Computer modeling of neuromodulation in the management of chronic pain

Summary of the thesis: Computer modeling of neuromodulation in the management of chronic pain by: Ljubomir Manola Neuromodulation is an important and frequent therapy applied, among others, in the management of chronic pain. Neuromodulation is defined as “a therapeutic alteration of activity in the central, peripheral or autonomic nervous systems, electrically or pharmacologically, by means of implanted devices��?. It encompasses a focal, minimal invasive and reversible approach. Due to its invasive character with some more risks than other modalities of pain treatment, neuromodulation is generally applied as the last option when all other therapies have failed. The overall success rate of this treatment modality is about 50%. It is often the only modality that may help a particular patient. However, there are still many chronic pain patients in whom neuromodulation was attempted with suboptimal or even less than satisfactory results. For the sake of this pool of patients it is an imperative to understand the effects occurring during a neuromodulation therapy and to design stimulation in such a way that a larger number of patients can be treated, while giving them an improved satisfaction with the therapy. This thesis is focused on spinal cord and motor cortex stimulation therapies for chronic, otherwise intractable pain management. The influence of various stimulation and volume conductor parameters on the neural response was investigated by means of realistic computer models. In Chapter 1 neuromodulation is presented as one of the treatment modalities for the management of various chronic pain syndromes. Further, the clinical methods named spinal cord stimulation and motor cortex stimulation are introduced. Details are given on working hypotheses (when existing), main indications, surgical procedures and technical aspects, difficulties and challenges encountered in the therapies. The chapter is completed by introducing the computer model as an invaluable tool to mimic/simulate the biophysical effects of neurostimulation. The technical performance of two different lead types commonly used in the clinical practice of spinal cord stimulation, the paddle (or surgical) lead and the percutaneous lead has been modeled and evaluated in Chapter 2. In addition to the thickness of the dorsal cerebrospinal fluid and the lead distance from the dura mater, the contact spacing on the lead was identified as the most important parameter influencing the performance. Despite a similar contact spacing and technical performance, a superior clinical performance of paddle leads was observed in clinical practice. This superiority is most likely attributable to the larger volume of a paddle and a consequent ventral displacement of the dura mater. The secure fixation of the paddle in close contact with the dura mater is most likely an additional factor determining the superior clinical performance as discussed in the chapter. In Chapter 3 the technical performance of percutaneous leads for spinal cord stimulation having various contact spacings is modeled and compared. While reducing the contact spacing the recruitment of dorsal column fibers is increased, though with larger energy expenditure. The effect on the recruitment of dorsal root fibers is opposite. Modelling predicts that a displacement of the lead dorsally or laterally in the dorsal epidural space results in a reduction of the ratio of dorsal root and dorsal column fiber thresholds and thereby in less paresthesia coverage (when the stimulus is just at the discomfort threshold). When two leads are connected in parallel modelling predicts that dorsal column recruitment is reduced. This prediction has recently been confirmed empirically. Owing to technical advances the concept of electrical field steering has recently emerged in spinal cord stimulation. A method to electronically shift the electrical field rostrocaudally along the spinal cord, resulting in predictable changes in the population of recruited sensory nerve fibres and thus the position of paresthesia is described in Chapter 4. It is predicted that a smaller contact spacing allows a better control of the stimulation field. This may be crucial in the treatment of pain syndromes such as chronic low-back pain, as discussed in this chapter. In Chapter 5, the first model of electrical stimulation of human motor cortex is introduced. The region of the precentral gyrus and surrounding sulci with an electrode array (paddle lead) on the overlying dura mater represents the core of the model’s volume conductor. A simple fiber model was taken from the spinal cord stimulation model and used to assess the neural response to the stimulation induced field. It was predicted that the thickness of the cerebrospinal fluid layer underneath the active contact influences the stimulation threshold considerably. In addition, nerve fibers oriented perpendicular to an anode above the cortical surface could be activated at a low amplitude. As a consequence, an anode cannot be considered an indifferent contact in motor cortex stimulation, contrary to common belief. A refined and extended model of motor cortex stimulation is described in Chapter 6. When the myelinated axon model was extended with a frustrum and another cylindrical membrane structure, representing a pyramidal cell and its apical dendrite, respectively, the stimulation threshold was generally reduced. By changing the polarity and/or position of the stimulating contact array, stimulation thresholds of pyramidal cells varied and the population of recruited cells was changed as well. The model predictions have been shown to be qualitatively in accordance with empirical data from the literature. Finally, the thesis is completed in Chapter 7 with a discussion of the most prominent effects in spinal cord and motor cortex stimulation that have been analyzed by computer modeling. In addition, recommendations for future theoretical and empirical studies are proposed.