Topography of cortical and subcortical connections of the human pedunculopontine and subthalamic nuclei

Abstract

Deep brain stimulation (DBS) of the subthalamic nucleus (STN) is the most common surgical therapy for Parkinson' s disease (PD). DBS of the pedunculopontine nucleus (PPN) is emerging as a promising surgical therapy for PD as well. In order to better characterize these nuclei in humans, we determined the anatomical connections of the PPN and STN and the topography of these connections using probabilistic diffusion tractography. Diffusion tractography was carried out in eight healthy adult subjects using diffusion data acquired at 1.5 T MRI (60 directions, b = 1000 s/mm², 2 × 2 × 2 mm³ voxels). The major connections that we identified from single seed voxels within STN or PPN were present in at least half the subjects and the topography of these connections within a 36-voxel region surrounding the initial seed voxel was then examined. Both the PPN and STN showed connections with the cortex, basal ganglia, cerebellum, and down the spinal cord, largely matching connections demonstrated in primates. The topography of motor and associative brain areas in the human STN was strikingly similar to that shown in animals. PPN Topography has not been extensively demonstrated in animals, but we showed significant topography of cortical and subcortical connections in the human PPN. In addition to demonstrating the usefulness of PDT in determining the connections and topography of small grey matter structures in vivo, these results allow for inference of optimal DBS target locations and add to our understanding of the role of these nuclei in PD.

Introduction

Deep brain stimulation (DBS) of specific nuclei can alleviate intractable Parkinson's disease (PD) symptoms. The subthalamic nucleus (STN) is currently the most common target and stimulation of this structure is particularly successful in reducing tremor and bradykinesia (Limousin et al., 1995, Krack et al., 1997, Limousin et al., 1998, Yokoyama et al., 1999, Bejjani et al., 2000, Stolze et al., 2001, Kleiner-Fisman et al., 2003, Rodriguez-Oroz et al., 2005). Recent evidence suggests that stimulation of the pedunculopontine nucleus (PPN) in the upper brainstem can be effective in ameliorating medically intractable akinesia (Jenkinson et al., 2004, Plaha and Gill, 2005, Mazzone et al., 2005, Galati et al., 2006). In both cases, the effects of stimulation are not limited to the targeted structure but will affect the distributed anatomical networks to which the target structure belongs. Therefore, understanding the anatomical connections of these deep brain structures will help elucidate treatment effects. In particular, establishing the topography of cortical and subcortical connections of STN and PPN in the human brain could help with accurate targeting of critical pathways in DBS.

The PPN plays an important role in controlling movement, particularly of the axial and proximal limb muscles employed in the maintenance of posture and locomotion (Masdeu et al., 1994, Lee et al., 2000, Pahapill and Lozano, 2000). While lesions to the PPN cause akinesia (Kojima et al., 1997, Munro-Davies et al., 1999, Matsumura and Kojima, 2001), stimulation (Jenkinson et al., 2004) or disinhibition (Nandi et al., 2002) of the PPN can dramatically improve motor behavior in monkeys made parkinsonian with MPTP, suggesting that the PPN is over inhibited by the basal ganglia in Parkinson's disease. Hence, deep brain stimulation of the PPN has been tested in Parkinsonian patients and early results show promising outcomes particularly in reducing akinesia and gait abnormalities (Plaha and Gill, 2005, Mazzone et al., 2005).

The need for accurate surgical targeting of the PPN for DBS procedures necessitates better anatomical characterization of the PPN. Currently, PPN connections have been best characterized using tracing and evoked-potential studies in rodents (Garcia-Rill et al., 1987, Rye et al., 1987, Garcia-Rill and Skinner, 1987a, Garcia-Rill and Skinner, 1987b, Spann and Grofova, 1989, Spann and Grofova, 1991, Spann and Grofova, 1992), but these connections differ greatly from those shown in primate tracing studies (Lavoie and Parent, 1994a, Lavoie and Parent, 1994b, Lavoie and Parent, 1994c, Pahapill and Lozano, 2000). Though both the rodent and primate PPN have strong connections with the subthalamic nucleus (STN), substantia nigra (SN), and globus pallidus (GP) (Lee et al., 2000), connections between the PPN and motor cortex have only been shown in primates, and projections from the PPN to the deep cerebellar nuclei and projections extending down the spinal cord have only been shown in rodents, though sparse projections from the deep cerebellar nuclei to the PPN have been shown in primates (Hazrati and Parent, 1992) and spinal cord connections are also suspected to exist in primates (Lee et al., 2000, Pahapill and Lozano, 2000, Matsumura et al., 2000). Recently, we have used probabilistic diffusion tractography (PDT) to trace connections of the PPN in humans (Muthusamy et al., in press). These results show that PPN connections in humans largely match those in primates, but also include descending cerebellar connections. These differing results illustrate the importance of verifying the existence of PPN connections across different species using the same methodological techniques.

In contrast to the PPN, the connections of the STN have been well characterized in both rodents and primates and there is little difference between the species (Hamani et al., 2004). The major connections of the STN are with the globus pallidus (GP) and substantia nigra pars reticula (SNpr). The STN also receives excitatory input from various parts of the cerebral cortex, including the motor cortex, SMA, and the dorsal and ventral PMC (dPMC and vPMC, respectively) (Hamani et al., 2004).

The current study investigated the topography of STN and PPN connections. Until now, PPN topography has been explored only in primates and has only been demonstrated for cortical connections in one study (Matsumura et al., 2000). However, PPN topography has not been demonstrated in humans or in other species. Distinct topography has also been demonstrated in the STN, separating STN regions connected to motor, associative, and limbic brain areas (Hamani et al., 2004) but has not previously been studied in humans.

We used diffusion tensor imaging (DTI) to test for topographic organization in the cortical and subcortical connections of human STN and PPN. DTI can be used to estimate anatomical connections in the human brain in vivo. Probabilistic methods for tractography (Parker et al., 2003, Behrens et al., 2003b) allow for tracking of probable fiber pathways between cortical and subcortical grey matter structures (Behrens et al., 2003a, Johansen-Berg et al., 2005, Sillery et al., 2005, Muthusamy et al., in press). For the purposes of this study, connections refer to the estimated anatomical pathways between brain regions (e.g., a fiber path estimated between the STN and a cortical brain region) while topography refers to the size and spatial distribution of the specific parts of a given grey matter structure that connect to other brain regions (e.g., the anterior–medial part of the STN connects with cortical areas). Note that here we are using the terms “connections” and “topography” in a purely anatomical sense. Although tractography cannot determine the direction or sign of functional projections between regions, it does allow for the estimation of the connections and topography of a given brain region, in vivo. Therefore, in order to further anatomically characterize human PPN and STN, we have employed PDT to determine the anatomical connections and topography of cortical and subcortical connections of the PPN and STN in healthy human subjects.