Poster #35 STUDYING EPISODIC MEMORY CEREBRAL NETWORK: A NOVEL EVENT-RELATED TASK FOR FMRI

Background: Converging findings from lesion and electrophysiological studies indicate that structures in the mesial temporal lobe, including the parahippocampus (PH) and the hippocampus (H), together with the prefrontal cortex (particularly the inferior frontal gyrus, IFG) are crucial for encoding and recall of episodic memory in animals and in humans. Deficits in episodic memory, especially in the encoding domain, are important features of schizophrenia. In patients, these deficits seem to be stable over time, independent of psychotic symptoms, less responsive to current pharmacological treatment and associated with impaired quality of life and pooroutcome. Aimof this study was to evaluate activity inH andPH aswell as in prefrontal cortex during encoding and retrieval of recognitionmemory by using a novel event-related paradigm. Methods: 23 healthy subjects (M/F = 11/12; mean age, years ± SD: 28±7.7) underwent 3T fMRI. The experimental paradigm consisted of encoding and retrieval of verbal stimuli. During encoding, 50 words were presentedat a rate of 2 seconds (s)with a variable inter-stimulus interval (mean ISI = 8 s).Participants were instructed to express a size judgment deciding if each word indicated a bigger or a smaller object than themselves.35 crosshairs were randomly intermixed with words presentation and served as a baseline against which the hemodynamic response was estimated.During retrieval, subjects viewed 50 words from the previous session plus 50 novel foils, one at a time, in a random order; again, crosshairs were randomly intermixed with words. For each item, participants were asked to express an old/newrecognition judgment. The task had an overall duration of 13.8 minutes.During scanning, all subjects responded by button presses with their right hand, allowing for determination of behavioral accuracy (% correct responses) and reaction time (msec). fMRI data were preprocessed and analyzed using Statistical Parametrical Mapping (SPM8, http://www.fil.ion.ucl.ac.uk). Random effects models, witha statistical threshold of p<0.05 Family-Wise-Error (FWE) correctedand a minimum cluster size (k=3), were used for statistical analysis. Finally, BOLD responses were extracted from significant clusters using MarsBar (http://marsbar.sourceforge.net/) to explore, outside SPM8, significant correlation with behavioral performance. Results: Encoding behavioral data indicated an accuracy (mean %± SD) of 95.03%±12.7. Retrieval behavioral data showed an accuracy for hits old of 86.95% ± 12.7. Imagingdata demonstrated bilateral activation of H (x=34, y=-33, z=-1, k=118, pFWE corr = 0.007; x=-22, y=-33, z=-1, k=132, pFWE corr= 0.006) and of PH (x=20, y=-7,z=-28, k=143,pFWE corr = 0.005; x=-26, y=-3, z=-35, k=122, pFWE corr = 0.007) during encoding, and a substantial activation of left H (x=-18, y=-7,z=-12 k=35, pFWE corr= 0.047) at retrieval. Remarkable bilateral activity resulted in the IFG at encoding (x=34, y=23, z=-5, k=448, pFWE corr = 0.0001; x=-37, y=23, z=-1, k=629, pFWE corr = 0.0001)and at retrieval (x=38, y=23, z=-1, k=370, pFWE corr = 0.002; x=-33, y=27, z=-1, k=509, pFWE corr = 0.0001). A negative correlation, at trend level, was found between accuracy at retrieval and BOLD signal in right IFG (p=0.063, Spearman R=0.393). Discussion: Findings from our study demonstrate bilateral activation of the H-PH complex during memory encoding, and left hippocampal activity during retrieval of verbal stimuli. Furthermore, they support the role of the prefrontal cortex in episodic memory. This novel fMRI paradigm will allow to investigate H-PH and prefrontal cortex activity in patients with schizophrenia, in whom episodic memory deficits represent a phenotypic vulnerability marker of the disorder.