
  were compared to 7 normal subjects on verbal
  fluency and sentence completion tasks
• PAX6 is a highly conserved homeobox
  neurodevelopmental control gene

9
Encompassing normal variabilityRegionally
variable genetic effects on brain structure
Path analysis can be used to estimate the
proportion of variance in brain structure
attributable to genetic effects (heritability) in
the context of classical twin designs. Wright et
al (2001), under review.
10
Encompassing normal variabilityStructural
associations with normal cognitive function
Spatial working memory errors negatively
correlated with fronto-parietal grey matter
anatomy in normal subjects
11
Imaging Brain Development with fMRI(K. Rubia et
al (2000) Neuroscience Biobehavioral Reviews
24, 13-19)

• Functional MRI is an unprecedented opportunity to
  image human brain development
• 17 normal controls, 9 adolescents and 8 young
  adults
• Two overt blocked periodic experiments
• motor timing
• motor response inhibition

12
Dysmaturation as a Cause of Hypofrontality?(K.
Rubia et al (2000) Neuroscience and
Biobehavioural Reviews 24, 13-19)
Left frontal opercular (solid) and middle frontal
(dashed)
Right inferior frontal (solid) and anterior
cingulate (dashed)
13
Antipsychotic Drug Effects fMRI(G. Honey et al
(1999) Proceedings National Academy of Science
96, 13432-13437)

• 10 patients with schizophrenia scanned twice on
  typical antipsychotic drugs
• overt verbal working memory task
• 6 week interval between scans
• 10 patients with schizophrenia scanned at
  baseline and after substitution of risperidone
• Risperidone treatment associated with
  significantly enhanced frontal cortical
  activation
• broadly compatible with previous functional
  imaging studies of dopaminergic drugs (Grasby et
  al 1992 Reed et al 1999)
• clinical implications?

14
Substitution of Risperidone Enhances Frontal
Activation by Working Memory(Honey et al (1999)
PNAS 96, 13432-13437)
15
Conclusions

• For psychiatry, the advent of safe functional
  brain imaging is an historically unprecedented
  opportunity to define neural substrates of
  disorder - this process may entail a revolution
  in definition of disorder
• The conjunction of fMRI and genomics is a major
  research opportunity for understanding causation
  of disorders
• Emergence of psychiatric disorders needs to be
  characterised in context of normal, variable,
  neurodevelopmental processes - which can be
  directly visualised by fMRI
• Pharmacological fMRI has promise scientifically,
  clinically and commercially.