Posterior Cerebellar Resting-State Functional Hypoconnectivity: A Neural Marker of Schizophrenia across Different Stages of Treatment Response

Urvakhsh Meherwan Mehta,Dhruva Ithal,Neelabja Roy,Shreshth Shekhar,Ramajayam Govindaraj,Chaitra T. Ramachandraiah,Nicolas R. Bolo,Rose Dawn Bharath,Jagadisha Thirthalli,Ganesan Venkatasubramanian,Bangalore N. Gangadhar,Matcheri S. Keshavan

DOI: https://doi.org/10.1016/j.biopsych.2024.01.027

IF: 12.81

2024-02-09

Biological Psychiatry

Abstract:Background Identifying stable and consistent resting-state functional connectivity (rsFC) patterns across illness trajectories has the potential to be considered fundamental to the pathophysiology of schizophrenia. We aimed to identify consistent rsFC patterns across heterogeneous schizophrenia groups defined based on treatment response. Methods In phase-1, we used a cross-sectional, case-control design to characterize and compare stable independent component networks from rs-fMRI scans between antipsychotic-naïve first-episode schizophrenia participants (N=54) and healthy subjects (N=43); we also examined their associations with symptoms, cognition, and disability. In phase-2, we examined the stability (and replicability) of our phase-1 results in four groups (N=105) representing a cross-sequential gradation of schizophrenia based on treatment response: risperidone responders, clozapine responders, clozapine non-responders, and clozapine non-responders following electroconvulsive therapy. Hypothesis-free whole-brain within- and between-network connectivity were examined. Results Phase-1 identified posterior and anterior cerebellar hypoconnectivity, and limbic hyperconnectivity in schizophrenia at family-wise error rate corrected cluster significance threshold of P < 0.01. These network aberrations had unique associations with positive symptoms, cognition, and disability. Phase-2 replicated the phase-1 results while comparing each of the four schizophrenia groups against the healthy participants. The two longitudinal sub-datasets did not demonstrate a significant change in these network aberrations following risperidone or electroconvulsive therapy. Posterior cerebellar hypoconnectivity (with thalamus and cingulate) emerged as the most consistent finding – replicated across different stages of treatment response (Cohen's d range -0.95 to -1.44), reproduced using different pre-processing techniques, and not confounded by educational attainment. Conclusions Posterior cerebellar-thalamo-cingulate hypoconnectivity is a consistent and stable state-independent neural marker of schizophrenia.

neurosciences,psychiatry