1. FIDMAG Research Foundation Hermanas Hospitalarias
Neuroimaging data help to clarify the nosological status of schizoaffective disorder (SAD)?
Mercè Madre Rull
FIDMAG Research Foundation Hermanas Hospitalarias
Barcelona, Spain

2. Overview Previous Neuroimaging data of SAD Functional Imaging SAD
VBM
fMRI FA
Previous Neuroimaging data of SAD
Functional Imaging SAD
Structural Imaging SAD
Discussion

3. Why neuroimaging in SAD?
Clinical studies comparing SAD with SZ and BP are not conclusive.
Differences in neuroimaging between SAD, SZ and BP might help to clarify their boundaries.
Useful for clarifying SAD nosology.
Findings could add valuable information to the diagnostic classification and might have consequences for treatment trials.
SAD might constitute a middle point of a continuum between SZ and BP disorder.
SAD is not an atypical form of SZ or BP, nor an independent mental disorder Cheniaux et al, Review
- SAD shares features of both SZ and BP.
SAD resembles more SZ than BP in most of the demographic and clinical categories Pagel et al, Meta-analysis

4. fMRI studies so far in SAD
Authors
Population
Method
Findings
Gruber et al 2006
SA = 14, SZ = 14, HC =14
fMRI
SA showed normal behavioral performance on tasks involving the articulatory loop component of working memory, but SZ did not.
Kalayci et al. 2012
SA =15, SZ= 15, BP = 15, HC = 15
H-MRS
Similar neurochemical changes in DLPFC and executive functions between groups, but more similarities between SA and BP.
Ongur et al. 2010
SA=7+SZ=7, BP = 14, HC = 15
(ICA compared the wider DMN network at rest)
In a sub-analysis, both SA and SZ showed similar reduction in resting state connectivity in the medial frontal cortex, which was greater than in BP.
Madre et al. 2013
SA =32, HC = 32
(working memory task )
SA showed frontal hypo-activations and a failure of de-activation in the medial frontal cortex (DMN dysfunction) compared to HC.
Madre et al. 2014
SA = 22, HC = 22
Similar DMN dysfunction in acute and remitted SA, as a trait-like feature of SAD.
Yuhui Du et al, 2014
SA = 31, SZ= 20, BP = 20, HC = 20
(ICA)
12 discriminating regions between diagnostic groups. SAD may be an independent disorder, although its depressive subtype
shares more similarity with SZ.

5. Study 1
Brain functional abnormality in schizoaffective disorder: an fMRI study
M. Madre, E. Pomarol-Clotet, P. McKenna, J. Radua, J. Ortiz-Gil, F. Panicali,
J. M. Goikolea, E. Vieta, S. Sarró, R. Salvador and B. L. Amann.
Psychological Medicine 2013 Jan;43 (1):
Study 2
Trait or state? A longitudinal neuropsychological evaluation and fMRI study in schizoaffective disorder
M Madre, J Radua, R Landin-Romero, S Alonso-Lana, R Salvador, F Panicali,
E Pomarol-Clotet, BL. Amann.
Schizophrenia Research 2014  Nov; 159(2-3):

6. Study 1
Study 2
32 acute Schizoaffective patients 16 schizomanic episode 16 schizodepressive episode Vs. 32 healthy subjects
22 Schizoaffective patients in clinical remission Vs. 22 healthy subjects
Follow-up
> 2 months of clinical remission
Inclusion criteria
Diagnosis of SAD bipolar type (DSM-IV, RDC)
Age years.
IQ ≥ 70
Right handed  
No history of brain trauma or neurological disease.
No Alcohol/substance abuse within 12 months prior to participation

7. MRI acquisitions in a 1.5 Tesla GE Signa scanner
N-back task
Working memory task: block design
Two levels of memory load: 1-back and 2-back *
Baseline A P F K
1-back
MRI acquisitions in a 1.5 Tesla GE Signa scanner A P F K U
2-back
Exclusion criteria: movement >3.0 mm, poor task performance
Results showed: 2-back vs. Baseline contrast

8. Activations (blue) De-activations (pink)
Study 1
Brain functional abnormality in schizo-affective disorder:
an fMRI study.
Whole brain analysis of healthy subjects and acute schizoaffective patients
Activations (blue) De-activations (pink)
Healthy subjects
(n=32)
Acute Schizoaffective patients (n=32) .
Madre et al, 2013

