heritability for the pathophysiology of emotions in schizophrenia. Aim of this study is to investigate with fMRI in unaffected siblings of schizophrenia patientsif emotional network activity previously associated with schizophrenia could represent a heritable pheno- type. Methods: We enrolled 19 patients with schizophrenia, 18 healthy siblings and 31 normalcontrols.All samples were wellmatched (p > 0.05) for a series ofdemographicalvariables.Schizophrenia patients were on stable treatment with antipsychotic drugs; symptoms were assessed using the Positive and Negative Syndrome Scale (PANSS).All subjects underwent fMRI at 3 T during implicit (gender discrimination) and explicit (approach or avoid evaluation) processing of faces with different expression (angry, fearful,happy, neutral).SPM5 was used for imaging analysis (p < 0.05, after small volume correction). Results: Repeated MeasuresANOVA revealed a main effect of diagnosis in right dorsolateralprefrontal cortex (DLPFC) (Brod- mann area 46; x 56, y 26 z 26; K = 4; Z = 3.43). In particular, DLPFC response was lower in patients than in controls, while it was intermediate in the siblings group. Furthermore,there was an interaction between task and diagnosis on anterior cingulate activity (Brodmann area 32; x 11, y 26, z 34; K = 4; Z = 2.90). In this area,activity was greater during explicitrelative to implicit processing in both siblings and controls, while the opposite pattern was present in schizophrenic patients. Discussion: These data suggest that DLPFC activity during emotion processing in schizophrenia may be heritable. On the other hand, altered activity in anterior cingulate does not seem to be present in siblings or a heritable trait. We conclude that state and trait variables may differentially impact functional responses in different brain areas during emotion processing in schizophrenia. doi: 10.1016/j.schres.2010.02.616 Poster 122 LEFT TEMPORAL DYSFUNCTION,ATTENTION CONTROL DEFICIT AND AUDITORY HALLUCINATION: AN FMRI STUDY Else-Marie Løberg 1,2 , Hugo A. Jørgensen 2,3 , Merethe Nygård 1 , Jan Øystein Berle 2 , Kenneth Hugdahl 1,2 1 Dept.Biol. Med. Psychology, University ofBergen,Bergen,Norway; 2 Div. Psychiatry,Haukeland University Hospital, Bergen,Norway; 3 Dept.Clin.Med.,University of Bergen, Bergen, Norway Background: Previous studies have related auditory hallucinations to dysfunctions of speech sound processing localized to the left temporal lobe. Furthermore,auditory hallucinationshave been suggested to be a failure of inhibiting internally generated speech perceptions.We used an fMRI paradigm with dichotic listening to consonant vowels (CV) -syllables and instructions to focus attention to either the right or left ear syllable in order to model such a dysfunction.We expected that the schizophrenia patientswith frequentauditory hallucinations would show dysfunctional tem- poral lobe activation patterns when listening to dichotic presenta- tions of CV-syllables. In addition,we examined possible effects of attention instructions,which could provide clues to failure of fronto-parietal attention control in hallucinating patients. Methods: Seventeen patientswith schizophrenia were grouped into two sub-groups based on their hallucinations score on the PANSS (hallucinatory behaviouritem) and a validation of the auditory quality of their hallucinations.This yielded a frequent hallucinatory group with scores of 4 or more (n = 4), and a non- frequent hallucinatory group with scores of 3 or less (n = 13). The schizophrenia patients and 16 healthy controls were scanned while listening to dichotic presentationsof CV-syllables.The subjects were scanned with a GE Signa 3.0 T scanner, and the data analyzed with the SPM5 software. The dichotic presentations during the scanning included both trials with and without attention instruc- tions. Results: As compared to the healthy controls, the patient group failed to show activation in the anterior cingulate cortex when instructed to focus attention (p < 0.5, corrected), and particularly in the situation with attention focused on the left ear syllable.In addition,the hallucinating patients failed to activate the speech areas in the upper posterior part of the temporal lobe, and on the left side (p < 0.001, uncorrected) while listening to the CV-syllables. Discussion: The patient group as a whole failed to activate the anterior cingulate cortex, which is a part of a generalized cognitive effort network, when instructed to focus attention. As expected, the frequent hallucinating group also failed to activate the speech areas in the temporallobe, indicating dysfunctional speech perception. Possibly,a combination ofattention controland speech dysfunc- tions may underlie the hallucination experience.The neuronal mechanism mediating this effect could be an inability to inhibit internally generated 'voices' in the form of speech mis-representa- tions. This is further enhanced by dysfunctional focusing of attention on the voices once they are elicited, as part of a dysfunctional fronto-parietal neuronal network. doi: 10.1016/j.schres.2010.02.617 Poster 123 PREFRONTAL ACTIVITY AND COMT GENOTYPE EFFECTS IN SCHIZOPHRENIA Pilar Lopez 1 , Leslie Young 1 , David Garcia 2 , Reyes Garcia de Eulate 2 , Juan Marin 1 , Felipe Ortuño 1 , Jose L.Zubieta 2 1 Department of Psychiatry, University of Navarra, Pamplona, Navarra, Spain; 2 Departmentof Radiology, University ofNavarra,Pamplona, Navarra,Spain Background: Dopamine levels in the prefrontal cortex (PFC) seem to play a crucial role in cognitive function in schizophrenia. The COMT enzyme has a functional polymorphism (val158met): val/val individuals have a higher functioning enzyme leading to lower dopamine levels in PFC and therefore to poorer cognitive perfor- mance.This genetic polymorphism could mediate the relationship between dopamine levels, cognitive functioning and neural activity of PFC. We used globalneuropsychological and specific cognitive assessments and fMRI to study the influence of COMT genotype on cognition and brain function in schizophrenia spectrum disorder patients,relatives and healthy control subjects. Methods: 73 schizophrenia spectrum disorder patients, 54 relatives and 42 healthy controls performed the MATRICS Consensus Cognitive Battery to evaluate several neuropsychological domains. A sample of 19 schizophrenia spectrum disorder patients, 17 relatives and 20 controls performed the Dot version of the expectancy AX continuous performance task (DPX task) to study context processing. We used functional Magnetic Resonance Imaging (fMRI) during the performance of the DPX task. Results: For the MATRICS battery, no group x genotype interaction was observed for any cognitive measure. There was a significant main effect of group for all neuropsychological subtests.Verbal learning (HVLT retention) showed a main effect of genotype (F = 3,28; p = 0,04). For the context processing task (DPX test) a geno- type effect was present behaviorally and in the brain activations. Abstracts 352