Rtms Treatment: Because the Stimulation Target in This Case Lay Within Medial Rather Than

METHODS

rTMS treatment: Because the stimulation target in this case lay within medial rather than lateral prefrontal cortex, motor thresholds were determined using stimulation of the medial aspect of primary motor cortex, based on activation of the EHL, as in previous studies involving stimulation of the DMPFC and adjacent ACC (Ciampi de Andrade et al, 2012; Harmer et al, 2001; Hayward et al, 2007). Under MRI guidance, the coil vertex was placed over the interhemispheric fissure, immediately anterior to the central sulcus. Previous studies have shown preferential activation of lower limb primary motor cortex with the coil oriented laterally, current flow directed towards the hemisphere to be stimulated (Terao et al, 2000, 2001). This approach was used for stimulation of the right and left primary motor cortex, and the resting motor threshold for contralateral hallux extension determined by visual inspection according to previously published methods (Schutter and Van Honk, 2006). These treatment parameters were drawn from those previously used safely in published work using high-dose rTMS (Hadley et al, 2011). As the authors of that study note, although these treatment parameters employ an inter-train interval shorter than that specified in the originally published safety guidelines for rTMS, they remain within those guidelines' general recommendation that the intertrain interval be twice as long as the stimulation interval (Chen et al, 1997).

Craddock Atlas: The parcellation map of Craddock et al. (Craddock et al, 2012)offers superior within-region homogeneity of activity as compared to anatomically-based templates whose regions have the potential to span several functional subregions and thus may conflate areas with heterogeneous patterns of activity. The map of Craddock et al. treats the brain as an undirected graph based on resting-state connectivity between gray matter voxels, and uses the NCUT algorithm to divide the graph into parcels that minimize the number of edges divided by the borders of the parcels. This approach has been shown to generate maps whose subregions have superior within-region homogeneity of activity.

DISCUSSION

In attempting to determine the specificity of these findings, it is instructive to compare our results to those of Kozel and colleagues (Kozel et al, 2011) who used a similar approach, examining baseline connectivity predictors of anti-depressant response to selective serotonin reuptake inhibitors (SSRI). Consistent with our findings, those authors observed that connectivity between dorsal and ventral regions of the cingulate cortex was associated with SSRI treatment response. However, with this type of treatment, Kozel et al. found that low, rather than high, baseline functional connectivity between these regions was associated with successful treatment outcome.

There are noteworthy methodological differences between the studies, perhaps the most prominent being anatomical differences between the ROIs: the dorsal cingulate region chosen by Kozel and colleagues (2011) spans three regions of cingulate thought to be functionally and structurally distinct from each other (Shackman et al, 2011; Vogt, 2005), while the ventral cingulate region in the current study is somewhat anterior to that of Kozel et al. Furthermore, the present study was conducted in a population of treatment-resistant patients, such that our sample would be considered non-responders in the treatment sample of Kozel and colleagues. It is therefore possible that the patients in this study are substantially different in baseline resting state connectivity, and that the relationship between treatment response and functional connectivity of given regions might be quite different depending on response profile. Further study of the relationship between resting-state connectivity and treatment outcome, across a variety of treatment modalities, will help to contextualize the differences observed between the study of Kozel and colleagues and the present study.
The final and perhaps most tantalizing possibility is that connectivity between dorsal and ventral cingulate regions could be developed into a clinically useful biomarker for treatment selection, allowing for the matching of patients to particular treatments based on relevant individual difference variables. The need for such markers is underscored by the limited recent progress in improving the efficacy of pharmacological, cognitive and behavioral approaches to depression. The fact that predictor regions were similar in the current study to those observed in previous studies confirms the role of regions like sgACC, aMCC, medial dorsal thalamus and striatum in depression and response to treatment (Kennedy et al, 2012). The finding that opposite patterns of resting state connectivity were associated with response to pharmacological and rTMS treatments suggests that these markers could find clinical utility in steering patients towards the treatment modality with the greatest likelihood of success, as per the framework of Seminowicz et al (2004).

