A potential neuromodulation target for PTSD in Veterans derived from focal brain lesions

Abstract

Neuromodulation trials for PTSD have yielded mixed results, and the optimal neuroanatomical target remains unclear. We analyzed three datasets to study brain circuitry causally linked to PTSD in military Veterans. After penetrating traumatic brain injury (n=193), lesions that reduced probability of PTSD were preferentially connected to a circuit including the medial prefrontal cortex (mPFC), amygdala, and anterolateral temporal lobe (cross-validation p=0.01). In Veterans without lesions (n=180), PTSD was specifically associated with connectivity within this circuit (p<0.01). Connectivity change within this circuit correlated with PTSD improvement after transcranial magnetic stimulation (TMS) (n=20) (p<0.01), even though the circuit was not directly targeted. Finally, we directly targeted this circuit with fMRI-guided accelerated TMS, leading to rapid resolution of symptoms in a patient with severe lifelong PTSD. All results were independent of depression severity. This lesion-based PTSD circuit may serve as a neuromodulation target for Veterans with PTSD.

Figure 1

61 participants met criteria for PTSD, 39 developed subthreshold PTSD symptoms, and 93 had neither. Lesion overlap from all three groups is depicted here.

Figure 2

(a) Each participant’s lesion was localized and mapped to a common MNI template (three examples shown here). (b) Using a normative connectome database (n=1000), we mapped the estimated connectivity profile of each lesion. (c) Across all participants, PTSD status was compared with lesion connectivity in order to identify which connections are associated with PTSD. (d) PTSD prevalence was lower in participants whose lesions were more connected to a network of regions that included the amygdala and the medial prefrontal cortex.

Figure 3

Split-half cross-validation confirmed that the PTSD circuit predicts clinical variance in an independent sample. (a) When the dataset was split into two subgroups, each subgroup yielded similar maps, even with 1000 randomly-sampled group assignments (mean spatial r=0.57). (b) Lesions in one subgroup were compared with a PTSD circuit derived from the other subgroup. The degree of overlap was significantly associated with PTSD status (p=0.01).

Figure 4. Individualized connectivity within the PTSD circuit is abnormal in PTSD patients.

(a) CWAS showed that voxels whose connectivity was abnormal in PTSD patients (blue/green) were significantly more likely to be within the lesion-derived PTSD circuit than outside the circuit. (b) PTSD symptoms were associated with abnormal connectivity within our circuit. This effect was stronger for the PTSD circuit than within or between seven canonical resting-state networks. Note: absolute t-values are depicted here to illustrate the overall strength of associations in each group (c) TMS-induced change in PTSD severity was associated with TMS-induced change within our circuit, but not other networks. Note: absolute t-values are depicted here to illustrate the overall strength of association in each group.

Figure 5. Comparison to published TMS targets.

(a) Three head-to-head trials conducted a direct comparison between EEG F3/F4 targets (which incidentally fell within our PTSD circuit) and sites that incidentally fell outside the circuit. All three of these trials found a significant advantage to stimulating the PTSD circuit. (b) Two studies reported increased fear when a fear conditioning task was paired with excitatory stimulation to our circuit. Five studies reported reduced fear when a fear conditioning task was paired with inhibitory stimulation or a fear extinction task was paired with excitatory stimulation. *Head-to-head comparison of TMS to the PTSD circuit vs. other circuits. **Results pooled across multiple groups/experiments that met inclusion criteria. (c) The PTSD circuit was highly similar to the previously-published circuit that is connected to TMS sites associated with improvement in “anxiosomatic” symptoms (blue) versus sites associated with improvement in dysphoric symptoms (white). (d) Amongst commonly-used TMS targets, the “5cm” and EEG F3/4 targets were positively correlated with our circuit, while the “6cm” and Beam F3/F4 targets were negatively correlated with our circuit. The peak TMS target likely to modify PTSD symptoms was in the mPFC. We identified six prior studies with five different stimulation targets in which TMS was applied with an extinction task or no task (yellow = positively connected targets, blue = negatively connected targets). Five of these studies were consistent with Hypothesis A, no studies were consistent with Hypothesis B, and one study was equivocal. LF = Low-Frequency TMS, HF = High-frequency TMS, MDD = major depressive disorder, iTBS = intermittent theta burst stimulation, cTBS = continuous theta burst stimulation

Figure 6. Personalizing and targeting the PTSD network with TMS.

The lesion-derived PTSD circuit (a) was personalized using a resting-state fMRI scan (b), in which the timecourse was extracted from the PTSD circuit (orange) and from other voxels in the brain (green). Timecourses with the strongest correlation to the PTSD circuit were identified, and are plotted here on a 3D reconstruction of the patient’s own brain (c). The peak voxel cluster on the prefrontal surface (circled in green) was chosen as a treatment target. (d) After 50 sessions of TMS treatment, the participant reported dramatic improvement, including persistent improvement in 17 of 20 PTSD symptoms across all four symptom domains.