Distinct Features of Cerebral Blood Flow and Spontaneous Neural Activity as Integrated Predictors of Early Response to Antidepressants.

Zhenghua Hou,Tong Li,Xiaofu He,Yuqun Zhang,Huanxin Chen,Wenhao Jiang,Yingying Yin,Yonggui Yuan
DOI: https://doi.org/10.3389/fpsyt.2021.788398
IF: 4.7
2022-01-01
Frontiers in Psychiatry
Abstract:AIMS:The purpose of this study is to explore whether pre-treatment features of brain function can discriminate non-responders to antidepressant medication in the early phase. METHODS:Forty-four treatment-responsive depressed (RD) patients, 36 non-responsive depressed (NRD) patients, and 42 healthy controls (HCs) were recruited. Regional cerebral blood flow (CBF) and amplitude of low-frequency fluctuation (ALFF) values were calculated for all subjects. Correlation analyses were used to explore the relationship between symptom improvement and CBF/ALFF. Receiver operating characteristics (ROC) and the 10-fold cross-validation support vector machine (SVM) classifier were applied for the discrimination of treatment response. RESULTS:Compared with the HCs, the RD and NRD groups exhibited lower CBF and ALFF in the right posterior lobe of the cerebellum. Compared with the NRD group, the RD group showed distinct CBF patterns in the left frontal striatal regions and right frontal cerebellar regions, as well as distinct ALFF features in the left frontoparietal striatum and right frontotemporal striatal cerebellar regions. The ROC and SVM classifier revealed the optimal power to distinguish the RD and NRD groups based on the combined measures (i.e., CBF and ALFF). CONCLUSION:Distinct features of CBF and ALFF in the frontal striatal network may serve as promising neuroimaging predictors for identifying patients with blunted responsiveness, which may facilitate personalized antidepressant treatment.
What problem does this paper attempt to address?