Chen R, Gerloff C, Classen J, Wassermann EM, Hallett M, Cohen LG (1997). Safety of different inter-train intervals for repetitive transcranial magnetic stimulation and recommendations for safe ranges of stimulation parameters. Electroencephalogr Clin Neurophysiol105: 415–421.

Ciampi de Andrade D, Galhardoni R, Pinto LF, Lancelotti R, Rosi J Jr, Marcolin MA, et al (2012). Into the island: a new technique of non-invasive cortical stimulation of the insula. Neurophysiol Clin42: 363–368.

Craddock RC, James GA, Holtzheimer PE 3rd, Hu XP, Mayberg HS (2012). A whole brain fMRI atlas generated via spatially constrained spectral clustering. Hum Brain Mapp33: 1914–1928.

Hadley D, Anderson BS, Borckardt JJ, Arana A, Li X, Nahas Z, et al (2011). Safety, tolerability, and effectiveness of high doses of adjunctive daily left prefrontal repetitive transcranial magnetic stimulation for treatment-resistant depression in a clinical setting. J ECT27: 18–25.

Harmer CJ, Thilo KV, Rothwell JC, Goodwin GM (2001). Transcranial magnetic stimulation of medial-frontal cortex impairs the processing of angry facial expressions. Nat Neurosci4: 17–18.

Hayward G, Mehta MA, Harmer C, Spinks TJ, Grasby PM, Goodwin GM (2007). Exploring the physiological effects of double-cone coil TMS over the medial frontal cortex on the anterior cingulate cortex: an H2(15)O PET study. Eur J Neurosci25: 2224–2233.

Kennedy SH, Downar J, Evans KR, Feilotter H, Lam RW, MacQueen GM, et al (2012). The Canadian Biomarker Integration Network in Depression (CAN-BIND): advances in response prediction. Curr Pharm Des18: 5976–5989.

Kozel FA, Rao U, Lu H, Nakonezny PA, Grannemann B, McGregor T, et al (2011). Functional connectivity of brain structures correlates with treatment outcome in major depressive disorder. Front Psychiatry2: 7.

Schutter DJLG, Honk J van (2006). A standardized motor threshold estimation procedure for transcranial magnetic stimulation research. J ECT22: 176–178.

Shackman AJ, Salomons TV, Slagter HA, Fox AS, Winter JJ, Davidson RJ (2011). The integration of negative affect, pain and cognitive control in the cingulate cortex. Nat Rev Neurosci12: 154–167.

Terao Y, Ugawa Y, Hanajima R, Machii K, Furubayashi T, Mochizuki H, et al (2000). Predominant activation of I1-waves from the leg motor area by transcranial magnetic stimulation. Brain Res859: 137–146.

Terao Y, Ugawa Y, Hanajima R, Machii K, Furubayashi T, Mochizuki H, et al (2001). A single motor unit recording technique for studying the differential activation of corticospinal volleys by transcranial magnetic stimulation. Brain Res Brain Res Protoc7: 61–67.

Vogt BA (2005). Pain and emotion interactions in subregions of the cingulate gyrus. NatRevNeurosci6: 533–544.

Supplementary Figure Legends

Figure S1 Change in Depression Scores: Black lines represent single subjects HAMD scores prior to and immediately after 4 week treatment with rTMS to dmPFC. The red line represents mean scores for the group at each time point.

Figure S2 Regions of Interest: The association between HAMD improvement connectivity of these ROIs were tested. Superior (red) and inferior (green) regions in the stimulated region were tested. No significant associations were found for the superior cluster, so the inferior (green) cluster is referred to as “dmPFC” throughout the manuscript. Connectivity of dmPFC with sgACC (blue) was significantly associated with HAMD, so given this region’s role in previous predictor studies, baseline connectivity and connectivity change of this region with HAMD was also tested.