Neuromodulation and Rehabilitative Interventions for the Spine

Past: Spinal Stimulation for Pain Control

Low back pain and leg pain are some of the most common conditions in the United States, affecting an estimated 15 million Americans. Furthermore, poorly controlled low back pain contributes to the opioid epidemic, which causes up to 60,000 deaths per year. One of the most accepted treatments of low back pain is spinal surgery, which itself is estimated to contribute over $80 billion toward total U.S. healthcare costs.

There are several treatment options for chronic low back pain and radiculopathy. Traditionally, spinal cord stimulation has been used when other treatments have failed. Indeed, the implantation of epidural electrodes to treat chronic back pain is the workhorse of modern clinical neuromodulation. The origin of spinal cord stimulation to treat back pain stems from studies performed as early as the 1970s that showed the analgesic effect of subdural spinal stimulation for chronic low back pain. ^, Indeed, the most often–quoted “first-in-human” use of spinal stimulation to treat chronic pain was carried out in 1967 by Shealy, who secured a stimulating electrode to the dorsum of the dura at T2 to treat a patient with recalcitrant cancer pain. The motivation behind the procedure at that time was the gate control theory of pain, which suggests that the stimulation activates a local interneuron pool within the spine, releasing local inhibitory neurotransmitters that dampen pain signals. Whether stimulation also initiates an ascending signal that mitigates the pain response in the brainstem or even the cortex has been postulated but is not currently known. The indications for spinal stimulation as an analgesic soon grew to incorporate recurrent radicular pain after failed back surgery, complex regional pain syndrome types I and II, and angina, among other sources of pain.

Regardless of mechanism, spinal stimulation proved efficacious to treat chronic pain. Specifically, for failed back surgery, which refers to continued low back pain after surgical treatment, several treatment options exist, including repeat surgery. To directly compare repeat surgery to spinal cord stimulation (SCS) for failed back surgery syndrome (FBSS), North and colleagues performed a randomized controlled trial (RCT) that showed that SCS led to more lasting pain control compared with repeat surgery. This seminal result was followed by other studies that further established SCS as a valid treatment for FBSS. Kumar and colleagues compared SCS to best medical therapy for pain analgesia and found that the stimulation group showed improvements on several quality of life metrics and other patient-reported outcomes, as well as other outcomes, including cost-effectiveness. ^,

The impact of these studies was widespread. From the perspective of chronic pain and FBSS, these data firmly established SCS in both the surgeon’s and the pain physician’s armamentarium. Indeed, a recent metaanalysis shows the robust effect of SCS for the treatment of chronic pain. From a societal standpoint, SCS fit well with the overall push to reduce opioid use secondary to the growing opioid epidemic. Finally, from a neuromodulation perspective, it was clear that spinal stimulation was a safe, efficacious, and cost-effective method of modulating spinal “circuits” for clinical purposes.

Present: Advances in Neuromodulation

From 2005 to 2010, several studies using randomized data established SCS as an effective method to control pain of spinal origin, resulting in an exponential increase in the number of studies investigating spinal stimulation ( Fig. 76.1 ).

The rapid rate of research in this area has led to several new technological developments that have increased the efficacy of SCS, and also expanded the indications for SCS beyond pain control. In the sections below, we summarize some of the challenges with SCS and how major pivotal trials have overcome these and improved outcomes for SCS therapy.

In the complex circuitry of the spinal cord, the types of targets that could be stimulated are in part dependent on the frequency and pulse width of stimulation, as well as lead location and cerebrospinal fluid (CSF) thickness between the lead and spinal cord. Normal daily activities such as sitting, standing, lying down, and even breathing can change the thickness of the CSF between the surface of the spinal cord and epidural electrodes, leading to significant variability in neural stimulation throughout the day, with overstimulation leading some patients to abandon the therapy or reduce the amplitude to subtherapeutic levels in a desire to avoid risking overstimulation. The larger the CSF layer, the less impact the electric field has on the spinal cord, as electricity is shunted away from the spinal cord, and the spinal cord moves within the electric field, leading to large variability in neural recruitment. To help clarify this point, we present a simplified circuit diagram that models the stimulation delivered to the spinal cord. Fig. 76.2 illustrates the amount of stimulation to the cord (I _cord ) will be determined by the resistance in the cord relative to the resistance in the CSF (R _CSF ). Specifically, lowering R _CSF causes current to more easily flow into the CSF, reducing I _cord . When the patient is supine, the cord rests adjacent to the dura, and the CSF space between the cord and the stimulator decreases, which increases the resistance of R _CSF , thereby increasing I _cord . However, when the patient is upright, R _CSF is reduced, which shunts current away from the cord and reduces I _cord . Thus, delivering a constant current through the stimulator will result in a highly variable amount of current delivered to the cord, depending on the position of the patient.

In addition to the variable CSF layer ( Fig. 76.3 ), the level of stimulation needed to control low back pain was also often complicated by stimulating thoracic rootlets that caused discomfort. Guarded electrode arrays, current fractionation, and tripole paddle leads provided a greater ability to focus the electric field aimed to stimulate dorsal afferent fibers while avoiding activation of the thoracic rootlets. This improved coverage and success in treating postlaminectomy back pain. Yet, several clinical trials still reported that only approximately 50% of patients were able to achieve 50% reduction in their back pain scores. ^,

Recently an RCT of high-frequency (10-KHz) versus low-frequency stimulation reported 80% and 79% success rates with 10-kHz stimulation at 12 and 24 months, respectively. Because 10 kHz is outside the neural response rate, patients did not experience thoracic root stimulation at amplitudes that provided relief of back pain.

More recently, a novel approach to addressing the challenges of CSF variability and undesirable rootlet stimulation has been to use a closed-loop controller designed to target the neural response instead of the device output. By measuring the amplitude of a patient’s individual evoked compound action potential (ECAP) that correlates with the patient’s comfortable neural stimulation level in a controlled clinic environment, the stimulator can be automatically programed to lock in this level of comfortable neural activation in a closed-loop fashion, regardless of movement of the spinal cord in the electric field ( Fig. 76.4 ).

Fig. 76.5 shows a raw tracing of the ECAP and device output current in a traditional open-loop setting and again in the closed-loop setting during various activities. As noted in the open-loop setting in Fig. 76.5 , the amplitude varies by as much as 5-fold between the supine position and prone position or coughing. This increase in ECAP creates perceptual changes that are uncomfortable, leading the patient to turn down the amplitude of the stimulator, resulting in decreased time in the therapeutic window, less SCS utilization, and ultimately as much as a 10% to 15% loss of efficacy of the SCS over time. ^, In the closed-loop setting these variations in neural activation are virtually eliminated by automatic adjustment of the output current. The closed-loop controller is fast enough to accommodate even the rapid increase in intrathoracic pressure and consequently overstimulation caused by a cough. A recent double-blind RCT revealed that the closed-loop system was able to keep patients within their targeted therapeutic range, with more than 80% of the patients achieving greater than 50% reduction in their pain scores. More generally, closed-loop stimulation represents an exciting development with several potential applications, including more complex feedback control of the spinal circuitry, to create highly patterned activity in the spine.