Tractography-assisted deep brain stimulation of the superolateral branch of the medial forebrain bundle (slMFB DBS) in major depression

Abstract

Background: Deep brain stimulation (DBS) of the superolateral branch of the medial forebrain bundle (slMFB) emerges as a - yet experimental - treatment for major depressive disorder (MDD) and other treatment refractory psychiatric diseases. First experiences have been reported from two open label pilot trials in major depression (MDD) and long-term effectiveness for MDD (50 months) has been reported.

Objective: To give a detailed description of the surgical technique for DBS of the superolateral branch of the medial forebrain bundle (slMFB) in MDD.

Methods: Surgical experience from bilateral implantation procedures in n = 24 patients with MDD is reported. The detailed procedure of tractography-assisted targeting together with detailed electrophysiology in 144 trajectories in the target region (recording and stimulation) is described. Achieved electrode positions were evaluated based on postoperative helical CT and fused to preoperative high resolution anatomical magnetic resonance imaging (MRI; Philips Medical Systems, Best, Netherlands), including the pre-operative diffusion tensor imaging (DTI) tractographic information (StealthViz DTI, Medtronic, USA; Framelink 5.0, Medtronic, USA). Midcommissural point (MCP) coordinates of effective contact (EC) location, together with angles of entry into the target region were evaluated. To investigate incidental stimulation of surrounding nuclei (subthalamic nucleus, STN; substantia nigra, SNr; and red nucleus, RN) as a possible mechanism, a therapeutic triangle (TT) was defined, located between these structures (based on MRI criteria in T2) and evaluated with respect to EC locations.

Results: Bilateral slMFB DBS was performed in all patients. We identified an electrophysiological environment (defined by autonomic reaction, passive microelectrode recording, acute effects and oculomotor effects) that helps to identify the proper target site on the operation table. Postoperative MCP-evaluation of effective contacts (EC) shows a significant variability with respect to localization. Evaluation of the TT shows that responders will typically have their active contacts inside the triangle and that surrounding nuclei (STN, SNr, RN) are not directly hit by EC, indicating a predominant white matter stimulation. The individual EC position within the triangle cannot be predicted and is based on individual slMFB (tractography) geometry. There was one intracranial bleeding (FORESEE I study) during a first implantation attempt in a patient who later received full bilateral implantation. Typical oculomotor side effects are idiosyncratic for the target region and at inferior contacts.

Conclusion: The detailed surgical procedure of slMFB DBS implantation has not been described before. The slMFB emerges as an interesting region for the treatment of major depression (and other psychiatric diseases) with DBS. So far it has only been successfully researched in open label clinical case series and in 15 patients published. Stimulation probably achieves its effect through direct white-matter modulation of slMFB fibers. The surgical implantation comprises a standardized protocol combining tractographic imaging based on DTI, targeting and electrophysiological evaluation of the target region. To this end, slMFB DBS surgery is in technical aspects comparable to typical movement disorder surgery. In our view, slMFB DBS should only be performed under tractographic assistance.

Keywords: CT, computed tomography; DBS, deep brain stimulation; DTI FT, DTI fiber tractography; DTI, diffusion tensor magnetic resonance imaging; Deep brain stimulation; Depression; Diffusion tensor imaging; EC, effective contact; FT, fiber tractography; Fiber tracking; HF, high frequency; Hz, Hertz [1/s]; IPG, internal pulse generator; MADRS, Montgomery-Åsberg Depression Rating Scale; MCP, mid-commissural point; MDD, major depressive disorder; MRI, magnetic resonance imaging; Medial forebrain bundle; OCD; RN, red nucleus; SNr, substantia nigra pars reticulata; STN, subthalamic nucleus; Stereotactic surgery; Tractography; VAT, volume of activated tissue; VTA, ventral tegmental area; mA, milli-ampere; slMFB; μs, micro second.

Graphical abstract

Fig. 1

Artistic representation of the slMFB and the stimulated region. The stimulated region is located (yellow sphere) between the mammillary-bodies, the red nucleus and the anterior most aspect of the subthalamic nucleus. Note the proximity of the target region and the occulomotor nerve that traverses the VTA, laterally. Structures: 1, Ventra tegmental area (black arrows); 2, superolateral branch of medial forebrain bundle; 3, occulomotor nerve (CNiii, white arrows); 4, substantia nigra; 5, subthalamic nucleus; 6, hyperdirect pathway; 7, corticospinal tract; 8, dentato-rubro-thalamic tract; 9, medial lemniscus; 10, red nucleus; 11, periaquaeductal grey; 12, mammillary body; 13, fornix; 14, inferomedial branch of the medial forebrain bundle.

Fig. 2

Typical slMFB DBS. A, axial slides showing deepest (left) and most superficial contacts (right) on T2-weighted anatomy. B, Outlines of functional structures given. C; left, outline shows how DBS electrode traverses the slMFB (green); right, three-dimensional view from lateral and left.

Fig. 3

Three-dimensional depiction of a typical bilateral slMFB-DBS implantation. A, implantation site as viewed from sub-mentally. The DBS electrodes are situated inside the slMFB (green bundles) in the corridor medial to the STN/SNr-complex. The tip of the electrode touches the ventral tegmental area (VTA). B, same as A but without fibers. C, view from superior and left.

Fig. 4

Electrophysiological synoptical graph. A, 144 trajectories and the differentiation in likelihood of occurrence of nuclear structures (STN = subthalamic nucleus; SNr = subtstantia nigra, Thal = thalamus; RN = red nucleus). B, occurrence of test stimulation with respect to the target region; C, typical DBS electrode position and relation to stimulation sites are given. Stimulation was typically performed on contact 1 (anodal) and 2,3 (cathodal). Orange lines indicate overlap with effective stimulation (1.5–6.5 mm above target).

Fig. 5

Left oculomotor nerve activation (B) as seen on the deepest stimulation contact and a stimulation current of 1.5 mA (milli-ampere).

Fig. 6

Representation of effective electrode contact (EC) positions in idealized atlas slices (coronal and axial) of the Schaltenbrand and Wahren atlas (Schaltenbrand & Hassler, 1977). Left Panel (A, axial; B, coronal): All EC = blue diamonds, target points (TP) = black dots (projected into the slide in A but in reality, more inferior below the axial plane, cf. B). The mean stimulation point is situated in the corridor between red nucleus and STN/SNr complex. Right panel (C, Axial; D, coronal): Same as A, B but responders yellow, non-responders grey.

Fig. 6

Representation of effective electrode contact (EC) positions in idealized atlas slices (coronal and axial) of the Schaltenbrand and Wahren atlas (Schaltenbrand & Hassler, 1977). Left Panel (A, axial; B, coronal): All EC = blue diamonds, target points (TP) = black dots (projected into the slide in A but in reality, more inferior below the axial plane, cf. B). The mean stimulation point is situated in the corridor between red nucleus and STN/SNr complex. Right panel (C, Axial; D, coronal): Same as A, B but responders yellow, non-responders grey.

Fig. 7

Therapeutic triangle (TT) definition (yellow) between mammillothalamic tracts STN/SNr and red nucleus, respectively. Definition of three stimulation levels for the determination of optimal EC position (A-C). D, sub-parcellation of the TT. E, EC of responders (yellow) are clearly located inside the TT (projection of EC in level 2, only for visualization purposes). A therapeutic effect is likely due to white matter modulation and not due to an inadvertent stimulation of grey matter structures (nuclei) in the proximity. However, this is not clearly defined by a certain position within the triangle but only by the DTI-FT rendition of the slMFB. For details and statistics see text.