Combining Transcranial Magnetic Stimulation With Behavioral Interventions for Posttraumatic Stress Disorder: Reasons for Optimism Despite Negative Findings

Posttraumatic stress disorder (PTSD) is a serious problem that impacts civilians and veterans worldwide. Despite significant strides in efficacious treatments, symptoms tend to persist in a large portion of individuals. This is sometimes due to the lack of completing often lengthy treatment protocols or because symptoms persist even after successful treatment completion. For first-line evidence-based psychotherapies (EBPs) for PTSD, such as cognitive processing therapy or prolonged exposure, studies have found that approximately one third of those seeking treatment drop out before completion, with higher rates in Veterans Healthcare Administration and Department of Defense settings (1). Many patients may respond to treatments including EBPs, medication, or a combination of the two; however, symptoms often do not fall below clinical thresholds even after completing a full course (2). Therefore, there is room for improvement, and one method that shows promise is transcranial magnetic stimulation (TMS).

Though TMS has been approved and used for treatment as a stand-alone method for treatment-resistant depression and obsessive-compulsive disorder, recent research has demonstrated a promising use in PTSD by combining it synergistically with behavioral treatments to augment its effects. In the current issue of Biological Psychiatry, Isserles et al. (3) report on a large, multisite randomized clinical trial (RCT) taking place in locations across the globe using a deep TMS (dTMS) system combined with a brief exposure for PTSD.

Isserles et al. (3) hypothesized that a brief trauma exposure combined with dTMS to the medial prefrontal cortex (mPFC) would contribute to extinction of fear memories and thereby reduce PTSD symptoms to a greater degree than brief exposure and sham dTMS. This was based on pilot data from the group that found benefit of exposure plus dTMS for PTSD compared with sham (4). In the present study, a new type of dTMS treatment coil was used, known as an H7 coil, on a BrainsWay device (BrainsWay sponsored the study). The procedures included conducting a brief exposure of the participants’ trauma followed by a session of dTMS (again to theoretically enhance extinction of the fear response induced by the exposure). The Clinician-Administered PTSD Scale for DSM-5 (CAPS-5) was used as the primary outcome. The study was well designed and carefully conducted. In addition, Isserles et al. (3) used a theoretical mechanism of TMS to guide study design, which is an important step forward for the field. Unfortunately, contrary to the study hypothesis, brief exposure followed by sham dTMS was associated with better outcomes compared with brief exposure followed by active dTMS (21-point vs. 16-point decrease on the CAPS-5). Because of this, the trial was discontinued early for futility.

These null effects are puzzling and disappointing, though several potential explanations exist. Isserles et al. (3) speculate that their findings may be due to the similarity of brief exposure to exposure EBPs for PTSD. In their pilot work, a neutral script was presented after the trauma script, which may have diminished the impact of the exposure and explained the greater effects of the exposure in the present study. Another explanation offered is that a negative influence of active dTMS, such as headache or application site discomfort, may have lowered the ability of patients to perform reprocessing and fear extinction. It is also noted that an H7 coil was used instead of an H1 as in the previous study, and this may have stimulated brain regions differently, including the dorsal anterior cingulate cortex (ACC), dorsolateral prefrontal cortex, and mPFC, contributing to a different outcome.

To the first point, though the brief exposure may share some similarities with EBPs for PTSD, there is no clinical evidence base to suggest that it may improve PTSD symptoms. Given that the premise of the study was to enhance fear extinction, it may have benefited the trial to use an established exposure treatment that is well known to contribute to fear extinction rather than an unestablished, experimental exposure. Several key differences between an experimental exposure and an EBP include EBPs having a trained clinician eliciting information from the patient to determine what their index trauma is, particularly if the patient has had multiple traumatic experiences, and participating in manualized, repeated exposure with treatment progression over several sessions that vary depending on protocol. For instance, extensive RCT work has determined that five progressive sessions in written exposure therapy and 12 to 18 sessions in prolonged exposure and cognitive processing therapy are efficacious for PTSD. Importantly, over the course of some EBPs for PTSD, is it expected that PTSD symptoms may actually worsen at the beginning of treatment and then improve later on (5). In the present study, the brief exposure differed from EBPs in that nonclinicians and the patient collaboratively decided on what the index trauma was, filled out a structured form, and subsequently wrote and recorded an audio script that was played before each of three weekly dTMS sessions for 4 weeks based on the procedures in the previous pilot. It may have been the case that inadequate experimental exposure or a mechanistic mismatch between the type of DBS and the behavioral therapy used negatively impacted the outcome.

Another potential reason the active treatment was not better than sham may have been that the order of study procedures was suboptimal. The study provided a brief trauma exposure prior to neurostimulation, predicated on the success of the prior pilot study with dTMS conducted after trauma imagery and resulting in improvement in intrusive items on the CAPS and a trend for total score improvement (4). However, successful studies in this vein often show that behavioral treatments after neurostimulation enhance the effect both in experimental work impacting emotion regulation in healthy samples (6) and in clinical studies for PTSD (7).

Order of procedures may have had an impact on mechanism. It is possible that by completing the behavioral intervention prior TMS, rather than after, the study missed out on possible focal and downstream neural activity taking place poststimulation that may aid in eliciting improvement. Kozel et al. (7) hypothesized that the greater effect in their RCT of active compared with sham TMS added to cognitive processing therapy immediately and 6 months posttreatment may have been due to enhancing metaplasticity. Metaplasticity is prior activity at the synaptic level leading to a persistent change in plasticity (8). Though the mechanisms of TMS are not known, previous work has demonstrated metaplasticity in the human motor cortex (9), and it is possible that metaplasticity in prefrontal cortex function and/or connectivity with limbic regions is a mechanism for TMS. Conducting TMS after behavioral intervention, as in the present study, may have missed out on this potential metaplasticity-inducing mechanism of TMS.

Finally, the present study used an H coil, which is a “deep” TMS coil stimulating regions 3 to 5 cm below the skull, as opposed to a more focal figure 8 coil, which stimulates 2 to 3 cm below the surface. Specifically, the study used an H7 coil, which is noted to be different and improved from the H1 coil used in the pilot because it stimulates the mPFC and ACC bilaterally as opposed to the mPFC in addition to unspecified “deep prefrontal brain regions” noted in the pilot study, with the differences demonstrated with a chart of the H1 and H7 field distributions [Figure 1 in Isserles et al. (3)]. Given the role of the ACC in emotion regulation and PTSD (10), it is a reasonable hypothesis that the additional ACC stimulation may be beneficial. However, as noted in the field distribution chart, not only are the ACC and mPFC stimulated with the H7 coil, but so are a host of other regions owing to the nonfocal nature of the H coil. It may be that the lack of specificity when stimulating brain regions affected PTSD symptoms in unanticipated ways.

In sum, the study was a valiant undertaking occurring over the course of many years with a high-quality RCT study design and laudable effort to determine if fear extinction may be a potential mechanism for PTSD treatment with a brief exposure added to dTMS. Determination of the mechanism of TMS for the treatment of PTSD is crucial, and Isserles et al. (3) made an impressive effort on this front, even in the context of null findings. Future PTSD treatment research studying the symbiotic effects of behavioral treatments with TMS is a promising avenue. Given the positive as well as the null findings informing this literature, achieving the best treatment protocol for symptom reduction in patients is a hopeful outcome.