9. Study 1 Left precentral cortex (DLPFC) Reduced activation in
ROIs derived from the differences between the whole brain analyses of acute schizoaffective patients vs. H.S
Mean BOLD response between manic, depressive and H.S.
Left precentral cortex (DLPFC)
Left middle temporal cortex
Bilateral parietal cortex
Reduced activation in
acute schizoaffective
Failure of deactivation in
acute schizoaffective
Anterior cingulate cortex
Madre et al, 2013

10. The brain’s Default Mode Network (DMN)
DMN anterior node
Medial prefrontal cortex
Anterior cingulate cortex
DMN posterior node
Posterior cingulate cortex Precuneus
Buckner et al, 2008

11. Mental illnesses and alterations of the DMN
Schizophrenia:
Failure of de-activation in the MPFC, using different tasks (Milanovic et al. 2011, Pomarol-Clotet et al. 2008, Whitfield-Gabrieli et al. 2009).
Some studies have additionally found failure to de-activate in the posterior cingulate cortex (Salgado-Pineda et al. 2011, Schneider et al. 2011).
Bipolar Disorder:
Failure of de-activation in the MPFC during mania (Pomarol-Clotet et al. 2011) depression (Fernandez-Corcuera et al. 2012) and euthymia (Pomarol-Clotet et al. 2014), using working memory tasks.
Failure of de-activation in the posterior cingulate cortex in euthymia, using a verbal fluency task (Allin et al. 2010)
DMN dysfunction as a trait-like feature.
Broyd et al, Review

12. No activation changes Study 2
Trait or state? A longitudinal neuropsychological evaluation and fMRI study in schizoaffective disorder
Acute episode vs. clinical remission in schizomanic and schizodepressive patients
No activation changes
Depressive patients vs. remission
(n=12)
Mean BOLD response in the ROIs during the acute phase and clinical remission in manic patients.
Whole-brain paired t-test.
Reversible frontal hypo-activation
Manic patients vs. remission
(n=12)
Madre et al, 2014

13. Study 2
Trait or state? A longitudinal neuropsychological evaluation and fMRI study in schizoaffective disorder
Schizoaffective patients in clinical remission (n=22)
Healthy subjects (n=22)
versus
Failure of deactivation in
remitted schizoaffective
Reversible frontal hypo-activation in acute manic SA that normalized in clinical remission, as a state-like feature.
DMN dysfunction is a trait-like feature of SAD.
Madre et al, 2014

14. HC = 20 SAD manic = 20 SZ = 20 SAD depressive = 13 BP = 20
Resting-state fMRI using ICA: 12 discriminating regions
(a) SupraMarginal L
(b) Frontal-Medl-Orb R
(c) Cingulatel-Ant L
(d) Cingulatel-Post L
(e) Frontall-Med_Ord R
(f) Parietall-Inf L
(g) Insula R
(h) Parietall-Inf L
(i) Precuneus R
(j) Cerebelluml-Crus2 L
(k) Suppl-Motor Area L
HC = 20
SZ = 20
BP = 20
SAD manic = 20
SAD depressive = 13 BP HC SZ
SADM
SADD
Yuhui Du et al, 2014

15. Structural studies so far in SAD
Authors
Population
Method
Findings
Rieder et al. 1983
SA=15, SZ=28, BD=19 CT
No differences in lateral ventricular volume or in index of cortical atrophy were found.
Getz et al. 2002
SA=12, BD=12, HC=12
vMRI
SA and BD had similar decrease in whole-brain volume compared to HC.
Smith et al. 2011
SA=15,SZ=47, HC=47
(thalamus)
SA and SZ showed similar volume reduction compared to HC.
Radonic et al. 2011
SA =15, SZ=15, BD=15,
HC=15
(hippocampus)
Reduction in hippocampal volume in SZ and SA compared to both BD and HC.
Arnold et al. 2014
SA=70, SZ=71,
Psy-BD=86, HC=145
SZ and SA had lower hippocampal volumes compared with psychotic BD and HC.
Mathew, Gardin et al. 2014
SA=142, SZ=219,
Psy-BD=188, HC=337
Comparable hippocampal volumes reductions within all the disorders.
Padmanabhan et al 2014
SA=117, SZ=181, Psy-BD=157, HC=352
Cortical Thickness
No analysis of SA group. No differences among diagnostic groups. CT associated with PANSS positive subscale.
Ivleva et al. 2013
SA=90, SZ=146,
Psy-BD=115, HC=200
VBM-MRI
SA and SZ showed gray matter reduction compared to psychotic-BD.
Landin et al (under review)
SA=45, HC = 45
Cortical Thickness- DTI
SA patients showed widespread areas of volume reduction compared to HC.
Amann et al. 2015
SA=45, SZ=45, BD=45,
HC=45
SA and SZ showed widespread and similar areas of volume reduction compared to BD and HC.
CT: Structural computed tomography, vMRI: volumetric magnetic resonance imaging, VBM: Voxel Based Morphometry, DTI: Diffusion Tensor Imaging

