Characterizing SCC Gray Matter in Depression: Implications for DBS Optimization

22 

Abstract Submission 

Carlos Alcocer^1,2, Helen Mayberg², Jungho Cha², Ha Neul Song², Martijn Figee², Brian Kopell², Ki Sueng Choi²

¹Weill Cornell Medical College, New York, NY, ²Nash Family Center for Advanced Circuit Therapeutics, Icahn School of Medicine at Mount Sinai, New York, NY

Carlos Alcocer, BA  
Weill Cornell Medical College|Nash Family Center for Advanced Circuit Therapeutics, Icahn School of Medicine at Mount Sinai
New York, NY|New York, NY

Helen Mayberg  
Nash Family Center for Advanced Circuit Therapeutics, Icahn School of Medicine at Mount Sinai
New York, NY

Jungho Cha  
Nash Family Center for Advanced Circuit Therapeutics, Icahn School of Medicine at Mount Sinai
New York, NY

Ha Neul Song  
Nash Family Center for Advanced Circuit Therapeutics, Icahn School of Medicine at Mount Sinai
New York, NY

Martijn Figee  
Nash Family Center for Advanced Circuit Therapeutics, Icahn School of Medicine at Mount Sinai
New York, NY

Brian Kopell  
Nash Family Center for Advanced Circuit Therapeutics, Icahn School of Medicine at Mount Sinai
New York, NY

Ki Sueng Choi  
Nash Family Center for Advanced Circuit Therapeutics, Icahn School of Medicine at Mount Sinai
New York, NY

Anatomical changes of the subcallosal cingulate cortex (SCC) have been observed in structural imaging studies demonstrating a volume reduction in depressed patients vs healthy controls (HC). The significance of this is unclear. Likewise, the left SCC volume reduction has been reported in deep brain stimulation (DBS) non-responders versus responders. Therefore, variation of neuroanatomy in the SCC, including structural asymmetry, may moderate differences in response to SCC DBS in patients with treatment-resistant depression (TRD). We aim to characterize pathological structural abnormality of SCC in TRD subjects versus HC, explore the effects of structural asymmetry on the lateralized behavioral response to DBS, and explore whether SCC gray matter (GM) modulates DBS response.

Participants included 47 subjects with TRD who underwent bilateral SCC DBS along with 16 healthy controls. Segmentation of SCC was performed with FreeSurfer (version 7.1.0, http://surfer.nmr.mgh.harvard.edu/) and further refined with a semi-automated correction. Volume, thickness, and laterality were extracted for all subjects. The volume of tissue activated (VTA) was generated with patient-specific stimulation parameters at 6 months. The volume of overlap between VTA and SCC GM (including overlap volumes of upper and lower banks) and the distance between VTA center of mass (COM) and SCC GM was also calculated for TRD subjects. ANCOVA was conducted to compare the structural asymmetry of the above variables between TRD subjects and HC and between DBS responders (> 50% decrease of HDRS-17 at 2 years) and non-responders. Lastly, multivariate linear regression was performed using these extract features and the following clinical measures: HDRS-17 baseline, HDRS-17 changes at 2 years, and time to stable response (TSR2, more than 50% HDRS-17 decrease for two consecutive weeks).

There was no significant GM volume reduction of left, right, or mean SCC volumes of TRD subjects versus HC nor between DBS non-responders and responders. Left SCC volume was significantly greater than right across TRD subjects [t (46) = 7.56, p < 0.0001] and HC [t (15) = 3.55, p = 0.003] with no difference in laterality between HC and TRD and responders and non-responders. There was no difference between responders and non-responders in VTA and SCC overlap volumes. There were also no between-group differences in overlap volumes of the SCC upper and lower banks. There were no differences in upper and lower bank volume of overlap laterality between responders and non-responders. However, an increased overlap, particularly of the right [t (27) = 2.36, p = 0.026], mean lower bank [t (27) = 2.27, p = 0.031] and lower right bank of SCC [t (27) = 2.62, p = 0.014], predicted an increase in TSR2. Finally, non-responders demonstrated a greater Euclidian [F (1, 44) = 9.52, p = 0.004] and horizontal distance [F (1, 40) = 5.33, p = 0.028] between the right VTA COM and SCC and increased right-sided Euclidian distance of VTA COM laterality index [F (1, 44) = 8.76, p = 0.005] than responders.

SCC volume and laterality may not serve as biomarkers to predict depression, depression severity, DBS response, or time to response. Left-sided SCC laterality likely represents normal anatomical asymmetry. While the volume of SCC gray matter included in the VTA does not differentiate DBS responders from non-responders, an increase in the volume of overlap of the right and lower banks of SCC predicts an increased time to respond. This suggests the importance of precise anatomical targeting in SCC DBS since placing the DBS lead too medial and inferior may slow the DBS response, and positioning the lead too lateral can prevent DBS response, likely due to missing critical white matter tracts. Therefore, optimal DBS response requires not only consideration of targeting WM tracts but also of the location of SCC GM in relation to the DBS leads.

Deep Brain Stimulation ¹

Psychiatric (eg. Depression, Anxiety, Schizophrenia) ²

Segmentation and Parcellation

Neuroanatomy Other

STRUCTURAL MRI

Other - Deep brain stimulation; Subcallosal cingulate; Depression