Modulation of dorsolateral prefrontal cortex functional connectivity after intermittent theta-burst stimulation in depression: Combining findings from fNIRS and fMRI

Abstract

Background: Resting-state functional magnetic resonance imaging (fMRI) can assess modulation of functional connectivity networks following repetitive transcranial magnetic stimulation (rTMS) in the treatment of depression. Functional near-infrared spectroscopy (fNIRS) is well suited for the concurrent application during rTMS treatment sessions to capture immediate blood oxygenation (oxy-Hb) effects, however limited in spatial resolution.

Objective: To understand the network effects behind such a prefrontal fNIRS response during rTMS, and to test whether the fNIRS signal may be predictive of treatment response, we linked data from fNIRS and fMRI within a clinical intervention study.

Methods: 42 patients with ongoing depression were recruited and randomized to receive active or sham intermittent theta-burst stimulation (iTBS) over the dorsomedial prefrontal cortex (dmPFC) twice daily for ten days at target intensity. Oxy-Hb was recorded with fNIRS during the first, fifth, and final day of iTBS, with the probe holders located laterally to the TMS coil over regions corresponding to the left and right dorsolateral prefrontal cortex (dlPFC). Resting-state fMRI scanning was performed before and after the whole iTBS treatment course. Functional connectivity analyses were then performed using dlPFC seeds from parcels of a brain atlas showing most overlap with the fNIRS probe locations during treatment.

Results: After active iTBS, left dlPFC-connectivity to the right insula/operculum was reduced compared to sham. The left insula showed a connectivity reduction to the left dlPFC that correlated with an improvement in symptoms. In addition, the posterior parietal cortex showed a connectivity reduction to the left dlPFC that correlated with the fNIRS signal following active iTBS. Finally, the fNIRS oxy-Hb signal from the left dlPFC-seed during the first treatment day was predictive of dlPFC-connectivity change to precentral and temporal cortex regions.

Conclusion: By linking findings from these two different methods, this study suggests that changes within both the salience network and the central executive network affect the fNIRS response to iTBS.

Keywords: Central executive network; Depressive disorder; Repetitive transcranial magnetic stimulation (rTMS); Resting-state; Salience network.

Fig. 1

1A: Placement of fNIRS optodes on the forehead, with the light emitter above the detector, aligned with the coil. 1B: fNIRS optode center locations (orange dots) onto the 400 parcellation atlas by Schaefer et al. (2018). The parcel with the highest optode center load per hemisphere was used as region of interest in the functional connectivity analyses, corresponding to regions within the left and right dorsolateral prefrontal cortex. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 1

1A: Placement of fNIRS optodes on the forehead, with the light emitter above the detector, aligned with the coil. 1B: fNIRS optode center locations (orange dots) onto the 400 parcellation atlas by Schaefer et al. (2018). The parcel with the highest optode center load per hemisphere was used as region of interest in the functional connectivity analyses, corresponding to regions within the left and right dorsolateral prefrontal cortex. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 2

Group means and individual data points depicting left dlPFC-connectivity changes to a cluster spanning the right insula and operculum (peak at X = 46, Y = 4, Z = 2, t = 4.44, p =.001, cluster size = 416) after active and sham iTBS. Note that a positive value indicates an increase in connectivity, and a negative value a decrease in connectivity following iTBS treatment. Error bars mark the 95 % confidence interval.

Fig. 3

Active compared to sham iTBS was followed by a stronger negative relationship between left dlPFC connectivity change and symptom change in the left insula (X = -36, Y = -10, Z = 24, t = 5.46, p =.022, cluster size = 226). Note that a positive symptom change value indicates a symptom reduction.

Supplementary figure 1

Supplementary figure 2