16. Gray matter volume differences within the DSM-IV Diagnosis among patients groups and healthy comparison subjects
SZ=146 vs. HC=200
SAD=90 vs. HC=200
Psychotic-BP=115 vs. HC=200
Ivleva et al. 2013

17. Study of gray matter volume reduction in SZ, SAD and BP compared to HC
45 SZ vs. 45 HC
45 SAD vs.45 HC
45 BP vs. 45 HC
Amann et al. Acta Psyc Scand, 2015

18. VBM comparison of the combined patient group with the healthy controls
Ventromedial Prefrontal
cortex
Bilateral frontal superior
Bilateral
temporo-insular-parietal
Cerebellum
Boxplots of the mean values of gray matter volume reduction in SAD (n=45), SZ (n=45) and BP (n=45).
Amann et al. Acta Psyc Scand, 2015

19. Surface-based brain morphometry and diffusion tensor imaging in schizoaffective disorder:
A multimodal approach
45 Schizoaffective vs. 45 Healthy subjects
Surface-based morphometry
Cortical Thickness
(CT)
Cortical Volume
(CV)
Surface Area
(SA)
Landín et al (under review)

20. Surface-based brain morphometry and diffusion tensor imaging in schizoaffective disorder:
A multimodal approach
45 Schizoaffective vs. 45 Healthy subjects
Diffusion Tensor Imaging (DTI)
White Mater
Corpus callosum
Bilateral anterior limb of internal capsule
Left superior longitudinal fasciculus FA
White Mater
Corpus callosum
Corona radiata
Grey Matter:
Frontal cortex
Temporal lobes
Left parahippocampal gyrus
right rectus
Left fusiform
Bilateral thalamic nuclei MD
Landín et al (under review)

21. Surface-based brain morphometry and diffusion tensor imaging in schizoaffective disorder:
A multimodal approach
45 Schizoaffective vs. 45 Healthy subjects
Multimodal analysis CV FA MD
White matter Grey matter
Surface-based morphometry
DTI
Landín et al (under review)

22. Functional MRI studies in SAD
Conclusions:
Functional MRI studies in SAD
Few fMRI studies in SAD
Acute SA patients show reduced prefrontal activation (DLPFC) implicated in working memory tasks, similar to the ‘hypofrontality’ described in SZ and in some studies of BP.
Reversible frontal hypo-activation in acute manic SA patients when compared to clinical remission, as a state-like feature of SAD.
DMN dysfunction is a trait-like feature of SAD. Described in both SZ and BP.
SAD is an independent disorder, although depressive type shares high similarities with SZ in functional network patterns.
(Madre et al, 2013; Madre et al, 2014; Yuhui Du et al, 2014 )

23. Structural MRI studies in SAD
Conclusions:
Structural MRI studies in SAD
Prior sMRI studies in SAD with important limitations.
Grey matter:
Widespread pattern of CV reduction (CT reduction)
Increased MD in both cortical and subcortical regions
White matter:
Widespread FA reduction
Increased MD
Two VBM studies comparing SAD with SZ and BP with similar results.
Structural changes in SAD are more similar to SZ than to BP.
(Landin et al, 2015; Ivleva et al, 2013; Amann et al, 2015)

24. What implications do the findings have for the nosological status of schizoaffective disorder?
SAD shares some fMRI similarities with both SZ and BP, supporting the continuum hypothesis
This is in line with genetic (Cardno and Owen, 2014) and neurocognitive data (e.g. Amann et al, 2013).
DMN dysfunction seems a common trait in all three pathologies.
Both fMRI and sMRI data support the hypothesis that SAD might be closer to SZ than to BP, suggesting it could be a subtype of SZ or an intermediate form of illness but skewed towards schizophrenia.
This in line with a recent clinical review (Pagel et al, 2013) and older literature which considers SAD as a subtype of SZ (Lehman, 1975; Welner et al, 1977).
Future studies are needed in SAD and comparing the three pathologies, using new techniques to confirm these hypotheses.

25. Acknowlegdement FIDMAG Research Foundation CIBERSAM
Hermanas Hospitalarias
CIBERSAM
Instituto Carlos III
Stabilization Contract grant (CES 12/024)
to B.L. Amann
CP06/00359
PI071278
PI10/02622