A1 Refereed original research article in a scientific journal
From Snapshots to Stable Outcomes: Resting-State Functional Magnetic Resonance Imaging–Based Prognosis of Functioning in Patients With Psychosis Risk or Recent-Onset Depression
Authors: Buciuman, Madalina-Octavia; Haas, Shalaila S.; Antonucci, Linda A.; Sarisik, Elif; Khuntia, Adyasha; Lichtenstein, Theresa; Rosen, Marlene; Kambeitz, Joseph; Pantelis, Christos; Lencer, Rebekka; Bertolino, Alessandro; Brambilla, Paolo; Upthegrove, Rachel; Wood, Stephen J.; Falkai, Peter; Riecher-Rössler, Anita; Ruhrmann, Stephan; Schultze-Lutter, Frauke; Meisenzahl, Eva; Hietala, Jarmo; Salokangas, Raimo K.R.; Borgwardt, Stefan; Dwyer, Dominic B.; Kambeitz-Ilankovic, Lana; Koutsouleris, Nikolaos
Publisher: Elsevier BV
Publication year: 2025
Journal:Biological Psychiatry
ISSN: 0006-3223
eISSN: 1873-2402
DOI: https://doi.org/10.1016/j.biopsych.2025.07.003
Web address : https://doi.org/10.1016/j.biopsych.2025.07.003
Self-archived copy’s web address: https://research.utu.fi/converis/portal/detail/Publication/500491337
Background: Early recovery of functioning is critical for favorable outcomes in psychotic and affective disorders. Transdiagnostic brain activity patterns may capture pathways for poor outcomes before clinical manifestation, thereby supporting timely prevention and intervention.
Methods: Using machine learning, we evaluated the transdiagnostic prognostic value of resting-state functional magnetic resonance imaging fractional amplitude of low-frequency fluctuations (fALFF) (slow-5 and slow-4 sub-bands) for functional outcomes in patients at clinical high risk for psychosis (n = 217) or with recent-onset depression (n = 198) from the multisite PRONIA (Prognostic Tools for Early Psychosis Management) study. Leave-site-out cross-validation assessed the geographic generalizability of models across disability and symptom domains, with outcomes defined as snapshots at 9- or 18-month follow-up or across both time points. We examined diagnosis-specific performance, generalization to recent-onset psychosis (n = 140), and negative symptoms and the added value of fALFF over clinical prognostication.
Results: Transdiagnostic models predicting stable good functioning across follow-ups showed up to 10% higher balanced accuracy (BAC) than snapshot models. Decreased slow-5 fALFFs in the default mode network, executive control network (ECN), and dorsal attentional network (DAN) and increased fALFF in the salience network, ECN, and DAN predicted impairment with BAC = 67% (sensitivity = 65%, specificity = 70%, p < .001). This model generalized to recent-onset psychosis (BAC = 62%, sensitivity = 64%, specificity = 59%, p < .001) and predicted (BAC = 65%, sensitivity = 66%, specificity = 65%, p < .001) and was mediated by negative symptoms. Slow-5-based models improved prognostic accuracy over expert ratings in disability (BACraters = 66%, BACraters+slow-5 = 75%, W = 1680, p < .001) and symptom (BACraters = 61%, BACraters+slow-5 = 71%, W = 1444, p < .001) domains.
Conclusions: We highlighted the prognostic value of fALFF for functional impairment in psychosis risk and early depression. Leveraging trajectorial information, we identified candidate imaging biomarkers to improve prognostication, thereby supporting personalized prevention and recovery strategies.
Downloadable publication This is an electronic reprint of the original article. |  |
Funding information in the publication:
PRONIA is a Collaboration Project funded by the European Union 7th Framework Programme under (Grant Agreement No. 602152). NK is supported through grants from the National Institutes of Health (Grant No. U01MH124639-01; ProNET), the Wellcome Trust, the Global Innovation Fund (CARE project), the German Federal Ministry of Education and Research (Grant Nos. 01ZX1904E and 01ZX2204A; COMMITMENT) and BEST projects, as well as European Partnership for Personalized Medicine (IMPLEMENT project). AK is funded through the COMMITMENT project.
AB reports speaker fees from Otsuka, Lundbeck, Angelini, and Rovi outside of the submitted work. JH reports personal fees from Orion Ltd., personal fees from Lundbeck, personal fees from Otsuka, and other from Takeda during the conduct of the study. NK, SR, and AR-R report grants from the European Union over the duration of the study. EM and NK hold patent US20160192889A1 (Adaptive pattern recognition for psychosis risk modeling). CP reports grants from the Australian National Health and Medical Research Council (NHMRC) during the study and personal fees from Lundbeck, Australia Pty. Ltd. outside the submitted work. RU reports speaker fees from Sunovion, Otsuka, and Vitaris outside the submitted work as well as unpaid officership with the British Association for Pharmacology (Honorary General Secretary 2021–2024). She serves as Deputy Editor for The British Journal of Psychiatry. RL reports honoraria for lectures or advisory activities from Boehringer Ingelheim, Janssen, Otsuka, and Recordati outside the submitted work. PF reports he has received research support/honoraria for lectures or advisory activities from Boehringer Ingelheim, Janssen, Lundbeck, Otsuka, Recordati, and Richter outside the submitted work. CP was supported by an Australian NHMRC L3 Investigator grant (Grant No. 1196508) outside the submitted work. LK-I has served on the advisory board of Boehringer Ingelheim outside the submitted work. NK reports speaker fees from Otsuka, Roche, and Angelini outside of the submitted work. All other authors report no biomedical financial interests or potential conflicts of interest.
