The anatomy of the human medial forebrain bundle: Ventral tegmental area connections to reward-associated subcortical and frontal lobe regions

Abstract

Introduction: Despite their importance in reward, motivation, and learning there is only sparse anatomical knowledge about the human medial forebrain bundle (MFB) and the connectivity of the ventral tegmental area (VTA). A thorough anatomical and microstructural description of the reward related PFC/OFC regions and their connection to the VTA - the superolateral branch of the MFB (slMFB) - is however mandatory to enable an interpretation of distinct therapeutic effects from different interventional treatment modalities in neuropsychiatric disorders (DBS, TMS etc.). This work aims at a normative description of the human MFB (and more detailed the slMFB) anatomy with respect to distant prefrontal connections and microstructural features.

Methods and material: Healthy subjects (n = 55; mean age ± SD, 40 ± 10 years; 32 females) underwent high resolution anatomical magnetic resonance imaging including diffusion tensor imaging. Connectivity of the VTA and the resulting slMFB were investigated on the group level using a global tractography approach. The Desikan/Killiany parceling (8 segments) of the prefrontal cortex was used to describe sub-segments of the MFB. A qualitative overlap with Brodmann areas was additionally described. Additionally, a pure visual analysis was performed comparing local and global tracking approaches for their ability to fully visualize the slMFB.

Results: The MFB could be robustly described both in the present sample as well as in additional control analyses in data from the human connectome project. Most VTA- connections reached the superior frontal gyrus, the middel frontal gyrus and the lateral orbitofrontal region corresponding to Brodmann areas 10, 9, 8, 11, and 11m. The projections to these regions comprised 97% (right) and 98% (left) of the total relative fiber counts of the slMFB.

Discussion: The anatomical description of the human MFB shows far reaching connectivity of VTA to reward-related subcortical and cortical prefrontal regions - but not to emotion-related regions on the medial cortical surface - realized via the superolateral branch of the MFB. Local tractography approaches appear to be inferior in showing these far-reaching projections. Since these local approaches are typically used for surgical targeting of DBS procedures, the here established detailed map might - as a normative template - guide future efforts to target deep brain stimulation of the slMFB in depression and other disorders related to dysfunction of reward and reward-associated learning.

Keywords: Brain; Deep brain stimulation; Depression; Human; Medial forebrain bundle; Normal anatomy; Obsessive compulsive disorder; TMS.

Fig. 1

Brodmann areae ((Brodmann, 1909), upper panel. Cortical masks for prefrontal cortex parcellation (Desikan/Killiany (Desikan et al., 2006)) were used in order to identify far-reaching VTA-connections. Cs, central sulcus. Dark grey tags represent PFC related Brodmann regions.

Fig. 2

Prototypical tracking result of the MFB in MNI group space. Global tracking performed on the HCP group level. Typical seed region (blue) selected in the triangle region (yellow, contralateral) between mtt, STN and RN (lateral VTA). (Legend: ALIC, anterior limb of internal capsule; DN, dentate nucleus; mtt, mammillo-thalamic tract; PFC, prefrontal cortex; RN, red nucleus; scp, superior thalamic peduncle; STN, subthalamic nucleus; vc, ventral capsule; VTA, ventral tegmental area).

Fig. 3

Results of global tracking (B–E) and segmentation regarding cortical parceling masks (A). B, view from dorsal; C, anterolateral right; D, antero-medial right. Fiber colors, segments: Pink, superior frontal; blue, rostral middle frontal; green, pars caudalis (of middle frontal gyrus); turquoise, medial orbitofrontal; yellow, pars triangularis; light green, pars orbitalis; not shown, pars opercularis and lateral orbitofrontal. E, view from posterior with directional color coding (blue superior/inferior, green anterior/posterior, red medial/latera). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Entire MFB (green) including main trunk in MNI152 space (a, axial; b, coronal; c, sagittal). Cortical parcellation masks (left panel) in juxtaposition. Color coding in green indicates probability of occurrence of fiber-streamlines in the entire group (in percent). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

(a–h): slMFB and its prefrontal white matter (WM) sub-segments addressing distinct cortical parcellations of the PFC parceling according to Desikan/Killiany (Desikan et al., 2006) (axial slices, MNI space). Color scale shows probability of fiber occurrence in [%] relative to entire MFB structure. Global tractographic approach used. Due to the dominating nature of the superior frontal, rostral middle and lateral orbitofrontal segmentations, in these the trunk region of MFB shows up with roughly 30%.

Fig. 5

(a–h): slMFB and its prefrontal white matter (WM) sub-segments addressing distinct cortical parcellations of the PFC parceling according to Desikan/Killiany (Desikan et al., 2006) (axial slices, MNI space). Color scale shows probability of fiber occurrence in [%] relative to entire MFB structure. Global tractographic approach used. Due to the dominating nature of the superior frontal, rostral middle and lateral orbitofrontal segmentations, in these the trunk region of MFB shows up with roughly 30%.

Fig. 6

(a, b): Three-dimensional white matter of MNI. slMFB and its prefrontal sub-segments addressing distinct (WM) of overlaying cortical Desikan/Killiany parcellation in the global tractographic approach. Numbers indicate Brodmann areae (Brodmann, 1909) (left panel prefrontal parcellations, 3D representation in MNI space). Color coding represents percentage of occurrence of fibers (cf. Fig. 5) relative to the entire group.

Fig. 7

Different tracking methods applied on the group level HCP data in MNI space. The MFB main trunk is visible in all three modalities (A, sagittal; B, axial). As expected, with more advanced visualization methods (CSD/FACT probabilistic, Global) more distal and especially lateral extensions of the MFB (red) become visible (middle and right).