In Reply

SIR: Schutter et al. present an intriguing and testable hypothesis regarding how and why slow rate (1 Hz) rTMS may be effective for treatment of depression. They suggest that a cortical circuit involving left frontal and right parietal regions is important in depression and that decreased functional connectivity between these regions occurs in depression. They suggest that slow rate magnetic stimulation rather than improving local left frontal cortical function enhances the connectivity of this region with the opposite parietal cortex.

Our finding of a therapeutic effect of slow left frontal stimulation is only “counterintuitive” if left frontal hypofunction is a marker of depression and slow rTMS further causes neural inhibition. In fact, results of functional imaging in depressed patients are not highly replicable and may differentiate subgroups of patients. For instance, a study using single photon emission computed tomography in refractory depressed patients showed hypoperfusion in the inferior frontal, anterior temporal, and anterior cingulate.¹ How does slow rTMS to the dorsolateral frontal cortex modulate this network of altered function? Our patients were suffering not only from refractory depression but also from long-standing posttraumatic stress disorder (PTSD); which networks (cortical-cortical, cortical-subcortical, cortical-limbic) have altered function? How do these networks interact with the dorsolateral frontal region? How does rTMS at slow or fast rates alter function in the stimulated region and in networks connected to the stimulated region?

As Schutter et al. suggest, data are just beginning to accrue on how rTMS may alter functional connectivity. Jing and Takigawa² showed that directed coherence between cortical regions increased after rTMS, particularly in the direction from frontal to parietal; however, they studied only normal subjects after limited rapid (10 Hz) rTMS, and their findings were larger within the hemisphere than between hemispheres. It remains a stretch from this observation to enhancement of left frontal to right parietal functional connectivity. In a more recent study, Strens et al.³ showed that subthreshold slow (1 Hz) rTMS over the motor cortex increased ipsilateral cortico-cortical coherence and increased coherence between left and right motor cortices. The effects lasted up to 25 minutes after a single session of 1,500 stimuli. However, there are no published data on changes in coherence after a typical course of rTMS treatment of depression (usually 10 treatments over 2 weeks).

We thank Schutter et al. for their hypothesis, which moves the concepts behind rTMS treatment beyond simply altering neuronal excitability in brain regions near the stimulator and moves them into the realm of altering cortical connectivity. We would merely point out the many questions that remain to be answered both with respect to the underlying pathophysiology of depression and PTSD and the mechanisms of brain function modulation by rTMS.