Predicting Trauma-Focused Therapy Outcome From Resting-State Functional Magnetic Resonance Imaging in Veterans With Posttraumatic Stress Disorder

Background Trauma-focused psychotherapy is the first-line treatment for posttraumatic stress disorder (PTSD) but 30-50% of patients do not benefit sufficiently. Here, we tested whether resting-state functional magnetic imaging (rs-fMRI) can predict treatment response for individual patients. Methods 44 male veterans with PTSD underwent baseline rs-fMRI scanning followed by trauma-focused therapy (EMDR or TF-CBT). Resting-state networks (RSN) were obtained using independent component analysis with 70 components on the basis of 28 trauma-exposed healthy controls, matched for age and gender. Dual regression was used to obtain subject-specific RSNs for the PTSD patients. All RSNs were individually included in a machine learning classification analysis using Gaussian process classifiers. Classifier performance was assessed using 10 times repeated 10-fold cross-validation. Results Patients were grouped into treatment responders (n = 24) and non-responders (n = 20), based on a 30% decrease in total clinician-administered PTSD scale for the DSM-IV (CAPS) score from pre- to post-treatment assessment. A network centered around the pre-supplementary motor area achieved an average accuracy of 81% (p < 0.001, based on a permutation test, corrected for multiple comparisons across 44 signal components), with a sensitivity of 84.5%, specificity of 77.5%, and area under receiver-operator curve (AUC) of 0.93. Conclusions Rs-fMRI recordings are capable of providing personalized predictions of treatment response in a sample of veterans with PTSD. It therefore has the potential to be useful as a biomarker of treatment response and should be validated in larger independent studies. Supported By ZonMw; AMC; Dutch Ministry of Defens